ANTIMONY AND ACETALDEHYDE MIGRATION FROM NIGERIAN

AND BRITISH PET BOTTLES INTO WATER AND SOFT DRINKS

UNDER TYPICAL USE CONDITIONS
concentration of migrants and some trace elements in polyethylene
terephthalate and in bottled contents

Aminu TUKUR

Submitted for the degree of Doctor of Philosophy

School of Engineering, Design and Technology

University of Bradford

OCTOBER 2011

Aminu Tukur
Antimony and acetaldehyde migration from Nigerian and British PET bottles into water
and soft drinks under typical use conditions:
Concentration of migrants and some trace elements in polyethylene terephthalate and
in bottled contents
Keywords: antimony, acetaldehyde, trace elements, polyethylene terephthalate, bottled
water, leaching, reuse, microwave digestion, ICP-MS, headspace GC-FID

ABSTRACT
Polyethylene terephthalate (PET) is an excellent material for bottling water, beverages,
edible oils and other liquids because it is light, tough and transparent. PET bottles are
also extensively reused for storage of drinking water, beverages and other liquids and
for solar disinfection of microbiologically unsafe drinking water in the tropics. In spite
of the usefulness of PET bottles earlier works have reported leaching of antimony and
acetaldehyde from the bottle matrix into the liquid contents. Both antimony trioxide
and acetaldehyde belongs to Group 2B (possible carcinogens) in the International
Agency for Research on Cancer (IARC) carcinogen classification. Additionally
acetaldehyde associated with alcoholic beverages (derived from alcoholic beverage and
formed endogenously) has recently been upgraded to IARC Group 1 carcinogen
(carcinogenic to humans).
The research aims to assess the pattern and extent of antimony and acetaldehyde
migration from British and Nigerian polyethylene terephthalate bottles into bottle
contents under typical use and reuse conditions. The research compares the assessed
extents of migration with the current regulations to determine whether the maximum
acceptable levels of antimony and acetaldehyde are being exceeded and whether
current regulations might need to be reassessed.
To achieve these goals the pattern and extent of PET bottle use and reuse in Britain
and Nigeria were appraised through survey. The survey revealed that new bottles with
contents are typically stored prior to use for periods ranging between one and 7 days,
with Nigerians storing for longer periods than British respondents. However storage
of up to one year was reported. The extent of bottle reuse was high and similar for
the two countries. Nevertheless Nigerian respondents reuse bottles for longer
periods than British respondents. The survey findings together with relevant literature
were used to design laboratory experiments that assessed the extent of antimony and
acetaldehyde migration from PET bottles into water/beverages.
A total of 82 brands of bottled water and soft drinks in plastic and glass bottles and in
cartons were collected. A few samples from Nigeria in plastic pouches were collected.
Materials used in bottling including glass and plastic bottle materials, metal and plastic
bottle cap materials and plastic cap lining materials were collected. All samples were
collected in supermarkets and shops in Britain and Nigeria except drinking water from
taps which was collected in Britain only. Some bottles were aged for the purpose of
studying the impact of bottle aging on chemical migration. Other bottles were stored

with their contents to study the impact of long term storage of bottle contents on
chemical migration.
Energy dispersive X-ray spectrometry (EDX) and Raman spectroscopy were used to
characterise PET bottle material and other materials associated with water and soft
drink bottling. Antimony and other trace metals in water and soft drinks were
determined using Inductively Coupled Plasma-Mass Spectrometry (ICP-MS). Antimony
content of PET and other plastics was determined by microwave digestion and ICP-MS.
Acetaldehyde content of water and soft drinks and PET were determined using
headspace gas chromatography with flame ionisation detection (GC-FID). Accuracy
and precision for determination of antimony and other trace elements in bottle
materials and bottle contents were good as recoveries were around 100% and
coefficients of variation were less than 15% for all analysis types. Accuracy and
precision for determination of acetaldehyde in bottle materials and bottle contents
were also good as recoveries were around 100% and coefficients of variation were less
than 15% for all analysis types. Impact of long term storage, elevated temperatures,
bottle thickness, carbonation, bottle aging and bottle size on migration of antimony and
acetaldehyde were also assessed.
All plastic bottle materials analysed were found to be PET. Bottle cap materials were
either polyethylene or polypropylene. All plastic cap lining materials from Britain and
some from Nigeria were found to be ethylene vinyl acetate/polypropylene copolymer.
Plastic cap lining materials from some Nigerian soft drinks were identified as polyvinyl
chloride. Glass bottle materials analysed were found to be soda-lime glass. Metal bottle
caps were identified as tinplate, tin-free-steel coated with chromium or aluminium
coated with chromium.
The antimony concentration in 32 PET bottle materials from Britain and Nigeria were
similar and ranged between 177 and 310 mg/kg with an average of 25030 mg/kg. The
concentration agrees well with the industry reported concentration of between 150
and 350 mg/kg. The concentration of residual acetaldehyde in 25 fresh PET bottle
materials from Britain and Nigeria ranged between 0.95 and 12.52 g/g. The average
concentration in British and Nigerian soft drinks PET materials are 4.76 and 2.17g/g
respectively. Concentration of residual acetaldehyde was higher in soft drinks and still
water PET materials than in sparkling water materials. The concentration of residual
acetaldehyde decreases as the bottle wall material becomes older. Also the thinner the
bottle walls the lower the concentration of residual acetaldehyde.
Antimony concentration in 47 freshly purchased British bottled water and soft drinks
ranged between 0.03 and 6.61g/L with only one sample going above the EU
acceptable limit. Concentrations of other trace elements measured were low except
titanium which was detected at part per million levels in soft drinks. Lead content of a
Nigerian soft drink in glass bottle stored for 2 months was above the EU acceptable
limit for lead. At realistic temperatures of 40 and 60C antimony concentration in the
water remained below the EU acceptable limit even after 48 hours of exposure but the
concentration exceeded the limit for most exposures at 80C. Concentration of
antimony in some Nigerian bottled water and soft drinks was above the EU limit after
11 months of storage at room temperature. Aged bottles leach lower amount of
antimony than new bottles. Similarly larger bottles leach lower amount of antimony
than smaller bottles.

ii

The average acetaldehyde concentrations found in British fruit juices, carbonated soft
drinks, sparkling water and still water were 5113, 1458, 22 and 8 g/L respectively.
Acetaldehyde was not detected in water bottled in glass. The concentration of
acetaldehyde in five fruit juice samples in PET bottles and carton was beyond the EU
specific migration limit (SML) of 6mg/kg. Also the tolerable daily intake of acetaldehyde
could be exceeded as a result of intake of some soft drinks and fruit juices.
Acetaldehyde content in soft drinks increase with storage but the increase cannot be
accounted for by the residual acetaldehyde in PET. Acetaldehyde was found to be
outgassing from some bottles. It was also found to be capable of migrating from soft
drinks into bottle wall. Without replenishment the concentration of acetaldehyde in
solution decreases with time.
The use of PVC cap lining in Nigeria as found in this study is a cause for concern as
PVC is associated with health risk issues. The study recommends actions to ensure
that antimony in fruit juices and other bottled products remain within the regulatory
standard from bottling to consumption for the purpose of safeguarding the health of
consumers. Glass used in bottling should be well scrutinized to ensure that it does not
contain high levels of lead or other chemical substances that can cause harm to
consumers through migration into contents. PET bottles can safely be used for solar
water disinfection without the risk of antimony intake at concentrations above safe
limits as water temperature achievable as the result of the technique doesnt go
beyond 60C. Also aged bottles are safer to use than new bottles because their
chemical leaching was found to be lower than that of new bottles. This study
recommends the reassessment of the absence of international guidelines for
acetaldehyde in water and foods. The study also recommends that the amount of
acetaldehyde that can be added to soft drinks as flavouring agent should be below the
specific migration limit (SML) for migration of acetaldehyde from PET bottle into bottle
contents. This is essential since the SML was designed to ensure that exposure to
acetaldehyde, as a result of intake of bottled water and soft drinks in PET bottles, is
below the tolerable daily intake (TDI) for acetaldehyde. As antimony was reported to
go beyond the safe limits in some Nigerian bottled water and soft drinks after 11
months of storage this study discourages the use of bottle contents stored for a very
long time.

iii

DEDICATION
This work is dedicated to my mother Fatima, my elder brothers Tukur, Uwaisu and
Ahmed, my elder sister Halima, my wife Fatima and my son Muhammad.

iv

ACKNOWLEDGEMENT
I express my profound appreciation to my supervisors in the persons of Dr Liz Sharp
and Dr Ben Stern whose guidance and contribution to this work remained beyond
compare. I owe great thanks to Dr Chedly Tizaoui for his initial supervision of this
work and the guidance he provided throughout the work. I am also thankful to Mr
Dennis Farwell, Mr Stuart Fox and Mrs Belinda Hill for their guidance, enthusiasm and
their readiness to assist at all times during the experimental component of this work. I
am highly indebted to M B Tijjani and Ibrahim Garba whom assisted with the survey
work in Nigeria. My deepest thanks extend to my family members whose
encouragement and moral support kept me going. Very special thanks to the
Commonwealth Scholarship Commission in the United Kingdom for the ample
sponsorship without which this study would not have been possible. Thanks to all that
contributed to the success of this work but whose names are not mentioned here.

ABBREVIATIONS
ACGIH
Al
ANOVA
ATSDR
Be
bis-HET
C
CAGR
Cd
CFR
Cl
CO
CO2
Cr
CWQG
DEHA
DEHP
DMT
DNA
EAWAG
EC
EDX
EFSA
EG
EPA
EU
EVA/PP
FAAS
FACS
FAO
FDA
Fe
FSA
GC-FID
GC-MS
GDP
Ge
HCl
HDPE
HPLC
IARC
ICP-MS
IDLH
IMF
INAA
IPCS
ISBM

American Conference of Industrial Hygienists

Aluminium
Analysis of variance
Agency for Toxic Substances and Disease Registry
Beryllium
Bis-hydroxyethyl terephthalate
Carbon
Cumulative annual growth rate
Cadmium
Code of Federal Regulations
Chlorine
Cobalt
Carbon dioxide
Chromium
Canadian Drinking Water Quality Guidelines
Bis(2-ethylhexyl) adipate
Bis(2-ethylhexyl)phthalate/ dioctyl phthalate
Dimethyl terephthalate
Deoxyribonucleic acid
Swiss Federal Institute for Environmental Science and
Technology)
European Commission
Energy Dispersive X-ray
European Food Safety Authority
Ethylene glycol
Environmental Protection Agency
European Union
Ethylene vinyl acetate/polypropylene copolymer
Flame atomic absorption spectrometry
Food Advisory Consumer Service
Food and Agriculture Organization
Food and Drug Administration
Iron
Food Standard Agency
Gas chromatographyFlame ionization detector
Gas chromatographymass spectrometry
Growth domestic product
Germanium
Hydrochloric acid
High density polyethylene
High Performance Liquid Chromatography
International Agency for Research on Cancer
Inductively coupled plasma mass spectrometry
Immediately Dangerous to Life and Health
International Monetary Fund
Instrumental Neutron Activation Analysis
International Programme on Chemical Safety
Injection Stretch Blow Moulding
vi

M
MAC
max
MCL
Mg
min
Mn
m/z
N
NAFDAC
nd
NIOSH
O
OSHA
P
Pb
PE
PEL
PET
PETRA
PP
ppb
ppm
PTFE
PVC
r
r2
Sb
SD
SEM
Si
SML
Sn
SODIS
t
TDI
TLV-C
TPA
Ti
UAE
US
USGS
UV
V
V
WRAP
W
WHO
wt
XRF

Mean
Maximum Admissible Concentration
maximum
Maximum Contamination Level
Magnesium
minimum
Manganese
mass-to-charge ratio
Nitrogen
National Agency for Food and Drug Administration and Control
not detected
National Institute for Occupational Safety and Health
Oxygen
Occupational Safety and Health Administration
Phosphorus
Lead
Polyethylene
Permissible Exposure Limit
Polyethylene terephthalate
PET Resin Association
Polypropylene
Part per billion
Part per million
Polytetrafluoroethylene
Polyvinyl chloride
Pearsons correlation
coefficient of determination
Antimony
Standard deviation
Scanning Electron Microscope
Silicon
Specific Migration Limit
Tin
Solar Water Disinfection
Student's t-test
Tolerable Daily Intake
Threshold Limit Value - Ceiling
Terephthalic acid
Titanium
United Arab Emirate
United States
United States Geological Survey
Ultra violet
Vanadium
Volt
Waste & Resources Action Programme
Watt
World Health Organisation
weight
X-Ray Fluorescence Spectroscopy
vii

Zn
Zr

Zinc
Zirconium

viii

SYMBOLS
cm-1
gcm-3
gmol-1
L
mg/kg
mg/L
mg/ml
ml
m
mm
ng/L
Pa.s
g/g
g/L
g/m3
g/kg/day
l
m
2
2

per centimetre
gram per centimetre cube
gram per mole
litre
milligram per kilogram
milligram per litre
milligram per millilitre
milliltre
metre
millimetre
nanogram per litre
pascal-second
microgram per gram
microgram per litre
microgram per metre cube
microgram per kilogram per day
microlitre
micrometre
Pearson's chi-square
Cramers phi coefficient

ix

TABLE OF CONTENTS
ABSTRACT .........................................................................................................................................................I
DEDICATION ................................................................................................................................................ IV
ACKNOWLEDGEMENT .............................................................................................................................. V
ABBREVIATIONS........................................................................................................................................... VI
SYMBOLS ........................................................................................................................................................ IX
TABLE OF CONTENTS ............................................................................................................................... X
LIST OF FIGURES .......................................................................................................................................XIV
LIST OF TABLES ....................................................................................................................................... XVII
CHAPTER 1: INTRODUCTION ................................................................................................................ 1
1.1
1.2
1.3
1.4

POLYETHYLENE TEREPHTHALATE BOTTLES AND CHEMICAL MIGRATION ................................... 1

JUSTIFICATION FOR RESEARCH ......................................................................................................... 1
AIMS AND OBJECTIVES OF RESEARCH ............................................................................................... 3
THESIS STRUCTURE ............................................................................................................................. 3

CHAPTER 2: MATERIALS ASSOCIATED WITH BOTTLING ........................................................... 5

2.1
2.2
2.3
2.4
2.5
2.6

INTRODUCTION ................................................................................................................................. 5
POLYETHYLENE TEREPHTHALATE (PET) .......................................................................................... 5
GLASS ................................................................................................................................................. 10
OTHER PLASTICS ............................................................................................................................... 11
SUMMARY........................................................................................................................................... 11
CONCLUSION ................................................................................................................................... 12

CHAPTER 3: BOTTLED WATER AND SOFT DRINKS USE AND BOTTLE REUSE ................ 13
3.1
3.2
3.3
3.4
3.5
3.6
3.7
3.8
3.9
3.10
3.11
3.12
3.13

INTRODUCTION ............................................................................................................................... 13
BOTTLED WATER AND SOFT DRINK USE ....................................................................................... 13
FACTORS INFLUENCING BOTTLED WATER AND SOFT DRINKS USE........................................... 16
BOTTLED WATER AND SOFT DRINKS REGULATIONS .................................................................. 19
BOTTLED WATER AND SOFT DRINKS CATEGORISATION............................................................ 20
THE SIGNIFICANCE OF SHELF LIFE ................................................................................................... 22
BOTTLE REUSE IN NIGERIA, BRITAIN AND OTHER COUNTRIES .................................................. 22
BOTTLE QUALITY REGULATION...................................................................................................... 26
FACTORS INFLUENCING PET BOTTLE REUSE PATTERN ............................................................... 26
PET BOTTLE REUSE AND BACTERIAL CONTAMINATION ............................................................. 27
PET BOTTLE USE AND REUSE AND THE SAFETY DEBATE ............................................................. 27
SUMMARY........................................................................................................................................... 28
CONCLUSION ................................................................................................................................... 29

CHAPTER 4: CHEMICALS MIGRATING FROM PET BOTTLES..................................................... 30

4.1
4.2
4.3
4.4

INTRODUCTION ............................................................................................................................... 30
ANTIMONY ........................................................................................................................................ 30
ACETALDEHYDE ................................................................................................................................ 33
BASIS AND IMPLICATIONS OF WATER QUALITY STANDARDS ..................................................... 36

4.5
4.6
4.7
4.8
4.9

THE IMPLICATION OF IARCS POSSIBLE CARCINOGEN STATUS ................................................. 37

LEACHING OF ANTIMONY AND ACETALDEHYDE INTO WATER AND FOODS.......................... 37
FACTORS AFFECTING THE LEACHING OF CHEMICALS ................................................................. 41
SUMMARY........................................................................................................................................... 45
CONCLUSION ................................................................................................................................... 46

CHAPTER 5: METHODOLOGY .............................................................................................................. 47

5.1
5.2
5.3
5.4
5.5
5.6
5.7
5.8
5.9
5.10

RESEARCH STRATEGY ....................................................................................................................... 47

SURVEY STRATEGY AND SAMPLING ................................................................................................ 51
SURVEY INFLUENCE ON LABORATORY ANALYSIS ......................................................................... 56
MATERIALS AND EXPERIMENTAL METHODS .................................................................................. 57
SAMPLES AND REAGENTS ................................................................................................................. 57
IDENTIFICATION OF PLASTIC MATERIALS BY RAMAN SPECTROSCOPY ...................................... 60
IDENTIFICATION OF METAL CAPS AND GLASS BOTTLE BY EDX ................................................ 64
QUANTITATION OF ANTIMONY IN WATER, SOFT DRINKS AND PET ....................................... 67
QUANTITATION OF ACETALDEHYDE IN WATER, SOFT DRINKS AND PET ............................... 75
RESEARCH TIMELINE ......................................................................................................................... 81

CHAPTER 6: SURVEY RESULTS AND DISCUSSION ........................................................................ 83

6.1
6.2
6.3
6.4
6.5
6.6
6.7
6.8
6.9
6.10
6.11
6.12
6.13
6.14
6.15
6.16

INTRODUCTION ............................................................................................................................... 83
REPORTED USE OF BOTTLED WATER AND SOFT DRINKS ............................................................ 83
STORAGE AND USE OF BOTTLED WATER AND SOFT DRINKS .................................................... 83
BOTTLE SIZE ...................................................................................................................................... 85
BOTTLE CONTENT TYPES ................................................................................................................ 86
STORAGE DURATIONS FOR PURCHASED BOTTLED WATER AND SOFT DRINKS ...................... 86
STORAGE PLACES .............................................................................................................................. 89
PET BOTTLE REUSE ........................................................................................................................... 90
PET BOTTLE REUSE DURATIONS ..................................................................................................... 92
NUMBER OF BOTTLES BEING REUSED IN PLACES OF RESIDENCE................................................. 95
SIZE OF REUSED BOTTLES................................................................................................................. 96
REUSE SAFETY PERCEPTION AND SAFETY DEBATE ........................................................................ 97
OTHER ISSUES AFFECTING REUSE .................................................................................................... 99
SUMMARY......................................................................................................................................... 101
IMPLICATION OF THE SURVEY RESULTS ON LABORATORY WORK ........................................... 101
CONCLUSION ................................................................................................................................. 105

CHAPTER 7: IDENTIFICATION OF MATERIALS USED IN BOTTLING .................................. 106

7.1
7.2
7.3
7.4
7.5
7.6
7.7
7.8
7.9

INTRODUCTION ............................................................................................................................. 106

PLASTIC BOTTLE MATERIALS.......................................................................................................... 106
POWDERY SUBSTANCE FROM PET MATERIAL DIGESTION ........................................................ 107
GLASS BOTTLE MATERIALS ............................................................................................................. 108
PLASTIC BOTTLE CAP MATERIALS.................................................................................................. 110
METAL CROWN AND SCREW CAP MATERIALS............................................................................ 111
BOTTLE CAP LININGS ..................................................................................................................... 113
SUMMARY......................................................................................................................................... 117
CONCLUSION ................................................................................................................................. 118

xi

CHAPTER 8: MIGRATION OF ANTIMONY AND OTHER ELEMENTS: RESULTS,

DISCUSSION AND SUMMARY.............................................................................................................. 119
8.1
8.2
8.3
8.4
8.5
8.6
8.7
8.8
8.9
8.10
8.11
8.12
8.13
8.14
8.15
8.16

INTRODUCTION ............................................................................................................................. 119

CALIBRATION CURVES ................................................................................................................... 119
LIMITS OF DETECTION.................................................................................................................... 121
SAMPLE CODING ............................................................................................................................. 121
ANTIMONY IN DIGESTED PET AND OTHER PLASTICS ............................................................... 121
ANTIMONY AND OTHER TRACE METALS IN WATER AND SOFT DRINKS................................. 128
STORAGE AND ANTIMONY MIGRATION...................................................................................... 136
WATER TEMPERATURE ELEVATION ON EXPOSURE TO SUNLIGHT ........................................... 140
ANTIMONY MIGRATION FROM PET AND GLASS AT ELEVATED TEMPERATURES .................... 141
DEPENDENCY OF ANTIMONY MIGRATION ON ITS CONCENTRATION IN PET...................... 147
DEPENDENCY OF ANTIMONY MIGRATION ON PET BOTTLE THICKNESS ............................... 147
BOTTLE AGING AND ANTIMONY MIGRATION............................................................................ 149
WATER PH AND ANTIMONY MIGRATION .................................................................................. 151
BOTTLE SIZE AND ANTIMONY MIGRATION................................................................................. 151
SUMMARY......................................................................................................................................... 152
CONCLUSION ................................................................................................................................. 154

CHAPTER 9: ACETALDEHYDE MIGRATION: RESULTS, DISCUSSION AND SUMMARY. 155

9.1
9.2
9.3
9.4
9.5
9.6
9.7
9.8
9.9
9.10
9.11
9.12
9.13
9.14
9.15
9.16
9.17
9.18

INTRODUCTION ............................................................................................................................. 155

CALIBRATION AND RETENTION TIME .......................................................................................... 155
LIMITS OF DETECTION.................................................................................................................... 157
ACCURACY AND PRECISION ......................................................................................................... 157
SAMPLE CODING ............................................................................................................................. 158
RESIDUAL ACETALDEHYDE IN PET BOTTLES MATERIALS ........................................................... 158
BOTTLE AGING AND ACETALDEHYDE CONCENTRATION IN PET .......................................... 162
BOTTLE THICKNESS AND ACETALDEHYDE CONCENTRATION IN PET ................................... 164
EFFECT OF CARBONATION ON ACETALDEHYDE QUANTITATION .......................................... 166
ACETALDEHYDE IN WATER AND SOFT DRINKS.......................................................................... 166
CONDUCTIVITY AND ACETALDEHYDE CONCENTRATION ...................................................... 171
STORAGE AND ACETALDEHYDE MIGRATION INTO BOTTLE CONTENTS ................................ 172
ACETALDEHYDE IN PET MATERIALS AND IN BOTTLE CONTENTS ........................................... 175
ACETALDEHYDE OUTGASSING FROM PET BOTTLES.................................................................. 176
ACETALDEHYDE MIGRATION FROM WATER MEDIUM INTO PET ............................................. 177
STABILITY OF ACETALDEHYDE WITH STORAGE .......................................................................... 179
SUMMARY......................................................................................................................................... 181
CONCLUSION ................................................................................................................................. 182

CHAPTER 10: CONCLUSION AND FURTHER WORK ................................................................ 184

10.1
10.2
10.3
10.4
10.5
10.6

INTRODUCTION ............................................................................................................................. 184

MAIN FINDINGS IN RELATION TO OBJECTIVES ............................................................................ 185
OTHER FINDINGS............................................................................................................................ 192
IMPLICATION OF RESULTS .............................................................................................................. 193
RESEARCH STRENGTHS AND LIMITATIONS.................................................................................. 196
FURTHER WORK ............................................................................................................................. 197

REFERENCES ............................................................................................................................................... 200

xii

APPENDIX 1 SURVEY QUESTIONS ..................................................................................................... 212

APPENDIX 2 ACETALDEHYDE INTAKE FROM SOFT DRINKS ................................................. 214

xiii

LIST OF FIGURES
Figure 1.1 Life cycle of PET bottle..................................................................................................... 2
Figure 2.1PET synthesis ..7
Figure 2.2 PET bottle injection stretch blow moulding process .......................................................... 8
Figure 2.3 PET bottle preforms.................................................................................................................... 8
Figure 2.4 Cleavage of ester bonds in PET ................................................................................................ 9
Figure 2.5 Acetaldehyde formation by regeneration of the PET hydroxyl end group .................... 9
Figure 2.6 Acetaldehyde formation by hydrolysis of vinyl end group ................................................. 9
Figure 2.7 Acetaldehyde formation through chain rebuild .................................................................... 9
Figure 2.8 Acetaldehyde formation through thermal scission ............................................................ 10
Figure 3.1 Global per capita consumption of bottled water in 1999 and 2004..14
Figure 3.2 Per capita consumption of bottled water in 2004 by regions ......................................... 15
Figure 3.3 Nigerian 500ml packaged water in polyethylene pouch ................................................... 18
Figure 3.4 Storage of drinking water in used PET bottles in a refrigerator in Nigeria ................. 24
Figure 3.5 Medicinal herbal concoctions vended in used PET bottles in Nigeria........................... 24
Figure 3.6 SODIS in an African setting ..................................................................................................... 25
Figure 3.7 SODIS application worldwide (Eawag, 2008) ...................................................................... 25
Figure 4.1 Relationship between antimony, PET and bottle contents39
Figure 4.2 Relationship between acetaldehyde, PET and bottle contents ....................................... 41
Figure 5.1 Vibrational energy level diagram61
Figure 5.2 Relationship between Raman spectra and Rayleigh scattering........................................ 62
Figure 5.3 Schematic of Raman spectroscopy ....................................................................................... 63
Figure 5.4 InVia Reex dispersive Raman microscope ......................................................................... 64
Figure 5.5 Schematic of SEM showing the position of the X-ray detector...................................... 65
Figure 5.6 Interaction of electron beam and sample............................................................................. 66
Figure 5.7 SEM-EDX Model Quanta 400 ................................................................................................. 67
Figure 5.8 Schematic of ICP-MS ................................................................................................................. 68
Figure 5.9 Quadrupole mass analyser ....................................................................................................... 68
Figure 5.10 Digested PET materials .......................................................................................................... 72
Figure 5.11 Schematic of Gas chromatography...................................................................................... 76
Figure 5.12 Schematic of Flame Ionisation Detector ............................................................................ 77
Figure 6.1 Bottled water and soft drinks availability...84
Figure 6.2 Sizes of bottles in places of residences ................................................................................. 85
Figure 6.3 Durations of storage for unopened bottles in places of residence................................ 87
Figure 6.4 Longest reported storage durations for unopened bottles ............................................. 88
Figure 6.5 Longest reported storage durations for opened bottles ................................................. 88

xiv

Figure 6.6 PET bottles reuse durations for British respondents........................................................ 93

Figure 6.7 PET bottles reuse durations for Nigerian respondents .................................................... 94
Figure 6.8 Number of bottles being reused in places of residence ................................................... 96
Figure 6.9 Sizes of reused bottles .............................................................................................................. 97
Figure 6.10 PET bottle reuse safety perception ..................................................................................... 98
Figure 6.11 Other issues influencing reuse ........................................................................................... 100
Figure 7.1 Raman spectra for colourless still water bottle material..106
Figure 7.2 Raman spectra for apple juice bottle material ................................................................. 107
Figure 7.3 Raman spectra for white precipitate from still water bottle digestion ....................... 108
Figure 7.4 Typical EDX spectra for soft drinks glass bottle materials ............................................ 109
Figure 7.5 Raman spectra for 3 bottled water bottle cap materials ............................................... 110
Figure 7.6 Raman spectra for 3 carbonated drinks bottle cap materials ....................................... 111
Figure 7.7 EDX spectra for British glass bottle screw cap ................................................................ 112
Figure 7.8 Typical EDX spectra for Nigerian glass bottle crown cap ............................................. 113
Figure 7.9 Typical EDX spectra for British glass bottle crown cap ................................................. 113
Figure 7.10 Raman spectra for British and some Nigerian cap lining materials ........................... 114
Figure 7.11 Raman spectra of a flexible Nigerian PET bottle cap lining material ......................... 115
Figure 7.12 Typical EDX spectra for PET bottle cap lining and a rigid PVC material ................. 115
Figure 7.13 Raman spectra of phthalate plasticisers ........................................................................... 116
Figure 7.14 Raman spectra for Nigerian glass bottle crown caps cap lining ................................. 117
Figure 8.1 Typical calibration curve for antimony.....119
Figure 8.2 Typical calibration curves for other trace elements........................................................ 120
Figure 8.3 Bar chart illustrating antimony concentrations in Nigerian PET bottles. ................... 126
Figure 8.4 Bar chart illustrating antimony concentrations in British PET bottles. ....................... 126
Figure 8.5 Bar chart illustrating antimony concentration in bottle contents ................................ 131
Figure 8.6 Antimony concentration in similar brands of still and sparkling water....................... 132
Figure 8.7 Conductivity and antimony concentration in bottle contents ...................................... 132
Figure 8.8 Relationship between pH and antimony concentration in bottle contents ............... 133
Figure 8.9 Change in antimony concentration with time in Nigerian samples ............................. 137
Figure 8.10 Change in lead concentration with time in a Nigerian soft drinks ............................ 138
Figure 8.11 Change in antimony concentration (g/L) with time in British samples .................. 139
Figure 8.12 Sunlight-assisted bottled water temperature elevation ................................................ 140
Figure 8.13 Impact of exposure temperature on antimony migration (sample UPA6) .............. 142
Figure 8.14 Impact of exposure time on antimony migration (sample UPA6) ............................. 142
Figure 8.15 Comparison of antimony migration in 2 PET and 2 glass bottles at 60C............... 143
Figure 8.16 Lead and antimony migration in 2 glass bottles at 60C .............................................. 143

xv

Figure 8.17 Antimony leaching propensity in PET and glass bottles at 60C for 6 hours ......... 144
Figure 8.18 Antimony leaching propensity for PET bottles at 40, 60 and 80C .......................... 146
Figure 8.19 Antimony migration in new and one year aged PET bottles (UPA12) ..................... 150
Figure 8.20 Antimony migration at 60C in new and one year aged PET bottles (UPA12) ...... 150
Figure 8. 21 Dependency of antimony migration on water pH........................................................ 151
Figure 8.22 Migration of antimony into bottles of different sizes at 70C .................................... 152
Figure 9.1 Calibration curve for acetaldehyde in bottled water...155
Figure 9.2 Calibration curve for acetaldehyde in soft drinks ............................................................ 156
Figure 9.3 Calibration curve for residual acetaldehyde in PET materials....................................... 156
Figure 9.4 Typical acetaldehyde GC-FID chromatogram................................................................... 156
Figure 9.5 Average acetaldehyde concentration in soft drinks PET materials .............................. 160
Figure 9.6 Average residual acetaldehyde in British PET bottles ..................................................... 161
Figure 9.7 Acetaldehyde concentration in new and aged PET bottle materials ........................... 162
Figure 9.8 Change in acetaldehyde concentration in Nigerian bottled water PET...................... 163
Figure 9.9 Change in acetaldehyde concentration in Nigerian soft drink PET ............................. 163
Figure 9.10 Relationship between bottle wall thickness and residual acetaldehyde.................... 164
Figure 9.11 Bottle wall thickness and residual acetaldehyde in British PET .................................. 165
Figure 9.12 Bottle wall thickness and residual acetaldehyde in Nigerian PET .............................. 165
Figure 9.13 Average acetaldehyde concentrations by bottle content type ................................... 167
Figure 9.14 Average acetaldehyde concentrations by bottle content and packaging type ........ 167
Figure 9.15 Acetaldehyde concentration in Nigerian and British still water ................................. 170
Figure 9.16 Acetaldehyde content in a cola drink brand from Nigeria and Britain ..................... 170
Figure 9.17 Acetaldehyde in 5 British brands marketed as still and sparkling water .................. 171
Figure 9.18 Conductivity and acetaldehyde concentration in bottled water ................................ 172
Figure 9.19 Change in acetaldehyde content in British bottled water ........................................... 173
Figure 9.20 Change in acetaldehyde content in a fresh Nigerian cola drink brand ..................... 175
Figure 9. 21 Acetaldehyde concentration in PET and its presence in bottled water .................. 176
Figure 9.22 Outgassing of acetaldehyde from empty PET bottles ................................................... 177
Figure 9.23 Absorption of acetaldehyde by PET .................................................................................. 177
Figure 9.24 Residual acetaldehyde in PET and the acetaldehyde in the carbonated drinks ...... 178
Figure 9.25 stability of 500g/L aqueous acetaldehyde solution ...................................................... 180
Figure 9.26 stability of 5000g/L aqueous acetaldehyde solution .................................................... 181

xvi

LIST OF TABLES
Table 3.1 British bottled water categorisation ....................................................................................... 21
Table 4.1 Properties and uses of antimony..30
Table 4.2 Antimony concentration in the environment....................................................................... 31
Table 4.3 Properties and uses of acetaldehyde ...................................................................................... 34
Table 4.4 Antimony in PET material ......................................................................................................... 38
Table 4.5 Antimony in bottled water ....................................................................................................... 39
Table 4.6 Acetaldehyde in bottled water and soft drinks .................................................................... 40
Table 5.1 The role of the survey in defining laboratory experiments.56
Table 5. 2 Summary of samples collected and analysis carried out ................................................... 59
Table 5.3 Elements of interest in certified water reference material ............................................... 71
Table 5.4 Bottles used in sunlight exposure experiments ................................................................... 73
Table 5. 5 Research timeline....................................................................................................................... 82
Table 6.1 British and Nigerian respondents PET bottle reuse status91
Table 6.2 British and Nigerian respondents PET bottle reuse durations ....................................... 93
Table 6.3 Reuse duration cumulative frequency table .......................................................................... 94
Table 6.4 Influence of survey results on laboratory experiments ................................................... 102
Table 7.1 Relative abundance of elements in glass bottle materials......109
Table 7.2 chemical constituents of metal crown and screw caps for glass bottles ..................... 112
Table 8.1 Limits of detection...121
Table 8.2 Accuracy determination using polyethylene reference material ................................... 123
Table 8.3 Precision determination using 14 replicates of PET sample UPA3 ............................... 123
Table 8.4 Terephthalic acid yield ............................................................................................................. 124
Table 8.5 Antimony in digestion blanks ................................................................................................. 124
Table 8.6 PET samples colours................................................................................................................. 125
Table 8.7 Antimony in other plastics ...................................................................................................... 127
Table 8.8 Recovery in spiked samples .................................................................................................... 128
Table 8.9 Accuracy determination usingTM-DWS.2 certified water reference........................... 128
Table 8.10 Antimony in tap water, bottled water and soft drinks .................................................. 129
Table 8.11 Antimony in PET- and glass-bottled contents .................................................................. 129
Table 8.12 Guidelines and standards for antimony and other trace metals in drinking water . 130
Table 8.13 Summary statistics for the measured trace elements .................................................... 134
Table 8.14 Concentration of antimony and trace elements in water and soft drinks ................ 135
Table 8.15 Change in antimony concentration (g/L) with time in Nigerian samples ................ 137
Table 8.16 Change in antimony concentration (g/L) with time in British samples .................... 139
Table 8.17 Dependency of antimony migration on its concentration in PET ............................... 147

xvii

Table 8.18 PET bottle thickness............................................................................................................... 148

Table 8.19 Dependency of antimony migration on PET bottle thickness ...................................... 149
Table 9.1 Limits of detection..157
Table 9.2 Accuracy and precision for acetaldehyde determination in water................................ 157
Table 9.3 Accuracy and precision for determination of residual acetaldehyde in PET .............. 158
Table 9.4 Residual acetaldehyde in fresh British and Nigerian PET bottle materials .................. 159
Table 9.5 Residual acetaldehyde in Nigerian PET ................................................................................ 161
Table 9.6 Effect of carbonation on acetaldehyde quantitation ......................................................... 166
Table 9.7 Acetaldehyde concentration in British and Nigerian bottled water and soft drinks 168
Table 9.8 Change in acetaldehyde content in stored British bottled water ................................. 173
Table 9.9 Change in acetaldehyde content in stored Nigerian bottled water .............................. 174

xviii

CHAPTER 1:
1.1

INTRODUCTION

Polyethylene terephthalate bottles and chemical migration

Polyethylene terephthalate (PET) bottles are excellent containers for water, beverages,
edible oils and other liquids because they are light, tough and clear. PET bottles are
also extensively reused for storage of drinking water, beverages and other liquids and
for solar disinfection of microbiologically unsafe drinking water in the tropics. In spite
of the usefulness of PET bottles earlier works have reported leaching of antimony and
acetaldehyde from the bottle into the liquid contents. Antimony leaches from PET
because its compounds are used as catalyst in the manufacture of the polymer at the
concentration of 150-350 mg/kg (EFSA, 2004; Thiele, 2004; Westerhoff et al, 2008).
Acetaldehyde is produced in the polymer matrix as a result of thermal degradation of
the polymer during the melt processing stage in bottle manufacture (El-Toufaili, 2006,
Ewender and Welle, 2008). Both antimony trioxide and acetaldehyde belongs to
Group 2B (possible carcinogens) in the International Agency for Research on Cancer
(IARC) carcinogen classification. Additionally, acetaldehyde associated with alcoholic
beverages (derived from alcoholic beverage and formed endogenously) has recently
been upgraded to IARC Group 1 carcinogen (carcinogenic to humans).

1.2

Justification for research

Migration of antimony and acetaldehyde from the wall of PET bottles into the bottle
contents is an issue that warrants thorough investigation for the purpose of
safeguarding the health of users. The PET bottle lifecycle is shown in Figure 1.1.
Potential exposure to leached material might occur at the stage when the bottle is
used to contain the beverage in which it was sold. Equally, this might occur during reuse. Antimony and acetaldehyde leaching propensity has been studied for PET bottles
from some countries but not for British and Nigerian samples. The study is unusual in
selecting laboratory conditions that mirror the way that these bottles are typically
used. This was achieved by using some of the survey information on the pattern and
extent of bottled water and soft drink use and PET bottle reuse to design the
laboratory investigations on the migration dynamics of the migrant chemicals. The
laboratory investigations assess likelihood of consumption of the migrant chemicals
above the safe limits base on typical usage behaviour. In spite of the importance of
information on bottled water and soft drink use and PET bottle reuse, this information

remains scanty in the literature. In particular only one American study carried out
some work on PET bottle usage patterns for the purpose of investigating the safety of
bottle reuse.

Figure 1.1 Life cycle of PET bottle

2

1.3

Aims and objectives of research

1.3.1

Aim of the research

The research aims to assess the pattern and extent of antimony and acetaldehyde
migration from British and Nigerian polyethylene terephthalate (PET) bottles into
bottle contents under typical use and reuse conditions in relation to current
regulations and controls.

1.3.2

Objectives of the research

1. To examine the pattern and extent of bottled water and soft drink use and PET
bottle reuse in Britain and Nigeria.
2. To assess the extent of antimony and acetaldehyde migration from PET into
water and soft drinks under conditions representative of real use and reuse in
Britain and Nigeria
3. Drawing on findings from 1 and 2, as well as existing regulations and controls
to assess whether acceptable limits of chemicals in foods and water are being
exceeded as a result of the chemical migrations
4. To generate recommendations about the extent to which existing regulations
and controls might merit re-examination

1.4

Thesis structure

This thesis is partitioned into 10 chapters, bibliography and appendices. Chapters 2-4
provide an introduction to the current knowledge about the nature and usage of PET
and glass bottles and about antimony and acetaldehyde migration, their health effect on
human body and how they are regulated in drinking water and foods. Chapter 5
(Methodology) gives an account of the approaches employed in identifying the sampling
frame for the survey, the different considerations that guided the survey questions
content and how the sampling process was carried out. The Chapter also describes
the experimental methods used to assess the concentration of antimony and
acetaldehyde in fresh bottled water and soft drink samples and samples exposed
different conditions and the experiments used to characterise the materials associated
with water and soft drinks bottling. Chapter 6 (Survey results) presented the analysis
of survey data and also the interpretation of the results obtained.

Chapter 7

(Identification of materials used in bottling) describes the chemical nature of the

different materials associated with still water and soft drink bottling based on the
3

experiments carried out in Chapter 5. Chapter 8 (Migration of antimony and other

elements) analyses and discusses the results obtained in experiments assessing the
migration antimony and other elements from PET and glass bottles into bottle
contents. Chapter 9 (Migration of acetaldehyde) analyses and discusses the results
obtained in experiments assessing the migration acetaldehyde from PET into bottle
contents. Chapter 10 (Conclusion) re-examines the objectives of the study and
discusses the findings and their implication. The Chapter reflects on the strengths and
limitations associated with the research and conclude by exploring potential areas for
further research.

CHAPTER 2:
2.1

MATERIALS ASSOCIATED WITH BOTTLING

Introduction

This Chapter describes the materials associated with bottling of water and soft drinks.
Particular emphasis is given to materials that are in contact with the bottle contents
namely bottling materials, cap materials and cap lining materials. The Chapter attempts
to explain the processes through which antimony and acetaldehyde become associated
with the bottle materials. Bottled water and soft drinks are principally bottled in PET
bottles and to lesser extent glass bottles. In Britain, for example, 93% of bottled water
is marketed in plastic bottle (Bottled Water Information Office, 2008). Polyethylene
bottles, metal cans and paperboard cartons are also widely used for soft drinks.
However this research is primarily interested in PET bottles and to lesser extent glass
bottles for comparison. Materials associated with the PET and glass bottling process
are PET and glass as bottle materials, plastics used as bottle caps, cap liners and label
materials, including polyethylene; polypropylene and polyvinyl chloride, paper used in
labelling, pigments and dyes for labelling and colouring of bottle parts, adhesives for
securing several components of bottles, metals as metal bottle caps, including
aluminium and steel usually coated with tin or chromium and lacquers applied to metal
caps to provide a durable finish. However the only materials that are in direct contact
with bottled contents are the bottle materials, the cap and the cap liner. Additionally
the only material documented to be releasing acetaldehyde and antimony into bottle
content is PET bottle material. While glass has not been documented to release
acetaldehyde Shotyk et al (2006) has associated it with leaching of antimony though to
lesser extent than PET.

2.2
2.2.1

Polyethylene terephthalate (PET)

Description of PET

PET is a long-chain thermoplastic polymer of the polyester family used in several

applications. It is one of the most important raw materials used in man-made fibres.
Bottle production accounts for around 30% of global demand of PET (McCarthy,
2007). In 1995, 2.9 million tonnes of PET were consumed worldwide in packaging
applications across a wide range of areas including bottles for carbonated drinks,
mineral water, edible oil, cosmetics, surfactants, films for thermoforming applications
and packaging tape, etc (Azapagic et al, 2003).

2.2.2

PET synthesis

PET homopolymer is synthesised from ethylene glycol (EG) and either dimethyl
terephthalate (DMT) or terephthalic acid (TPA) (Figure 2.1), all of which are derived
from crude oil. Ethylene glycol is generated from ethylene. Para-xylene (P-xylene) is
either oxidized to terephthalic acid or reacted with methanol to produce dimethyl
terephthalate (Azapagic et al, 2003). In some cases cyclohexane dimethanol and
isophthalic acid substitutes some EG and DMT/TPA molecules respectively to generate
a PET copolymer with lesser crystallinity, greater barrier properties, etc.
The first step in the synthesis of PET is the formation of a prepolymer, bishydroxyethyl terephthalate (bis-HET). Subsequent polymerization of this material (with
the removal of ethylene glycol) forms the polymeric polyethylene terephthalate
(Nexant, 2008). Antimony-based catalysts (principally antimony oxide, to lesser extent
antimony acetate or antimony glycolate) in the concentration range of 150 350mg/kg
(0.015 0.035%), catalyses the polycondensation of the intermediate prepolymer to
PET (Thiele, 2004, EFSA, 2004). The most common catalyst is antimony trioxide, but
salts of titanium, germanium, cobalt, manganese, magnesium and zinc are also used
(Matthews, 2000). In general catalysts other than antimony trioxide are either less
efficient or more expensive than antimony or even both (International Antimony
Oxide Industry Association, 2006). Antimony catalyzes the chain prolongation reaction
by ligand exchange mechanism within its coordination sphere (El-Toufaili, 2006).
Coordination sphere of a metal ion in a coordination complex is the set of ligands
immediately attached to the ion. As a rule some of the catalysts remain encapsulated
into the polymer matrix or in the polymer chain itself (Matthews, 2000). Blue toners
including cobalt compounds are sometimes used to mask undesirable colours in PET
(El-Toufaili, 2006)

Figure 2.1 PET synthesis (Adapted from El-Toufaili, 2006)

2.2.3

PET bottle manufacture

PET bottles are produced by a two-stage process known as Injection Stretch Blow
Moulding (ISBM). The process depicted in Figure 2.2 involves the production of an
injection moulded PET bottle blank or preform followed by subsequent reheating,
stretching and blow-moulding to produce a full-sized bottle. PET naturally absorbs
water from its surroundings. Before processing it is usually heated to reduce its

moisture content to less than 50 parts per million to avoid hydrolytic reactions which
reduces its quality by generation of acetaldehyde.

Figure 2.2 PET bottle injection stretch blow moulding process (Adapted from Visy Pty
Ltd, 2008)
After about 4 hours of drying at a temperature of not more than 160C (to avoid
thermal degradation), the PET is melted and injected into the preform mould, resulting
in the production of the PET bottle preform (Figure2.3). The PET bottle preform is
heated to the correct profile for blowing, after which it is introduced into the blow
mould. The hot preform is simultaneously stretched and blown in the blow mould to
form the bottle (Kenplas, 2008, PET Planet Insider, 2001, VISY, 2008).

Figure 2.3 PET bottle preforms (Source: Aeco-Pack Corp, no date)

8

2.2.4

Acetaldehyde formation in PET material

Thermo-degradative generation of acetaldehyde in PET material occurs by a number of

ways (El-Toufaili, 2006, Ewender and Welle, 2008). At high temperatures the ester
bonds cleave via a cyclic transition state to generate acids and vinyl end groups (Figure
2.4). Then the vinyl end groups generate acetaldehyde through transesterification with
ethylene glycol (Figure 2.5), through hydrolysis (Figure 2.6) or through chain rebuild by
reaction of the vinyl end group with hydroxyl end group (Figure 2.7).

Figure 2.4 Cleavage of ester bonds in PET

Figure 2.5 Acetaldehyde formation by regeneration of the PET hydroxyl end group

Figure 2.6 Acetaldehyde formation by hydrolysis of vinyl end group

Figure 2.7 Acetaldehyde formation through chain rebuild

Acetaldehyde can also be formed through thermal scission of the PET hydroxyl end
group (Figure 2.8).

Figure 2.8 Acetaldehyde formation through thermal scission of the PET hydroxyl end
group
Acetaldehyde concentrations in PET matrix can be reduced by either use of PET with
low intrinsic viscosity or by addition of acetaldehyde scavenger additive to the PET.
Low intrinsic viscosity which implies low molecular weight is associated with lower
levels of acetaldehyde production during melt-processing stage. On the other hand
lower intrinsic viscosity implies less stiff PET material (Kenplas, 2008).

2.3

Glass

Glass is brittle, and often optically transparent non-crystalline solid material primarily
made from silica. Several types of glass including silica-free glass are available. The most
common glass utilised in bottle making is soda lime glass. Based on Seward and Vascott
(2005) the composition of soda-lime glass for containers is 74% silica, 13% sodium
oxide, 10.5% quick lime, 1.3% alumina, 0.3% potassium oxide, 0.2% sulphur trioxide,
0.2% magnesia, 0.04 ferric oxide and 0.01% titania. In the course of glass melt
processing tiny pockets of air from the atmosphere and from constituents
decomposition tend to get caught in the melt resulting in bubbles that can cause
performance and aesthetic issues in the final product. Trapped bubbles are removed by
a process referred to as fining, which may be physical or chemical. According to Shelby
(2005) arsenic and antimony oxides at 0.1 1% by weight (1000 10,000mg/kg) are
the most efficient chemical fining agents for glass. These oxides are probably not used
as fining agents for glass to be used for bottle manufacture because of toxicity issues. If
used however, the risk of the elements migrating from the bottle matrix into the
bottle content becomes a possibility. In addition to its use as fining agent antimony also
serves as opacifier agent in glass. Shotyk et al (2006) reported antimony concentrations

10

of 7.6 and 10.1ppm from 2 glass bottles for bottling of water and cola drink
respectively.

2.4

Other plastics

Polyethylene (PE) and polypropylene (PP) are olefin polymers of ethylene and
propylene, synthesised using Ziegler-Natta catalysts. These catalysts are usually formed
by the reactions of transition metal compounds of groups 4-10 (mainly Ti, V, and Zr)
with alkyls or hydrides of groups 1, 2, 13, or 14 (Corradini et al. 2004). Polyvinyl
chloride (PVC) is a polymer of vinyl chloride containing as much as 57% chlorine by
mass. A vast array of additives including plasticizers, heat stabilizers, UV stabilizers, etc
are usually added to PVC before it is made into finished products. Plasticizers are
specifically added to the hard and brittle PVC to make it flexible. Phthalates, which are
the most widely used plasticizers in PVC have been reported to be associated with
allergies in children and decrease in anogenital distance among male infants exposed
before birth, inducement of less male typical play behaviour in boys and other
manifestations related to mimicry of human hormones (Bornehag et al. 2004, Swan et
al, 2005, Swan et al, 2010). Additionally phthalates have been repeatedly reported to
affect various life forms including fish and invertebrates adversely. As a result of the
safety debate associated with the use of PVC several major corporations including
Microsoft, Wal-Mart, and Kaiser Permanente are said to have announced efforts to
eliminate PVC from products and packaging in 2005 (Center for Health, Environment
and Justice, 2009). For similar reasons the State of California is currently considering a
bill that would ban the use of PVC in consumer packaging (Californians Against Waste,
2010).

2.5

Summary

While many different materials are associated with bottled water and soft drink
bottling only the bottle material, which may be PET or glass, the bottle cap and/or the
lining of the bottle cap are in contact with the bottled content. PET bottling material
has been established to release both antimony and acetaldehyde into bottle contents.
Glass bottling material may only release antimony into bottle content and even then in
quantities much lesser than in PET. Although the cap and cap lining materials have not
been reported to release either of antimony and acetaldehyde these materials may be
associated with other safety issues depending on the type of material used.
11

2.6

Conclusion

This Chapter explained the processes involved in the manufacture of PET and the
stages at which antimony and acetaldehyde become associated with the PET material.
The Chapter also talked about glass material and the processes through which
antimony may become associated with the material. From the information on plastics
other than PET it is clear that these plastics are not manufactured using antimony.
Consequently these plastics will not be expected to release antimony.

12

CHAPTER 3:
3.1

BOTTLED WATER AND SOFT DRINKS USE AND

BOTTLE REUSE

Introduction

The Chapter explores the pattern and known extent to which bottled water and soft
drinks PET bottle are used and reused in different countries including Nigeria and
Britain. Factors influencing bottled water and soft drinks use and PET bottle reuse are
discussed. Bottled water and soft drinks regulations and categorizations and how these
regulations and categorizations define the different bottled water and soft drink types
are examined. The measures usually taken to ensure that bottles used in packaging are
safe are outlined and the impact of bottle quality regulation on bottle reuse is
discussed. Bottled water and soft drinks shelf life and best before dates and the
significance of such dating to migration are stated. The Chapter also look at the
discourse on the safety of bottle reuse.

3.2

Bottled water and soft drink use

Water must be consumed by human beings either in its pure form or mixed with
other constituents. Regardless of its form it has to be clean otherwise it will not
guarantee the wellbeing of human beings. Of the more than 6 billion people in the
world, more than one billion have no access to improved drinking water (National
Environmental Service Centre, 2006). The WHO Joint Monitoring Programme for
Water Supply and Sanitation defined access to improved water-supply services as the
availability of at least 20 litres per person per day from a source that is likely to
provide "safe" water (household connection, a borehole, etc,) within one kilometre of
the user's dwelling (Joint Monitoring Programme, 2006). While the water supply
coverage was 99% in 2004 in the developed world (Europe, North America, Japan and
Australia) it was only 56 and 50% in sub-Saharan Africa and Oceania respectively (Joint
Monitoring Programme, 2006). Lesser availability of the most affordable sources of safe
drinking water supply in developing countries together with other factors translates
into greater need for bottled water as an alternative safe drinking water (Rothschild
and Nzeka, 2005). Conversely in most of these countries utilisation of bottled water as
a source of safe drinking water may be severely delimited by cost relative to level of
prosperity. As a result of this bottled water and soft drink consumption tends to be
much higher in the developed world than in developing countries.

13

3.2.1

Bottled water

The world consumed 154 billion litres of bottled water in 2004, a 57 % increase from
the 98 billion litres consumed five years earlier (Arnold and Larson, 2006). US, the
highest consumer in terms of total annual consumption consumed 25.8 billion litres.
Italy on the other hand was the highest per capita consumer. Even though Britain was
not the highest in Europe in 2007, with a cumulative annual growth rate (CAGR) of 8%
between 2002 and 2007, the rate increase in consumption was the highest in Europe.
With a projected CAGR of around 7% between 2007 and 2012, the per capita
consumption is expected to move from 41 litres in 2007 to 57.8 litres in 2012 (JustDrinks, 2008). Figures 3.1 and 3.2 show the per capita annual consumption of bottled
water for Britain, Nigeria and other nations in 1999 and 2004 (for Nigeria bottled
water here refers to both water in bottles and water in pouches) and for different
regions of the world in 2004. In France, Germany and Italy close to 90% of the
population patronises bottled water in comparison to about 50% in Britain
(Finewaters, 2009).
200
180
160

Litres

140
120
100
80

1999

60

2004

40
20
0

Figure 3.1 Global per capita consumption of bottled water in 1999 and 2004 (Adapted
from Arnold and Larson, 2006)

14

97.5
74.7

33.2
24.2
9.7

North America

Europe

South America

Asia

4.2
Africa/Middle Global average
East/Oceania

Figure 3.2 Per capita consumption of bottled water in 2004 by regions (Gleick, 2008)
While developing countries overall bottled water consumption is comparatively lower
than that of developed countries, these countries are still important consumers of
bottled water. US, the highest consumer in 2004, was followed by emerging and
developing countries (Mexico, China and Brazil) and then by Italy. In terms of per
capita consumption Italy was followed by Mexico and United Arab Emirates. Of the
top 15 per capita consumers of bottled water, Lebanon, the United Arab Emirates, and
Mexico have the fastest growth rates, with consumption per person increasing by 44
50 percent between 1999 and 2004 (Arnold and Larson, 2006).
In 2004 the estimated market share of packaged water in Nigeria was 1.4 billion litres
valued at approximately $500 million (Rothschild and Nzeka, 2005). According to the
authors water packaged in plastic pouch, termed pure water, accounts for 68
percent of total commercial water and is consumed by the low-income group. The
estimated per capita consumption, based on the 2004 Figures and the Nigerian
population of 130 million, was 10.77 litres in 2004. Nigerias low per capita
consumption of bottled water in comparison to the world average and most of the
countries in Figure 3.1 probably results from the fact that Nigerias per capita GDP is
lower than the world average and is low in comparison to the most countries in the
Figure. The water packaged in plastic pouches will not be fully investigated in this study
because the packaging for this water is polyethylene rather than polyethylene
terephthalate.

15

3.2.2

Soft drinks

In terms of soft drinks consumption Germany was Europe's largest market with Britain
following fast behind (Food navigator, 2004). On average a Briton consumed 156 litres
of soft drinks a year. Unlike bottled water some soft drinks are not bottled in glass or
plastic bottles. However a Waste & Resources Action Programmes (WRAP) estimate
put the number of PET bottles utilised annually in soft drinks in Britain at 6.5 billion
(WRAP, 2008) In Nigeria the market share of soft drinks (fruit juice and carbonated
drinks) was approximately 1.27 billion litres in 2004 (Rothschild and Nzeka, 2005). If
the market share volumes were the same as consumed volumes, the estimated per
capita consumption would have been 9.77 litres. According to the authors Nigerias
soft drink sub-sector maintained a 1% growth in five years mainly due to the increasing
consumer health concern over products sugar content and consumers demand for
nutritious beverages. It must be emphasised that in Nigerian carbonated soft drink
market refillable glass bottles are more widely used in comparison to disposable PET
bottles and aluminium cans. Among other reasons soft drinks are cheaper in refillable
glass bottles than in disposable bottles and cans.

3.3
3.3.1

Factors influencing bottled water and soft drinks use

Country economic status

The economic status of a country influences the use of bottled water and beverages.
This can evidently be seen from Figure 3.2 where the per capita consumption of North
American and European regions by far surpasses the per capita consumption in Asia
and Africa/Middle East/Oceania regions. Bottled water is more expensive than pipeborne municipal water, even though it is perceived to be safer. Consequently in lowincome countries there will be lesser tendency for bottled water use to be as
widespread as in high income countries. Even though there is no similar data for
bottled soft drinks there is no reason to believe that the trend is dissimilar. The extent
to which availability of clean drinking water influences bottled water consumption is
discussed in subsection 3.3.4.

3.3.2

Climatic conditions

In arid and tropical climates of the world the degree of transpiration in humans is
higher than in milder climates, consequently the need to drink water and beverages
will be higher. However hotter climatic conditions alone may not necessarily translate

16

into higher consumption of bottled water. Most likely the influence of climate on
bottled water usage will be linked to other factors including economic status of
countries. For example United Arab Emirate (UAE) and Oman are Middle Eastern and
very hot countries that share a common boundary. Whereas the former has a per
capita bottled water consumption of 164 litres in 2004, the later has just 12.6 litres
(Gleick, 2008). This could be justified to some extent by the fact that the UAE GDP
per capita in 2004 almost triples that of Oman according to the IMF (IMF, 2010).

3.3.3

Environmental awareness

Environmental awareness is another factor that has some influence on the extent of
use of bottled water and drinks. As mentioned elsewhere, plastic bottles are used in
bottling of water and drinks more than glass and aluminium cans due to their superior
qualities. Nonetheless the manufacture and utilisation of plastics is associated with
release of toxic pollutants and greenhouse gases, littering and accumulation of plastics
in world waters which is detrimental to marine life. Consequently increasing numbers
of individuals and communities may prefer not to use plastics as a result of growing
awareness campaigns by environmental organisations. Notable examples in Britain
include Modbury and Hebden Bridge towns where a voluntary agreement between
local shop owners and the communities saw to the banning of plastic shopping bags in
the towns. This issue is however more prominent in developed countries. Questions
are increasingly being asked recently on why so many people should be drinking
bottled water rather than tap water. Additional concern is excessive withdrawal of
natural mineral water or spring water to produce bottled water (Li, 2008).

3.3.4

Safety and health

In large number of developing countries safe drinking water is scarce. In such countries
demand for bottled water and soft drinks will be raised by the scarcity (Rothschild and
Nzeka, 2005; Li, 2008), especially if a cheaper versions of packaged water and soft
drinks are available. In Nigeria for example 68% of packaged water is packaged in
plastic pouches of 500ml capacity (Figure 3.3) rather than plastic bottles (Rothschild
and Nzeka, 2005). By implication the scarcity of safe drinking water coupled with the
availability of the cheaper water in pouches raises the per capita consumption of
packaged water. In low-income countries scarcity of safe drinking water in the absence
of cheap packaged water is not likely to increase packaged water consumption.

17

Conversely, availability of safe drinking water as in high-income countries may have

only small effect on reduction of bottled water consumption. The impact of increasing
desire for healthier lifestyle is probably more on soft drinks than for bottled water. In
both developing and developed countries people are concerned that the sugar content
of soft drinks makes them unhealthy to consume on regular basis.

Figure 3.3 Nigerian 500ml packaged water in polyethylene pouch

3.3.5

Taste

In some areas groundwater, which is usually used as drinking water, tastes bad as a
result of dissolution of chemicals from underlying rocks, contaminants reaching the
water from surface or leakage of briny seawater into aquifers especially in coastal
areas. According to Li, (2008) in such areas patronage of packaged water can be
expected to be high. In some countries people do not like drinking tap water because
of its aftertaste which is associated with use of chlorine as disinfectant. In such cases
bottled water, which is mostly treated using ozone, provides an alternative to tap
water.

3.3.6

Other factors

Other factors that influences bottled water and drinks patronage include idolization of
bottled water as fashion accessory (Royte, 2008), aggressive marketing strategies by
the manufacturers, office working environments (Rothschild and Nzeka, 2005; Li,

18

2008). The fact that bottled water and soft drinks can be carried conveniently in a
backpack or in a car may promote their consumption. Rural-urban divide may possibly
influence bottled water and soft drinks patronage to greater extent in developing
countries than in developed countries. Income inequality is more prevalent in the
developing countries than in the developed countries with people in rural areas been
poorer. As mentioned earlier better economic status promote bottled water and soft
drinks use. Consequently people from rural areas in developing countries will probably
be much less likely to be using bottled water and soft drinks than people from rural
areas in developed countries

3.4

Bottled water and soft drinks regulations

Packaged water and drinks together with their packaging are usually regulated by
government agencies principally in charge of food safety and in some cases together
with drugs and related consumables. Such agencies in Britain, US and Nigeria are Food
Standard Agency (FSA), Food and Drug Administration (FDA) and National Agency for
Food and Drug Administration and Control (NAFDAC) respectively. In Britain bottled
water is regulated under The Natural Mineral Water, Spring Water and Bottled
Drinking Water Regulations 2007 together with subsequent amendments (separately
for England, Wales, Scotland and Northern Ireland). Additionally Bottled water must
also comply with Directive 2009/54/EC of the European Parliament and of the Council
of 18 June 2009 on the exploitation and marketing of natural mineral waters. Fruit
Juices are covered by European Commission Directive 2009/106/EC of 14 August 2009
and The Fruit Juices and Fruit Nectars Regulations 2003. Unlike fruit juices and bottled
waters, there is no formal legal definition or compositional standard for flavoured
water and soft drinks other than fruit juices and nectars. However the colours
preservatives, sweeteners and other additives used are all covered by separate EU
directives and British regulations. In the US bottled water is covered by the bottled
water standard of identity and quality regulations (21 CFR 165.110) and current
Good Manufacturing Practice regulations for the processing and bottling of bottled
drinking water (21 CFR part 129). Additional regulations are the labelling regulations
(21 CFR part 101) and current Good Manufacturing Practice regulations (21 CFR part
110) for all other foods are also applicable bottled water. Canned fruit juices are
covered by Title 21 CFR part 146. In Nigeria soft drinks and fruit juices are regulated
under Soft Drinks Regulations 2005 and Fruit Juice and Nectar Regulations 2005
19

together with Pre-Packaged Food (Labelling) Regulations 2005. For packaged water
(bottled water and water in pouches) registration with NAFDAC is mandatory as
required by the provisions of ACT CAP F33 LFN 2004 (formerly decree 19 of 1993)
which also covers other foods. In addition to the mandatory registration, NAFDAC
guideline documents NAFDAC/EID/003/00 and NAFDAC/RR/005/00 specifies the
guidelines for establishment of packaged water plant in Nigeria and the guidelines for
registration of food and water manufactured in Nigeria. Generally bottled water and
soft drinks regulations are rules and restrictions meant to fully define and differentiate
the different types of bottled water and soft drinks, treatments allowed, packaging and
labelling, levels of contaminants allowed, good manufacturing practice and also define
what constitutes breach of these rules and restriction and sanctions associated with so
doing.

3.5

Bottled water and soft drinks categorisation

As earlier mentioned flavoured water and soft drinks other than fruit juices and
nectars are not defined by law in Britain. Consequently, legally binding categorisation is
non-existent. Nevertheless the British Soft Drinks Association defined flavoured water
and soft drinks as "a manufactured drink, optionally sweetened, acidulated, which may
contain fruit, fruit juice and other salts; the flavour may derive from vegetable extracts
or flavourings" (British Soft Drink Association, no date). In Nigeria however the Soft
Drinks Regulation 2005 defined soft drinks as non- alcoholic carbonated or noncarbonated ready to drink beverages. In both Britain and US the name of a fruit or
fruits followed by juice can only legally be used to describe a product which is 100%
pure juice. If diluted (to a degree limited by regulations) with water and/or
contain additives besides fruit juice, including natural and artificial sweeteners,
and preservatives, it is then referred to as nectar. Fruit juices may be categorised into,
freshly squeezed, short life and long life juices if they have a shelf life of no more than
14 days, up to 30 days and between 6 and 12 months respectively (British Soft Drink
Association, no date)
The British Bottled Drinking Water Regulations 2007 together with its subsequent
amendments categorised bottled water into natural mineral water, spring water and
bottled drinking water. While all the water type must meet safety criteria as specified
in schedule two of the regulations, they differ in other attributes as described in Table

20

3.1. Bottled water categorisation in the US is generally similar to that in Britain.

However in the US other categories exist in addition to the categories defined in
Britain. These include artesian water, ground water, sterilised water and well water. In
Canada bottled water is legally either mineral water also called spring water or just
bottled water if it is not labelled as the former (Health Canada, 2009).There appears to
be no legislation categorising packaged water in Nigeria. From their labels bottled
water in Nigeria are either categorised as table water or spring water. Water in plastic
pouches is sometimes referred to as sachet water and less formally but popularly
pure water. An additional categorisation for bottled water is whether they are still
or sparkling. While this categorisation does exist in Britain, US and other countries, it
is non-existent in Nigeria as carbonated water is not available in the bottled water
market.
Table 3.1 British bottled water categorisation
Attribute
Source
Recognition
process

British bottled water type

Natural mineral water
single non-polluted
ground water source
must undergo a 2-year
recognition process

Stable
chemical composition
composition must be stable except
for an inevitable
permissible variation
Treatment
altering chemical or
microbiological quality
not permitted, removal
of unstable elements
permitted
Bottling
must be bottled at
source
Labelling
mineral analysis, name of
source and place of
exploitation must be on
the label

Spring water
single non-polluted
ground water source
no formal recognition
process required
chemical composition
does not have to be
stable
may undergo permitted
treatments to meet the
microbiological criteria
in the Drinking Water
Regulations
must be bottled at
source
name of source and
place of exploitation
must be on the label

Drinking water
may come from
many sources
no formal
recognition
process required
chemical
composition
does not have to
be stable
permitted

not restricted to
source
No restriction as
for spring and
natural mineral
water

The above categorisations have some implications on the presence and or migration of
chemicals including acetaldehyde and antimony. For example concentrations of

21

antimony and acetaldehyde were reported to be higher in carbonated drinks than in

still drinks (Nawrocki et al 2002, Keresztes et al 2009). Also Mutsuga et al (2006)
reported higher amount of acetaldehyde in sterilised water than unsterilized water
supposedly because the degradative activity of bacteria on acetaldehyde is absent in the
former.

3.6

The significance of shelf life

Bottled water produced in accordance with current good manufacturing practice and
quality standard regulations is considered to have an indefinite safety shelf life if stored
in an unopened, properly sealed container (FDA, 2009). Bottled water is still labelled
with 'best before' dates even though this is voluntary and unrelated to interaction
between bottle content and bottle material. In fact according to Foods Standard
Agency (no date) for most food products 'best before' dates are more about quality
than safety. Bottled water and carbonated drinks in PET bottles from Nigeria have a
shelf life of one year and six months respectively based on the production and 'best
before' dates stamped on the bottles. In the US and Canada bottled water's stamped
shelf life is usually two years (Environmental Health & Safety Online, 2006; Health
Canada, 2009). Actually Health Canada (2009) suggested storing bottled water for
emergency use for as long as one year. While still water does not expire, the 'best
before' dates on carbonated water and soft drinks are probably in part related to their
vulnerability to loss of carbon dioxide with time. Bottled water and soft drinks are
most likely typically consumed before their 'best before' dates. However information
on the actual time span between purchase and consumption of bottled water and
drinks is unavailable in the literature. Even though shelf lives labelled on bottled water
and drinks are not related to migration of chemicals, concentrations of migrants have
been reported to rise with storage time as will be discussed later.

3.7

Bottle reuse in Nigeria, Britain and other countries

PET bottle reuse in the context of this work refers to putting the PET bottles in any
use other than the original intended use. In developed countries health authorities
sometimes discourage the reuse of the single-use PET bottles. For example Health
Canada, the department of the government of Canada with responsibility for national
public health, does not recommend the reuse for the sole reason of doing away with
microbiological risk (Health Canada, 2009). PET bottle reuse may be a wide spread
22

practice and may vary in application depending on where it is practiced. However data
on PET bottle reuse is scarce in the literature. Reuse will probably be more
widespread in developing countries than in developed countries in large part due to
lesser need to reuse bottles in developed countries as a result of greater prosperity.
On the other hand greater availability of empty bottles as a result of higher use may
elevate extent of reuse in developed countries. Interestingly Lilya (2001) in a
preliminary survey of the university of Idaho community in the US found that 88% of
the participants reused polyethylene terephthalate (PET) plastic bottles used for
bottled water, in some cases, for as long as six months.
PET bottles are not only used in packaging of water and soft drinks but also for
packaging of edible oils, cosmetics, detergents, etc (Azapagic et al, 2003). In terms of
reuse PET bottles initially used for products other than water and soft drinks are
probably less likely to be reused at all because reusing them will require more vigorous
cleaning than for bottles used for water and soft drinks. While reuse may not
necessarily be limited to drinking water, reuse for drinking water will almost certainly
turn out to be the most common form of reuse. In the temperate developed countries
like Britain where the weather is generally cool, there may be only marginal need to
store drinking water in the refrigerator. In such countries reuse of PET bottles will
probably be limited to drinking water storage by people on the move and in work
places than in homes. Nevertheless reuse at home may not be ruled out completely. In
Nigeria in almost all households that can afford to own a refrigerator, empty bottles,
mostly PET bottles serve as a means of storing water in the refrigerators to make the
water cool (Figure 3.4). Additionally local beverages and medicinal herbal concoctions
are vended in used PET bottles (Figure 3.5).

23

Figure 3.4 Storage of drinking water in used PET bottles in a refrigerator in Nigeria

Figure 3.5 Medicinal herbal concoctions vended in used PET bottles in Nigeria
A wide spread use of used PET bottles in developing countries is in the disinfection of
drinking water by solar radiation, a process called Solar Water Disinfection (SODIS)
(Figure 3.6). SODIS, a low-cost technology with a great potential to improve the health
of those without access to safe drinking water, was developed by the Swiss Federal
Institute for Environmental Science and Technology (EAWAG) and its Department for
Water and Sanitation in Developing Countries (SANDEC). SODIS utilizes solar UV-A
radiation and temperature to inactivate pathogens in water. According to EAWAG
(2008) SODIS is used for the daily treatment of drinking water by over 2 Million users
in more than 20 countries including Ecuador, Bolivia, Peru, Brazil, Congo, Uganda,
24

Tanzania, Mozambique, Cameroon, Ivory coast, Pakistan, India, Vietnam, Cambodia and
Sri Lanka (Figure 3.7). Additionally it has been shown that SODIS, combined with
improved hygiene behaviour, can reduce diarrhoea incidence by 20 to 70% (Wegelin,
2006). SODIS can be achieved using both glass and PET bottles however as PET bottles
are more readily available than glass bottles they are almost certainly more commonly
used in SODIS.

Figure 3.6 SODIS in an African setting (Source: The Water School, 2008)

Figure 3.7 SODIS application worldwide (Eawag, 2008)

25

3.8

Bottle quality regulation

Even though the many plastics used in packaging of foods are generally inert, some may
still transfer substances which can present a risk to human health. Directive
2002/72/EC of the European Union came into force for the purpose of ensuring that
plastic materials used in packaging of foods are not releasing chemicals at levels that
can unacceptably change the quality of packaged foods and/or present a risk to human
health. The directive specified the maximum migration of antimony and acetaldehyde
from PET bottling material into bottled contents at 40 and 6000 g/kg respectively.
The specific migration limits (SML), as they are called, are derived from tolerable daily
intakes for antimony and acetaldehyde and by law they should not be exceeded in any
interaction between plastic packaging and the packaged foods. As the provision of this
directive deals with safety, the provision can also be extended to PET bottle reuse
situation. The provision may not be legally binding in the case of bottle reuse as single
use PET bottles are not originally meant to be reused after initial use. Consequently,
for the sake of safety, reused bottles should not transfer antimony and acetaldehyde
into water or any other content for consumption at concentration greater than the
SMLs.

3.9
3.9.1

Factors influencing PET bottle reuse pattern

Country economic status and climate

People from high income countries are probably less likely to reuse PET bottle than
those from low income countries simply because they have greater ability to buy
bottled water and they have greater access to clean water. PET bottle reuse will
probably be more widespread in hot low-income countries than in other countries.
Currently SODIS is promoted and practiced only in tropical developing countries
(Figure 3.7).

3.9.2

Safety debate

At present there is on-going debate mainly in developed countries about the safety of
reusing PET bottle. PET bottles and other plastic containers have been demonstrated
to leach chemicals at low concentrations. Improperly cleaned bottles have also been
shown to harbour indicator bacteria which point to possible presence of pathogenic
microorganisms. In some quarters people went to the extent of alleging that reuse of

26

bottles can cause cancer as will be discussed later. Due to these on-going debates
some people may opt not to reuse PET bottles.

3.9.3

Cost

Cost influence reuse of PET bottle because it is cheaper to refill a bottle with tap
water than to buy a new bottle of water. Conversely other people especially in
developed countries may prefer to buy new bottles seeing them as cheap and safer.

3.9.4

Other factors

Other factors that may influence the reuse of PET bottles include the availability of
used bottles, age of bottles, environmental concern, original content of bottles,
convenience due to portability of bottles, knowledge of and access to SODIS and the
availability and pattern of supply of potable pipe-borne water supply to places of
residence.

3.10 PET bottle reuse and bacterial contamination

Oliphant et al, (2002) in a study involving the assessment of bacterial water quality of
elementary students PET water bottles found that the Canadian Drinking Water
Quality Guidelines (CWQG) criterion was exceeded for total coliform in 13.3% of 75
samples. Faecal coliform and total heterotrophic bacteria criteria were exceeded in
8.9% (of 68 samples) and 64.4% (of 76 samples) respectively. The presence of faecal
coliforms in water implies possible contamination of the water with faecal material of
human or animal origin and hence elevated possibility of the presence pathogenic
organisms in the water. Identification of pathogens in PET bottles should clearly be
related to introduction of the contaminants by the human handlers through repeated
usage. Usage of a bottle by single person and thorough washing of the bottles with
detergent before use can do away with microbial contamination hazard. However
studies on the impact of long-term rigorous washing on bottle behaviour are
unavailable.

3.11 PET bottle use and reuse and the safety debate
Water quality guidelines and standards provide a benchmark for measuring safety in
terms of the amount of chemicals in drinking water including chemicals that may
migrate from plastic materials into the water. However according to Lichter (2009) in
online material titled Are chemicals killing us? different interest groups portray the
27

risk associated with chemical migration possibilities differently. While environmental

organisations and media are believed to be overstating the risk associated with release
of chemicals from many products, industry sources are seen as understating these
risks. Government agencies and professional associations are on the other hand
believed as providing mainly accurate portrayals of associated risk.

In an online

document authored by an architectural firm (PM Architecture) with interest in

sustainable portable water consumption and titled Bottled water do we need it? a
purported carcinogen named bisphenol A was alleged to be leaching from PET bottles
into the water content. In reality this chemical is not in any way associated with PET
and has not been established as a carcinogen. Another chemical widely publicised in
the media as a carcinogen migrating from PET bottles is Bis(2-ethylhexyl) adipate
(DEHA) based on a Masters degree thesis of a university of Idaho student (Lilya 2001).
According to South African Food Advisory Consumer Service (FACS 2009) DEHA
commonly used as a plasticizer in other plastics, has not been identified either as a raw
material in PET or as its decomposition product. Additionally even if DEHA migrates
from PET it falls in Group 3 in IARC carcinogen classification rendering it not
classifiable as to its carcinogenicity to humans. There is also the same stance for
antimony; the chemical established to be leaching from PET bottles. Even though this
chemical has only been established as a possible carcinogen (IARC Group 2B), in many
web pages it is usually referred to as if it is a confirmed carcinogen. On the other hand
many documents released by industry stakeholders including International Antimony
Oxide Industry Association, International Bottled water Association and PET Resin
Association (PETRA) attempt to completely exclude the potential for harm in terms of
chemicals migrating from PET even though some studies have established
concentrations beyond the regulatory levels.

3.12 Summary
Unavailability of safe drinking water in the developing countries promotes bottled
water use in those countries. In spite of that usage is still higher in developed
countries. Regulations and categorisations may have some implications on the presence
and migration of chemicals from bottle material into content. Even though shelf lives
labelled on bottled water and drinks may be voluntary and unrelated to migration of
chemicals, concentrations of migrants have been reported to be influenced by storage
time. Reuse is probably, as widespread in developed countries as in developing
28

countries. However, impact of long-term reuse and rigorous washing of bottles on

bottle behaviour is not well documented in the literature. As a result of the varied
interests on the subject of PET bottle use and reuse safety by different interest groups
a lot of debate still remains on the subject.

3.13 Conclusion
Chapter 3 presented background information on the pattern and extent of bottled
water and soft drinks use and PET bottle reuse and the factors influencing bottled
water and soft drinks use and PET bottle reuse. The Chapter discussed bottled water
and soft drinks regulations and categorizations and how the regulations and
categorizations define the different bottled water and soft drink types. The Chapter
revealed that bottles must meet some quality criteria in term of antimony and
acetaldehyde migration before they can be used in bottling. From the information in
the Chapter it is clear that the best before dates on water are unrelated to safety
implying that bottle contents may not be unsafe to use even after the best before
dates. Because PET bottle reuse information is scanty in the literature this study will
attempt to establish the bottle reuse pattern in Nigeria and Britain to further enrich
the literature. Impact of long-term reuse and rigorous washing of bottles on bottle
behaviour, which is not well documented in the literature, will be studied in this work.

29

CHAPTER 4:
4.1

CHEMICALS MIGRATING FROM PET BOTTLES

Introduction

This Chapter presents an overview of antimony and acetaldehyde in terms of their

abundance in the environment, their health effects on human beings and how they are
regulated for the purpose of safeguarding human health. The Chapter review the
literature on the migration of antimony and acetaldehyde from bottle wall into bottle
contents. Factors affecting the migration of chemicals from bottle wall into bottle
contents are also discussed.

4.2
4.2.1

Antimony
Background information

Antimony is silvery lustrous grey metalloid occurring principally as sulphide ores

(stibnite and jamesonite) and to lesser extent as oxide ores (senarmontite and
valentinite). Important properties and uses of antimony are summarised in Table 4.1.
Identified world resources of antimony are estimated to be in the region of 4 to 6
metric tons and are located mostly in China, Bolivia, Mexico, Russia, and South Africa
(Butterman and Carlin, 2004).
Table 4.1 Properties and uses of antimony
Chemical symbol
Atomic number
Electronic configuration
Group, period, block
Atomic mass
Density
Natural isotopes
Melting point
Boiling point
Oxidation numbers
Uses

4.2.2

Sb
51
2,8,18,18,5
15, 5, p
121.76gmol-1
6.684gcm-3
121
Sb (57.21%), 123Sb (42.79%)
631C
1587C
-3, 0, 3, 5
Lead-acid batteries, bearing metal, solders, flame
retardants, ceramics and glass, plastic stabilizers, plastic
catalyst and pigments

Abundance in earth crust and world waters

At an estimated concentration of 0.2 to 0.5 parts per million, antimony accounts for
between 0.00002 and 0.00005% of the earth crust (USGS, 2010). Filella et al (2002) in
an extensive review of literature on occurrence of antimony in the environment

30

reported typical concentrations of dissolved antimony in unpolluted waters as being

less than 1g/L. In two Polish studies involving the antimony concentration in 49
groundwater samples from southern and eastern Wielkopolska (Niedzielski et al, 2001)
and in groundwater samples from Pozna city (Niedzielski and Siepak, 2005) the
maximum antimony concentrations reported were 1.25 and 0.71 g/L respectively. In
the earlier study the lowest and average concentrations were 0.2 and 0.53 g/L
correspondingly. Shotyk et al (2006) reported antimony concentrations of as low as
0.0022 0.0012g/L for pristine ground water from Canada. Antimony concentrations
have been reported to reach up to 100 times the natural levels in the proximity of
anthropogenic sources. A typical example of this elevation due to anthropogenic
activities was reported in the work of Fu et al, (2010) where the average antimony
concentrations in water bodies (including drinking water reservoirs) located between
one and 8 kilometres away from an antimony mining area in Hunan, China ranged
between 8.7 1.2 and 156 4 g/L with an average of 53.6 46.7g/L. An earlier
reported case of high elevation of antimony concentration in water as a result of
human activity as reported by Grimes et al. (1995) was near a gold deposit containing
antimony and arsenic minerals in Nevada, USA. In this case the antimony concentration
in groundwater was found to rose up to 260g/L. For antimony in sediments and
unpolluted soils the review by Filella et al (2002) reported the concentrations as
being in the order of a few g/g. Table 4.2 gives concentration ranges of antimony
in fresh water systems, oceans, estuaries, sediments and soils as reported by these
authors.
Table 4.2 Antimony concentration in the environment (Filella et al. 2002)
Environmental system
Freshwater
Seawater
Estuarine systems
Sediments
Soils

range
consulted publications
0.0001 96 g/L
62
<0.007 17.045 g/L
52
0.0047 3.25 g/L
9
0.04 12500 g/g
21
0.1 5000 g/g
47

Antimony concentration in environmental waters and soils has an important bearing

on the quality of bottled water and beverages as it is an important determinant of the
ultimate antimony concentration in the finished products.

31

4.2.3

Health effects of antimony

Antimony trioxide and antimony pentoxide are the most important compounds of
antimony with antimony pentasulfide, antimony chloride, antimony potassium tartrate,
antimony trichloride, antimony trisulfide and antimony hydride been of lesser
importance (Australian Department of the Environment, Water, Heritage and the Arts
(2006). Antimony trioxide, which is the catalyst in the production of PET, is the form
in which most antimony emission into the environment occurs (WHO, 2003).
Antimony compounds are hardly absorbed in the gastrointestinal tract, resulting in less
hazard of acute poisoning. Long-term animal studies have reported liver damage and
blood changes when animals ingested antimony (ATSDR, 2007). Meglumine
antimoniate is a pentavalent antimonial drug, used for the treatment of leishmaniases
for over half a century. In a 21-day study to investigate its developmental toxicity as
well as the transplacental transfer of antimony in rats by Miranda et al. (2006) no
adverse effect was noted on the mothers at any dose level and no embryotoxicity was
observed at the lowest dose. At the highest dose, Meglumine antimoniate increased
embryo lethality, reduced foetal weight and augmented the occurrence of some softtissue and skeleton variations. In a 60-day short-tailed field vole antimony trioxide
ingestion experiment no harmful effects were evident even though elevated organ
concentrations were observed (Ainsworth et al, 1991). Similarly Kirkland et al (2007)
detected no clinical signs of toxicity in rats exposed to antimony trioxide for 21 days
except for some reductions in body-weight gain in the top dose group. The fact that
antimony is not well absorbed in the gastrointestinal tract (ATSDR, 2007a) and does
not bioaccumulate (WHO, 2003) could be an explanation for its low toxicity on shortterm exposure in laboratory animals. Ainsworth et al (1991) observed rapid
establishment of equilibrium between uptake and excretion with no subsequent
occurrence of progressive increase in organ concentrations. Additionally the
researchers observed rapid clearance of antimony on termination of dietary intake.
Long-term inhalation of high levels of antimony irritates eyes and lungs and causes
heart and lung problems, stomach pain, diarrhoea, vomiting, and stomach ulcers. In rats
lung cancer and problems with fertility were reported when rats breathed very high
levels of antimony (ATSDR, 2007). However, except for corneal irregularities and
dose-related increase in cataracts, no adverse clinical observations were attributed to
antimony trioxide in a subchronic and chronic inhalation rat study by Newton et al
32

(1994). In spite of lung cancer development in rats as reported by other studies

Newton et al (1994) did not find antimony trioxide to be carcinogenic in an inhalation
study involving rats. These researchers concluded that previously reported studies,
which found antimony trioxide to be a carcinogen, were run at higher lung burdens.
Even though WHO made a pointer on some existing evidence on the carcinogenicity
of certain antimony compounds by inhalation (WHO, 2003), the International Agency
for Research on Cancer (IARC), and the Environmental Protection Agency (EPA) did
not classify antimony as human carcinogen (ATSDR, 2007). However based on
inadequate evidence for the carcinogenicity of antimony trioxide in humans and
sufficient evidence for its carcinogenicity in experimental animals the IARC categorised
antimony trioxide as being possibly carcinogenic to humans (IARC Group 2B) (IARC,
1989). Additionally arsenic, a chemical element that shares some chemical and

toxicological properties with antimony, is a proven carcinogen (Gebel, 1997).

4.2.4

Guideline and regulatory standards for antimony in water and

foods

The tolerable daily intake (TDI) for humans and the corresponding WHO guideline
value for antimony in drinking water used to be 0.86 g/kg/day and 5 g/L before they
were changed to 6 g/kg/day and 20 g/L respectively in 2003 (WHO, 2003). The
elevated values imply increased margins of consumer safety for antimony. While the
Japanese maximum admissible concentration changed from 2 to 15g/L (Wakayama,
2005), the EU maximum admissible concentration and the EPA maximum
contamination level remain unchanged at 5 and 6g/L respectively (European
Commission, 2003; EPA, 2010). The EU specific migration limit (SML) for antimony
from PET into foods is 40g/kg of food (EFSA, 2004).

4.3
4.3.1

Acetaldehyde
Background information

Acetaldehyde is a colourless, volatile, flammable, organic liquid with a pleasant, fruity

odour at low concentrations and a pungent, suffocating odour at high concentrations.
Important properties and uses of acetaldehyde are summarised in Table 4.3.

33

Table 4.3 Properties and uses of acetaldehyde

Chemical group
Molecular formula
Molecular mass
Melting point
Boiling point
Solubility in water
Density
Viscosity
odour threshold
Flash point
Autoignition
temperature
Uses

4.3.2

Aldehydes
CH3CHO
44.05 gmol1
123.5 C
20.2 C
Miscible in all proportions
0.788 g cm3
2.456 X 10-4 Pa.s
0.05 ppm
39 C
185C
intermediate in the manufacture of acetic acid, pyridine and
pyridine bases, and esters, manufacture of disinfectants, drugs,
perfumes, explosives, lacquers and varnishes, also as flavouring
agents in foods including milk products, baked goods, fruit
juices, candy, desserts, and soft drinks

Abundance in the environment and use as food additive

According to EPA (2007) Acetaldehyde is ubiquitous in the ambient environment

occurring naturally in bread, and ripe fruit, as an intermediate product of higher plant
respiration and as a product of incomplete wood combustion in fireplaces and
woodstoves, coffee roasting, burning of tobacco, vehicle exhaust fumes, and coal
refining and waste processing. In water, wide range of foodstuff and in air,
concentrations are generally believed to be less than 0.1 g/L (ppb), less than 1 g/g
(ppm) and averaging about 5 g/m3 (2.78ppb) respectively (IPCS 1995). In Los Angeles,
California, levels of acetaldehyde up to 32 ppb have been measured in the ambient
environment (EPA 2007), and in smoky indoor atmospheres its concentrations may
reach as much as 100 ppb (Morris 1997). In some foodstuffs particularly some fruit
juices and vinegar concentrations up to several 100 ppm were reported (IPCS 1995).
While food and beverages, cigarette smoke and, to a lesser extent air are the major
source of exposure to human population, the main source of human exposure to
acetaldehyde is said to be through the metabolism of alcohol (IPCS 1995).
In addition to its natural presence in fruit juices acetaldehyde is added to some soft
drinks as a flavouring agent (Miyake and Shibamoto 1993, Food Safety Commission
2005, National Toxicology Program 2010). Acetaldehyde is one of the chemical
substances in the US FDAs Everything Added to Food in the United States (EAFUS)

34

database (FDA 2011). According to FDA the EAFUS list of substances contains
ingredients added directly to food that FDA has either approved as food additives or
listed or affirmed as Generally Recognized as Safe (GRAS).

4.3.3

Health effects of acetaldehyde

Acetaldehyde vapour at moderate concentrations causes eye irritation in animals and

humans (EPA 1994). In humans signs of eye irritation manifest at 50 ppm, and at 200
ppm red eyes and transient conjunctivitis develop (Clayton and Clayton, 1993).
Repeated exposure may lead to chronic irritation of the eyes with resultant permanent
damage (New Jersey Department of Health and Senior Services, 2010) In liquid form
acetaldehyde can cause skin irritation, painful burning sensation and possibly skin
allergy (ACGIH, 2001; New Jersey Department of Health and Senior Services, 2010).
Vapour inhalation has been reported to cause adverse respiratory tract effects in both
animals and humans (USEPA 1994). In a 4-week study involving acetaldehyde inhalation
in rats Appleman et al. (1982) reported increased lung weights, and severe
degenerative, hyperplastic and metaplastic changes of the nasal, laryngeal and tracheal
epithelium at concentrations of 5000ppm. While increased blood pressure and
decreased heart rate were observed in acetaldehyde inhalation study in rats (Egle,
1971), the author concluded that concentrations of acetaldehyde producing significant
changes of blood pressure and heart rate are somewhat higher than those that would
be encountered in cigarette smoking. Limited evidence links acetaldehyde with adverse
developmental and neurological effects in animals, including central nervous system
depression and neural degeneration (EPA 1994). Acetaldehyde was not found to be
mutagenic in Salmonella or in E. coli, however it was reported to be positive for
chromosome aberrations and sister chromatid exchange both in vitro and in vivo
mammalian assays (EPA 1994). Additionally Singh and Khan (1995) observed an
irreversible breakage of both single and double stranded DNA in addition to significant
cell loss in a study evaluating its cytotoxicity and genetoxicity in human lymphocytes. In
the concentration range of greater than 1 to 100ppm acetaldehyde has been linked to
moderate acute toxicity to aquatic organisms (EPA 1994).
Studies have implicated acetaldehyde in causation of tumours in experimental animals
(Feron et al. 1982, Woutersen et al. 1984, Woutersen et al. 1986). Acetaldehyde has
also been suggested to be an important factor in alcohol-associated carcinogenesis of

35

the upper aero-digestive tract of humans (Seitz and Meier, 2007). Based on inadequate
evidence for the carcinogenicity of acetaldehyde in humans and sufficient evidence for
its carcinogenicity in experimental animals the International Agency for Research on
Cancer (IARC) categorised acetaldehyde as being possibly carcinogenic to humans
(IARC Group 2B) (IARC, 1999). However, an IARC working group of 30 scientists
from 10 countries reassessing the carcinogenicity of some carcinogens and identifying
additional tumour sites and mechanisms of carcinogenesis concluded that acetaldehyde
associated with alcoholic beverages (derived from the alcoholic beverages and formed
inside human body as a result of alcohol metabolism) is carcinogenic to humans
(Group 1)(IARC 2009)

4.3.4

Guideline and regulatory standards for acetaldehyde in water and

foods

International guidelines and standards for acetaldehyde in drinking-water have not been
established (IARC 1999). The EU specific migration limit (SML) for acetaldehyde in
foods is 6mg/kg (6000 ppb) as specified in the Commission Directive 2002/72/EC of 6
August 2002 (European Commission, 2002). The tolerable daily intake (TDI) for
acetaldehyde is 0.1mg/kg body weight per day in humans (European Commission,
1998) The Occupational Safety and Health Administrations (OSHA) Permissible
Exposure Limit (PEL), the American Conference of Governmental Industrial Hygienists
(ACGIH) Threshold Limit Value - Ceiling (TLV-C) and the US National Institute for
Occupational Safety and Healths (NIOSH) Immediately Dangerous to Life and Health
(IDLH) limit are 200, 25 and 2000ppm respectively (National Toxicology Program
2010).

4.4

Basis and implications of water quality standards

Water quality guidelines and standards for drinking water refers to maximum levels or
concentrations of chemical, microbiological and physical contaminants that are allowed
in drinking water based on evidence that such concentrations do not result in any
significant risk to health over a lifetime of consumption. The primary purpose of both
guidelines and standards is protection of the public health. While guidelines are nonlegally binding recommendations, standards are legally enforceable national regulations
and thus infringement can attract prosecution (Radojevi and Bashkin, 1999). WHO
guidelines for drinking-water quality are international norms on water quality and

36

human health meant to be used as the basis for regulation and setting of standards in
all countries. Consequently standards tend to differ between countries as exemplified
by drinking water antimony standard for EU, US and Japan as mentioned in 4.2.4.
Standards also tend to be more stringent than the WHO guidelines.

4.5

The implication of IARCs possible carcinogen status

The IARC categorised antimony trioxide and acetaldehyde in Group 2B based on

inadequate evidence of carcinogenicity in humans and sufficient evidence of
carcinogenicity in experimental animals. The basis and implication of this categorisation
is explained in IARC (2010). According to IARC inadequate evidence of carcinogenicity
in humans may either imply insufficient quality, consistency or statistical power of the
available studies to permit a conclusion regarding the presence or absence of a causal
association between exposure and cancer, or absence of data on cancer in humans. On
the other hand sufficient evidence of carcinogenicity in experimental animals implies a
situation where a causal relationship has been established between the agent and an
increased incidence of malignant neoplasms or of an appropriate combination of benign
and malignant neoplasms in (a) two or more species of animals or (b) two or more
independent studies in one species carried out at different times or in different
laboratories or under different protocols. Additionally an increased incidence of
tumours in both sexes of a single species in a well-conducted study, ideally conducted
under Good Laboratory Practices, can also provide sufficient evidence. Furthermore a
single study in one species and sex might be considered to provide sufficient evidence
of carcinogenicity when malignant neoplasms occur to an unusual degree with regard
to incidence, site, type of tumour or age at onset, or when there are strong findings of
tumours at multiple sites.
The overall implication of this categorisation is that the 2 chemicals have been
confirmed to be carcinogenic in experimental animals but that there is high degree of
uncertainty with regard to their carcinogenicity in humans. However, as mentioned
earlier acetaldehyde in alcoholic drinks and acetaldehyde formed in the body as a
result of their metabolism has recently been upgraded to Group 1 carcinogen.

4.6

Leaching of antimony and acetaldehyde into water and foods

Chemicals documented to leach from PET bottles/containers into water and/or food
include antimony (Hansen and Pergantis, 2006; Shotyk et al, 2006; Shotyk and Krachler,
37

2007; Westerhoff et al, 2008), carbonyl compounds including acetaldehyde,

formaldehyde and acetone (Sugaya et al, 2001, Nawrocki et al, 2002, Matsuga et al,
2006), PET cyclic oligomers (Matsuga et al, 2005; Nasser et al, 2005) and UV stabiliser
Tinuvin (Begley et al, 2004). Bis(2-ethylhexyl) adipate (DEHA) and Bis(2ethylhexyl)phthalate (DEHP), the plasticisers used with PVC but not with PET, have
also been reported to migrate from PET bottles (Lilya, 2001, Nathan and Philip, 2009)

4.6.1

Antimony

In PET polymerisation germanium-, titanium-, antimony-, cobalt-, manganese-,

magnesium- and zinc-based catalysts are employed but consequent of lower cost of the
antimony-based catalysts in relation to their efficiency more than 90 % of the world
PET production is made by addition of 150-350 ppm antimony mostly as antimony
trioxide and to lesser extent as antimony acetate or antimony glycolate (EFSA, 2004;
Thiele, 2004; Westerhoff et al, 2008). Concentration of antimony in PET materials
reported by different authors using different analytical methods mostly falls within the
industry reported concentrations as shown by Table 4.4 below. Antimony leaches into
bottle contents because it is not chemically bonded to PET material. The relationship
between antimony, PET bottle material and bottle contents is shown in Figure 4.1
Table 4.4 Antimony in PET material
Concentration (ppm)
397, 351
98 - 190
213
357 8, 326 6
210, 290

method
INAA
FAAS
ICP-MS
ICP-MS, XRF
ICP-MS

author
Shotyk et al, 2006
Lopez-Molinero et al, 2007
Westerhoff et al, 2008
Alt et al, 2008
Keresztes et al, 2009

As earlier said in Subsection 4.2.2 the typical concentration of antimony in unpolluted

water is less than 1g/L. Consequently the antimony content of freshly bottled water
will generally not be expected to go beyond this concentration significantly. As
antimony has been established to migrate from PET bottles into content the antimony
concentration in water bottled in PET cannot confirm the actual antimony content of
the source water. Additionally antimony migration cannot be confirmed unless
experiments varying parameters of interest are carried. The antimony content of
water from Hungarian well used by bottlers was reported to be below the detection
limit of 0.7ng/L in the work of Keresztes et al (2009). In an earlier study Shotyk et al
38

(2006) reported antimony concentrations of 3.8 0.9 ng/L for bottled water source in
Germany. The antimony concentrations of bottled water samples as reported by
different authors are given in Table 4.5. From the Table it can be seen that the EU
maximum admissible of 5g/L has not been exceeded in any of the studies, though it
has been attained in the Turkish study. Factors reported to influence antimony
migration and the degree to which they effect the migration will be discussed in
Section 4.7.

Figure 4.1 Relationship between antimony, PET and bottle contents

Table 4.5 Antimony in bottled water
Concentration detection Brands author
(ng/L)
limit (ng/L)
8.9 2570
0.35
69
Shotyk et al 2007
7 249
2
158
Cicchella et al 2010
nd 931
10
Reimann et al 2010
2.15 2350
2
Birke et al 2010
95 521
4
9
Westerhoff et al 2008
nd - 5000
189
Gler 2007

4.6.2

country
16 countries
Italy
many
Germany
US
Turkey

Acetaldehyde

As mentioned earlier in Subsection 2.2.4, acetaldehyde is generated as a result of

thermal degradation during the melt-processing stage of PET bottle manufacture.
Acetaldehyde formed in the plastic matrix can either diffuse outward into the
atmosphere or inward into the contents of the PET bottle (Kenplas, 2008). The

39

relationship between acetaldehyde, PET bottle material and bottle contents is shown in
Figure 4.2. The odour and taste threshold of acetaldehyde in water is reported to be
2040 g/L (Nijssen et al, 1996; Schrder, 2001). In beverages bottled in PET
acetaldehyde the taste is masked by the flavour of the beverages. In bottled water
however very small amounts of acetaldehyde can be tasted and smelt as the result of
the low odour and taste threshold (Kenplas, 2008, Federal Institute for Risk
Assessment, 2007). Consequent of such differences PET materials with minimal levels
of acetaldehyde are desired for water bottling. The acetaldehyde concentrations in 38
PET bottle materials in the work of Matsuga et al, (2005) were 8.4-25.7 g/g in
Japanese bottles, 5.0-13.1 g/g in French and Italian bottles, and 9.1-18.7 g/g in US and
Canadian bottles, respectively. From the work of the same author the acetaldehyde
concentration in 10 different bottle- and sheet-making PET pellets was 3.5 12.4 g/g.
Bashir et al (2002) reported much lower concentrations (0.3 0.8 g/g) in an earlier
study involving nitrogen-cooled ground PET material. While acetaldehyde does migrate
from PET bottle materials into bottled water it has also been found in water not in
contact with PET. Nawrocki et al (2002) reported concentrations ranging of 1.4, 0.1
and 4.5 g/L for distilled water, deionised water and tap water respectively. Sugaya et
al (2001) reported a maximum concentration of 1.1g/L for tap water samples.
Concentration found in bottled water and beverages are given in Table 4.6
Table 4.6 Acetaldehyde in bottled water and soft drinks
Type

Concentration detection
(g/L)

Brands author

limit (g/L)

sample
country

Bottled water 0.9 317.8

0.1

14

Nawrocki et al 2002 Poland

Bottled water nd 260

0.5

32

Sugaya et al 2001

Many

Bottled water nd 107.8

20

Mutsuga et al 2006

Many

Beverages

Miyake and

US

460 - 101900

Shibamoto 1993
Carbonated
beverages

18.5 358.5

10

40

zlem 2008

Turkey

Figure 4.2 Relationship between acetaldehyde, PET and bottle contents

4.7
4.7.1

Factors affecting the leaching of chemicals

Elevated temperatures

Temperature is a measure of kinetic energy of molecules, hence the higher the

temperature the faster the molecular movements and chemical reactions. Consequent
of higher kinetic energy of molecules at elevated temperatures degradation of PET
material and leaching of chemicals from PET bottles into the environment is expected
to increase. Westerhoff et al, (2008) reported faster rates of antimony leaching from
PET bottle into the water with increasing storage temperatures. After 7 days at 80C,
for example, the authors recorded antimony concentration of 14.4 ppb in the water, a
concentration more than twice the EPA Maximum Contamination Level (MCL). The
authors also extrapolated that for exposure temperatures of less than 58C, exposure
durations of greater than 1 year were required to reach the MCL. For exposure

41

temperatures of 60, 65, 70, 75, 80, and 85 C, however, the exposure durations
required to reach the MCL decrease rapidly to 176, 38, 12, 4.7, 2.3, and 1.3 days,
respectively. In the work of Al-Malack (2001) migration of lead and cadmium from PVC
pipes into water was not found to be significantly affected by the increase in water
temperature. In contrast, tin, barium, and calcium concentrations were found to
increase when the water temperature was raised from 35 to 45C by 42, 85 and 29%,
respectively. Nawrocki et al (2002), Ahmad and Bajahlan (2007) and Le et al (2008) also
reported higher migrations of carbonyl compounds from PET bottles into water,
styrene monomer from Styrofoam cups and bisphenol A from polycarbonate bottles
into drinking water respectively at elevated temperatures. Elevated temperatures not
only promote migration of chemicals but also the degradation of the plastic material
itself. By implication exposing food, water and beverages in plastic containers to higher
temperatures could result in consumption of elevated levels of the different chemicals
that leaches from the plastic materials into the foods, water, beverages and the
environment. In a SODIS study by Tukur et al (2006) a temperature of 58.3 C was
reported to be achievable on exposure of water in glass bottles for 2.5 hours from
midday.

4.7.2

Sunlight and UV radiation

Sunlight is made up of infrared, visible and ultraviolet radiations. Light and

electromagnetic waves of any frequency can bring about heating effect on surfaces that
absorb them. The suns infrared radiation accounts for 49% of the heating of the Earth
(Arizona Solar Center, 2008). Consequently, direct exposure to sunlight can bring
about the same effect as heat on the rate of molecular movements and chemical
reactions. In SODIS, for example, sunlight treats contaminated water through the
synergetic effects of DNA alterations by UV-A radiation, photo-oxidative disinfection
and, heat inactivation (Borucke et al., 2001). In addition to the sunlights ability to
accelerate leaching of chemicals as does elevated temperatures the ultraviolet
component of solar radiation is known to bring about degradation of plastic materials.
It is this UVs degradative ability on plastics that prompts the addition of UV-stabilizing
additives to plastics during processing. In the work of Westerhoff et al, (2008) UV
irradiation increased leaching of antimony significantly even though as the authors
reported separating the effects of UV irradiation from those of heating on antimony
leaching from the plastic is difficult. Nawrocki et al (2002) observed the enhancement
42

of acetaldehyde migration as a result of synergistic effect of sunlight and elevated

temperatures. Al-Malack (2001) also reported increased migration of metallic
stabilizers from PVC pipes into water as time of exposure to UV radiation increased.

4.7.3

Carbonation

Carbonation is the addition of carbon dioxide into water or aqueous solutions.

Carbonation lowers the pH of sparkling water or soft drinks by raising the hydrogen
ion concentration through formation of carbonic acid. In the reports of Nijssen et al
(1996), Schrder (2001) and Nawrocki et al (2002) carbonation of water was reported
to enhance formation and/or migration of acetaldehyde from PET bottles into water.
No report is available on the mechanism by which carbonation enhance formation
and/or migration of acetaldehyde. However the acidification of water or aqueous
solution brought about by the carbonation process may play some role. Carbonation
was not reported to affect migration of other plastic-associated chemicals into water.

4.7.4

Storage

Prolonged storage of bottled water and soft drinks may occur at the supply chain stage
(production, wholesale and retail stages) or at consumer stage. In theory the longer
the duration of contact between water, soft drinks or foods and the packaging material
(PET, glass, etc), the higher the amount of migrant chemicals to be found in the water
or soft drinks. In PET bottled water stored for 9 months the content of aldehydes
gradually increases over the period of 89 months and then begins to decrease
(Nawrocki et al, 2002). The gradual decrease of aldehydes concentration was
associated with the gradual loss of dissolved CO2, a gas whose presence in bottled
water enhances the formation and/or desorption of acetaldehyde. Both CO2 and
aldehydes are believed to diffuse through the bottle wall into the environment. Le et al
(2008) described noticing the release of bisphenol A from polycarbonate bottles
increasing with time at room temperature. Hansen and Pergantis (2006) and Shotyk
and Krachler (2007) also reported increased leaching of antimony from PET bottles
into citrus juices and water respectively with greater duration of storage. Al-Malack
(2001) in an investigation on metal stabilizers leaching from PVC pipes into water
reported observing increase in metal stabilizer concentrations with respect to time of
exposure. After 10 h of exposure, lead concentration reached a value of 0.43 mg/l, and
by the end of the experiment (48 h), it increased to 0.78 mg/l. Tin was found to

43

increase to 0.27 and 0.31 mg/l after 24 and 48 h of exposure, respectively. Barium and
cadmium were found to increase to 0.42 and 0.1 mg/l, respectively, after 48 h of
exposure to double distilled water. Moreover, calcium concentration increased to 46
and 49 mg/l after 24 and 48 h of exposure, respectively.

4.7.5

Bottling material

Acetaldehyde migration is expected to take place only in PET bottle as it cannot be

generated in glass. In the work of Sugaya et al (2001) acetaldehyde concentrations in
mineral water bottled in both glass and paper containers were below detectable levels.
Because antimony is used in glass industry as fining agent and as pacifying agent as
stated earlier in Section 2.3, its migration from glass bottles will not be unexpected
depending on the nature of the glass material. Water bottled in glass was reported to
show antimony concentration higher than the source water but lower than for water
bottled in PET in the work of Shotyk et al (2006). For the source water concentration
of 3.80.9 ng/L the concentration in glass and PET bottled water were 11.54.4 and
35954ng/L respectively. In the work of Hansen et al (2010) the highest concentration
of antimony (13.6g/L) was found in a fruit juice bottled in glass.

4.7.6

Plastic aging

Plastic aging in the context of this study refers to noticeable changes that occur in
plastic materials over time due to degradation. Degradation of polymeric materials
denotes changes in physical, mechanical, optical, thermal and other characteristics
brought about by chemicals, heat, microbial attack and mechanical handling and light.
PET bottle ageing brings about yellowing of bottle surfaces, loss of elasticity and some
degree of opacity in bottles. Information on the effect of PET bottle aging on leaching
of chemicals from the bottles into the water is scanty. However in the work of
Nawrocki et al (2002) the concentrations of acetaldehyde in 1-month old bottle
extract were evidently lower than those from the fresh bottle extract. Such was
probably because acetaldehyde content of the bottles diminishes with time and that
the older bottle was not exposed to conditions that enhance the formation of
acetaldehyde. While Howdeshell et al (2003) reported an increased rate of bisphenol A
leaching from polycarbonate plastic with age, Le et al (2008) did not observe significant
leaching difference between new polycarbonate bottles and bottles used under normal
conditions for 1 to 9 years. The expectation is that concentration of additives and

44

other plastic chemicals that are not replenished and their ability to leach will diminish
with time. On the other hand new chemicals could be generated as a result of the
degradative action of the environment on the plastic material.

4.7.7

pH

Low pH and high total dissolved solids were associated with increased migration of
metallic stabilizers from PVC pipes into water (Al-Malack, 2001). However in the work
of Westerhoff et al (2008) pH had no effect on antimony leaching into bottled water
within the range of 6.3 to 8.3. In the report of Dental Abstracts (2007) pH values of 20
soft drinks ranged from 2.39 to 4.04. In that report the pH values of non-cola drinks
were found to be significantly higher than those of cola drinks and the sugared cola and
non-cola drinks had lower pH values than the non-sugared versions.

4.7.8

Bottle colour

In an experiment to evaluate the leaching potential of two PET bottles of different

colours (clear and blue-coloured) Westerhoff et al, (2008) found that the clear plastic
PET released four times more antimony than the blue-coloured PET plastic. This
experiment was conducted with only one brand each of the clear and the bluecoloured bottles and was not extended to plastic bottles of other colours.

4.7.9

Other factors

Other factors that are thought to possibly influence leaching of chemicals from PET
bottles and other plastic containers are washing with alkaline detergents in the case of
polycarbonates (Biedermann-Brem et al, 2008), bottle size (Keresztes et al, 2009), and
aggressive washing to eliminate potential bacterial contamination.

4.8

Summary

Antimony and acetaldehyde are the most important chemicals reported to be

migrating from PET bottle material into bottle contents. Antimony leaches out of PET
because it is used as a catalyst in the manufacture of PET at concentrations of 150
350mg/kg. Acetaldehyde is generated in PET as a product of PET degradation at
elevated temperatures. The European Union maximum admissible concentration of
antimony in drinking water is 5g/L. The tolerable daily intake (TDI) is 6g/kg body
weight per day in humans. Typical concentrations of dissolved antimony in unpolluted
waters are less than 1g/L. Nonetheless concentrations can reach up to 100 times the

45

natural levels in the proximity of anthropogenic sources. Concentrations of antimony

reported in bottled water ranged from low levels that cannot be detected by analytical
instruments to 5 g/L. Acetaldehyde, a volatile organic compound occurring naturally
in ripe fruits, is used as a flavouring agent in soft drinks and other foods. International
guidelines and standards for acetaldehyde in drinking-water have not been established.
The tolerable daily intake (TDI) is 0.1mg/kg body weight per day in humans. The odour
and taste threshold of acetaldehyde in water is reported to be 20-40 g/L.
Concentrations of acetaldehyde are generally less than 0.1 g/L, less than 1 g/g and
averaging 5g/m3 in water, wide range of foodstuff and in air respectively. Maximum
concentrations of acetaldehyde reported in bottled water and soft drinks are 317.8
g/L and 101.9 mg/L respectively. Factors reported to influence the migration of
antimony and acetaldehyde from PET bottle materials into bottle contents include,
temperature, sunlight and UV radiation, carbon dioxide content, storage, bottling
material, age of plastic material, pH and bottle size.

4.9

Conclusion

Typical concentrations of antimony and acetaldehyde in unpolluted water in the

environment and in bottled water and soft drinks have been revealed in this Chapter.
The Chapter also revealed the maximum concentrations of antimony and acetaldehyde
allowed in bottled water and soft drinks for the purpose of safeguarding human health.
The typical concentrations together with the maximum allowable concentrations will
be used as the basis for assessment of the concentrations of antimony and
acetaldehyde in bottled water and soft drinks in this study.

46

CHAPTER 5:
5.1
5.1.1

METHODOLOGY

Research strategy
The role of literature in streamlining the research objectives

The literature reviewed has helped in streamlining the different objectives this study
focuses on. The first objective sought to examine the pattern and extent of bottled water
and soft drink use and reuse in Britain and Nigeria. The objective was to be achieved by
quantifying bottled water and soft drinks use and PET bottle reuse, assessing storage
periods and PET bottle reuse periods and establishing the number and sizes of bottles
being used and reused in the two countries. The Objective also sought to assess how
PET bottle reuse is been perceived and also the factors influencing reuse in the two
countries. From the literature bottled water and soft drinks use have been shown to
be higher in developed countries than in developing countries. However the literature
gave little information in terms of bottle reuse pattern and extent in developing and
developed countries. This observation points to a need to investigate the pattern of
reuse in both developing and developed countries and also to see whether the factors
influencing this behaviour are similar in the two countries. Because bottled water does
not technically expire, Health Canada recommended a period of up to one year for
storage of emergency bottled water. Considering the fact that some freshly purchased
Turkish bottled water was reported to contain antimony at a the EU MAC (5g/L)
(Gler, 2007), it will be worthwhile to know how long people store purchased bottled
water and soft drinks before consumption and what happens during extended storage
periods in terms of chemical migration. From the literature review some internet
pages associate PET bottle reuse with consumption of carcinogenic chemicals. A
question arise here as to whether carcinogenic chemicals actually migrate from reused
bottles. It is also worth knowing the extent to which this information in the internet
does influence PET bottle reuse.
The second objective sought to assess the extent of antimony and acetaldehyde migration
from PET into water and soft drinks under conditions representative of real use and reuse in
Britain and Nigeria. The literature review outline the stages involved in the manufacture
of PET including how antimony is added in the manufacture of PET and how
acetaldehyde is generated. Based on the information in the literature acetaldehyde is
only formed in PET at temperatures above 160C. Consequently acetaldehyde will not

47

be expected to be formed in the PET in any other stage apart from the bottle
manufacturing process. This information also revealed that a temperature of up to
150C can be used in determination of acetaldehyde in PET material using headspace
GC-FID. Information on the concentration of acetaldehyde and antimony in the
environment, in PET and in bottled contents from previous works has been revealed in
the literature. This presents an opportunity for comparison of what has been found in
the study and what the literature revealed. For example the review revealed that
antimony concentration in unpolluted water is below 1g/L. The literature review has
also revealed that acetaldehyde is naturally present in fruit juices at concentration up
to several 100mg/L and that it is added to processed foods including soft drinks as
flavouring agent. Consequently this study will attempt to quantify the acetaldehyde
contents of bottled fruit juices and other soft drinks. Bottle caps and caps linings, are
the other bottling materials in contact with bottled contents in addition to the bottle
material. These bottle components are not expected to release antimony and
acetaldehyde as they are made from constituents different from those used in making
PET. Additionally unlike in PET, chemicals other than antimony are the catalysts used in
the manufacture of these plastics. Notwithstanding, these expectations, this study will
analyse these material for the purpose of confirming their identity (i.e. whether they
are made up of PP, PE, etc) and also to rule out the presence of antimony in the
materials. It has been established from the review that antimony may be used in the
course of glass manufacture for the purpose of removing bubbles trapped in the glass.
A single study has also found antimony in glass bottles at much lower concentrations
than in PET (Shotyk et al, 2006). Another study also revealed migration of antimony
from glass bottles at lower rate in comparison to PET (Cicchella et al, 2010). The
current study did not quantify the antimony concentration in glass bottle materials.
However the study looked at antimony concentration in bottled water and soft drinks
bottled in glass. The study also assesses the migration of antimony from some glass
bottles at elevated temperatures. The review discussed factors that are reported to
have some influence on the migration of chemicals from bottles into their contents.
Some of these factors and other factors not reported in the literature were assessed
in this study for the purpose of monitoring how these factors may influence migration.

48

The third objective draw on findings from 1 and 2, as well as existing regulations and
controls to assess whether acceptable limits of chemicals in foods and water are being
exceeded as a result of the chemical migrations. The review clearly stated the different
regulations currently in force regarding the presence of antimony and acetaldehyde in
foods and water, the degree of chemical migration permitted and quantities of
antimony and acetaldehyde that can be taken on daily basis without the risk of suffering
from harm over the lifetime. This information from the review form the basis for the
assessment of what is happening in terms of migration of antimony and acetaldehyde
into foods and water against the existing regulations.
The fourth objective sought to generate recommendations about the extent to which
existing regulations and controls might merit re-examination. Useful information obtained
from the review that may have some significance on the achievement of this objective
is the recent decision by IARC to upgrade acetaldehyde in alcoholic drinks to the
status of group one carcinogen (Human carcinogen).

5.1.2
5.1.2.1

Issues informing the adoption of the strategies used in the research

Sampling locations

Bottled water and soft drinks are consumed worldwide and PET bottles are possibly
reused in large number of countries. Any study relating bottled water and soft drinks
use and PET bottle reuse with presence and migration of chemicals, will be better if it
involves many countries. Behaviour may differ with countries and regions. Also
regulations guiding the use of chemical ingredients in the manufacture of bottles and
the ambient quality of ground water utilized in bottling may differ. However feasibility
is the main issue that should guide a study of this nature. Consequently this study
resolved to use Nigeria and Britain for the purpose of the survey and collection of
samples for laboratory analysis. Nigeria and Britain were chosen based on the evidence
of differences in terms of weather and prosperity. Apart from these two factors there
could also be other socio-cultural differences between the 2 countries. The two
countries can be seen as representation of developing sub-Saharan countries and
developed western countries. An additional advantage associated with the selection is
the fact that the researcher is familiar with the two countries, and consequently the
relative ease of basing the research on the countries. Thus, the selection of the two
case study countries provides a route towards exploring how the differences between
the countries may influence the answers to the research questions. It is believed that
49

the results of the survey could be generalized to a large extent to cover other
developing sub-Saharan countries and developed western countries. Laboratory
findings may or may not differ for the two countries. In general laboratory findings may
not be easily generalizable to other countries. However generalization is by and large
more likely to be possible between Britain and other developed western countries
than between Nigeria and other developing sub-Saharan countries. This is in large part
due to the existence of European Union, a platform through which a standardised
system of laws and regulations are generated and applied in all member states.

5.1.2.2

Use of survey

The decision to embark on the survey was as a result of the quest to understand how
the behaviour of respondents regarding bottled water and soft drink use and PET
bottle reuse may affect the migration of antimony and acetaldehyde. The survey also
has the potential to enrich some of the user behaviour information that is currently
scanty in the literature. For example only a single study was found revealing the
pattern and extent of PET bottle reuse (Lilya 2001). A survey was selected as a good
means of collecting this information because the data needed was not complex, it
included relatively few issues of interpretation, and data could be collected from a
larger number of people than would have been possible through interviews or other
more in-depth social science investigation techniques.

5.1.2.3

Bottle samples

The samples primarily targeted by this study are bottled water and soft drinks in PET
bottles. These samples are targeted because antimony and acetaldehyde are expected
to migrate only from PET plastic. However a resolve was made to also have a look at
some samples bottled in glass, cartons and plastic pouches for comparison.

The

laboratory analytical equipment utilized in this study includes ICP-MS, GC-FID, EDX
and Raman spectrometer. These instruments were selected because of the strength
they have in analyzing the analytes of interests. EDX and Raman spectrometer were
not as essential as ICP-MS and GC-FID. However their use helped in confirming the
identity of plastic samples analyzed in this study in addition to providing an interesting
secondary information as will be discussed in Section 7.7. Use of household microwave
oven to digest PET for the purpose of quantifying antimony has not been described in

50

the literature. The study explores this method based on earlier report involving the
digestion of polyethylene for quantification of metals. (Sakurai et al, 2006)

5.2

Survey strategy and sampling

As earlier stated the first objective of this research is to examine the pattern and
extent of bottled water and soft drink use and PET bottle reuse in Britain and Nigeria.
This objective will be addressed through survey. The objective is further divided into
sub-objectives listed below:
i.

To establish the pattern and extent of typical bottled water and soft drink use
and PET bottle reuse including, the approximate proportion of individuals/
households using and reusing plastic bottles, bottled water and soft drinks
storage periods prior to use, bottle sizes most often used and reused, bottle
reuse periods, the number of bottles being used and reused by an individual
or a household at any one time, etc.

ii.

To determine public perceptions of the safety of reusing plastic containers to

store drinking water, beverages, etc

iii.

To find out factors influencing peoples preferences with respect to reuse of

plastic water containers

This Section describes the approach employed in identifying the sampling frame for the
survey, the different considerations that guided the survey questions content, how the
sampling process was carried out and how the data was analysed.

5.2.1

Sampling strategy

Ideally, this survey would have sampled a representative proportion of the population
in Nigeria and Britain.

However, for a study such as this, identifying such a

representative sample poses significant issues, namely:

Not enough is known about plastic bottle use and reuse to identify what factors
(age, household size, gender, social class) should be used to achieve a
representative sample;

The resources devoted to this part of the study are not sufficient to sample
across multiple locations, in particular, it is not viable to visit many households
individually;

51

Attention needs to be given to the likelihood of invited participants choosing to

fill in the survey.

In response to these issues, the study was conducted in university environments.

Specific advantages associated with this sampling strategy are as follows:

While University populations cannot be said to be completely representative of

the wider population, there is no reason to think their plastic bottle use and/or
re-use is atypical of the societies concerned.

The data will yield indicative results, enabling identification of general patterns of
bottle use and re-use, and perhaps identifying significant factors that contribute
to its variation. As such, the study could guide the dimensions of future better
resourced studies of plastic bottle use and re-use.

The sample population is easily accessible, and because based in a University

environment, are more likely to be sympathetic to the objectives of carrying out
research;

Multiple households can be accessed within a small number of locations (i.e. the
semi-public spaces within University campuses).

5.2.2

Questionnaire development

This Subsection discusses the different considerations that have guided the
development of the questionnaire used in the survey. Initially the survey was intended
to be in the form of an interview containing open ended questions. Interviews were
agreed to be more efficient in extracting information from respondents than
questionnaires. Among other issues questions not understood by respondents could
easily be rephrased and potential compromise to response could be avoided.
Interview questions were developed in collaboration with supervisors. Ten copies of
the interview questions were piloted among fellow research students between 24th and
27th of November 2008. One of the supervisors was interviewed on Thursday 27th of
November, 2008 and feedback on the suitability of the survey was obtained. The
supervisor suggested more detailed introduction of the survey to prospective
respondents. The supervisor also advised on the need to avoid reading direct from
document when interviewing respondents and also the need for greater confidence

52

and composure in the course of future interviews. However, the other supervisor
advised that the introduction needs to explain the survey without telling so much as to
influence responses. The Interview was found to last for an average of 20 minutes.
To reduce the amount of time required to collect survey information from one
respondent a suggestion was made to explore the possibility of using questionnaires
rather than interviews. Questionnaires were seen as been more effective in gathering
large quantity of data within smaller time duration in comparison to interviews.
Additionally questionnaires were perceived as a better means of minimizing prestige
bias, which could be compounded in person-to-person interviews. Prestige bias is the
tendency for respondents to answer in a way that make them feel superior. The survey
questions were modified for use in questionnaires. Questions were reduced from 31 in
3 pages to 24 in 2 pages. In the course of the questionnaire development both closed
ended and open ended systems were considered. Open ended questions were finally
adopted because in such situation response is not restricted to options. With open
ended questions situations where no category fits the position of a respondent is
avoidable. Even though the questionnaire was meant to be the open format type in
some instances options, including none and dont know, were typed immediately after a
question. This was meant to remind the respondents that if applicable none and dont
know are also valid answers. The ultimate aim was to reduce the occurrence wrong or
ambiguous answers. Additionally a decision was made to restrict questionnaire
collection to only 2 universities rather than the 6 universities earlier proposed. The 2
agreed universities were the University of Bradford in England and Ahmadu Bello
University in Nigeria. Ten questionnaires were piloted in the Hub (a central student
oriented area of the University of Bradford) on 2nd of December 2008. Out of the 13
people approached only three declined because they did not have time. Nine out the
ten respondents filled the 2-page questionnaire fully. One respondent stopped halfway.
A quick look at the filled questionnaires revealed that they are as good as interview in
collecting the required information.
Appendix 1 shows the survey questions asked. The preamble aimed to set participants
at ease, and to explain the purpose of the survey to encourage them to participate.
Against this objective, the preamble was designed not to tell the participants so much
about the survey as to influence their answers. A picture of some bottled water brands

53

was pasted on the top right-hand corner of the questionnaire to further help the
respondents to differentiate between PET bottles and other plastic bottles. The first
question in the survey is to give an idea of the proportions of the respondents that
uses different bottled liquids. The survey categorised bottles into unopened bottles,
opened bottles with original liquid content and reused bottles. Those questions under
unopened bottle aim to reveal the proportion of respondents that usually have
unopened bottle in their places of residence, and the average time the bottles remain
unopened. As mentioned in Subsection 4.7.4 quantities of chemicals found in liquids
bottled in PET bottles increase with duration of storage. The questions on proportions
of the different sizes of bottles used and reused will reveal what bottle sizes are most
commonly used and reused. The study will look at the relationship between bottle
size and migration of chemicals. The answer to the question on bottle storage places is
also important because migration of chemicals is generally accelerated by light and
elevated temperatures. Questions under bottle reuse are intended to reveal the
proportion of respondents that are in the habit of reusing PET bottles at home, at
work and when on the move, the number of bottles being reused and the average and
longest times of bottle reuse. Questions at the end of the questionnaires relate to
respondents awareness of the safety debate on use and reuse of PET bottles and to
reveal what factors influence respondents choices to reuse PET bottles. Factors that
are thought to influence reuse are discussed in Section 3.9. The demographic data at
the end of the questionnaire were meant to provide a means for understanding the
demographic characteristics of the samples. For example behaviour of British
respondents can be compared to the behaviour of non-British respondents; the
behaviour of males can be compared to the behaviour of females, etc.

5.2.3

Sampling procedure

The survey aimed to achieve 1000 responses. However, a total of 995 questionnaires
were collected from the University of Bradford (464 questionnaires) and Ahmadu
Bello University (531) in Nigeria. In the University of Bradford the questionnaires were
collected between 2nd of December 2008 and 7th of April 2009. At the beginning a
questionnaire consisted of 2 sheets of paper printed on one side. After collection of
168 questionnaires from the University of Bradford the questionnaire was converted
into back to back print to make it look less lengthy in response to complaints from
some respondents. In Ahmadu Bello University the questionnaires were collected
54

between 21st of April 2009 and 5th of May 2009. The respondents included students
(undergraduate and postgraduate) and members of staff (academic and non-academic).
Data from students was collected by approaching the student in public spaces and
issuing them with the questionnaires. In some instances data was collected from
lecture halls. Data from members of staff was collected from offices. Of the 464
respondents from the University of Bradford 399 disclosed their country of origin
(54% British, 46% non-British) and 412 disclosed their gender (59% males, 41%
females). Of the 531 respondents from Ahmadu Bello University 495 disclosed their
country of origin (98.8% Nigerians, 1.2% non-Nigerians) and 494 their gender (68%
males, 32% females).

5.2.4

Data analysis

Survey data collected was analysed using SPSS and EXCEL. Results are presented using
descriptive statistics (measures of central tendency, tables, charts, etc) and inferential
statistics (statistical significance testing). Pearson's chi-square test was used to test
whether 2 or more qualitative variables are homogeneous. The ability of Chi square to
establish the status of similarity or dependency between variables depends on the
strength of the relationship between these variables and the sample size. So with large
samples it is possible to find significance even when the differences or associations are
very small (Morgan et al, 2004). This potential problem is taken care of by a statistic
called squared Cramers phi coefficient (2), a measure similar to correlation
coefficient in its interpretation. Phi statistics eliminate the effect of sample size by
dividing chi-square by the sample size. According to Cohen (1998) the relationship
between 2 variables is small if 2 approximates 0.01, medium if 2 approximates 0.09
and large if 2 approximates 0.25. None of the quantitative data encountered in this
study were normally distributed as shown by Kolmogorov-Smirnov and Shapiro-Wilk
tests of normality. Most of the data were skewed to the right (positively skewed).
Consequently nonparametric (Mann-Whitney U test) rather than parametric test (t
test) was used to test whether 2 or more of these quantitative data are similar. Unlike
their parametric counterparts nonparametric tests make no assumptions about the
probability distributions of the variables being assessed and can thus be used for data
that is not normally distributed.

55

5.3

Survey influence on laboratory analysis

The survey carried out in this study was in most part meant to provide information on
bottled water and soft drink use and PET bottle reuse patterns. The information from
the survey was expected to be used in designing some of the laboratory experiments
carried out in this study. Table 5.1 gives some insight on the role played by the
different questions in the survey towards designing some of the laboratory
experiments carried out in the study. The laboratory experiments influenced by the
survey questions are explained in Sections 5.8 and 5.9. These Sections primarily deal
with quantification of antimony and acetaldehyde in PET materials and bottled water
and soft drink samples.
Table 5.1 The role of the survey in defining laboratory experiments
Group Question Topic of
number(s) question
1
2

1, 2, 7,
11, 15
and 16
4, 5, 9
and 10

6 and 9

3, 8 & 14

12, 13
and 17

18 and 19

20

21, 22, 23
and 24

Influence on
laboratory
component

Location of
laboratory
experiment

laboratory
result /
discussion

5.8.12,
5.9.8

8.7, 8.13,
9.12

5.8.16

8.14

5.8.14,
5.9.10

8.12, 9.7

status of use
and reuse
storage of
bottles with
contents

influence of storage
on antimony and
acetaldehyde
migration
storage places storage at room
of bottles
temperature
with contents
sizes of bottle influence of bottle
being
size on antimony
used/reused
migration
Bottle aging
influence of bottle
aging on antimony
and acetaldehyde
migration
reuse safety
perception
Factors
influencing
bottle reuse
Demographic
information

56

Answers to other questions provided background information forming the basis for
going ahead with laboratory experiments. Other questions provided information which
is useful without influencing the laboratory experiments. For example misinformation
on the danger of PET bottle reuse is common on the internet. So the question on
reuse safety perception provided information on respondents opinion regarding the
safety of bottle reuse.

5.4

Materials and experimental methods

In this Section samples, chemicals, instruments and laboratory methods employed in

the research and where applicable the basis for employing them will be described. The
origin and purity of chemicals and condition of instruments will also be described. This
component of the research was achieved principally using ICP-MS and GC-FID. Raman
spectroscopy and Energy Dispersive X-ray spectrometry were utilised in authenticating
the identity of plastic and glass bottle materials, plastic bottle cap material, plastic cap
lining materials and metal crown and screw cap materials. Raman Spectroscopy and
Energy Dispersive X-ray spectroscopy are either not sensitive enough to detect the
low levels of antimony and acetaldehyde or they are entirely not meant for detection
of these analytes. Access to ICP-MS was initially delayed and infrequent essentially
because there was not much demand for the instrument to warrant the purchase of
argon gas. The research also wanted to compare the microwave digestion-ICP-MS,
used in this study to quantify antimony in PET, with laser ablation-ICP-MS. However
that was not achieved due to unavailability of the laser ablation system. At the early
stages of the laboratory work GC-MS was explored for quantitation of acetaldehyde.
Difficulties were encountered in separating acetaldehyde peak from a peak due to
nitrogen. In the long run the research resorted to GC-FID because of the greater
sensitivity of flame ionization detector (FID) to volatile organic compounds in
comparison to GC-MS and its insensitivity to nitrogen and other non-combustible
gases.

5.5
5.5.1

Samples and reagents

Samples

A total of 82 brands of bottled water and soft drinks in plastic and glass bottles and in
cartons were collected. A few samples from Nigeria in plastic pouches were collected.
Materials used in bottling including glass and plastic bottle materials, metal and plastic
57

bottle cap materials and plastic cap lining materials were collected. Plastic bottle
material identified as polyvinyl chloride based on plastic identification code was
collected for use as a reference in identification of materials made using PVC. The
summary of samples collected is given in Table 5.2. Drinking water from taps in the
university was collected four times. All samples were collected in supermarkets and
shops in Britain and Nigeria except drinking water from taps which was collected in
Britain only. Samples from Nigeria were collected on 14th of August 2009 and 1st of
August 2010. British samples were usually collected few days to the dates of analysis.
The decision to collect bottled water and soft drinks and their bottling materials was
based on the aim of this research to quantify antimony and acetaldehyde in bottled
water and soft drink samples and also to study the migration of these chemicals from
bottle wall into bottle content under different conditions.

5.5.2

Certified reference materials

Certified reference materials used in this research include

Trace element fortified water TW-DWS.2 (Environment Canada)
Polyethylene reference material ERM-EC681k (IRMM, Belgium).

5.5.3

Chemicals and reagents

The following chemicals and reagents were used in the experiments:

Nitric acid TraceSELECT (69.5%) for trace analysis (Sigma-Aldrich, Britain), nitric acid
(70%) analytical reagent grade, 10006ppm antimony ICP/DCP standard solution in 7.7
wt% hydrochloric acid (Sigma-Aldrich, USA), 1000ppm cadmium in ca. 1M nitric acid
(Fisher Scientific), 1000ppm germanium in water (BDH Laboratory Supplies, England),
1000ppm zinc in 2% nitric acid (CPI International), 1000ppm aluminium in 0.5M
hydrochloric acid (ROMIL, England), 1000ppm beryllium in 2% nitric acid (CPI
International), 1000ppm titanium in ca. 2M hydrochloric acid (Fisher Scientific),
1000ppm cobalt in 0.5M nitric acid (ROMIL, England), 1000ppm lead in 0.5M nitric acid
(ROMIL, England), Hydrochloric acid TraceSELECT (37%) for trace analysis (SigmaAldrich, Britain), indium ICP standard solution for ICP (Sigma-Aldrich, Britain),
deionised water from Direct-Q 3 water purification system (Millipore, USA), sodium
chloride extra pure (Acros Organics, USA), acetaldehyde - puriss. p.a., anhydrous,
99.5% (Sigma-Aldrich, Britain), sodium hydroxide solution, 1M (Fisher Scientific,
Britain), terephthalic acid (Sigma Aldrich, Germany, 98%).

58

10

Raman, ICP-MS, GC-FID

sparkling
PET
water/GB
Soft drinks/GB PET

Raman, ICP-MS, GC-FID

13

10

Raman, ICP-MS, GC-FID

Glass
Glass

2
5

2
2

EDX
EDX

Glass
PE
Carton
NA
PET
PET
Glass
PE pouch

7
1
2
4
11
5
5
5

5
1
NA
NA
11
5
1
1

EDX
ICP-MS, Raman
NA
NA
ICP-MS, Raman
ICP-MS, GC-FID, Raman
EDX
ICP-MS, Raman

Analysis
type

17

Cap lining

PET

Analysis
type

Container
(bottle
or pouch)

still water/GB

still water/GB
sparkling
water/GB
Soft drinks/GB
Soft drinks/GB
Soft drinks/GB
Tap water/GB
still water/N
Soft drinks/N
Soft drinks/N
still water/N

bottling material analysed

Cap
(plastic
or metal)

Container

Analysis
type

Contents/
country

Brands
collected

Table 5.2 Summary of samples collected and analysis carried out

Bottled
water and
soft drink
Brands
analysed

Analysis type

ICP-MS, NA NA
Raman
Raman NA NA

17

ICP-MS, GC-FID

ICP-MS, GC-FID

Raman

13

ICP-MS, GC-FID

2
5

ICP-MS,
Raman
Raman
Raman

2
2

Raman
Raman

2
5

ICP-MS, GC-FID
ICP-MS, GC-FID

3
4
NA
NA
5
4
2
NA

Raman
Raman
NA
NA
Raman
Raman
Raman
NA

5
1
NA
NA
NA
5
5
NA

NA = sample not analysed or not applicable

59

Raman
7
Raman
1
NA
2
NA
4
NA
11
Raman, EDX
5
ICP-MS, Raman
3
NA
lost in storage

ICP-MS, GC-FID
ICP-MS, GC-FID
ICP-MS, GC-FID
ICP-MS, GC-FID
ICP-MS, GC-FID
ICP-MS, GC-FID
ICP-MS, GC-FID
NA

5.6
5.6.1

Identification of plastic materials by Raman spectroscopy

Introduction

Raman spectroscopy is a non-destructive analytical method that can be used for

quantitative and qualitative analysis of organic, inorganic and biological materials.
Raman spectroscopy results in sharp spectral bands whose pattern and intensity is
dependent on the type of atoms or molecules present in a sample and the
concentration of the chemical. In this work Micro-Raman spectroscopy, a procedure
integrating microscopy and Raman spectroscopy, was used to characterize plastic
materials associated with bottling of water and soft drinks. Samples analysed include
bottle material, bottle cap material and cap lining material.

5.6.2

Principle of Raman spectroscopy

As monochromatic light impinges upon molecules of a sample the photons which make
up the light may be absorbed, transmitted without interacting with the molecules or
may interact with the molecules and consequently get scattered. If interaction occurs,
molecules absorb photon energy and begin to vibrate. The vibration results in the
movement of the molecules from ground state or a vibrational level of ground state to
an unquantized virtual level between the ground state and the first electronic excited
state (Figure 5.1). Usually most of the molecules will return to their original level
without overall gain or loss of energy. In such circumstance the incident photons will
be scattered elastically without any change in energy through a process referred to as
Rayleigh scattering. Notwithstanding, a small fraction of the molecules will return to
levels lower or higher than their original levels as a result of exchange of energy with
the incident photons. Accordingly the photons that exchange energy with the
molecules - approximately 1 in 10 million - are shifted to higher or lower frequencies
relative to the incident photons. This kind of scattering is called Raman scattering and
the spectrum of the wavelength-shifted electromagnetic radiation is called the Raman
spectrum. Vibrations that occur as a result of interaction of photons and molecules can
be stretching vibrations (symmetrical or asymmetrical) or bending vibrations
(scissoring, rocking, wagging or twisting).

60

Figure 5.1 Vibrational energy level diagram showing the transitions involved in Raman
signal generation (adapted from Skoog et al, 1998)
Because energy is transferred either from the molecules to the photons or vice versa
after electronic relaxation, two kind of Raman shifts centred around Rayleigh
scattering usually occur. Energy shift of the photon to blue region of the spectrum is
observed when virtual state molecules originally from excited vibrational levels transfer
energy to the photons and thus returning to lower energy ground state. This process
is called anti-Stokes shift. Conversely, the energy of the photon will shift to the red
region when virtual state molecules originally from ground level gain energy from the
photons. These more energetic molecules return to an excited vibrational level rather
than their original ground level in a process referred to as Stokes shift. Under normal
conditions most molecules are in the ground vibrational level thus stokes shift is more
likely to happen than anti-Stokes shift. Consequently in a Raman spectrum the less

61

energetic stokes lines are usually more intense than the more energetic anti-Stokes
lines (Figure 5.2). Stokes Raman shift is more commonly utilised in Raman
spectroscopy and is generally simply called Raman shift.

Figure 5.2 Relationship between Raman spectra (stokes and anti-stokes) and Rayleigh
scattering (source: author)
Fluorescence, which at times constitutes a nuisance to Raman spectroscopy, occurs
when the energy of the excitation photon approaches the transition energy between
the two electronic states. It differs from a type of Raman spectroscopy termed
resonance Raman spectroscopy in that relaxation to the ground state is preceded by
prior relaxation to the lowest vibrational level of the excited electronic state. It is
usually avoided by careful selection of appropriate laser excitation wavelength to
ensure that either the excitation photon does not provide enough energy to the
molecule as to elicit fluorescence or the fluorescence so generated differs remarkably
in energy from the Raman signal in such a way that it cannot interfere with the Raman
spectrum. Another way of eliminating fluorescence is to expose a sample to the laser
beam until the fluorescence decays.

62

A typical Raman system consist of a powerful laser in the visible region of

electromagnetic spectrum, a sample illuminating chamber, a narrow-band rejection
filter to minimise the intensity of Rayleigh scattered light, a spectrometer to disperse
the inelastically scattered light and a detector (Figure 5.3)

Figure 5.3 Schematic of Raman spectroscopy (source: author)

5.6.3

Instruments

inVia Reflex Raman microscope equipped with Renishaw 785nm near infrared Diode
laser (Renishaw plc, Gloucestershire, Britain), glass slides.

5.6.4

Method

Raman spectra were collected using the Renishaw InVia Reex dispersive Raman
microscope (Figure 5.4). The Raman scattering was excited with a 785 nm nearinfrared diode laser (Renishaw HPNIR laser) and a 50X objective lens giving a laser
spot diameter of 5m. Spectra were obtained for a 10s exposure of the CCD detector
in the wavenumber region 1003200 cm-1 using the extended scanning mode of the
instrument. Up to 60s exposure time was used where fluorescence is encountered and
this longer exposure time was found useful in eliminating the fluorescence. With 100%

63

laser power, one, nine or twenty accumulations were collected for samples. Spectral
acquisition, presentation, and analysis were performed with the Renishaw WIRE
(service pack 9) and GRAMS AI version 8 (Thermo Electron Corp, Waltham, MA,
USA) softwares.

Figure 5.4 InVia Reex dispersive Raman microscope

5.7
5.7.1

Identification of metal caps and glass bottle by EDX

Introduction

Energy Dispersive X-ray (EDX) spectroscopy was used to characterise glass bottles
material and metal crown caps. The analytical technique was also used together with
Raman spectroscopy to verify the identity of PVC bottle cap lining material. EDX, a
variant of X-ray fluorescence spectrometry (XRF), is a non-destructive analysis based
on the spectral analysis of the characteristic X-ray radiation emitted from the sample
atoms upon irradiation by the focussed electron beam of a Scanning Electron
Microscope (SEM). EDX analyzes the top two microns of the sample with spatial
resolution of one micron. The minimum detection limits is about 0.1 weight percent
equivalents to about 1000 ppm (Kuisma-Kursula, 2000). The schematic of SEM showing
the position of the X-ray detector is in Figure 5.5. EDX could be employed for
quantitative analysis if appropriate external standards are available. In the absence of
appropriate external standards EDX can be used in qualitative and semi-quantitative
elemental analysis. In this research EDX was used as qualitative analysis tool.

64

Figure 5.5 Schematic of SEM showing the position of the X-ray detector (adapted from
Skoog et al, 1998)

5.7.2

Principle of EDX

In EDX the incident beam electrons from the SEM excite electrons in a lower energy
states, prompting their ejection and resulting in the formation of electron holes within
the atoms electronic structure. Electrons from an outer, higher-energy shell then fill
the holes, and the excess energy of those electrons is released in the form of X-ray
photons. The release of these X-rays creates spectral lines that are highly specific to
individual elements. In this way the X-ray emission data can be analyzed to characterize
the sample in question (Walther-Meiner-Institute for Low Temperature Research,
2007). The interaction of the electron beam and the atoms of the sample is shown in

65

Figure 5.6. The EDX data is presented as atomic and weight percent of the elements
contained in the specimen. The data is at most semi-quantitative as the approximate
concentrations of the elements in the sample are presented as ratios to each other
rather than directly as percentages.

Figure 5.6 Interaction of electron beam and sample (source: Sanama 2008)

5.7.3

Instruments

SEM-EDX Model Quanta 400 (FEI) with INCAx-sight detector and INCAEnergy EDS
software (Oxford Instruments), EMITECH K 450 high vacuum carbon-coating unit
(Quorum Technologies, East Grinstead, Britain), 12mm extra smooth self-adhesive
carbon discs (Aldermaston, Britain).

5.7.4

Method

Glass, metal and plastic samples were mounted on 12mm self-adhesive carbon discs
attached to metal stubs. In order to minimize charging effects and improve electrical
conductivity before examination all glass and plastic samples were carbon coated using
EMITECH K 450 high vacuum carbon-coating unit. Printed metal caps are usually
chiselled to reveal the metal. The chemical composition of the samples were then
determined using the energy dispersive X-ray of the SEM-EDX unit (Figure 5.7).

66

Figure 5.7 SEM-EDX Model Quanta 400

5.8
5.8.1

Quantitation of antimony in water, soft drinks and PET

Introduction

ICP-MS couples two components namely an inductively coupled plasma ion source
(ICP) and a mass spectrometer (MS). The ICP is radio-frequency generated argon
plasma (partially ionised electrically conductive argon) which can reach the
temperature of up to 10,000 K. A mass spectrometer is an instrument that separates
ions according to their mass-to-charge ratios (m/z). ICP-MS which is the instrument of
choice for the determination of a range of metals and several non-metals in water or
solid samples is so sensitive that it can measure elements at concentrations below one
part per trillion as reported by Shotyk et al (2006). The schematic of ICP-MS is shown
in Figure 5.8. In this study ICP-MS is used to measure antimony concentration in water
and soft drinks and in digested PET samples. As the instrument has the capability to
measure several elements simultaneously, concentrations of cadmium, germanium,
zinc, aluminium, beryllium, titanium, cobalt and lead were also determined. Indium was
used as internal standard.

67

Figure 5.8 Schematic of ICP-MS (Skoog et al, 1998)

5.8.2

Principles of ICP-MS

The sample is introduced into the ICP by flow injection or other means depending on
the state of sample. In the ICP the introduced sample is nebulised, atomised and
ionised after which it is passed into the mass spectrometer. In the mass spectrometer
the electric and magnetic fields in the analyser (in this case a quadrupole shown in
Figure 5.9) deflects the ions depending on their m/z ratios with lighter ions getting
more deflected by the electromagnetic force than heavier ions. As the voltage is varied
ions of different m/z are brought into focus on the detector which builds up a mass
spectrum by recording the relative abundance of each ion type.

Figure 5.9 Quadrupole mass analyser (Skoog et al, 1998)

68

5.8.3

Glassware preparation

To do away with problem of contamination all glassware used in antimony analysis

using ICP-MS were initially soaked in 10% nitric acid for 30days after which they were
washed with detergent and rinsed with deionised water. In subsequent works the
glassware were soaked overnight in 10% nitric acid after which they were washed with
detergent and rinsed with deionised water.

5.8.4
5.8.4.1

Preparation of standards and blanks

Multielement stock standard

To make 250ml of 10g/ml multielement stock standard, 2.5ml of each of the stock
standards of the elements at 1000ppm (including indium used as internal standard) and
0.25ml of antimony stock standard at 10006ppm were added into a 250ml volumetric
flask containing 50ml of reagent water (1% nitric acid for trace analysis) and the
mixture was made up to 250ml using the reagent water.

5.8.4.2

Multielement working standard

To make 100ml of 1g/ml multielement working standard 10ml of the multielement

stock standard was diluted to 100ml using reagent water.

5.8.4.3

Calibration standards, calibration blanks and method blank

Calibration standards ranging from 0 to 5g/L (0, 1, 2, 3, 4 and 5 g/L) of the trace
metals analysed were used in this study. Calibration standards ranging from 0 to
20g/L were initially used. However due to high level of concentration disparity
between samples and highest calibration standard a memory effect was encountered
which resulted in elevated results. The calibration standards 1, 2, 3, 4 and 5 g/L were
prepared by making 0.1, 0.2, 0.3, 0.4 and 0.5mL of the 1 g/ml multielement working
standard to 100ml using reagent water. Reagent water was used as calibration blank.

5.8.4.4

Rinse blank

To reduce possible memory interference between samples runs to minimal 2% nitric

acid was used as serves as a rinse blank for flushing the system between successive
samples.

69

5.8.4.5

Internal standard

Indium at 3g/L was used as internal standard in samples. 1g/ml of indium internal
standard solution was formed by making 0.1ml of the indium stock solution (1g/L) to
100ml.

5.8.4.6

Method blank

To monitor contamination during sample preparation and analysis a method blank was
run at intervals. Reagent water containing the indium internal standard at 3g/L was
used as method blank. Detection limit was determined by multiplying the standard
deviation of 10 method blank results by 3.

5.8.5

Instruments

PlasmaQuad 3 quadrupole ICP-MS instrument (Thermo Fisher Scientific, USA),

Cookworks MM717CFA microwave oven (Argos, Britain), Gallenkamp thermostat
water bath, 60ml PFA digestion vessels and caps with wrench sets (Savillex
Corporation, USA), KX+syringe filters PTFE 25mm, 0.45 m (Kinesis, Britain), 12ml
Norm-ject Luer lock syringes(Sigma Aldrich, USA), 125ml PLASTIBRAND narrowmouth polypropylene bottles, with screw cap (Sigma Aldrich), Powder free latex gloves
(Microflex, Austria), Hand-held pH/mV/temperature/RS232 meter pH 11 series
(Oakton/Eutech Instruments (Nijkerk, The Netherlands), HI-9033 conductivity meter
(Hanna instruments, Leighton Buzzard, Britain), Mettler AE 200 pan electronic balance
(Mettler-Toledo Ltd., Leicester, Britain),

Pocket digital thermometer model 314

(Comark Instruments, Britain).

5.8.6

Conductivity and pH of samples

In order to assess the possibility of a relationship between water conductivity and pH

in one hand and antimony concentration in water on the other hand the conductivity
and pH of samples were measured using the hand-held pH meter and the HI-9033
conductivity meter prior to any analysis.

5.8.7

PET bottle thickness

To assess whether the thickness of PET bottle has any influence on antimony migration
bottles thickness were measured using a digital calliper. For each sample three
measurements were taken and averaged.

70

5.8.8

Antimony and other trace elements in water and soft drinks

Bottled water brands (still and sparkling) and soft drinks in PET bottles from Britain
and Nigeria and tap water were analysed for antimony as in Shotyk and Krachler
(2007). One aliquot of sample was diluted with four aliquots of reagent water to
reduce the concentrations of alkaline and earth alkaline elements, to avoid clogging of
the cones, to add the internal standard, and to reduce the amount of dissolved CO 2 in
the samples. All dilutions were carried out using reagent water. Certified water
reference material and reagent water spiked with the elements of interest at 1g/L
were also analysed to monitor accuracy of analysis. The certified water reference
material contains the elements of interest at concentrations given in Table 5.3.
Table 5.3 Concentration of the elements of interest in certified water reference
material
Element
Concentration (g/L)
Antimony
3.2
Cadmium
4.2
Zinc
379
Aluminium
58.3
Beryllium
13.4
Titanium
15.1
Cobalt
64.2
Lead
7.82

5.8.9

Sample digestion using domestic microwave oven

Ideally a purpose-built laboratory microwave digestion system should have been used
for digestion of plastic samples. However as a result of the unavailability of a
microwave digestion system domestic microwave oven placed in a fume chamber was
used based on the modification of a method explained by Sakurai et al. (2006). With
domestic microwave oven digestion vessels containing samples have to be opened and
degassed at short interval of times to avoid build-up of pressure.

5.8.9.1

Optimization of digestion parameters

To digest plastic materials and the polyethylene reference material the use of nitric
acid alone and a mixture of nitric and hydrochloric acids (5:1) were investigated. Use
of heating powers of 120, 230 and 385W was also assessed. About 150mg of samples
were usually added to digestion vessels containing 3ml of either nitric acid alone or a
mixture of the 2 acids. Addition of 0.5ml HCl to 2.5ml nitric acid resulted in the
71

formation of brown colour possibly due to formation of oxides of nitrogen. Digestion

with nitric acid resulted in formation of brownish gas for some samples. Digestion was
found to be faster with nitric acid alone. Consequently use of HCl was discontinued.
Digestion for 5 minutes at 230W was found to be efficient for PET, PP, EVA/PP and PE.
For the PVC and unidentified plastic cap lining materials digestion was achieved
after about 8 minutes.

5.8.9.2

Digestion of PET samples

Digestion of PET samples resulted in yellowish or colourless liquid with white

precipitate (Figure 5.10). The liquid remained clear even after dilution with deionised
water. The white precipitate persisted with addition of excess nitric acid and additional
heating in the oven. The precipitate was allowed to settle after which the liquid was
decanted into an acid cleaned beaker. The precipitate was washed twice with deionised
water and the water decanted into the beaker. The liquid in the beaker were filtered
into acid cleaned 100ml volumetric flask using PTFE syringe filters attached to 12ml
Norm-ject Luer lock syringes. The contents of the flask were then made to 100ml with
deionised water. The precipitate was dried and analysed using Raman spectroscopy
and EDX as explained in Sections 5.6.4 and 5.7.4.

Figure 5.10 Digested PET materials

5.8.9.3

Digestion of other plastics

Digestion of polyethylene, EVA/PP copolymer polypropylene and unidentified cap lining

material resulted in a clear yellowish solution which became milky with addition

72

deionised water. The milky liquid was filtered into acid cleaned 100ml volumetric flask
using PTFE syringe filters attached to 12ml Norm-ject Luer lock syringes. The contents
of the flask were then made to 100ml with deionised water.

5.8.9.4

Digestion method blank

To monitor contamination during sample digestion, handling and analysis, 3ml of the
nitric acid for trace analysis was digested. After digestion a slightly yellowish liquid
containing no precipitate was obtained. This liquid remained clear on addition of
deionised water. The liquid was then treated exactly as for samples.

5.8.10 Antimony in digested samples

The filtrate from digested samples, digestion reference and digestion blank were
analysed in the same way as water samples using indium at 3g/L as internal standard.

5.8.11 Bottled water temperature elevation on exposure to sunlight

As earlier stated in the literature a temperature of up to 58.3 C was reported to be
achievable on exposure of water in glass bottles for 2.5 hours from midday in Nigeria
(Tukur et al, 2006). In this study experiments were carried out to find water
temperatures achievable after exposure of water in PET and glass bottles of different
colour and size to the sun on a British summer day. In the experiment PET and glass
bottles filled with water at initial temperature of 19.1C were exposed to the sun on a
clear summer day with brilliant sunlight. Change in water temperature was monitored
on hourly basis from 11am to 6pm using a handheld digital thermometer. Bottles used
in the study are given in Table 5.4.
Table 5.4 Bottles used in sunlight exposure experiments
Bottle material Bottle size (ml)
PET
500
PET
500
PET
500
PET
500
PET
250
PET
750
Glass
1000
Glass
750
Glass
750

73

Bottle colour
bluish tint
colourless
green
blue
colourless
bluish tint
colourless
green
crimson

5.8.12 Bottled water storage and antimony migration

In the literature storage has been reported as one of the factors that elevates the
concentration of migrants from PET and other plastic into water. In this study a survey
was conducted for the purpose of establishing the average and maximum storage
periods prior to use for bottled water and soft drinks. The concentration of antimony
in bottled water and soft drinks samples stored for periods chosen based on the
survey results were measured.

5.8.13 Antimony migration at different temperatures and times

To monitor the behaviour of PET and glass bottles of different colours from different
brands of water and soft drinks in terms of migration of antimony at different
temperatures and exposure times, 300ml of deionised water were added to ten 500ml
PET bottles of different colour and from different brands. The bottles were then
heated at 40, 60, and 80C using a thermostat water bath for 6, 24 and 48 hours.
Additional 8 PET bottle brands were heated at 60C for 6 hours. Green and colourless
750ml glass bottles containing 500ml of deionised water were also heated at 60C for
6, 24 and 48 hours. Temperatures of 40, 60 and 80C were chosen based on the
results of the experiments in subsection 5.8.11 and the information in the literature.
Tukur et al (2006) reported a water temperature of 58.3 C in Nigeria on exposure of
water in glass bottles for 2.5 hours from midday.

5.8.14 Bottle aging and antimony migration

From the literature PET bottles were found to be reused for storing drinking water by
88% of respondents in the US for as long as 6 months. Additionally, a survey carried
out in this study attempted to establish the proportion of respondents reusing PET
bottles in Nigeria and Britain and the average and maximum periods of bottle reuse.
To assess the effect of PET bottle aging on migration of antimony six 500ml clear
bluish PET bottles were aged for a period chosen based on the result of the survey.
For simple aging experiment three bottles were filled with water stored at room
temperature, emptied on weekly basis, rinsed with water and refilled. For aging
experiments involving washing with detergent and hot water the same procedure was
followed except that bottles were scrubbed using brush with dilute solution of
detergent in hot water. After the aging period an aged bottled, detergent/hot water
aged bottle together with fresh bottle were filled with deionised water and sparkling

74

water. The filled bottles were stored at room temperature for 283 days after which
the contents were analysed for antimony. To monitor the antimony migration
tendency of the aged bottles at elevated temperatures 300ml deionised water were
added to an aged bottled, detergent/hot water aged bottle and a fresh bottle. The
bottles were then heated at 60C for 6 hours using a thermostat water bath.

5.8.15 pH and antimony migration

To observe the effect of pH on antimony migration pH values of deionised water were
adjusted to 2, 5 and 8 using sodium hydroxide and hydrochloric acid. Three 500ml
green PET bottles were then filled with the deionised water (range chosen to reflect
the pH range 2.43 8.01 obtainable from bottled water and soft drinks). The pH
values of the water were adjusted using sodium hydroxide and hydrochloric acid. The
water contents were analysed after 266 days. For comparison with the original
antimony concentration each of the pH-adjusted water samples was stored in acidcleaned polypropylene bottles.

5.8.16 Bottle size and antimony migration

Bottled water and soft drinks are bottled in bottles of different sizes. The survey
carried out in this study attempted to establish the extent to which bottles of different
sizes are used. Westerhoff et al (2008) suggested the possibility of an inverse
relationship between bottle size and achievable antimony concentration in water due
to migration. This suggested phenomenon was related to the relationship between
available contact area of PET bottle and the volume of the liquid in the bottle. Also
Keresztes et al (2009) reported observing higher antimony concentration in water
bottled in smaller packages. To assess this at elevated temperatures 5 clear colourless
bottles of different sizes from a brand of bottled water were used. The sizes of the
bottles used were 2000, 1500, 750, 500 and 330ml. The bottles were filled to half their
nominal volumes with deionised water and then heated at 70C for 9 hours after
which the contents were analysed for antimony.

5.9
5.9.1

Quantitation of acetaldehyde in water, soft drinks and PET

Introduction

Headspace GC-MS was initially employed in this study for the quantitation of
acetaldehyde in aqueous samples and in PET material. However GC-FID was later
adopted because of the greater sensitivity of flame ionization detector (FID) to volatile
75

organic compounds in comparison to GC-MS and its insensitivity to nitrogen and other
non-combustible gases. Additionally FID has a large dynamic range and low noise.
According to Skoog et al (1998) flame ionization detector (FID) is the most useful
general detector for the analysis of most organic samples. In headspace GC the
headspace vial allows for concentration of organics even from dilute solutions.

5.9.2

Principle of headspace GC-FID

In headspace gas chromatography polar organic volatiles dissolved in aqueous medium

are concentrated into the headspace of vials by addition of inorganic salts and by
heating. The salting out effect, together with elevated temperatures, lowers the
partition coefficients of organic volatiles in the sample matrix and promotes their
transfer into the headspace. The concentrated volatile analytes from the headspace are
injected into the GC column where they partition between a solid or liquid stationary
phase and a gaseous mobile phase in the column. The differential partitioning into the
stationary phase allows the compounds in the sample to be separated in space and
time. The schematic of Gas chromatography is shown in Figure 5.11

Figure 5.11 Schematic of Gas chromatography (adapted from Sheffield Hallam

University, no date)
In FID the partitioned compounds from the column are mixed with air and then ignited
by a hydrogen flame. A large electrical potential typically 100 300V is applied at the
burner tip, and a collector electrode is placed above the flame. The increased current

76

due to electrons emitted by burning carbon particles is then measured. The schematic
of FID is shown in Figure 5.12.

Figure 5.12 Schematic of Flame Ionisation Detector (adapted from Sheffield Hallam
University, no date)

5.9.3

Instruments

GC-FID HP 6890 (Hewlett Packard, USA), Headspace screw top 20ml clear glass vials
and ultra-clean 18mm screw caps with septa (Agilent Technologies, Germany), 2.5ml
gastight syringe (SGE, Australia).

5.9.4
5.9.4.1

Preparation of standards and blanks

Oxygen-free deionised water

Acetaldehyde solutions and all dilutions were made using boiled oxygen-free deionised
water. Deionised water were boiled to expel air and to kill microorganisms that may
contribute to degradation of acetaldehyde in solution. Nitrogen flushing was carried
out to expel oxygen which also contributes in degradation of acetaldehyde. To prepare
oxygen-free deionised water deionised water was boiled using electric kettle,
transferred into clean one litre capped glass bottles, allowed to cool, and then flushed
with nitrogen.

77

5.9.4.2

Stock and calibration standards

A 10 mg/ml acetaldehyde stock standard was made by dissolving 5grams of

acetaldehyde in 500ml of cold oxygen-free deionised water. The stock standard was
stored in a refrigerator to keep it cool to minimise evaporation and degradation of
acetaldehyde. Calibration standards of different concentrations were formed using the
cold oxygen-free deionised water depending on the predicted acetaldehyde
concentration in samples.

5.9.4.3

GC-FID conditions

Column used was a Zebron ZB-1 30m x0.32mm id x 0.25m film thickness 100%
methyl polysiloxane. The injector and detector temperatures were maintained at 150
and 200C respectively. Oven temperature was programmed at 40C for 1 minute
increasing by 10C to 70C. While the retention time of acetaldehyde differed with the
split/splitless mode, it generally eluted in less than 1 minute. Nitrogen was used as
carrier gas.

5.9.5

Acetaldehyde in water and soft drinks

Acetaldehyde determination in bottled water and soft drinks was made based on a
variation of the method described by zlem (2006). In this method the relative volume
of the headspace compared to volume of the sample in the sample vial (phase ratio)
was one. For the salting out effect a salt concentration of 39.1g/100ml was targeted
which is the maximum solubility of NaCl at 100C. Bottled water or soft drink sample
of 10 ml was pipetted in to a clean nitrogen-flushed headspace 20ml vial containing 4 g
of sodium chloride. The vials were then closed with the ultra-clean 18mm screw caps
with septa. The vials were heated in a hotplate for 25minutes at 95C. To further
ensure efficient partitioning vials were agitated slightly at 5minutes interval. Headspace
samples of 2.5ml were withdrawn and injected into the GC instrument manually using
a gastight headspace syringe heated at 95C. While bottled water samples were
analysed at a splitless mode, all soft drinks samples were analysed using 1:50 split ratio.
Calibrations for quantitation of acetaldehyde in soft drinks were from 0 to 10mg/L.
Calibrations for determination of acetaldehyde in bottled water were 0 to 500g/L.
Detection limit at 3 x the standard deviation of 10 samples was determined by analysis
of 10 samples containing acetaldehyde at concentration of 10g/L. To determine

78

recovery acetaldehyde spikes of 50 and 100g/L were analysed 10 and 4 times

respectively.

5.9.6

Residual acetaldehyde in PET materials

Acetaldehyde determination in PET material was made based on a variation of the

industry standard French National Acetaldehyde generation test as explained in Howell
and Ponasik (2006). The acetaldehyde desorption conditions of 150C for 60 minutes
are chosen because at this conditions no measurable acetaldehyde is regenerated by
the sample during the desorption process. In the bottle making, heating of the
hygroscopic PET granulate at 160C for 4 hours reduces its moisture content to less
than 50ppm without eliciting acetaldehyde-generating hydrolytic reactions.
Approximately 1 gram of the PET bottle material cut into very small pieces was placed
into 20ml clear glass headspace vial. The vial was immediately flushed with nitrogen and
capped with the ultra-clean 18mm screw cap with septa. Sample vials were heated to
150C for sixty minutes to desorb the acetaldehyde from samples into the vial. Vial
samples of 2.5ml were withdrawn using a heated gastight syringe and injected into the
GC-FID operating at a split ratio of 1:50. Calibration was achieved by use of clean;
nitrogen-flushed 520.5ml glass bottles sealed using several layers of PVC cling film. To
make calibration standards of between 96.06 to 960.65g/L, 5, 10 , 20, 30, 40 and 50l
of 10mg/ml solution of acetaldehyde were added to first, second, third fourth, fifth and
sixth bottles. The amounts of acetaldehyde added to the bottles being 50, 100, 200,
300, 400 and 500g respectively. A nitrogen-flushed sample containing no acetaldehyde
was used as calibration blank. The bottles were warmed on a hotplate and shaken
vigorously to disperse the solution and vaporise the acetaldehyde. A sample of 2.5ml of
the content of each of the bottles were analysed by GC-FID at a split ratio of 1:50.
Detection limit at 3 times the standard deviation of 10 samples was determined by
analysis of 10 samples in 1.157L bottles containing acetaldehyde vapour at 8.64g/L.
The concentration of 8.64g/L was achieved by addition of 10l of 1mg/ml
acetaldehyde solution into the bottles. To determine recovery at a concentration
within the range found in bottle material analysis, acetaldehyde vapour spike at
385.80g/L was analysed 5 times. The concentration was achieved by addition of 20l
of 10g/l (10mg/ml) solution into bottles having capacity of 0.5184L.

79

5.9.7

Effect of carbonation on quantitation

Sparkling water and carbonated drinks differ from still water in having dissolved carbon
dioxide. To determine the effect of carbon dioxide dissolve in sparkling water on
acetaldehyde determination by headspace still and sparkling water samples of similar
brand were spiked with acetaldehyde solution at about 500g/l. The samples were
then analysed in triplicate for acetaldehyde.

5.9.8

Storage and acetaldehyde in PET materials and bottle contents

As mentioned in Subsection 4.7.4 storage elevates the concentration of migrants from

PET bottles into their contents. To determine the effect of storage on acetaldehyde
content of PET bottle material and bottled water/soft drinks, bottle materials and
contents of freshly purchased bottles were analysed together with bottle materials and
contents of bottles stored for extended period.

5.9.9

Bottle thickness and acetaldehyde concentration in PET

Bottle thickness may affect acetaldehyde concentration in bottles because thicker

bottle walls theoretically trap more acetaldehyde than thinner ones. To determine
whether bottle thickness has any influence on bottle material residual acetaldehyde
content bottle thickness earlier measured with a digital calliper were correlated with
the acetaldehyde content of bottle materials.

5.9.10 Bottle aging and acetaldehyde concentration in PET

To assess the effect of PET bottle aging on acetaldehyde concentration in the bottle
material clear green and clear bluish 500ml PET bottles were aged for periods chosen
based on the results of the survey. For simple aging experiment three bottles were
filled with water stored at room temperature, emptied on weekly basis, rinsed with
water and refilled. For aging experiments involving washing with detergent and hot
water the same procedure was followed except that bottles were scrubbed using
brush with dilute solution of detergent in hot water. After aging the acetaldehyde
content of the PET materials were determined as in Subsection 5.9.6.

5.9.11 Acetaldehyde outgassing from PET bottles

Fruity smell resembling that of acetaldehyde at low concentration was perceived when
some empty bottled water bottles left to stand for several months were sniffed.
Analysis of the gaseous content of the bottles revealed acetaldehyde. To observe the

80

behaviour of different bottles allowed to stand for different period of times, bottles
from 3 different brands were used. Two of the brands used were clear bluish still
water bottles the other brand was green sparkling water bottle. Two green bottles
aged for 266 days were also assessed.

5.9.12 Acetaldehyde migration from water medium into PET

To assess whether acetaldehyde in aqueous solution can migrate into PET materials,
three different experiments were carried out using clear bluish and clear green PET
bottle materials. Three portions of each of the PET materials weighing about half gram
were soaked into 0.1 and 10mg/ml acetaldehyde solutions and into neat acetaldehyde
for 24 hours. After 24 hours the PET materials were thoroughly washed with tap
water and dried by blotting with paper tissue. The acetaldehyde content of the PET
materials was then determined as in Subsection 5.9.6.

5.9.13 Stability of acetaldehyde solutions with storage

To assess the stability of acetaldehyde with storage, three acetaldehyde solutions in
propylene bottles with concentrations of 50, 500 and 5000g/L were stored at room
temperature and inside a refrigerator at temperatures slightly above zero. The
acetaldehyde concentrations were measured 5 times over 30 days.

5.10 Research timeline

This research work spanned a period of about 3 year and 9 months. The timeline for
the activities carried out in the research is given in Table 5.5. The timeline designed at
the beginning of this research spanned a period of three years. However the timeline
could not be adhered to in large part due to unforeseen delay in accessing some
instruments used in this research.

81

Table 5.5 Research timeline

Period
Activity

2008
J

F M A M J

2009
J

A S O N D J

2010

F M A M J

Start date
Literature review/
Research plan
Survey

Survey development
Survey in Britain
Survey in Nigeria
Survey report
MPhil to PhD transfer
Change of supervision
Laboratory work

Sample collection Nigeria

SEM-EDX
ICP-MS
GC-MS/FID
Raman
PWG conference
SLS poster presentation
Writing up

First draft
82

A S O N D J

F M A M J

2011
J

A S O N D J

F M A M J

CHAPTER 6:
6.1

SURVEY RESULTS AND DISCUSSION

Introduction

Chapter 5 explained that a survey had been developed with the aim of obtaining an
understanding of typical use and re-use patterns for plastic bottles in the UK and
Nigeria. This Chapter presents and discusses the results obtained as the result of the
analysis of the data obtained from that survey. The Chapter also discusses the
implication of the survey results on the laboratory work.

6.2

Reported use of bottled water and soft drinks

Almost all respondents reported using water and/or soft drinks bottled in PET bottles
(Nigeria 96%, Britain 98%), additionally 95% of Nigerian respondents reported using
water in plastic pouches. It was initially thought that using water in pouches would
reduce the likelihood of using bottled water. However use of water in pouches and
use of bottled water have been shown to be independent of each other (2 (1, n =
514) = 2.054; p = 0.152). Consequently use of water in pouches, which is about ten
times cheaper than bottled water, does not reduce the usage status of bottled water.
One possible explanation for this is that even though water in pouches is cheaper than
bottled water it is less portable than bottled water because the packaging is flaccid and
has no cap. Although usage status of water in pouches does not reduce the usage
status of bottled water, it may reduce the quantity of bottled water used especially
among the low-income group. As mentioned earlier (Rothschild and Nzeka, 2005)
water in pouches accounts for 68 percent of total packaged water consumed in
Nigeria.

6.3

Storage and use of bottled water and soft drinks

As mentioned in Section 3.3, prosperity and climatic conditions of a country have an

effect on the degree of usage of bottled water and soft drinks in the country, and
consequently the degree to which the bottled water and soft drinks would be found in
places of residence. In this work the degree to which unopened and opened bottled
water and soft drinks in PET bottles were found in places of residence of British and
Nigerian respondents differed significantly [2 (1, n = 918) = 104.538; p < .001, 2 =
0.11]. The results (Figure 6.1) implied that respondents country of residence to
moderate extent predicts the availability bottled water and soft drinks in places of

83

residence. Bottled water and soft drinks are accordingly more commonly found in the
homes of British respondents than those of Nigerian respondents. Greater
consumption of bottled water/drinks in Britain results from the fact that Britain is
economically more prosperous than Nigeria. According to Finewaters (2009) in
France, Germany and Italy close to 90% of the population patronises bottled water in
comparison to about 50% in Britain. In this work the availability of bottled water in
places of residence of British respondents is 58%. As mentioned in Subsection 5.2.3
54% of British respondents were native Britons with the remaining 46% been visitors.
However the two groups were found to be similar in terms of possession of unopened
and opened bottles at the time of the survey (unopened bottles native Britons-74%,
visitors-72%; opened bottles native Britons-70%, visitors-72%; all types of bottles

Bottled contents availbility (%)

native Britons-86%, visitors-87%).

100
90
80
70
60
50

Britain

40

Nigeria

30
20
10
0

Unopened

Opened

All types

Bottled content
Figure 6.1 Bottled water and soft drinks availability in British and Nigerian places of
residence
Many factors including country economic status, climatic conditions, safety and health,
environmental awareness, taste, idolization of bottled water as fashion accessory , etc
were hypothesised to have some influence on consumption of bottled water and soft
drinks. The identification of bottled water and soft drinks in more households in
Britain than in Nigeria in this study is consistent with literature about developing and
developed countries. Economic status is clearly a factor playing a significant role in the
observed pattern. However manner and extent to which all other factors may be
influencing the observed pattern is not clear. For example the harsher climatic

84

conditions in Nigeria will be expected to translate into higher consumption, even

though this was not observed.

6.4

Bottle size

Bottle size has been mentioned as one of the factors that may be influencing migration
of chemicals from bottle wall into bottle content (Subsection 4.7.9). Smaller bottles
were reported to release more antimony than bigger ones (Keresztes et al (2009)). In
places of residence of British respondents 2L bottles were found to be the most
commonly available bottles followed by 0.5L bottles and then 1L (Figure 6.2). In places
of residence of Nigerian respondents 0.5L bottles were the most common bottles
followed by 1L bottles and then 0.33L (Figure 6.2). It is worth mentioning that the
bottle size data for Nigerian respondents showed some degree of discrepancy in the
sense that bottle sizes (330ml) that are almost unobtainable in Nigeria were reported
in the questionnaire. This finding raises a question about the accuracy of all the bottle
size data for Nigeria. The misreporting of bottle sizes in Nigeria may have happened
either because Britsh respondents are more conversant with bottle sizes than Nigerian
respondents or because in Nigeria unlike in Britain centiliter (cl) is more commonly
used on bottle labels than milliliter (ml). All the questionnaires used in the survey used
milliliter and liter as units of bottle liquid volumes. Further evidence to support the
supposition that Nigerian bottle size data is inaccurate is that unlike British
respondents, the Nigerian respondents did not specify the size of about 16% of the
bottles they reported.

Bottle availability (%)

40
35
30
25
20

Britain

15

Nigeria

10
5
0

Bottle size
Figure 6.2 Sizes of bottles in places of residences
85

6.5

Bottle content types

Sparkling water is not marketed in Nigeria. Consequently still water accounted for 82%
of bottled liquids stored and used in places of residences in Nigeria, with the remaining
18% accounted for by soft drink. In Britain where sparkling water is consumed it
accounted for 5% of bottled liquids reported in places of residence. Bottled drinks and
still water accounted for 50 and 45% respectively. Most bottled water from Nigeria is
sterilized table water in contrast to Britain where most bottled water is either spring
water or natural mineral.

6.6

Storage durations for purchased bottled water and soft drinks

Storage duration for purchased bottled water and soft drinks is an important
parameter in terms of chemical migration because as mentioned in Subsection 4.7.4,
duration of contact between bottle material and bottle content is one of the factors
that influence concentration of migrants in bottle content. In this work the durations
of storage of unopened PET bottled water and soft drinks at the time of the survey
differed significantly between British and Nigerian respondents (Mann-Whitney U =
11115.5, n1 =235, n2 =131, p < 0.01). The median period of storage in Britain and
Nigeria were 7 and 10 days respectively. Median is given as a measure of central
tendency because the data are positively skewed (many low values and few high
values). While 79% of British respondents stored for between 1 and 7 days only 50%
of Nigerian respondents stored for the same period (Figure 6.3). Nevertheless, for
both countries the storage period ranged between 1 day and 1 year. For British
respondents the durations are similar between native Britons and visitors. The median
period was 7 days for both and the proportion of respondents storing between 1 and
7 days are 84% for native Britons and 76% for visitors. In this case the storage
behaviour of the British visitors is more like that of native British respondents than
that of Nigerian respondents.

86

80

Respondents (%)

70
60
50
40
Britain

30

Nigeria

20
10
0
1-7

8 -14

15-30

31-60

61-90

>90

Storage duration (days)

Figure 6.3 Durations of storage for unopened bottles in places of residence

Respondents were also asked a question on the longest time both unopened and open
bottles ever remained in their possession as a worst-case scenario. In the case of
unopened bottles the storage durations were found to be similar for both British and
Nigerian respondents (Mann-Whitney U = 12816, n1 = 231, n2 =121, p = 0.197). The
median period in both Britain and Nigeria was 14 days. The longest storage period for
about 41% of all respondents was between 1 and 7 days (Figure 6.4). For respondents
from both countries the longest unopened bottles storage periods ranged between 1
day and more than a year. It is not known whether unopened bottle contents are still
consumed after long storage. For British respondents the median was 14 days for both
native Britons and visitors and the durations were similar (Mann-Whitney U = 4642, n1
=126, n2 =87, p = 0.055). The proportion of respondents that stored for between 1
and 7 days were 42% for native Britons and 47% for visitors. The longest periods
contents of opened bottles last before been used up are similar for British and
Nigerian respondents (Figure 6.5). Respondents from both countries reported having
opened bottles that lasted for periods ranging between 1 day and 1 year. Even so for
both countries more than three-quarter of respondents reported consuming bottle
contents within 7 days (Figure 6.5).

87

45

Respondents (%)

40
35
30
25
20

Britain

15

Nigeria

10
5
0
1-7

8 -14

15-30

31-60

61-90 91-180 >180

Storage duration (days)

Figure 6.4 Longest reported storage durations for unopened bottles in places of
residence
90

Respondents (%)

80
70
60
50
40

Britain

30

Nigeria

20
10
0
1-7

8 -14

15-30

>30

Storage duration (days)

Figure 6.5 Longest reported storage durations for opened bottles in places of
residence of residence
A time interval ranging from few hours to several months exists between purchase of
bottled water and soft drinks and their consumption. Additionally best before dates
which, as earlier mentioned, are unrelated to chemical migration, exists for all bottled
contents. Bottled water and carbonated drinks in PET bottles from Nigeria have a shelf
life of one year and six months respectively based on the production and 'best before'
dates stamped on the bottles. In the US and Canada bottled water's stamped shelf life
is usually two years (Environmental Health & Safety Online, 2006; Health Canada,

88

2009). From the results in this work between 1 and 6% of British and Nigerian
households were found storing bottled water and/or soft drinks for a period greater
than 3 months prior to use. Similarly small percentages (between1 and 8%) of British
and Nigerian households reported ever storing bottled water and soft drinks for a
period greater than 6 months prior to use. Consequently even if the best before
dates have any relevance to risk of consumption of liquids with leached bottle
contents, the proportion of households to be affected by this will be very low. The risk
associated with storage of bottled water and soft drinks will be discussed later in
relation to actual antimony and acetaldehyde migration patterns.

6.7

Storage places

As discussed in Subsections 4.7.1 and 4.7.2, both heat and light can speed up the rate
at which migration of chemicals from the plastic materials into the contents ensues. In
this study no statistically significant difference was found between British and Nigerian
respondents in terms of places of storage of unopened bottles [2 (1, n = 453) = 2.639;
p = 0.104]. Unopened bottles are stored inside and outside the refrigerator at
approximately equal frequencies for both countries. Storage outside the refrigerator
was reported by 56 and 48% of British and Nigerian respondents respectively. While it
is very cool and mostly dark inside refrigerators, conditions outside the refrigerator
could range from dark and cool, dark and warm, to bright and cool and bright and
warm depending on the section of place of residence, season, and heating and lighting
in a residence. In Leeds the most populous city in West Yorkshire in the UK the mean
ambient temperature ranges between 0.2C (February) and 19.9C (July/August). In
Abuja the capital city of Nigeria the range is between 15.5C (December) and 36.9C
(March). In Niamey the capital city of neighbouring Niger Republic the mean maximum
temperature is as high as 40.9C (April) (World Weather Information Service, no
date). It is not unusual for ambient temperatures in northern Nigerian cities bordering
Niger republic to reach 40C in hot season.
Reported storage places other than refrigerator include cupboard, carton, kitchen,
living room, locker, storeroom, wardrobe, window sill, pantry, garage, bedroom,
basement and attic. Storage places for opened bottles depend on season. During the
coldest seasons of the year more Nigerians stores opened bottles outside refrigerator
than British respondents [2 (1, n = 363) = 3.652; p = 0.056 marginal significance, 2 =

89

0.01]. During these seasons, 55 and 66% of British and Nigerian respondents stores
outside refrigerator. During the hottest seasons of the year opened bottles are stored
inside and outside refrigerator at approximately equal frequencies for both countries
[2 (1, n = 360) = 2.599; p = .110]. Storage outside refrigerator was 29 and 21% for
British and Nigerian respondents respectively.

6.8

PET bottle reuse

PET bottles are reused in places of residence, on the move and at work. For both
British and Nigerian respondents the extent of reuse in places of residence and the
overall reuse were high and were not significantly different (Table 6.1). Proportions of
British and Nigerian respondents reusing PET bottles are 80% and 83% respectively.
No statistically significant difference could be observed between native British
respondents and visitors for all reuse situations. An important observation made was
that the reuse information revealed by the first 144 questionnaires collected before
the PhD upgrade report was similar to the information revealed by the 320 samples
collected after the report. The proportion of respondents reusing PET bottles was
similar for all reuse situations. What this implies is that 144 questionnaires are as
effective as the 464 questionnaires in terms of revealing the information on PET bottle
reuse. Interestingly, the extent of reuse revealed by the study for the 2 countries is
similar to what was observed by Lilya (2001) in a preliminary survey of the university
of Idaho community in the US. In that study the author found that 88% of the
participants reused polyethylene terephthalate (PET) plastic bottles used for bottled
water.
Nigeria is a low-income tropical country, on the other hand Britain is a high-income
temperate country and Idaho is a temperate region in a high-income country. Higher
reuse was initially thought to be more associated with low-income tropical countries
than high-income temperate countries in large part due to higher need to drink fluids
and presumed lesser need to reuse bottles as a result of greater prosperity. However
it is also appreciated that greater availability of empty bottles in high-income countries
as a result of higher use may elevate reuse. From the figures obtained PET bottle reuse
in places of residence and overall PET bottle reuse are independent of country, and by
implication, independent of economic status and climate.

90

Table 6.1 Comparison between British and Nigerian respondents in terms of PET
bottle reuse status
Reuse
location
Places of
residence
On the
move
At work
Overall
reuse

Country % reusing
PET bottles
Britain
67
Nigeria
73
Britain
68
Nigeria
50
Britain
53
Nigeria
39
Britain
80
Nigeria
83

Chi square test of homogeneity

2 (1, n = 813) = 3.536; p = 0.06, no statistically
significant difference
2 (1, n = 926) = 32.144; p < 0.01, 2 = 0.04,
statistically significant difference
2 (1, n = 902) = 18.029; p < .01, 2 = 0.02,
statistically significant difference
2 (1, n = 961) = 1.620; p =0 .203, no statistically
significant difference

Conversely statistically significant difference was found between the 2 countries in

terms of reuse on the move and at work (Table 6.1). For both reuse on the move and
at work, more British respondents reuse PET bottle than Nigerian respondents. This
observation refuted the assumption that low-income tropical countries reuse PET
bottles more than high-income temperate countries. However for both reuse
situations the strength of the associations (2) lies between small and medium implying
that country of respondent only weakly predicts the degree of reuse. In these
situations the observed association between degree of reuse and country can be
explained more by lifestyle than by economic status and climate.
While bottle reuse is an environment-friendly activity there is concern that it may not
be safe as a result of chemical migration and bacterial contamination. In fact in some
developed countries health authorities sometimes discourage the reuse of the singleuse PET bottles due to risk of bacterial contamination. Health Canada, the department
of the government of Canada with responsibility for national public health, does not
recommend the reuse for this reason (Health Canada, 2009). As mentioned in Chapter
3, information is scanty in the literature on bottle reuse pattern in developed and
developing countries. However this study provides a useful insight into the pattern of
reuse in both developing and developed countries. As mentioned in Section 3.9, some
of the factors thought to be influencing bottle reuse include economic status and
climate of a country, safety debate, cost, bottle availability, age of bottles,
environmental concern, original content of bottles and convenience. The results
obtained in this study clearly indicate that economic status and climate of a country
may not be factors that influence bottle reuse. This is because the extent of reuse for

91

both Nigeria and Britain were found to be similar at 83 and 80% respectively. While
these results are similar for the 2 countries, it is possible that the factors that brought
about this behaviour synergistically impacted differently to give the similar results. This
will be discussed further in relation to the actual factors found from the survey to be
influencing reuse. The results in this study together with the earlier results from the
US (88% reuse reported by Lilya, 2001) suggest that PET bottle reuse is similar in both
developing and developed countries.

6.9

PET bottle reuse durations

Tables 6.2 and 6.3 together with Figures 6.6 and 6.7 shows the reuse duration
attributes for Nigerian and British respondents for different reuse situations. Even
though both mean and median are given as measures of central tendency, the median is
much more useful in revealing the centres of the distributions than the mean because
the distributions are positively skewed (many low values and few high values). The
Mann-Whitney U test was used to ascertain whether the British and Nigerian samples
are drawn from similar populations in terms of reuse period. The test was also used to
see whether reuse durations of the native Britons and those of visitors are similar. The
results in Table 6.2 showed that the durations of reuse in places of residence and at
work differ significantly for British and Nigerian samples. On the other hand the reuse
durations on the move are similar. For British natives and visitors reuse durations
were similar in all situations save for reuse at work. The average reuse duration at
work was 37 and 16 days for native Britons and visitors respectively. However the
median duration was 7 days for both respondents. In places of residence and at work,
Nigerian respondents reuse PET bottles longer than British respondents. Nigerian
respondents are also at the forefront in terms of longest reported bottle reuse periods
for all reuse locations. For Nigerian respondents the longest reported bottle reuse
durations are between 4 and 6 years for all reuse situations. For British respondents
these periods are between 1 and 2 years. Lilya (2001) in the University of Idaho
reported 6 months (approximately 180 days) as longest reuse period.

92

Table 6.2 Comparison between British and Nigerian respondents in terms of PET
bottle reuse durations
Reuse
location
Place of
residence
On the
move
At work

Country Reuse duration attributes

Mean Median Range
(days) (days)
(days)
Britain
39
14
1 360
Nigeria

150

35

Britain

27

Nigeria

72

Britain

38

Nigeria

98

28

Mann-Whitney U test

Mann-Whitney U = 12797.5, n1
=179, n2 =229, p < 0.01 (two1 2160 tailed), distributions in the two
groups differed significantly
1 720 Mann-Whitney U = 7894.5, n1 =
183, n2 =87, p = 0.912 (two1 1440 tailed), distributions in the two
groups similar
1 720 Mann-Whitney U = 5024, n1 =141,
n2 =87, p < 0.05 (two-tailed)
1 1440 distributions in the two groups
differed significantly

70

Respondents (%)

60
50
Places of residence
On the move
At work

40
30
20
10
0
1-7

8 -14

15-30

31-60

61-90

91-180 181-360 >360

Reuse period (days)

Figure 6.6 PET bottles reuse durations for British respondents

93

60

Respondents (%)

50
40
30

Places of residence
On the move

20

At work

10
0
1-7

8 -14

15-30

31-60

61-90 91-180 181-360 >360

Reuse period (days)

Figure 6.7 PET bottles reuse durations for Nigerian respondents

Table 6.3 Reuse duration cumulative frequency table
Country

Britain

Nigeria

Reuse location

Proportion of respondents found reusing within

1-7

1-14

1-30

1-60

1-90

1-180 1-360

days

days

days

days

days

days

days

Places of residence

41%

57%

74%

86%

91%

98%

100%

On the move

61%

70%

84%

89%

93%

97%

98%

At work

60%

69%

84%

91%

95%

98%

99%

Places of residence

17%

26%

48%

63%

73%

86%

94%

On the move

55%

67%

79%

88%

88%

94%

97%

At work

41%

47%

62%

77%

82%

90%

97%

The longer bottle reuse periods in Nigeria are probably partly attributable to lower
availability of used bottles as a result of lower use of bottled water and soft drinks in
comparison to Britain. Lesser availability of bottles translates into longer bottle reuse
duration. Availability in this context may refer to availability of funds to purchase used
94

bottles in addition to physical availability of the bottles. In developing countries like

Nigeria used bottles are in many instances not available for free. These bottles are
generally available from collectors who collect them for the purpose of selling them to
recyclers and to people wanting to reuse them. The longest reuse durations reported
in this work (6 years) may look unrealistic. However, a look at the data will reveal that
these long durations were reported by only very small proportions of respondents.
Additionally a bottle has been collected from an office in the University of Bradford
that had been reused for a period longer than 1 year. Because of their clarity and
rigidity it is possible to use PET bottles for an extended period of time without
noticing a change that may elicit the need for replacement.

6.10 Number of bottles being reused in places of residence

Figure 6.8 gives details on the number of bottles that were being reused in places of
residence at the time of the survey. More bottles were being reused in places of
residence of Nigerian respondents than in places of residence of British respondents
(Mann-Whitney U = 12957.5, n1=218, n2=256, p < 0.01). For instance, 38% of British
respondents were reusing one bottle in comparison to only 14% of Nigerian
respondents. The median number of bottles being reused was 2 and 6 for British and
Nigerian places of residence respectively. Nigerian respondents also recorded the
highest number of PET bottles being reused in places of residence. The number of
bottles being reuse by native Britons and visitors differed. While only 25% of visitors
were found reusing one bottle, the proportion of native Britons reusing one bottle was
45%. However the median number of bottles being reused by both groups is two. It
follows that native Britons reuses less bottles in their places of residents than visitors
in Britain or Nigerians. One of the factors believed to influence the number of bottles
being reused at any given time is the state of potable pipe-borne water supply in places
of residence. Supply of potable water is continuous in places of residence of British
respondents in contrast to the places of residence of Nigerian respondent where the
supply is intermittent. The intermittent supply logically calls for storage of potable
water in larger quantities for use when supply ceases.

95

40
35

Respondents (%)

30
25
20
Britain

15

Nigeria

10
5
0
1

6-10

>10

Bottles being reused

Figure 6.8 Number of bottles being reused in places of residence

6.11 Size of reused bottles

The most commonly reused bottles in places of residence of British respondents were
500ml bottles followed by 250/330ml bottles and then 1L and 2L bottle (Figure 6.9). It
is not known whether reused bottles will behave in the same way as new bottles
considering the effect of age on them. From Figure 6 the commonly reused bottle in
places of residence of Nigerian respondents is 1L bottle. As for unopened and opened
bottles (section 6.4) the accuracy of bottle size data for all Nigerian reused bottles is in
question as the 250/330ml bottles reported by respondents are unavailable in Nigerian
bottled water and soft drink market. Again some evidence to this observation comes
in the form of the proportion of users whom reported not knowing the sizes of the
bottles they reuse. For reuse in places of residence, which is high for both British and
Nigerian respondents, as much 17% of Nigerian respondents reported not knowing
the sizes of the bottles they were reusing while only 1% of British respondents
reported not knowing the sizes of the bottles they were reusing in places of residence.

96

50

Bottle availability (%)

40

30
Britain

20

Nigeria
10

0
0.33L

0.5L

0.75L

1L

2L

others

unspecified

Bottle size
Figure 6.9 Sizes of reused bottles

6.12 Reuse safety perception and safety debate

Safety concerns about PET bottle reuse exist in both developing and developed
countries. However the degree and type of concern may vary with country. In this
survey the extent to which respondents reported safety concern on PET bottle reuse
differ between Nigerian and British respondents (2 (1, n = 926) = 25.076; p < 0.01, 2
= 0.03). At 28 and 43% for British and Nigerian respondents respectively, more
Nigerians than British respondents are concerned that reusing PET bottle is unsafe.
Reusing and non-reusing respondents in both countries were found to have similar
levels of safety concern. Also native Britons and visitors have similar levels of safety
concern. While some of the concerns reported are similar for the 2 countries others
are unique to individual country (Figure 6.10). Eighty and 71% of the concerns
reported by British and Nigerian respondents were contamination-related. These
concerns include unspecified contamination, chemicals, germs, hygiene and cancer
causation. Concerns not related contamination includes water remaining for too long
in bottle, age of bottle and water source.

97

Respondents (%)

40

30

20

10

Nigeria
Britain

Reasons why bottle reuse is though to be unsafe

Figure 6.10 PET bottle reuse safety perception
Perceiving bottle reuse as unsafe is one thing and the reuse being unsafe is another. As
will be discussed later aged bottles were found to release lesser quantities of migrants
than new bottles in this study. In spite of this the results of the survey carried out in
this study have confirmed that some respondents are actually concerned that bottle
reuse can be unsafe. However the fact that both reusing and non-reusing respondents
in both countries have similar levels of safety concern implied that safety concern is
not an important determinant of bottle reuse. Even though Nigerian respondents
reuse bottles for a longer period of time, they are still concerned that reusing old
bottles is health risk. This can be seen as evidence that unavailability of new bottles is
part of the reason behind longer reuse of PET bottles in Nigeria. Also as a reflection of
insufficiency of clean drinking water in Nigeria only Nigerian respondents reported
water source as a concern in bottle reuse. Both British and Nigerian respondents are
concerned about presence of harmful chemicals in water contained in reused PET
bottles. However British respondents are over four times more concerned about the
presence of harmful chemical contaminants in water contained in bottles than Nigerian
respondents (Figure 6.10). Chemicals in water contained in reused PET may originate
from the water source. In this study most British respondents that mentioned chemical
contamination as a concern mentioned something about plastic releasing chemical into

98

water. In contrast none of the Nigerian respondents specifically mentioned the issue
chemicals being released from plastic material. This clearly shows that only British
respondents nurture the concern about plastics releasing chemicals into the bottle
contents. Further evidence to support this interpretation is that 5% of the British
respondents believe that there is some risk of getting cancer as a result of drinking
water from a reused PET bottle. In other words these respondents believe that reused
PET bottles are releasing carcinogenic chemicals at concentrations that can result in
the user developing cancer. Risk of bacterial contamination and infection has been
mentioned as a safety hazard by both Nigerian and British respondent (Figure 6.10). As
stated in Subsection 3.8 bacterial contamination is a hazard associated with reuse of
PET bottles, especially if a bottle is being reused by more than 1 person and thorough
washing of bottle is not carried out. Usage of a bottle by single person and thorough
washing of the bottles with detergent were suggested as a solution. However the
impact of long-term rigorous washing on bottle behaviour is unknown.
The extent to which chemical migration risk is overstated by environmental
organisations, media and other interest group is believed to have some influence on
the bottle reuse safety perception observed in this study. As mentioned in Section 3.11
chemicals that are not associated with PET bottle and are not carcinogenic are in many
cases termed as carcinogens leaching from PET bottles. The greater concern about
chemical migration and risk of cancer from bottle reuse in Britain than in Nigeria could
be connected to greater access to internet in Britain than in Nigeria.

6.13 Other issues affecting reuse

As mentioned earlier factors thought to be influencing bottle reuse include economic
status and climate of a country, safety debate, cost, bottle availability, age of bottles,
environmental concern, original content of bottles and convenience. Economic status
and climate of a country has already been ruled out as a factor influencing bottle reuse.
Other issues influencing the respondents tendency to reuse plastic bottles apart from
safety concern are shown in Figure 6.11. The degree to which the British and Nigerian
respondents bottle reuse is influenced by issues other than safety concern was found
to statistically significantly differ (2 (1, n = 806) = 4.407; p < 0.05, 2 = 0.005).
However as can be seen the actual strength of this difference is very small and is
deemed negligible. Consequently in spite of the statistical significance of the difference

99

British and Nigerian respondents are practically not very different in terms of the
degree to which their tendency to reuse bottle is affected by issues other than safety
concern. Additionally issues other than safety concern influence the bottle reuse status
of native Britons and visitors equally.

Respondents (%)

40

30

20
Nigeria

10

Britain

Other reasons influencing reuse

Figure 6.11 Other issues influencing reuse
For respondents from both countries the tendency to reuse bottle is affected by
environmental concern, the desire to save money, convenience of reusing bottle and
original content of bottles as shown in Figure 6.11. For the British respondents the
single most important motivation for reusing bottle is the desire to preserve the
integrity of the environment followed by the need to save money. For the Nigerian
respondents convenience associated with reuse of PET bottle is the most important
motivating factor followed by the need to save money. Cost was initially thought to
influence reuse both positively (need to save money) and negatively (preference to buy
new bottles because they are cheap). Contrary to this assumption only 1 British
respondent reported preferring to buy new bottles because according to this
respondent new bottle is cheap and it is easier to dispose the used bottles than to
reuse them. Original content of bottle influences bottle reuse to the same degree for
100

both Nigerian and British respondents. The reuse status of both the British and
Nigerian respondents is influenced by the age of bottle being reused. However while
Nigerian respondents reported the issue of old bottles as a safety concern the British
respondents reported the issue under concern not related to safety (Tables 6.10 and
6.11).

6.14 Summary
While British respondents use more bottled water and soft drinks than Nigerian
respondents which agree with the fact that Britain is more prosperous than Nigeria,
the later stores unopened bottles for longer durations before use. The pattern of
storage in terms of storage places is similar for the two countries. However, possibility
of migration of chemicals from plastic material into the content is likely to be higher in
Nigerias bottled water/drinks as a result longer storage duration and harsher climate.
For both countries the possibility of accumulation of chemicals beyond international
guidelines and standards is likely in only few cases where storage periods are long. The
extent of reuse was similar for both countries, nevertheless Nigerian respondents
reuse bottles for longer duration than British respondents. In case of reuse, the risk of
accumulation of chemicals beyond regulatory levels will depend on the influence of age
on the consistency of bottle material and migration activity of the chemicals in bottle
material. While bottle reuse has not been established as a risk factor in chemical
poisoning, perception that this is so has been observed from among a small proportion
of British respondents. However, the perception that reuse is risky behaviour does not
appear to reduce reuse. For the British respondents the most important motivation to
reuse bottle is the desire to preserve the integrity of the environment followed by the
need to save money. For the Nigerian respondents convenience associated with reuse
of PET bottle is the most important motivating factor followed by the need to save
money.

6.15 Implication of the survey results on laboratory work

The survey carried out in this study provided information which was utilised in three
ways. Consequently, questions asked in the questionnaire can be grouped into three
categories. The first category provided information on issues that were inadequately
addressed in the literature. In some cases questions in the first category helped in
identifying the eligibility of respondents to answer subsequent questions in the
101

questionnaire. The second category provided information used in designing some

aspect of the laboratory work in this study. The third category provided demographic
information on respondents including country of origin, gender, age and university
status. Questions in the first category include questions 1, 2, 7, 11, 15 and 16. Answers
to these questions revealed the extent to which bottled water and soft drinks are
consumed in the 2 countries studied and also the extent to which PET bottles are
reused. This information acted as an additional justification for this study in that the
higher the use of bottled water and soft drinks and the higher the reuse of PET bottle
the greater the need to study migration of chemicals from bottles into contents. The
remaining questions in the first category are questions 18, 19 and 20. These questions
provided information on the issues influencing respondents choice to reuse PET
bottle.
Table 6.4 Influence of survey results on laboratory experiments
Group Question
number(s)
1
1, 2, 7,
11, 15
and 16
2
4, 5, 9
and 10
3

6 and 9

3, 8 and
14

12, 13
and 17

18 and 19

20

21, 22, 23
and 24

Topic of
question
status of use
and reuse

Relevance

storage of
bottles with
contents
storage places
of bottles
with contents

laboratory
work,
literature
laboratory
work,
literature

influenced the choice of

unopened bottle storage
period
influenced the choice of
storage temperature (room
temperature)

sizes of bottle
being
used/reused
Bottle aging

laboratory
work,
literature
laboratory
work,
literature
literature

influenced the choice of

bottle sizes for experiment

reuse safety
perception
Factors
influencing
bottle reuse
Demographic
information

Influence on laboratory
component

literature

literature

102

influenced the choice of

bottle aging periods

Questions in the second category include questions 3, 4, 5, 6, 8, 9, 10, 12, 13, 14 and
17. Answers to these questions influenced the selection of parameters used in the
design of the laboratory work as shown in table 6.4. It is worth mentioning that
inadequate and irregular access to the laboratory equipment meant that the answers
to questions in the second category influenced the laboratory work only partially. For
example 41, 57, 74, 86, 91, 98 and 100% of British respondents reported reusing
bottles in places of residence at maximum periods of 7, 14, 30, 60, 90, 180 and 360
days respectively. Ideally bottles should have been aged for these periods or a
selection of these periods for the purpose of assessing the influence of bottle aging on
chemical migration. However rather than aging the bottles based on these periods the
bottles were aged based on the availability of the instruments.

6.15.1 Storage and chemical migration

Storage is one of the factors that elevate the concentration of migrants from PET
bottles into bottle contents. In the survey conducted in this study both British and
Nigerian respondents reported storing bottled water/soft drink at room temperature
or in the refrigerator for as long as 1 year after purchase. The typical storage duration
for unopened bottles prior to use ranges between one and 7 days. Also the typical
period contents of opened bottles last before been used up is between one and 7 days.
These periods were chosen as typical storage periods because for all cases (unopened
and opened bottles) 50% or more of all observations fell within these period. Based on
this information laboratory experiments were designed to store Nigerian and British
bottled water and soft drinks in PET and glass bottle at room temperature for 7 days,
3 months, 6 months and one year to assess the migration of antimony and
acetaldehyde from bottle wall into the contents. However as a result of instrument
availability and sample storage problems Nigerian samples for assessment of antimony
migration were stored for two and 11 months and British samples were stored for 19
months. For assessment of acetaldehyde migration Nigerian samples were stored for
12 and 25 months and British samples were stored for 20 months. Even though some
British and Nigerian respondents reported storing bottles in refrigerator, chemical
migration at low temperatures associated with refrigerator was not assessed.

103

6.15.2 Bottle age and chemical migration

Concentration of migrant chemicals in PET bottle wall are expected to decrease if the
chemicals are not been regenerated in the bottle wall. From the literature 88% of
respondents in a University of Idaho study reported reusing PET bottle for up to 6
months. In the results of the survey carried out in this study PET bottles were found
to be reused for storing drinking water by 80% of respondents in Britain and Nigerian.
In few cases bottles were reported to be used for over a year. The typical bottle reuse
durations in Britain are between one and 7 days for reuse on the move and at work
and between one and 14 days for reuse in places of residence. In Nigeria the typical
bottle reuse durations are between one and 60 days, one and 7 days and one and 30
days for reuse in places of residence, on the move and at work respectively. Typical
reuse durations are durations covering 50% or more of all observations for a particular
reuse situation. Based on this information laboratory experiments were designed to
age Nigerian and British PET bottles for 7 days, 14 days, 2 months, 3 months, 6 months
and one year for the purpose of assessing the influence of bottle aging on the
concentration of antimony and acetaldehyde in PET bottle wall and on the extent of
migration of these chemicals into bottled water and soft drinks. However as a result of
instrument availability and sample storage problems clear bluish 500ml PET bottles for
assessment of antimony and acetaldehyde concentration in bottle walls and their
migration into contents were aged for 368 days only. Clear green 500ml PET bottles
also used to assess the influence of aging on the concentration of acetaldehyde in
bottle wall, were aged for 266 days.

6.15.3 Bottle size and chemical migration

The ratio between contact area of PET bottle and the volume of the liquid in the
bottle increases as the bottle size decreases. Consequently greater build up of
chemical migrants was proposed in smaller bottles than in bigger bottles. In the survey
carried out in this study the bottle sizes reported are 330, 500, 1000, 1500, 2000, 3000
and 5000ml. According to the survey 61% of bottles in places of residence were 2 litre
bottles or 500ml bottles. To assess the influence of bottle size on antimony migration
5 clear colourless bottles within the range of sizes reported in the survey (2000, 1500,
750, 500 and 330ml) were used.

104

6.16 Conclusion
One of the objectives of the survey was to assess whether the durations of bottled
water and soft drinks storage and use and PET bottle reuse could result in migration of
chemicals into the content to levels beyond international guidelines and standards. The
survey has revealed the typical storage durations and bottle reuse durations in the two
countries. The durations played a vital role in the subsequent chapters in assessing
whether antimony and acetaldehyde are migrating above acceptable limits as a result of
bottled water and soft drinks storage and use and PET bottle reuse.

105

CHAPTER 7:
7.1

IDENTIFICATION OF MATERIALS USED IN BOTTLING

Introduction

Plastic bottles and in some cases plastic bottle caps are coded with the Society of
Plastic Industrys three-chasing arrow recycling symbol (plastic identification code) for
the purpose of recycling. However bottle cap linings, glass bottles and metal crown and
screw caps are not usually coded with these symbols. For the purpose of knowing the
identity of the samples used in this research with absolute certainty all samples used
were characterised by use of Raman Spectroscopy, Energy Dispersive X-ray
spectroscopy or a combination of both techniques. It is worth mentioning that these
techniques were only used for characterisation of samples as they are either not
sensitive enough to detect the low levels of antimony and acetaldehyde being assessed
in this study or they are entirely not meant for detection of these analytes. The results
obtained from the characterisation exercise are presented in this Chapter.

7.2

Plastic bottle materials

Forty transparent Nigerian and British bottled water and soft drink bottle materials
and one whitish British apple juice bottle material were analysed as explained in
Subsection 5.6.4. Thirty one clear bottled water and soft drink materials were
colourless, eight were green and one sample was blue in colour. All bottled water and
soft drink materials were coded as PET. The apple juice material was coded as HDPE.

Figure 7.1 Raman spectra for (a) colourless still water bottle material and (b) PET
reference spectra at 500 - 2000 cm-1

106

Spectra were compared against PET and PE spectra from Thermo Fisher Scientifics
Spectraonline database and Hendra and Agbenyega (1993). Based on the comparison
bottled water and soft drink materials were confirmed to be PET and the apple juice
material PE. Figure 7.1 and 7.2 show the Raman spectra of the materials analysed
together with the reference spectra from Fisher Scientifics Spectraonline database. In
few cases samples suffered from high level of fluorescence.

Figure 7.2 Raman spectra for (a) apple juice bottle material and (b) PE reference
spectra at 600 1800 cm-1
PET bottles are the main bottle types used for commercial bottling of water in large
part because they are light, tough and clear. Use of other plastics including polyvinyl
chloride and polypropylene is not common but has been reported in the works of
Benfenati et al (1991) and Shotyk and Krachler (2007). Soft drinks are bottled in PET,
PE and other non-plastic materials.

7.3

Powdery substance from PET material digestion

As mentioned in Subsection 5.8.9 white precipitate was obtained in the microwave

digestion of PET material. The identity of the white substance could not be ascertained
immediately as digestion of PET material using domestic microwave oven has not been
reported in the literature. The white substance was initially analysed using EDX. As
only oxygen and carbon were detected in significant amounts the substance was

107

presumed to be organic and was thus reanalysed using Raman spectroscopy. A Raman
spectra similar to that of terephthalic acid was obtained. Figure 7.3 show the Raman
spectra obtained for the white substance and from pure terephthalic acid powder used
as reference.

Figure 7.3 Raman spectra for (a) white precipitate from still water bottle digestion and
(b) terephthalic acid powder reference spectra at 200 - 1750 cm-1
The white substance was confirmed to be terephthalic acid. It was thus verified that
the microwave-assisted nitric acid digestion designed in this study hydrolysed the PET
into terephthalic acid (TPA) and possibly ethylene glycol, (the other constituent in the
synthesis of PET), oxalic acid (by product of ethylene glycol oxidation by nitric acid),
other organic compound(s) or carbon dioxide and water. Yoshioka et al (2003)
reported a process for the depolymerisation of PET powder from waste bottles by
using nitric acid. In that process the ethylene glycol generated was simultaneously
oxidized to oxalic acid.

7.4

Glass bottle materials

Glass bottle materials were analysed as explained in Subsection 5.7.4. Figure 7.4 shows
a typical EDX spectra obtained for soft drinks glass bottle materials from both Britain
and Nigeria. Table 7.1 shows comparison of the relative abundance of the elements in
the samples with the weight proportions reported for soda lime glass by Seward and
Vascott (2005). This revealed good agreement confirming the glass materials as soda
lime glass.
108

Figure 7.4 Typical EDX spectra for soft drinks glass bottle materials
No arsenic and antimony were detected. Although as mentioned in Section 2.3 arsenic
and antimony oxides at 0.1 1% by weight may be used as fining agents for glass.
Conclusion cannot be made on their presence or otherwise because the actual weight
percentage range for elemental arsenic and antimony in glass containing 0.1 1% by
weight of arsenic and antimony oxides may not be detected by EDX. According to
Kuisma-Kursula (2000) the minimum detection limits of EDX is about 0.1 weight
percent. In addition, due to the well-known toxicity of arsenic oxide, they are not
likely to be used as fining agents in glass for beverage bottles.
Table 7.1 Relative abundance of the different constituent elements in glass bottle
materials
Element

O
Na
Mg
Al
Si
K
Ca
Fe
Ti
S

British soft
drink bottle
(% wt)
47.73
9.70
0.49
0.79
32.76
0.32
6.4
0.25

Nigerian soft
drink bottle
(% wt)
47.03
9.63
0.30
0.68
34.91
0.34
6.9
0.23

109

Soda lime glass for containers

(Seward and Vascott, 2005) (% wt)
49.25
7.67
0.12
0.69
34.00
0.25
7.46
0.028
0.006
0.08

7.5

Plastic bottle cap materials

Plastic bottle cap materials of different colours (light blue, dark blue, green, white,
black, light grey, red and orange) were analysed. Raman spectra were compared against
Thermo Fisher Scientifics Spectraonline database spectra and Hendra and Agbenyega
(1993). Based on the comparison bottle cap materials were found to be either
polypropylene-based or polyethylene-based plastics. Most caps for carbonated soft
drinks from Nigeria and Britain were made up of polypropylene or a polypropylene
based copolymer. Caps for bottled water were found to be made up of polyethylene
or a polyethylene based copolymer.

Figure 7.5 Raman spectra for 3 bottled water bottle cap materials at 100 and 3000 cm-1
Figure 7.5 shows the Raman spectra of 3 bottled water bottle cap materials between
100 and 3000 cm-1. The bands at 1061, 1128, 1294, 1401-1547 and 2827-3028 cm-1 are
common to all samples whilst other bands were found to be sample-specific. The white
cap spectra matched exactly with the reference spectra of polyethylene (Figure 7.2b).
The blue cap spectra has an additional strong band at 139 cm-1. The spectra for the
dark blue cap have 7 additional bands at 254, 477, 678, 739, 951, 1340 and 1526 cm-1.
The presence of the additional bands especially for the middle spectra may indicate a
polyethylene-based copolymer rather than pure polyethylene. Moreover as the caps
are of different colours the additional bands may simply be due to the colorants used.
Figure 7.6 shows the Raman spectra of 3 carbonated drinks bottle cap materials
together with the polypropylene reference spectra. While all bands from the reference

110

are present in all samples some bands not present in the reference spectra were found
to be sample-specific. These bands appear between 660 and 794cm-1 and between
1509 and 1626 cm-1. These bands are probably attributable to colorants or other
additives in the samples.

Figure 7.6 Raman spectra for (a) 3 carbonated drinks bottle cap materials and (b)
polypropylene reference spectra

7.6

Metal crown and screw cap materials

Metal crown and screw caps from bottled water and soft drinks glass bottles were
analysed using EDX (Section 5.7). The results obtained are shown in Table 7.2. Figures
7.7, 7.8 and 7.9 show typical EDX spectra obtained for British glass bottle screw cap,
Nigerian glass bottle crown cap and British glass bottle crown cap respectively.
According to industry sources crown caps are usually manufactured using printed
tinplate or tin-free steel. While the former is tin-coated steel, the letter is steel coated
with chromium. As an additional measure to prevent corrosion crown caps are as a
rule varnished with lacquer, which is a varnish that dries by solvent evaporation.
Aluminium screw caps are manufactured in a similar manner.

111

Table 7.2 chemical constituents of metal crown and screw caps for glass bottles
Cap

Part

British
bottle
screw cap

major
elements
detected

minor elements
detected

Remark

outer chiselled surface Al

C, O, Mg, Mn, Fe

Aluminium

Inner smooth surface

Al, O, Si, P, Cr, Mn

British
outer chiselled surface Fe
crown cap
Inner smooth surface C

C, Sn, N

Nigerian
crown
caps

outer chiselled surface Fe

C, Cr

Inner smooth surface

O, Fe, Cr, Cl, P

Tinplate (TP)

O, Sn, Fe, P, Cl

Tin-free steel
(TFS) coated
with Chromium

The EDX spectroscopy findings revealed that Nigerian crown caps are made from tinfree steel while the British crown caps are made from tinplate. Glass bottle metal
screw caps from Britain were found to be aluminium.

Figure 7.7 EDX spectra for British glass bottle screw cap

112

Figure 7.8 Typical EDX spectra for Nigerian glass bottle crown cap

Figure 7.9 Typical EDX spectra for British glass bottle crown cap

7.7

Bottle cap linings

The metal and plastic caps of most PET and glass bottles for soft drinks from both
Nigeria and Britain have a plastic material lining the inner part of the cap to help
preserve the carbon dioxide and the aroma in the drinks. Cap lining material is as
important as the bottle material in chemical migration studies because it comes in
direct contact with the bottled liquid even though it does not present larger surface
area as for the bottle material. While the linings for all the British bottles and some
Nigerian bottles were slightly rigid, those found in some Nigerian bottles were flexible
resembling plasticized polyvinyl chloride. The spectra obtained for all the British cap
linings and some Nigerian cap linings match that of ethylene vinyl acetate (EVA)
copolymer. The plastic materials lining the cap of carbonated drinks in PET bottles
from a multinational bottling company in Nigeria were found to be made up of PVC

113

plasticised with a chemical that strongly matches the identity of di(2-ethylhexyl)

phthalate also called dioctyl phthalate (DEHP/DOP). The lining for Nigerian glass
bottles crown caps from the same company revealed a plastic material of unknown
identity.
Typical Raman spectra for British and some Nigerian cap lining and the EVA reference
spectra are shown in Figure 7.10. EVA spectra are virtually similar to that of
polyethylene. According to Shimoyama et al (1997) PE and EVA differ in that EVA has
extra bands at approximately 1740 and 629 cm-1 (C=O stretching and O-C=O
deformation modes) arising solely from vinyl acetate. In Figure 7.10 these two bands
are visible at the extreme ends of the spectra. The size of these bands relative to the
bands of PE indicates the proportion of the vinyl acetate polymer relative to PE. The
Raman bands at about 806 and 839 cm-1 in the cap liner spectra also indicates the
presence of some polypropylene in the material. The two bands are usually the most
intense bands in polypropylene Raman spectra. EVA-based copolymers are used as cap
lining materials in place of pure PE probably because of their superior softness and
flexibility both of which are good attributes in terms of sealing ability.

Figure 7.10 Raman spectra for (a) British and some Nigerian cap lining materials and (b)
Ethylene vinyl acetate reference spectra
Figure 7.11 shows the spectra of a flexible Nigerian PET bottle cap lining material
together with the spectra of a rigid hand wash bottle material coded as PVC. The

114

typical EDX spectra for these samples (Figure 7.12) confirm the presence of chlorine
which makes up as much as 57% of the weight of pure PVC. The relative abundance of
chlorine in the EDX spectra of the cap lining and that of the bottle material were
about 25 and 48% respectively an indication that both materials contains other
chemicals in addition to the PVC with the former containing higher proportion of the
extra chemicals. The 2 C-Cl stretching bands (about 635 and 698 cm-1) typical of PVC
are common to both spectra.

Figure 7.11 Raman spectra of a flexible Nigerian PET bottle cap lining material and a
rigid hand wash PVC bottle material

Figure 7.12 Typical EDX spectra for flexible Nigerian PET bottle cap lining material and
a rigid hand wash PVC bottle material

115

Polyvinyl chloride is usually a rigid plastic. To make it flexible for use in applications
including cap lining, plasticizers are added to it. Plasticized PVC may contain 30 40%
of plasticizers in the form of the phthalate plasticiser, DEHP (Tickner et al, 1999). PVC
may also be plasticised with other types of phthalate plasticisers, adipates, trimellitates,
benzoates, citrate esters, etc. Consequently the plasticizer content of a PVC material
will have some influence on the Raman spectra of the PVC material. The band at about
1728 cm-1 due to C=O stretching, the 2 bands at about 1038 and 1601cm-1 and the
very weak band at 3075 cm-1 (not shown) due to aromatic ring vibrations, occurs in
the cap linings spectra but not in the rigid PVC spectra. This confirms the presence of
aromatic ester plasticizer in the cap lining material. In fact a closer look at the cap
lining spectra reveals it to look very much like a superimposition of the bottle material
spectra over the spectra in Figure 7.13 labelled as dioctyl (DEHP). In other words the
strong resemblance of the cap lining spectra to the DEHP spectra strongly suggests the
presence of DEHP as the used plasticiser in the cap lining material. However because
phthalates esters as a group may show similar Raman bands as can be seen in Figure
7.13, it cannot be said with absolute certainty that the aromatic ester in the cap lining
is DEHP.

Figure 7.13 Raman spectra of phthalate plasticisers (adapted from IDES 2011)

116

The Raman spectra in Figure 7.14 was obtained from the analysis of the cap lining for
Nigerian glass bottles crown caps from the multinational bottling company mentioned
earlier. The spectra has bands at 1083, 1301, 1439 and 2848 cm-1. While this material
looks like plasticised PVC physically, it was not possible to identify the material as
comparable spectra could not be obtained from the literature.

Figure 7.14 Raman spectra for Nigerian glass bottle crown caps cap lining

7.8

Summary

The principal purpose of the characterisation of the bottle materials was to confirm
the identity of samples being used in this research. These characterisation experiments
carried out were not expected detect or quantify antimony and acetaldehyde, the
principal chemicals being investigated in this study. In addition to confirming the
identities of the bulk materials studied here an important finding was made. It was
found that a bottling company uses ethylene vinyl acetate/polypropylene copolymer, a
plastic not associated with any health risk, as bottle cap lining material in Britain.
However in Nigeria the company uses plasticised polyvinyl chloride, a plastic material
which is associated with health risk issues.
PVC is a polymer of vinyl chloride which is an established carcinogen. In the work of
Al-Malack (2001) concentration of vinyl chloride polymer was found to be higher than
the U.S. Environmental Protection Agency permissible level in drinking water (2g/L)
after 30 days exposure of water to rigid PVC pipe in sunlight. As mentioned in Section
2.4 major corporations including Microsoft, Wal-Mart, and Kaiser Permanente have
117

announced efforts to eliminate PVC from products and packaging. Also a bill is being
considered in the State of California for banning the use of PVC in consumer
packaging. Plasticised PVC like the one identified in this study may contain phthalate
plasticisers which are associated with health risks as reported by Bornehag et al. (2004)
and Swan et al, (2005). Due to the potentiality of phthalate plasticisers to cause harm
in children an EU directive (Directive 2005/84/EC of The European Parliament and of
the Council of 14 December 2005) has banned the marketing of plastic toys and
childcare article containing more than 0.1% by mass of some 6 phthalates including
DEHP, the phthalate plasticiser most likely associated with the Nigerian PVC cap lining
material.
It is not clear why the bottling company uses plasticised PVC as cap liner in Nigeria but
not in Britain. However legislation and greater consumer awareness in developed
countries than in developing countries on the PVC health risk issues could be the
reason. The need for a cap sealing material with superior sealing ability to counteract
the effect of the warmer Nigerian climate could also be a reason.

7.9

Conclusion

This Chapter characterised the different materials related to bottling and in the
process narrow down the bottle components expected to contain and release
antimony and acetaldehyde.

118

CHAPTER 8:
8.1

MIGRATION OF ANTIMONY AND OTHER ELEMENTS:

RESULTS, DISCUSSION AND SUMMARY

Introduction

This Chapter presents the results obtained from the experiments in Section 5.8. The
experiments quantified antimony in PET and other plastic materials and in drinking
water from tap, freshly purchased bottled water and soft drink samples in PET bottles,
glass bottles and other plastic bottles. The experiments also assessed migration of
antimony under different conditions. The Chapter interprets and discusses the results
in conjunction with the reviewed literature and the results in preceding chapters.

8.2

Calibration curves

The analysis of antimony, lead, titanium, cadmium, cobalt, germanium, beryllium,

aluminium and zinc in water and soft drinks and the analysis of antimony in PET
materials were based on calibration curves as explained in Section 5.8. The curves for
antimony and the other trace elements are presented in Figures 8.1 and 8.2
respectively. The regression coefficients for all the curves were greater than 0.99, an
indication of linearity and hence accuracy of response within the concentration ranges
used.
4000

Sb

Count

3000

y = 729.1x + 28.101
R = 0.9999
2000

1000

0
0

Concentration (g/L)
Figure 8.1 Typical calibration curve for antimony

119

Figure 8.2 Typical calibration curves for other trace elements

120

8.3

Limits of detection

The limits of detections for antimony and the other trace metals determined in this
study are presented in Table 8.1. The limits of detection are based on three times the
standard deviations of analyte concentrations in 10 analyses of method blanks.
Table 8.1 Limits of detection
Element limit of detection
(g/L)
Sb
0.02
Pb
0.03
Cd
0.02
Co
0.01
Ti
0.16
Ge
0.02
Be
0.003
Zn
0.13
Al
0.08

8.4

Sample coding

To conceal the identity of samples all samples were coded using a 4-point coding
system consisting of 3 letters and a number as typified by UPA1. The code is similar for
bottle material and the content of the bottle. The first letter represents the country of
origin of sample with U standing for Britain and N for Nigeria. The second letter
represents bottling or packaging material with P standing for PET, G for glass, C for
carton and O for other plastic materials. The third letter stands for bottle/carton
content (i.e. bottled water or soft drink) with A standing for still water, B sparkling
water and C soft drinks. The number is a brand number. It is thus similar for similar
brands bottled in different type of packaging or presented as still and sparkling water.

8.5

Antimony in digested PET and other plastics

An important step toward achieving the aims of this study is to quantify the antimony
contents of the bottles themselves. Antimony will not be expected to migrate if it is
absent in bottle material. Moreover, antimony in bottled water and soft drinks can only
be attributed to migration if it can be identified in bottle material.

121

8.5.1

Digestion method

The most commonly reported method of antimony determination in PET involves

complete acid mineralisation of PET followed by analysis by ICP-MS. One of the
drawbacks associated with this method is that a purpose-built laboratory microwave
digestion system has to be available. A new method for PET digestion using domestic
microwave oven has been developed and validated in this study. The method, which is
a modification of the method described by Sakurai et al (2006) for digestion of
polyethylene, is based on total hydrolysis of PET into terephthalic acid (TPA) and
ethylene glycol, ethylene glycol degradation products, carbon dioxide and water. The
digestion procedure described in Subsection 5.8.9 involved putting about 150mg of
PET into a digestion vial containing 3ml of concentrated trace metal grade nitric acid.
The PET material is then digested in the microwave oven for 5 minutes at 230W.

8.5.2

Digestion accuracy and precision

Accuracy of the method was assessed by digestion of polyethylene reference material

ERM-EC681k. Ideally a certified PET reference material should have been used.
However because there is currently no certified PET reference material containing
antimony the polyethylene reference material was used instead. Dobney et al (2002)
has reported using similar reference material in a study involving the forensic analysis
of pressure sensitive adhesive (PSA) tape. An additional indication on the effectiveness
of this method is the resultant total hydrolysis of the PET matrix as evidenced by the
Raman spectra obtained Section 7.3, which completely matched the description of TPA
and also the TPA yield which is in good agreement with the expected yield as will be
discussed in the next subsection. Since antimony is a catalyst, it is not expected to be
bonded to PET molecules. Thus total hydrolysis of the PET material is expected to
completely release the antimony content of the material.
According to the certificate of analysis of the reference material a result is unbiased if
the absolute difference between the mean measured value and the certified value is
less than or equal to the expanded uncertainty of the difference between the measured
and certified value. The measurement is unbiased for antimony and zinc but not for
lead and cadmium as shown in Table 8.2. One sample t-test also revealed a statistically
significant difference between certified values for lead and cadmium and the measured
values. However if the FDA (2001) criteria as stated in the Guidance for

122

Industry/Bioanalytical Method Validation is to be followed, then the difference

between the certified values and measured values for lead and cadmium is not
significant because mean measured values are within 15% of the certified values.
Precision of measurement was determined by measurement of antimony in 14
replicates of PET sample UPA3 as shown in Table 8.3. Precision of measurements is
good as the coefficient of variation is not up to 15%. According to FDA (2001) the
coefficient of variation for the determination of precision should not exceed 15%.
Table 8.2 Accuracy determination using polyethylene reference material ERMEC681k
Element
Sb (ppb) Pb(ppb)
Cd(ppb) Zn(ppb)
Measurement 1
104.43
106.58
122.03
1237.97
Measurement 2
94.76
105.32
130.70
1257.59
Measurement 3
104.26
104.56
131.27
1232.28
Measurement 4
103.87
104.88
129.13
1277.50
Measurement 5
87.42
106.25
133.63
1248.75
Measurement 6
91.17
104.69
119.81
1253.75
Mean single standard deviation
988
105.40.9 1286
125116
996
certified value
986
1374 125070*
Absolute difference between mean
1.35
7.38
9.24
1.31
measured and certified value
Expanded uncertainty
8.59
6.04
6.04
71.21
Recovery (%)
99
108
93
100
Coefficient of variation (%)
8.1
0.9
4.7
1.3
*Indicative value
Table 8.3 Precision determination using 14 replicates of PET sample UPA3
Mean (ppb)
standard deviation (ppb)
Standard error (ppb)
Coefficient of variation (%)

8.5.3

247.81
15.27
4.08
6.16

Terephthalic acid yield in PET digestion

Terephthalic acid yield was determined in PET digestion in part to see if the hydrolysis
of PET was complete. Based on the molecular weights of ethylene glycol (62) and
terephthalic acid (166) and their ratio in PET (1:1), about 27% of PET by weight is
composed of ethylene glycol and about 73% terephthalic acid.

Getting a TPA

proportion of about 73% will thus be an indication of complete hydrolysis of PET. The
results obtained for 5 Nigerian and 5 British samples are presented in Table 8.4. Onesample t-test revealed no statistically significant difference between the mean TPA

123

value observed (76%) and the expected value [t(10) = 1.743; p = 0.115] giving an
indication of complete hydrolysis of PET.
Table 8.4 Terephthalic acid yield
S/no PET material Weight of
Weight of %of TPA
digested PET(g) TPA (g)
by weight
1
NPA8
0.249
0.218
88
2
NPA9
0.267
0.203
76
3
NPA10
0.253
0.208
82
4
NPA11
0.255
0.201
79
5
NPC12
0.265
0.184
69
6
UPA1
0.267
0.192
72
7
UPB5
0.233
0.182
78
8
UPB8
0.249
0.186
75
9
UPC24
0.234
0.182
78
10
UPC28
0.212
0.142
67
Mean
76

8.5.4

Antimony in digestion blanks

The measured antimony in digestion blanks gave a good indication of the absence of
contamination in the determination of antimony in PET as shown in Table 8.5. The
amount of antimony in 5 digestion blanks is less than 0.2% of the amount of antimony
in a digested reference sample.
Table 8.5 Antimony in digestion blanks
Sample
PE reference
digestion blank
digestion blank
digestion blank
digestion blank

8.5.5

mass dilution antimony

(g)
factor
concentration (g/L)
0.171
50
357.2
0
50
0.45
0
50
0.27
0
50
0.40
0
50
0.28

proportion of Sb in digestion
blanks relative to reference
0.13%
0.08%
0.11%
0.08%

Antimony in PET

PET and glass as bottle materials, other plastics (polyethylene, polypropylene and
polyvinyl chloride) used as bottle caps, cap liners and label materials, paper used in
labelling, pigments and dyes for labelling and colouring of bottle parts, adhesives for
securing several components of bottles, metals (aluminium and steel usually coated
with tin or chromium) as metal bottle caps, and lacquers applied to metal caps to
provide a durable finish are all associated with bottled water and soft drink bottling.

124

However, only the bottle material (PET or glass) and to lesser extent the bottle cap
and the lining of the bottle cap, are in constant contact with the bottled content. Only
PET materials were expected to contain antimony as it is used as catalyst in the
synthesis of most PET materials. Even though antimony may be used as fining agent in
glass, glass bottles were not expected to contain antimony at concentrations
obtainable in PET material.
Antimony concentration in Nigerian and British PET materials are presented in Figures
8.3 and 8.4. Table 8.6 shows the colour of the analysed samples. All PET bottles are
transparent. Two-third of still water PET materials analysed have a bluish tint while the
remaining one-third are colourless. Most sparkling water PET materials are green in
colour. The antimony concentration in the 32 PET bottle materials ranged between
177.89 and 310.86 mg/kg with an average of 25030 mg/kg. The concentration agrees
well with the industry reported concentration of between 150 and 350 mg/kg.
Table 8.6 PET samples colours
Nigerian samples
S/no Code
Sample colour
1
NPA1 colourless
2
NPA2 bluish tint
3
NPA3 bluish tint
4
NPA4 colourless
5
NPA5 bluish tint
6
NPA6 bluish tint
7
NPA7 bluish tint
8
NPA8 bluish tint
9
NPA9 bluish tint
10
NPA10 bluish tint
11
NPA11 bluish tint
12
NPC12 colourless
13
NPC13 colourless
14
NPC14 green

British samples
S/no Code
1
2
2
UPA3
3
UPB3
4
UPA4
5
UPB5
6
UPA6
7
UPB8
8
UPB10
9
UPA11
10
UPA12
11
UPA13
12
UPA15
13
UPA17
14
UPC22
15
UPC24
16
UPC25
17
UPC28
18
UPC44

125

Sample colour
colourless
bluish tint
green
colourless
green
colourless
colourless
green
colourless
bluish tint
bluish tint
bluish tint
bluish tint
colourless
colourless
green
green
blue

NPC12

NPC13

268.27

247.27

183.65

250.18

276.42

271.04

246.17

281.66

188.63

177.89

200

262.76

300

275.48

287.2

257.27

Concentration (mg/kg)

400

100

NPC14

NPA11

NPA10

NPA9

NPA8

NPA7

NPA6

NPA5

NPA4

NPA3

NPA2

NPA1

UPB5

UPC25

UPA4

UPC24

UPB3

UPC22

UPA3

249.87

265.67

239.87

221.72

250.09

195.23

263.14

288.4

304.67

272.65

200

202.09

250.85

250.92

309.19

252.56

250

242.07

300

247.81

350

310.86

Concentration (mg/kg)

Figure 8.3 Bar chart illustrating antimony concentrations (mg/kg) in fourteen Nigerian
PET bottles.

150
100
50
UPC44

UPC28

UPA17

UPA15

UPA13

UPA12

UPA11

UPB10

UPB8

UPA6

UPA1

Figure 8.4 Bar chart illustrating antimony concentrations (mg/kg) in eighteen British
PET bottles.
Antimony concentration in glass bottle materials was not determined as the method
used for PET is not meant to be used with glass. As earlier mentioned however,
Shotyk et al (2006) reported antimony concentrations of 7.6 and 10.1ppm from 2 glass
bottles for bottling of water and cola drink respectively. The average antimony
concentration in Nigerian and British PET materials are 25040 and 26030 mg/kg
respectively. The antimony concentrations in PET from these two countries were
shown to be not significantly different (Mann-Whitney U, U = 122, exact p = 0.896)
126

implying the materials to be similar in terms of the amount of antimony catalyst added
during synthesis. Additionally a Kruskal-Wallis test revealed the antimony
concentration to be similar in bottles of different colours [bluish - mean = 248.19
mg/kg, colourless - mean = 265.12 mg/kg and green PET material mean = 239.16
mg/kg) ( (2) = 2.33; p = 0.312] and in bottles having different contents [still water mean = 271.60 mg/kg, sparkling water - mean = 236.48 mg/kg and soft drinks PET
material mean = 245.44 mg/kg) ( (2) = 4.7; p = 0.092].

8.5.6

Antimony in other plastics

As mentioned earlier bottle-related materials other than antimony are not expected to
contain antimony in quantities that could leach into the bottle contents. Table 8.7
shows the antimony concentration found in plastic bottle caps, cap liners and two
polyethylene potable water pouches from Nigeria. As can be seen from the Table the
level of antimony found in these materials is very small in comparison to what is
obtained in PET. This is expected as none of these plastic materials is manufactured
using antimony as catalyst. The low levels of antimony in these samples provide
evidence that the caps do not contribute in the migration of antimony into bottle and
pouch contents. However, the levels of aluminium and titanium were found to be high
in these samples. For the polypropylene and polyethylene-based polymers the levels of
aluminium and titanium are most likely because of the use of Ziegler-Natta catalyst in
their synthesis. Ziegler-Natta catalyst usually contains titanium and organoaluminum.
Table 8.7 Antimony in other plastics
Sample

plastic material

Sb

Al

Ti

(mg/kg)

(mg/kg)

(mg/kg)

UPC24 plastic cap

PP

0.07

43.19

1.12

UPA6 plastic cap

PE

0.03

15.12

10.10

NGC13 plastic cap liner

unidentified plastic

0.03

97.53

13.24

UPC24 plastic cap liner

EVA/PP

0.01

146.08

20.47

NOA17 plastic pouch

PE

0.04

7.12

1.26

NOA18 plastic pouch

PE

0.03

42.86

3.44

127

8.6
8.6.1

Antimony and other trace metals in water and soft drinks

Analytical accuracy and precision

Analytical accuracy and precision for determination of antimony and other trace metals
was monitored by analysis of TM-DWS.2 certified water reference material and by
spiking at 1g/L. The spikes cover germanium which is not contained in the reference
material. From the results in Tables 8.8 and 8.9 analytical accuracy and precision were
good.
Table 8.8 Recovery in spiked samples
Element replicates added *found
recovery coefficient of
(g/L) (g/L)
(%)
Variation (%)
Sb
6
1.00
1.020.07
102.1
6.9
Pb
6
1.00
0.970.03
97.2
3.1
Cd
6
1.00
1.020.05
101.6
4.9
Ge
6
1.00
1.110.04
110.6
3.6
Zn
6
1.00
1.080.08
108.1
7.4
Co
3
1.00 1.0250.007
102.5
0.7
Ti
3
1.00
0.980.03
98.2
3.1
Be
3
1.00 1.0270.003
102.7
0.3
Al
3
1.00
1.050.05
105.4
4.8
*Mean and standard deviation
Table 8.9 Accuracy determination usingTM-DWS.2 certified water reference material
1
Element replicates certified
found
recovery coefficient of
(g/L)
(g/L)
(%)
Variation (%)
2
Sb
10
3.200.05
3.40.3
105.3
8.8
Pb
15
7.80.1
7.90.3
100.4
3.8
Cd
15
4.200.05 4.260.19
101.4
4.5
Zn
15
3793
3774
99.4
1.1
Co
15
64.20.5
67.31.2
104.9
1.8
Ti
7
15.100.14
162
106.4
12.5
Be
15
13.400.14 14.30.4
107
2.8
Al
15
58.30.6
62.30.4
106.8
0.6
1
2
Mean and standard deviation, 95% confidence interval

8.6.2

Antimony in water and soft drinks

The concentration of antimony in drinking tap water and 47 freshly purchased British
bottled samples was determined (Table 8.14). Freshly purchased Nigerian samples
could not be analysed due to sample storage and instrument availability. The antimony
concentration in bottled water and soft drinks ranged between 0.033 and 6.61g/L,

128

with only one juice drink sample going above the EU maximum admissible
concentration and US maximum contaminant level (5 and 6g/L respectively) for
antimony in drinking water (Table 8.12). In the works of Shotyk et al (2006), Shotyk
and Krachler (2007) and Westerhoff et al (2008) the antimony concentration ranges
reported for bottled water were 0.112 0.375 g/L, 0.0089 2.57 g/L and 0.095
0.521 g/L respectively. In this work the average concentration in drinking tap water
samples analysed at 3 different times was 0.270.015 g/L. Only 5 out of the 47
analysed samples have antimony below the concentration found in the tap water. The
concentration of antimony was shown to be similar in still and sparkling water using
Mann-Whitney U test (U = 96, exact p = 0.896). However a significant difference was
found between bottled water (regardless of type) and soft drinks (U = 61, exact p <
0.01). The same could be observed from the average concentrations in Table 8.10. The
antimony concentration in the samples analysed can thus be summarised by the
expression below.
Sb in tap water < Sb in bottled water < Sb in soft drinks
Table 8.10 Antimony in tap water, bottled water and soft drinks
Content type

number

of minimum maximum average number above

samples

(g/L)

(g/L)

(g/L)

acceptable level

still water

19

0.04

2.10

0.89

sparkling

13

0.03

1.92

0.70

soft drinks

15

0.73

6.61

1.98

tap water

0.25

0.29

0.27

water

The average concentrations in water and soft drinks bottled in different bottle types is
given in Table 8.11. The concentrations were found to differ (U = 61, exact p = 0.059)
with contents in glass bottles showing lower concentrations than contents in PET.
Table 8.11 Antimony in PET- and glass-bottled contents
Bottle
type
PET
Glass

number
samples
35
11

of minimum maximum average number above

(g/L)
(g/L)
(g/L)
acceptable level
0.15
6.61
1.31
1
0.03
2.06
0.75
-

129

Table 8.12 Guidelines and standards for antimony and other trace metals in drinking
water
Contaminant EU
Council
directives
98/83/EC and
2003/40/EC

Britain

US

WHO

The Water Supply

Title 40 CFR

(Water Quality)

part 141

Regulations 2000

Maximum

Maximum

Maximum

Secondary

Guideline

Admissible

Admissible

Contaminant

standards

values

Concentration Concentration

Level

(guidelines) (g/L)

(g/L)

(g/L)

(g/L)

(g/L)

Antimony

5 (5)

5(5)

20

Cadmium

5 (3)

5 (3)

Germanium

Zinc

5000

Aluminium

200

200

50 - 200

Beryllium

Titanium

Cobalt

10 (10)

252 (10)

151

10

Lead
1

Action level, 2to change to 10g/L after 25 December 2013, () maximum limit for
constituents naturally present in natural mineral water at source

8.6.3

PET versus glass

To further study the relationship between bottling material and antimony

concentration in bottled content 4 bottled water and 2 soft drinks bottled in both glass
and plastic bottles were studied. A wilcoxon matched-pairs, signed ranks test showed
that the difference between the median concentrations for contents bottled in PET (M
= 1.231g/L, SD = 0.543g/L) and contents bottled in glass (M = 0.754g/L, SD =
0.865g/L) was significant beyond the 0.05 level (exact p < 0.05, two-tailed). In fact the
glass to PET antimony concentration ration in sample UB8 is 1:33. The data in Figure
8.5 also shows the same picture. However the difference appears to be insignificant for
the 2 soft drinks. It is not clear why the concentrations are higher in PET contents for
bottled water but similar for both bottle types for the 2 soft drink brands. The most
likely explanation is that most of the antimony in the soft drinks may have come from

130

the drinks themselves rather the bottles. Cicchella et al (2010) and Reimann et al
(2010) have also reported finding higher concentration of antimony in bottled water in
PET than in similar brands in glass. However these researchers did not extend their
studies to soft drinks. The lower antimony concentration in the water bottled in glass
probably results from the lower antimony concentration observed in glass in
comparison to PET as reported by Shotyk et al (2006).

Concentration (g/L)

2.5
2
1.5

PET
1

Glass

0.5

UC22

UC23

UB10

UB8

UB3

UA3

Figure 8.5 Bar chart illustrating antimony concentration in PET- and glass-bottled
contents

8.6.4

Still versus sparkling water

In Subsection 8.6.2 the concentration of antimony in freshly purchased still and

sparkling water samples were shown to be similar based on Mann-Whitney U test. In
this Subsection wilcoxon matched-pairs, signed ranks test on similar still and sparkling
water brands also showed the difference between the samples to be insignificant (exact
p = 0.57, two-tailed). However in the work of Keresztes et al (2009) antimony
concentration was reported to be higher in 3 sparkling water brands studied than in
similar still water brands. From Figure 8.6 it can be seen that the results are mixed for
the 9 paired samples in this study. It should be appreciated that freshly purchased
British samples are not the best samples to be used to compare antimony migration
from different containers or water types because the actual bottling dates for samples
cannot be ascertained. With Nigerian samples however it is easy to know if samples
are freshly bottled because bottling dates are usually embossed on bottles together
with Best before dates.

131

Concentration (g/L)

1.5

still
sparkling

0.5

UG8

UP7

UP6

UP5

UP4

UG3

UP3

UP2

UP1

Figure 8.6 Bar chart illustrating antimony concentration in similar brands of still and
sparkling water

8.6.5

Conductivity, pH and antimony concentration

In this study bottled water and soft drinks with higher conductivity tend to have higher
concentration of antimony based on the results of Pearsons correlation (r(30) = 0.48;
p < 0.01, r2 = 0.23). According to Cohen (1988) an r value of around 0.10 denotes
small (weak) relationship, around 0.30, medium relationship and around 0.50 or more,
large (strong) relationship. The scatter plot is shown in Figure 8.7.

Conductivity (S/cm)

1600
1200
800
400
0
0

0.5

1.5

2.5

Sb concentration
Figure 8.7 Relationship between conductivity and antimony concentration in bottle
contents
From Figure 8.8 and the results of the Pearson correlation (r(23) = 0.14; p = 0.517, r2
= 0.02) there is little or no relationship between pH and antimony concentration in

132

freshly purchased samples. Reiman et al (2010) also reported observing an almost

independent relationship between pH and antimony migration in an antimony leaching
experiment.

pH

6
4
2
0
0

0.5

1.5

2.5

Sb concentration
Figure 8.8 Relationship between pH and antimony concentration in bottle contents

8.6.6

Other trace elements in water and soft drinks

Concentrations of cadmium, germanium, zinc, aluminium, beryllium, titanium, cobalt

and lead in bottled water and soft drinks are presented in Table 8.14. Summary
statistics for the trace metals are presented in Table 8.13. Concentrations of all the
elements were higher in soft drinks than in bottled and tap water as demonstrated in
Table 8.11. The higher concentrations are believed to be as a result of the ingredients
usually added to soft drinks but not to bottled water. Similarly concentrations of all the
elements except antimony were higher in glass bottled contents than in PET bottled
contents as demonstrated in the Table. The concentrations of lead, cadmium and
beryllium are all within the EU MAC and/or US MCL. Guidelines and standards for
cobalt, titanium, germanium and zinc in drinking-water have not been established. In
the US secondary non-enforceable guidelines regulating contaminants that may cause
cosmetic or aesthetic effects in drinking water covers zinc and aluminium. The
concentration of zinc in all the samples falls within the US MCL. The concentration of
aluminium in 2 soft drinks is however greater than the British MAC. Concentration of
titanium was found to be much higher in soft drinks than in bottled water. The
concentration was found to be highest in cola drinks. Titanium in the form of titanium
dioxide is a food additive approved for use in water-based flavoured drinks as stated in
Codex Alimentarius General Standard for Food additives online database (FAO/WHO
2010). It is reported to be used in powdered concentrate mixes for fruit beverage
133

drinks in India at concentrations not exceeding 100mg/kg (Kuznesof, 2006). It is also

used as a catalyst in the manufacture of PET to very small extent. Its presence at
relatively higher concentrations in soft drinks in this study is believed to be as a result
of its use as food additives in the soft drinks because it was identified in soft drinks
bottled in glass, PET and other plastics at similar concentrations. In spite of its
approved usage as food additive it has recently been reclassified by IARC from Group
3 (not classifiable as to its carcinogenicity to humans) to Group 2B (possibly
carcinogenic to humans) based on inadequate evidence for its carcinogenicity in
humans and sufficient evidence for its carcinogenicity by inhalation in experimental
animals (IARC, 2010).
Table 8.13 Summary statistics for the measured trace elements
Element

Min
(g/L)

Mean
(g/L)
all

Median
(g/L)
water

soft
drinks

PET

Max
(g/L)

SD
(g/L)

glass

Cd

< 0.02

0.40

0.29

0.67

0.36

0.62

0.07

4.41

0.82

Ge

< 0.02

0.21

0.16

0.35

0.21

0.24

0.09

1.14

0.29

Zn

< 0.13

14.91

2.64

41.95

7.72

34.00

2.89

160.49

32.18

Al

< 0.08

28.79

3.46

84.52

11.82

59.01

2.72

373.97

74.94

Be

< 0.003

0.03

0.01

0.07

0.02

0.05

0.004

0.39

0.06

Ti

< 0.16

141.6

2.32

447.78

115.42

224.84

2.46

1650.85

367.02

Co

< 0.01

0.32

0.16

0.66

0.23

0.52

0.16

1.43

0.38

Pb

< 0.03

0.51

0.31

0.95

0.32

1.06

0.21

3.89

0.72

134

Table 8.14 Concentration of antimony and trace elements in water and soft drinks
Brand

Concentration (g/L)
Sb

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48

UPA1
UPA2
UPA19
UPA4
UPA20
UPA18
UPA17
UPA12
UPA13
UPA14
UPA7
UPA5
UPA3
UPA6
UPA15
UPA16
UGC23
UGC22
UPC22
UPC23
UPC24
UPC25
UPC26
UPB2
UPB6
UPB4
UPB3
UPB10
UPB5
UGB8
UGA8
UGB10
UGB9
UPB1
UGB3
UGA3
UPB8
UPB7
UPA21
UPC29
Tap water
UPC30
UGC31
UPC32
UPC33
UGC34
UOC35
UGC36

1.11
0.87
0.15
0.94
0.53
1.04
1.83
1.55
0.32
1.35
0.97
1.24
0.82
0.47
0.75
2.10
1.61
2.06
2.14
1.63
2.35
1.32
2.41
0.28
1.30
0.48
0.74
0.97
0.50
0.03
0.04
0.51
0.61
0.57
0.12
0.19
1.10
1.92
0.60
1.75
0.27
1.90
0.98
1.08
6.61
1.32
1.75
0.73

Cd
<0.02
0.02
1.93
<0.02
0.07
0.04
<0.02
0.05
0.25
0.02
<0.02
2.03
0.12
0.35
0.78
1.12
4.41
0.09
2.40
0.81
1.02
0.05
0.09
0.12
0.07
<0.02
0.38
0.10
0.04
0.05
1.29
0.19
0.04
<0.02
0.08
0.06
0.04
0.04
0.07
0.04
0.08
<0.02
<0.02
0.34
<0.02
0.62
0.12
<0.02

Ge

Zn

Al

Be

<0.02
<0.02
0.76
0.93
1.14
0.12
0.07
0.22
0.29
0.27
0.12
0.04
0.14
0.03
0.06
0.16
<0.02
0.14
0.32
0.07
0.1
<0.02
0.12
<0.02
<0.02
0.39
<0.02
0.05
<0.02
<0.02
<0.02
0.06
<0.02
<0.02
0.11
<0.02
0.03
<0.02
<0.02
0.16
0.08
0.57
0.46
0.45
0.46
0.76
0.57
0.98

4.55
4.62
2.77
2.89
1.72
5.62
1.00
0.74
7.67
0.72
0.94
1.64
3.94
4.08
0.54
0.18
3.38
3.99
4.14
1.19
14.72
0.65
12.50
11.44
2.00
<0.13
0.30
<0.13
<0.13
0.98
<0.13
0.53
2.84
1.76
6.93
2.90
<0.13
2.44
6.58
11.83
4.26
57.51
117.39
34.93
63.97
160.49
68.10
74.41

4.51
1.55
1.46
3.42
5.60
4.27
1.08
2.09
35.44
4.88
0.29
0.16
2.06
1.96
2.72
1.43
15.66
14.99
6.39
5.17
27.25
7.60
2.18
1.44
3.56
0.94
2.72
0.30
0.09
4.48
1.50
1.85
3.82
<0.08
<0.08
<0.08
<0.08
<0.08
<0.08
1.07
20.13
20.44
170.58
172.85
88.36
62.09
299.35
373.97

0.003
<0.003
<0.003
0.013
<0.003
0.005
<0.003
<0.003
<0.003
<0.003
0.008
0.005
<0.003
<0.003
<0.003
0.004
<0.003
0.07
0.066
0.045
0.026
0.036
<0.003
0.053
<0.003
<0.003
<0.003
0.004
<0.003
<0.003
<0.003
<0.003
<0.003
<0.003
0.02
0.011
0.018
0.022
0.009
0.023
0.019
<0.003
0.016
0.187
0.06
0.024
0.066
0.391

135

Ti
1.10
1.75
2.22
1.92
2.12
2.94
5.69
2.43
13.76
2.49
2.15
3.29
2.03
2.89
5.59
5.58
12.16
1650.78
1319.31
16.89
98.06
15.17
1255.41
1.47
1.66
1.47
1.18
1.18
1.11
0.96
1.12
0.81
1.27
<0.16
1.25
0.75
0.56
1.06
0.49
208.80
5.84
72.42
148.07
909.29
76.10
478.11
278.25
177.96

Co
0.04
0.05
0.08
0.03
0.06
0.17
0.01
0.29
0.18
0.11
0.20
0.09
0.13
<0.01
0.09
0.08
0.18
0.18
0.10
0.15
0.19
0.09
0.50
0.09
0.09
0.05
0.12
0.55
0.14
0.20
0.26
0.53
0.46
0.50
0.11
0.10
0.21
0.16
0.11
0.70
0.11
0.94
0.99
0.54
1.39
1.43
1.24
1.30

Pb
0.20
0.61
0.18
0.14
0.08
0.20
0.15
0.13
0.26
0.60
0.62
<0.03
<0.03
<0.03
0.05
0.03
0.76
0.26
0.22
0.22
0.16
0.34
0.49
0.62
0.21
0.11
<0.03
0.04
<0.03
1.21
0.84
1.11
0.35
0.12
0.04
<0.03
1.83
0.43
<0.03
<0.03
<0.03
0.42
1.95
1.01
1.41
1.26
1.80
3.89

8.7

Storage and antimony migration

As stated in Subsection 4.7.4 the longer the duration of contact between water, soft
drinks or foods and the packaging material (PET, glass, etc), the higher the possibility of
accumulation of migrants from the bottle in the water, soft drinks or foods. From
Section 6.6 unopened bottles were found to be stored for as long as 1 year. In this
study Nigerian and British brands of bottled water and soft drinks collected at different
times (different bottling times) were studied to observe the influence of storage on
antimony migration. All samples were stored at room temperature. The best way of
studying migration is to observe changes of antimony concentration with time in the
same sample. However in this work that was not possible due to instrument
availability. It was found that the antimony concentration in 7 Nigerian samples stored
for two months failed to reach 3g/L (Table 8.15 and Figure 8.9). Yet for similar
samples stored for 11 months the concentration has gone beyond the EU MAC in four
of the 19 samples (two soft drinks and two bottled water). Eleven British samples were
also analysed immediately after collection and after 19 months storage (Table 8.16 and
Figure 8.11). For these samples the highest concentration achieved after 19 months is
2.95g/L. The higher concentrations in Nigerian samples could probably be related to
initial exposure to high tropical temperatures in Nigeria before purchase. In the
Nigerian and British samples the highest percentage increases observed were 730 and
584% respectively. It must be noted that the samples with the highest percentage
increases were not the samples with highest concentration as percentage increase
depends on initial concentration and the amount of increase. In the work of Keresztes
et al (2009) antimony concentration in two bottled water samples studied in similar
way approached but failed to reach 1g/L even after three years. Additionally
according to Welle and Franz (2011) maximum migration levels caused by roomtemperature storage of water in PET will not be expected to go higher than 2.5g/L
even after three years. However, the overall concentration of antimony as result of
migration depends appreciably on the initial concentration and ambient weather
conditions. For example (Niedzielski et al, 2001) has reported an antimony
concentration of 1.25g/L from a ground water sample. Additionally in this study and
in the works of Westerhoff et al (2008) and Keresztes et al (2009) different PET
materials have been shown to behave differently in terms of antimony migration.

136

Table 8.15 Change in antimony concentration (g/L) with time in Nigerian samples
Brand
NPA5
NPA2
NPA8
NPA4
NPA6
NGC16
NPC16
NPA1
NPA3
NPA7
NPA9
NPA10
NPA11
NPC12
NPC13
NGC13
NPC15
NGC15
NPC14
Mean

Storage period
2 months 11 months
2.0
2.9
0.3
2.6
1.5
5.1
0.9
1.2
0.7
3.4
1.4
5.1
2.9
5.5
2.6
2.5
1.7
3.4
4.9
5.1
1.2
2.1
4.4
1.6
2.0
3.4
1.4
3.2

% increase
46
730
250
31
362
258
88

Concentration (g/L)

6
5
4
2 months

11 months
2
1
0
NPA5

NPA2

NPA8

NPA4

NPA6

NGC16 NPC16

Figure 8.9 Bar chart illustrating change in antimony concentration with time in Nigerian
samples (dotted line shows the EU MAC)
An interesting observation was made for samples NPC16 and NGC16, which are
similar Nigerian soft drinks bottled in glass and PET. For these samples antimony and
137

lead migration was observed from both the glass and PET bottles in a converse pattern
(Figures 8.9 and 8.10). The initial and final concentrations of antimony were lower in
glass content than in PET content. While the increase in antimony is less than 100% in
PET it was over 200% in glass. Conversely the initial and final concentrations of lead
were much higher in glass than PET. In fact the lead concentration in the 2 and 11
months old soft drinks were higher than the EU and British MAC for lead respectively.
Also the lead concentration in another 2 month old Nigerian soft drink brand from the
same bottling company approached the British MAC. The current British MAC for lead
is 25g/L up to 25th of December 2013 after which it should be reduced to 10g/L
based an EU directive. Some glass bottle materials have earlier been shown to leach
antimony as vigorous as PET materials. In a report for antimony migration Reimann et
al (2010) observed the highest antimony leaching value was from a glass bottle even
though the median antimony concentration for the waters sold in PET bottles was 21
times higher than for the same water sold in glass bottles.

Concentration (g/L)

30
25
20
2 months
15

11 months

10
5
0
NGC16

NPC16

Figure 8.10 Bar chart illustrating change in lead concentration with time in a Nigerian
soft drink sample bottled in glass and PET(dotted line shows the EU MAC)

138

Table 8.16 Change in antimony concentration (g/L) with time in British samples
Storage period
new
19 months
0.32
2.16
0.97
2.88
0.74
1.65
0.82
2.76
0.50
0.76
1.83
3.77
1.30
2.21
1.55
2.95
0.48
0.64
0.75
0.82
0.93
2.06

Brand
UPA13
UPB10
UPB3
UPA3
UPB5
UPA17
UPB6
UPA12
UPB4
UPA15
Mean

% increase
584
196
123
239
53
106
71
90
34
9

Concentration (g/L)

new
19 months

UPA15

UPB4

UPA12

UPB6

UPA17

UPB5

UPA3

UPB3

UPB10

UPA13

Figure 8.11 Bar chart illustrating change in antimony concentration (g/L) with time in
British samples
From the results in Chapter 6 79% of British respondents reported that bottled water
and soft drinks were consumed within 7 days of purchase and 95% reported that they
were consumed within 30 days. Only about 1% of respondents reported consuming
contents in time periods greater than 3 months. Even in Nigeria where contents are
consumed over a longer period of time (7 and 30 days by 50 and 85% of respondents)
only about 6% of respondents reported consumption in time periods greater than 3
months. Consequently the likelihood of consuming soft drinks containing antimony
above the current standard levels is very small. In case of lead the concentration going
beyond the EU MAC after 2 months for bottles that were most likely reused several

139

times is worthy of further investigation In Nigeria glass bottles are used more
frequently in bottling the most popular soft drinks than PET bottles. Bottles are usually
reused by the bottling companies but it is not known many times these bottles are put
into reuse.

8.8

Water temperature elevation on exposure to sunlight

Experiments were carried out to determine the maximum water temperature

achievable on exposure of bottled water to sunlight in Britain in the summer. The
results of these experiments are presented in Figure 8.12. The experiments revealed
that water in green bottles (PET or glass) displayed highest temperature gain while the
water in colourless bottles displayed lowest temperature gain. The temperature gain
for water in blue PET is midway between green and colourless PET. This observation is
believed to be related to differences in absorption, transmission and reflection of light
by the bottles of different colours. Water in PET heats up faster within the first 2 to 3
hours after that the water in glass heats up faster and thus becoming hotter. In general
the experiments revealed that water temperature of up to 46C is achievable if water
in PET or glass bottles is exposed to brilliant sunlight on a clear British summer day
and that a temperature of at least 40C is sustained for up to 6 hours. Attempts to
repeat the experiments for the purpose of monitoring antimony migration were
unsuccessful because of bad weather.

Temperature (C)

50
45

Atmospheric
temperature
Clear PET

40

Bluish PET
Clear glass

35

Blue PET
30
Green PET
25

Crimson Glass

20
15
1000

Green glass
1200

1400

1600

1800

Time of the day

Figure 8.12 Sunlight-assisted bottled water temperature elevation

140

8.9

Antimony migration from PET and glass at elevated temperatures

The results in Section 8.9 confirms that water temperature of up to 46C is achievable
if water in PET or glass bottles is exposed to brilliant sunlight on a clear British
summer day and that a temperature of 40C or more can be sustained for up to 6
hours. In Nigeria a temperature of up to 58.3 C was reported to be achievable on
exposure of water in glass bottles for 2.5 hours from midday (Tukur et al, 2006).
Consequently temperatures of 40 and 60C were chosen for the purpose of assessing
the antimony leaching propensity of eighteen PET bottles and two glass bottles over
forty eight hours exposure time. To also see the behaviour of these materials at
extreme non typical use conditions temperature of 80C was included.
Typical results obtained showing the impact of exposure temperature and exposure
time, on antimony migration are shown in Figures 8.13 and 8.14. As can be seen in
Figure 8.13 the impact of temperature is not uniform over the temperature range
studied even though the temperature increase is uniform. Also from Figure 8.14 a
sharp rise in antimony migration is only observed from about 55C for all the 3
exposure times. Keresztes et al (2009) reported observing a PET bottle material
whose antimony migration rate was not affected by either exposure temperature or
exposure time. However in this work the results obtained for all the ten samples
studied at three exposure temperatures and three exposure times shows similar
pattern as in Figures 8.13 and 8.14. The results for all the ten samples are shown in
Figure 8.18. Figure 8.15 compares the migration propensity of two PET bottles and
two glass bottles at 60C. While the migration pattern is similar for the ten PET
bottles (Figures 8.13, 8.14 and 8.18) the migration intensity differs. The reason for the
difference in antimony migration intensity is probably related to the concentration of
antimony in the PET materials.

141

Concentration (g/L)

25
40C

20

60C
80C

15
10
5
0
0

10

20

30

40

50

Hours
Figure 8.13 Impact of exposure temperature on antimony migration (sample UPA6)

Concentration (g/L)

25
6 hours

20

24 hours
48 hours

15
10
5
0
40

45

50

55

60

65

70

75

80

Temperature (C)
Figure 8.14 Impact of exposure time on antimony migration (sample UPA6)
As discussed earlier some glass bottles may leach as much antimony as PET bottles and
in some cases even more. From Figure 8.15 it can be seen the two British glass bottles
studied in the leaching test leaches very little antimony in comparison to the PET
bottles. In fact for these bottles the leaching stopped after just 6 hours. These bottles
appeared to behave in the same way as the PET bottle material reported by Keresztes
et al (2009) as mentioned earlier. The migration of lead from these two British bottles
has similar pattern and intensity as the leaching of antimony (Figure 8.16). In terms of
antimony and lead leaching these bottles can be said to be clean in comparison to the
142

Nigerian glass bottle (NGC16) discussed in storage experiments (Section 8.7). Similar
experiments should have been carried out using the Nigerian bottle. However the
experiment was not carried out because only one NGC16 bottle sample was collected
from Nigeria.

Concentration (g/L)

3.5
UPA6

UPC24

2.5

UGC36

UGB10

1.5
1
0.5
0
0

10

20

30

40

50

Hours
Figure 8.15 Comparison of antimony migration in 2 PET and 2 glass bottles at 60C

Concentration (g/L)

0.4

0.3
UG36Sb

0.2

UGB10Sb
UGC36Pb

0.1

UGC10Pb
0
0

10

20

30

40

50

Hours
Figure 8.16 Lead and antimony migration in 2 glass bottles at 60C
Figure 8.17 shows the antimony migration results at 60C and 6 hours exposure time
for all the twenty bottles studied. In addition to migration in deionised water migration
into a soft drink was also assessed for ten out of the twenty bottles. Two PET bottles
showing the highest levels of antimony migration are two still water bottles (UPA1 and
UPA12) from France and Norway. Two PET bottles showing lowest level of migration
at these conditions are for still water samples from Britain and Turkey (UPA4 and

143

UPA15). The quantity of antimony that migrated from these bottles is similar to the
quantity from the two glass bottles. It is worth mentioning that even though these
bottles released small amount of antimony relative to other bottles the antimony
concentrations in the bottle materials is similar to average antimony concentration for
the PET bottles analysed in this study. These bottles would probably behave in the
same way as the PET bottle from the work of Keresztes et al (2009) as mentioned
earlier. However because these bottles were studied only at 60C for 6 hours it could
not be ascertained with clarity as to whether they would have behaved in similar way.
The migration pattern observed for bottles filled with soft drink is similar to the
pattern observed for deionised water even though the final concentrations were
higher.

Concentration (g/L)

deionised
water
soft drink

UGB10

UGC36

UPB5

UPB8

UPC25

UPB10

UPA6

UPA1

UPB3

UPC28

UPC24

UPC22

UPA11

UPA13

UPA17

UPA3

UPA15

UPA12

UPA4

UPC44

Figure 8.17 Bar chart illustrating antimony leaching propensity for 18 PET and 2 glass
bottles at 60C for 6 hours
The EU specific migration limit (SML) for antimony from PET bottle wall in foods and
water and the EU maximum admissible concentration (MAC) of antimony into drinking
water are 40g/kg and 5g/L respectively. Because the density of liquid water at 4C is
about 1kg/L it follows that the SML can also be presented as 40g/L. In this study the
release of antimony into deionised water and soft drinks at all the exposure conditions
was lower than the EU SML (Figure 8.18). The implication of this is that all the studied
PET and glass materials from both countries met the requirements of European
Commission Directive 2002/72/EC on plastic materials and articles intended to come
in contact with food. For all experiments at 40 and 60C for up to the 48 hours
maximum exposure time the antimony concentration remained below the EU MAC.
144

The highest concentration recorded at 60C exposure for 48 hours was 4.08g/L.
However at 80C for 6 hours the antimony concentration exceeded the EU MAC in 8
out the10 samples studied. At 80C for 24 and 48 hours the EU MAC was exceeded in
all the ten samples. The highest antimony concentration achieved is close to five times
the EU MAC. Westerhoff et al, (2008) recorded antimony concentration of 14.4g/L
after exposing water in a PET bottle to a temperature of 80C for 7 days. In this work
the antimony concentration attained in some bottles after 24 hours of exposure at
80C is similar to what was obtained by Westerhoff et al, (2008) after 7 days exposure
as can be seen in Figure 8.18. But for other bottles antimony concentration could not
reach the concentration reported by Westerhoff et al, (2008) even after 48 hours of
exposure at 80C. From these results it is clear that different bottle materials behave
differently with regards to antimony leaching in water at elevated temperatures.
Additionally it is likely that the experiments of Westerhoff et al, (2008) carried out for
a fewer days would have achieved concentrations close to what they obtained for 7
days because antimony migration from PET generally decline with time as can be seen
in Figures 8.13 and 8.15. Westerhoff et al, (2008) extrapolated the exposure durations
required to reach the US Maximum Contaminant Level of 6g/L for exposure
temperatures of 60 and 80C as 176 and 2.3 days respectively. This extrapolation is
not applicable to all bottle materials because from Figure 8.18 it can be that seen for a
particular bottle the antimony concentration was close to 5g/L after just 48 hours of
exposure at 60C and for other bottles the US MCL was exceeded after 6 hours of
exposure at 80C.
As mentioned earlier bottled water temperatures at 40 and 60C are temperatures
that could be realistically encountered in PET bottle use or reuse in Nigeria and
Britain. Consequently the risk of consuming water containing antimony above the EU
MAC is removed for exposures within the time span studied. In solar water
disinfection the minimum water treatment duration is 6 hours in sunny days and for
days that are 50% cloudy. An important point worth mentioning is that in tropical
developing countries like Nigeria it is still possible for bottled water to be exposed to
the scorching tropical sun for several days in their life cycle. As the maximum
exposure period used in this study is 48 hours it is not clear if the EU MAC could be
exceeded after exposure at 60C for several days.

145

146
UPB5 80C

UPB5 60C

UPB5 40C

UPB8 80C

UPB8 60C

UPB8 40C

UPC25 80C

UPC25 60C

UPC25 40C

UPB10 80C

UPB10 60C

UPB10 40C

UPA6 80C

UPA6 60C

UPA6 40C

UPA1 80C

UPA1 60C

UPA1 40C

UPB3 80C

UPB3 60C

UPB3 40C

UPC28 80C

UPC28 60C

UPC28 40C

UPC24 80C

UPC24 60C

UPC24 40C

UPC22 80C

UPC22 60C

UPC22 40C

Antimony concentration (g/L)

25

20

15

0 hours

10

6 hours

24 hours

48 hours

Figure 8.18 Bar chart illustrating antimony leaching propensity for 10 PET bottles at 40, 60 and 80C for 6, 24 and 48 hours (dotted line shows
EU MAC)

8.10 Dependency of antimony migration on its concentration in PET

The concentration of antimony used as catalyst in PET synthesis varies from 150 and
350mg/kg as mentioned earlier and as observed in this study. According to Fick's law
the rate of diffusion in a given direction is directly proportional to the concentration
gradient. Thus if antimony migration from PET follows the fickian diffusion kinetics the
migration will be directly proportional to antimony concentration in PET at any given
temperature. Experiments were carried out to assess whether there is any correlation
between the concentration of antimony in PET and its migration in deionized water as
explained in Sections 5.8. From the results in Table 8.17 it can be seen that at 60C
antimony migration follows the fickian diffusion kinetics for all exposure times. At this
temperature antimony migration is directly proportional to the initial antimony
concentration in PET for all the samples. For exposure at 40C for 6 hours very high
but negative correlation was observed implying that at these exposure conditions the
higher the concentration in PET the lower the migration into the water. At 80C the
dependency is uniform but low for all exposure time. The low dependency at 80C is
probably indicates that different PET materials respond to the elevated temperature
differently. Different PET materials could differ in their intrinsic viscosity and
crystallinity.
Table 8.17 Dependency of antimony migration on its concentration in PET
Temperature Exposure
time (hours)
40C
6
24
48
60C
6
24
48
80C
6
24
48

Pearsons
correlation (r)
-0.75
0.38
-0.12
0.60
0.65
0.64
0.37
0.23
0.25

8.11 Dependency of antimony migration on PET bottle thickness

To assess whether there is any relationship between bottle thickness and antimony
migration a digital calliper was used to measure the thickness of PET bottles for
different content types from both Britain and Nigeria. The thickness statistics are
147

shown in Table 8.18. Both t and Mann-Whitney U tests proved the thickness of the
samples from the two countries to be statistically significantly different. The thinner
bottles in Britain are not unrelated to the efforts in Britain to reduce packaging waste
as reported by WRAP (2007). According to WRAP drinks sector contributes onethird of all packaging arising in the household waste stream. Bottles for carbonated
drinks and sparkling water were also found to be thicker than bottles for still water in
both countries most likely because of the greater need for them to be stronger for the
purpose of withstanding the elevated pressures due to carbonation of contents.
Table 8.18 PET bottle thickness

Average

Nigerian bottle

British bottles

carbonated

still

carbonated

still

drinks

water

drinks

water water

sparkling

407

253

272

197

273

Maximum (m)

462

350

302

253

340

Minimum(m)

348

215

195

125

233

thickness (m)

The thickness of British bottles was tested against their antimony migration ability at
different temperatures. The results are presented in Table 8.19. During the first 6
hours of exposure at 40C the relationship is weak, implying that thickness is not
influencing migration. The absence of any relationship at this condition is most likely
because migrating antimony is mainly from the outermost surface of the bottle wall.
This can be further substantiated by the fact that for all temperatures rate of migration
is highest within the first 6 hours (Figures 8.13 and 8.15). After 24 hours a direct
proportionality can be observed between bottle thickness and antimony migration,
implying that as thickness increases migration increases. The relationship became
inverse after 48 hours of exposure. At 60C positive but medium dependency was
observed for all exposure times. At 80C no relationship could be observed between
bottle thickness and antimony migration. Again this could be due to different
behaviours of different PET materials in response to elevated temperatures. Overall it
can be said that antimony migration shows some dependency on bottle thickness at 40
and 60C but not at 80C.

148

Table 8.19 Dependency of antimony migration on PET bottle thickness

Temperature Exposure
Pearsons
time (hours) correlation (r)
40C
6
0.13
24
0.67
48
-0.47
60C
6
0.27
24
0.39
48
0.36
80C
6
-0.02
24
-0.03
48
-0.04

8.12 Bottle aging and antimony migration

From the survey carried out in this study and information from the literature as much
as 80% of respondents in Nigeria, Britain and United States reported being in the habit
of reusing PET bottles. Also a few respondents reported reusing bottles for more than
a year. The results for experiments carried out to assess the antimony leaching
propensity of new and aged PET bottles are presented in Figures 8.19 and 8.20. The
bottles labelled as aged were aged by filling and emptying with tap water on weekly
basis over a period of one year. The bottles labelled aged 2 were treated in similar way
except that after each emptying the bottles are scrubbed using bottle brush with mild
detergent in hot water. In Figure 8.19 the three bottles from a French still water brand
(UPA12) were filled with deionised and sparkling water and allowed to stand for 9
months. In Figure 8.20 the bottles were subjected to antimony migration test at
elevated but realistic temperatures. Since the amount of antimony for migration is
supposed to be depleted with sustained use of bottle it was expected that reused
bottles will leach antimony to a lesser extent in comparison to new bottles. The
results obtained agreed with the expected results. The greater concentration of
antimony in sparkling water than in still water in Figure 8.19 should be in part due to
higher initial concentration in sparkling water. In all the antimony migration
experiments new bottles released more antimony into water than aged bottles and in
all cases antimony concentration fell short of reaching the EU MAC for antimony in
drinking water. An important point worth mentioning here is that only a proportion of
antimony from the bottle wall is available for migration. According to Alt and

149

Haldimann (2008) only about 60% of the antimony in PET bottle wall is available for
migration.
3

Concentration (g/L)

2.5
2

new

1.5

aged
aged 2

1
0.5
0

Sparkling water

Deionised water

Figure 8.19 Bar chart illustrating antimony migration in new and one year aged PET
bottles (UPA12) filled with deionised water for 9 months

Concentration (g/L)

1.6

1.2

0.8

0.4

New bottle

Aged bottle 1

Aged bottle 2

Figure 8.20 Bar chart illustrating antimony migration at 60C in new and one year aged
PET bottles (UPA12)

150

8.13 Water pH and antimony migration

In Subsection 8.6.5 antimony concentration in freshly purchased bottled water and soft
drink was independent of pH. In this experiment a slight difference in antimony
concentration was observed in three water samples of different pH stored in green
sparkling water PET bottle (UPB5) over a period of 266 days. The results in Figure
8.21 confirmed that in stored water sample the lower the pH of the water the higher
the migration of antimony.

Concentration (g/L)

0.6
0.5
0.4
0.3

pH2
pH5

0.2

pH8
0.1
0
0

266

Days of storage
Figure 8. 21 Bar chart illustrating the dependency of antimony migration on water pH

8.14 Bottle size and antimony migration

From the survey results 2L and 500ml bottles accounted for 61% of the freshly
purchased bottled water and soft drinks reported in places of residence of British
respondents. Also close to 60% of bottles being reused are bottles with a volume of
500ml or less. Based on the results in their study Westerhoff et al (2008) hypothesized
a possible relationship between antimony release in PET bottles and contact area to
liquid volume ratio. Keresztes et al (2009) also observed a direct relationship between
antimony concentration in bottle content and surface area to liquid volume ratio for 3
bottled water samples bottled at the same time in PET bottles of different volumes.
Further study of this phenomenon at elevated temperatures using 5 clear still water
bottles of the same brand (UPA3) but having different sizes gave the results in Figure
8.22. The results agree with the observation of Keresztes et al (2009) except for the

151

smallest bottle. For all the bottles used the contact area to liquid volume ratio
increases as bottle size decreased. While the antimony concentration in the biggest
bottle was 2.22g/L the concentration in the bottle next to the smallest one was
beyond the EU MAC. These results confirmed the existence of an inverse relationship
between bottle size and build-up of antimony in bottle contents. It is however worth
mentioning that the migration temperature used is not typically encountered in normal
bottle usage. So for all bottles of all sizes concentration of antimony will not be
expected to go beyond the EU MAC if bottles are heated at realistic temperatures.

Concentrartion (g/L)

6
5
4
3
2
1
0

330ml

500ml

750ml

1500ml

2000ml

Bottle size
Figure 8.22 Bar chart illustrating migration of antimony into bottles of different sizes at
70C (dotted line shows EU MAC)

8.15 Summary
The concentration of antimony and some trace metals in PET bottle materials and in
bottled water and soft drinks from Nigeria and Britain were determined. Migration of
antimony from PET and glass bottles at different conditions was also assessed. All
Nigerian and British PET bottle materials contained antimony within the concentration
range reported by industry sources implying the use of antimony catalyst in their
synthesis. Antimony concentration in Nigerian and British PET materials were similar
containing 25040 and 26030 mg/kg respectively. Antimony content in plastic bottle
caps and cap liners was found to be too low to contribute in migration of antimony

152

into bottled contents. A new method for PET digestion involving the hydrolysis of PET
using domestic microwave oven has been described and validated.
Antimony concentration in British drinking tap water was 0.270.015 g/L. The
concentration ranged between 0.033 and 6.61g/L in 47 freshly purchased British
bottled water and soft drinks with only one sample going above the EU acceptable
limit. While the concentration in still and sparkling water were similar the
concentration in soft drinks was found to be higher than in bottled water regardless of
type. The antimony concentration of liquids contained in PET was higher than in glass
bottles. Bottled water and soft drinks with higher conductivity tend to have higher
concentration of antimony, however little or no relationship exists between pH and
antimony concentration in freshly purchased samples. But lower pH appeared to be
associated with higher antimony concentration in storage experiments.
Concentrations of trace elements investigated (Cd, Ge, Zn, Al, Be, Ti, Co, Pb) were
higher in soft drinks than in bottled and tap water. Similarly concentrations of all the
elements except antimony were higher in glass bottled contents than in PET bottled
contents. High levels of titanium were detected in soft drinks from both countries. The
detected titanium is believed to be in the soft drinks as a result of its usage as food
additive. Concentration of antimony in some Nigerian bottled water and soft drinks
was above the EU MAC after 11 months of storage at room temperature. For 10
British bottled water samples the concentration remained below the EU MAC even
after 19 months of storage. A Nigerian glass bottle for soft drink leached both
antimony and lead above EU MAC after 2 months. However 2 British glass bottles
subjected to antimony migration test at elevated temperatures demonstrated low level
of antimony and lead migration.
Antimony concentrations in water exposed at 40, 60 and 80C for up to 48 hours in
PET and glass bottles remained below the EU specific migration limit for antimony
from plastic materials and other articles intended to come in contact with food. At
realistic temperatures of 40 and 60C antimony concentration in the water remained
below the EU MAC even after 48 hours of exposure but the concentration exceeded
the EU MAC for most exposures at 80C. Antimony migration into water was found
to be directly proportional to the antimony concentration in PET and to bottle
thickness for some exposure conditions. British bottles were generally thinner than
153

Nigerian bottles. Aged bottles leach lower amount of antimony than new bottles.
Similarly larger bottles leach lower amount of antimony than smaller bottles.
Antimony concentration in freshly purchased bottled water and soft drinks and in
migration experiments at realistic conditions are mostly within acceptable levels.

8.16 Conclusion
In this Chapter antimony concentration in PET bottle materials and in bottle contents
were analysed. From the results most freshly purchased bottled water and soft drinks
were found to be safe for consumption in terms of the content of antimony and the
other trace elements analysed. The results also showed that extended storage of
bottled water and soft drinks bottled in PET can lead to the concentration of antimony
going above the safe limit. Exposure of water to PET at realistic temperatures did not
result in antimony concentrations going above the safe limits. However exposure at
atypical extreme temperatures resulted in antimony concentration in the water going
above the safe limits.

154

CHAPTER 9:
9.1

ACETALDEHYDE MIGRATION: RESULTS, DISCUSSION

AND SUMMARY

Introduction

This Chapter presents the results obtained from the experiments explained in Section
5.9 of Chapter 5 (Methodology). The experiments quantified acetaldehyde in PET
materials, drinking water from tap, freshly purchased bottled water and soft drink
samples in PET bottles, glass bottles and other plastic bottles. The experiments also
assessed migration of antimony under different conditions. The Chapter interprets and
discusses the results in conjunction with the reviewed literature and the results in
preceding chapters.

9.2

Calibration and retention time

The analysis of acetaldehyde in water and soft drinks and in PET materials was based
on calibration curves as explained in Section 5.9. The curves for acetaldehyde
determination in water, soft drinks and PET materials are presented in Figures 9.1, 9.2
and 9.3 respectively. The regression coefficients for all the curves were greater than
0.99, an indication of linearity and hence accuracy of response within the concentration
ranges used. Retention time of acetaldehyde differed with the injection mode, in all
cases the retention time was less than 1 minute. A typical acetaldehyde chromatogram
is shown in Figure 9.4.
35
30

y = 0.0628x + 0.2428
R = 0.9978

Peak area

25
20
15
10
5
0
0

100

200

300

400

500

Acetaldehyde concentration (g/L)

Figure 9.1 Calibration curve (splitless injection mode) for acetaldehyde in bottled water

155

Peak area

45
40
35
30
25
20
15
10
5
0

y = 4.2063x - 0.0061
R = 0.9999

10

Acetaldehyde concentration (mg/L)

Figure 9.2 Calibration curve (split injection, 1:50 split ratio) for acetaldehyde in soft
drinks
120

Peak area

100
y = 0.1132x + 0.7412
R = 0.9997

80
60
40
20
0
0

200

400

600

800

1000

Acetaldehyde concentration (g/L)

Figure 9.3 Calibration curve (split injection, 1:50 split ratio) for residual acetaldehyde in
PET materials

Figure 9.4 Typical acetaldehyde GC-FID chromatogram (retention time: 0.5 minutes)

156

9.3

Limits of detection

The limits of detections for acetaldehyde in water and residual acetaldehyde in PET are
presented in Table 9.1. The limits of detection are based on three times the standard
deviations of analyte concentrations in 10 analyses of 10 g/L acetaldehyde solution
and acetaldehyde vapour at 8.64 g/L.
Table 9.1 Limits of detection
limit of detection (g/L)
acetaldehyde in water
2.16
Residual acetaldehyde in PET
0.86

9.4

Accuracy and precision

Accuracy and precision for the determination of acetaldehyde in water and soft drinks
were achieved by analysis of spikes of acetaldehyde at 50 and 100 g/L (Table 9.1).
Accuracy and precision for determination of residual acetaldehyde in PET was achieved
by analysis of acetaldehyde vapour at concentration of 385.80 g/L (Table 9.2). In all
cases accuracy and precision were good as recoveries were around 100% and
coefficients of variation were less than 15%.
Table 9.2 Accuracy and precision for acetaldehyde determination in water
Added
Found

Mean
standard deviation
Standard error
Coefficient of variation (%)
Recovery (%)

Spikes (g/L)
50
52.54
51.20
51.20
50.77
50.70
51.13
50.63
50.35
52.11
50.56
51.12
0.70
0.22
1.38
102.24

157

100
92.82
101.48
100.70
105.14

100.04
5.18
2.59
5.18
100.04

Table 9.3 Accuracy and precision for determination of residual acetaldehyde in PET
Added
Found

Mean
standard deviation
Standard error
Coefficient of variation (%)
Recovery (%)

9.5

Spike (g/L)
385.80
394.21
376.17
385.60
370.68
382.07
381.75
9.00
4.02
2.36
98.95

Sample coding

Similar sample coding was used as in Section 8.3.

9.6

Residual acetaldehyde in PET bottles materials

The concentration of residual acetaldehyde in freshly purchased British and Nigerian

PET bottle materials is given in Table 9.4. Samples placed on the same rows in the
Table are samples for similar soft drinks obtainable in both Britain and Nigeria. The
minimum and maximum concentrations for British samples are 0.95 and 12.52 g/g
respectively. The total average and the average for still water, sparkling water and soft
drinks PET materials for British samples are 4.76, 4.1, 1.55 and 6.7g/g respectively.
The average concentration for Nigerian soft drinks PET materials is 2.17g/g.
As can be seen from Figure 9.5 and Table 9.4 the acetaldehyde concentration is higher
in British soft drink PET materials than in the corresponding Nigerian PET materials.
This difference is believed to be related to the concentration of acetaldehyde in the
contents of the bottles. As will be discussed later acetaldehyde concentration was
found to be much higher in a cola drink from Nigeria than in similar brand marketed in
the UK. Additionally PET material has been shown in this study to have the ability to
absorb acetaldehyde in solution. The average concentrations reported by Matsuga et al
(2005) for Japanese, European and American bottles were 14.9, 7.8 and 12.3g/g
respectively. Comparison of the results obtained in this study with the results of
Matsuga et al (2005) will imply that the residual acetaldehyde in PET bottle materials is
lower in British and Nigerian bottles than in Japanese, European and American bottles.

158

Table 9.4 Residual acetaldehyde in fresh British and Nigerian PET bottle materials
British PET
samples
UPC38
UPC33
UPC30
UPC32
UPC41
UPC25
UPC22
UPC26
UPC24
UPC23
UPA12
UPB10
UPA5
UPB5
UPA4
UPB4
UPA3
UPB3
UPA6
UPB6

Acetaldehyde
concentration
(g/g)
3.06
8.56
9.24
6.78
3.67
4.47
3.79
4.05
12.52
10.89
1.08
0.95
4.49
2.12
8.39
1.19
4.55
1.93
1.97
1.56

Nigerian PET
samples

Acetaldehyde
concentration
(g/g)

NPC14
NPC13
NPC16
NPC12
NPC15

2.61
1.29
2.54
2.72
1.69

It is however worth mentioning that Matsuga et al (2005) used a quantitation method

different from the one used in this study and it is not known if quantitation method
difference has any impact in determination of residual acetaldehyde in PET. The
method used by Matsuga et al (2005) Involved simultaneous dissolution of PET in
trifluoroacetic acid and derivatization of the acetaldehyde content with 2,4dinitrophenylhydrazine

followed

by

precipitation,

solvent

extraction

with

dichloromethane, evaporation, reconstitution in acetonitrile and analysis by HPLC. The

acetaldehyde concentration range as reported by Linssen et al (1995) and Bashir et al
(2002) were 1.1 3.8 and 0.3 0.8 g/g respectively. It is unclear what method was
used by Linssen et al (1995), however Bashir et al (2002) used a method differing
slightly from the method used in this study. In order to generate greater surface area
which will result in greater efficiency of extraction Bashir et al (2002) ground the PET
after cooling in liquid nitrogen. While this method generate greater surface area for

159

efficient desorption of acetaldehyde it also presents the risk of evaporation of the

volatile acetaldehyde before the ground PET can be placed in vials.

Acetaldehyde concentrations
(g/g)

8
7
6
5
4
3
2
1
0

British PET

Nigeria PET

Figure 9.5 Average acetaldehyde concentration in British and Nigerian soft drinks PET
materials
PET material for making containers for bottled water needs to contain lower amounts
of acetaldehyde than material for making bottles for carbonated soft drinks because
carbonated drinks can mask the taste and odour of acetaldehyde more than bottled
water. As shown in Figure 9.6, the residual acetaldehyde content in PET differs for PET
materials used for bottling of different contents. The acetaldehyde content was found
to be higher in soft drinks PET followed by still water and sparkling water PET. The
difference in concentration was found to be statistically significant using Kruskall-Wallis
chi-square test (p < 0.01: 2 (3) = 9.351) and one-way ANOVA (p < 0.05, adjusted R2
= 0.329). However Tukeys HSD range test revealed a statistically significant difference
in acetaldehyde concentration between sparkling water PET and soft drink PET but not
between still water PET and any of the 2 samples. As suggested earlier the higher
concentration in soft drinks PET is believed to be related to the concentration in the
soft drinks. The concentration difference between still and sparkling water PET is
believed to be due to the higher acetaldehyde extraction ability in sparkling water than
in still water.

160

Acetaldehyde concentrations
(g/g)

7
6
5
4
3
2
1
0

Soft drinks

Still water

Sparkling water

PET bottle material

Figure 9.6 Average residual acetaldehyde in British PET bottles used for different soft
drinks (n=10), still water (n=5) and sparkling water (n=5)
The concentration of residual acetaldehyde in 2, 13 and 26 months old Nigerian PET
bottle materials is given in Table 9.5. The minimum and maximum concentrations are
0.46 and 3.09g/g respectively. As for British samples the acetaldehyde concentration
was found to be higher in soft drinks PET (mean = 1.5 g/g) than in bottled water PET
(mean = 0.83 g/g). The difference was found to be statistically significant using MannWhitney U test (exact p = 0.009 U = 5). Change of acetaldehyde concentration in
Nigerian PET with time is discussed in Subsection 9.6.
Table 9.5 Residual acetaldehyde in Nigerian PET
PET samples Acetaldehyde concentration (g/g)
2 months
13 months
26 months
NPA2
0.90
1.05
0.950
NPA6
0.87
0.46
NPA8
1.63
0.60
NPA4
0.97
0.63
NPA5
1.64
0.64
0.60
NPA1
1.85
0.63
NPA11
0.51
NPA9
0.93
NPA10
0.69
NPA3
1.06
NPA7
0.68
NPC14
1.10
NPC13
1.63
NPC16
3.09
1.67
NPC12
1.53
NPC15
1.56
161

9.7

Bottle aging and acetaldehyde concentration in PET

As mentioned earlier, acetaldehyde is formed in PET by thermal degradation at

temperatures above 160C. As PET bottles are not typically exposed to such high
temperature, the formation of acetaldehyde will not be expected to happen in PET
bottle material after manufacture. As mentioned in Subsection 4.7.6. Nawrocki et al
(2002) reported lower acetaldehyde concentration in one month old bottle material
than in new bottle material. These authors also observed increases in the acetaldehyde
content of bottled water with time. In addition, as will be discussed later outgassing of
acetaldehyde has been observed from some PET bottles in this study. Based on these
observations, the acetaldehyde content of PET bottle material will be expected to
decrease with time rather than increase. Figure 9.7 compares the acetaldehyde
contents of new and aged materials for two different brands of PET bottles. The
samples UPB5 and UPA12 were aged for 266 and 368 days respectively. The aging
period was selected in part because bottle reuse period of up to one year was
reported by both British and Nigerian respondents in the survey carried out in this
study. As explained in Subsections 5.9.10 the aging process was made to as much as
possible imitates actual bottle reuse. Also while the bottles labelled as aged bottle 1
were just rinsed with water before refilling, the bottles labelled aged bottle 2 were
scrubbed using brush with dilute solution of detergent in hot water prior to refilling.

Acetaldehyde concentrations
(g/g)

2.5

New bottle
Aged bottle 1
Aged bottle 2

1.5
1

0.5
0

UPB5

UPA12

Figure 9.7 Acetaldehyde concentration in new and aged PET bottle materials
As expected, the results revealed that that the acetaldehyde content of PET material
decreases as the bottle material ages. However the amount of decrease observed is
lower than what was expected. This most likely implies that only the acetaldehyde in
162

the upper layers of the PET material is available for diffusion under the aging
conditions.
The results in Figures 9.8 and 9.9 further substantiate the observation made in Figure
9.7. In both cases the concentration of acetaldehyde in bottles allowed to remain with
their contents for 13 months was plotted against the concentration of acetaldehyde in
newer bottles. With the exception of NPA2 and NPC13, the concentration of
acetaldehyde is higher in newer bottle materials than in bottle materials allowed to
remain with contents for 13 months.

Acetaldehyde concentrations
(g/g)

1.8

2 months
13 months

1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
NPA2

NPA6

NPA8

NPA4

NPA5

Figure 9.8 Change in acetaldehyde concentration in Nigerian bottled water PET with
time

Acetaldehyde concentrations
(g/g)

fresh

2.5

13 months
2
1.5
1
0.5
0
NPC14

NPC13

NPC16

NPC12

NPC15

Figure 9.9 Change in acetaldehyde concentration in Nigerian soft drink PET with time

163

9.8

Bottle thickness and acetaldehyde concentration in PET

Bottle wall thickness was presumed to be positively associated with the concentration
of residual acetaldehyde in PET. British PET bottles for water were used to test this
hypothesis as newly collected samples were available and unlike PET bottles for soft
drinks these bottles have not been in contact with liquids that may contain
acetaldehyde. Pearsons correlation for the data showed strong statistically significant
but negative correlation (r(8) = -0.823; p = 0.012. r2 = 0.68). The r2 value which is
referred to as coefficient of determination gives the proportion of variability of
thickness that can be explained by the linear relationship between the two variables
being studied. In this case it implies that 68% of the total variation in bottle wall
thickness can be explained by the linear relationship between concentrations of
residual acetaldehyde and bottle wall thickness. The negative correlation implies that
the thinner the bottle the higher the concentration of residual acetaldehyde. The
scatter plot for the relationship is shown in Figure 9.10. The results obtained are
clearly opposite to what was predicted. However as can be seen in Figure 9.6 and 9.11
the acetaldehyde concentration in the thicker sparkling water PET is lower than in the
thinner still water PET. The acetaldehyde concentrations in the two materials were
expected to be similar. The lower acetaldehyde content in sparkling water PET is in
large part attributed to sparkling waters greater ability to extract acetaldehyde from
PET bottle wall than still water. As stated in Subsection 4.7.3, carbonation of water
was reported to enhance formation and/or migration of acetaldehyde from PET bottles
into water.

Bottle thickness (m)

300
250
200
150
100
0

Residual acetaldehyde (g/g)

Figure 9.10 Relationship between bottle wall thickness and residual acetaldehyde
(British PET)
164

Residual AA in PET
wall thickness

still

sparkling

Figure 9.11 Bottle wall thickness and residual acetaldehyde in British still and sparkling
water PET
To test the hypothesis using PET materials from still water only, Nigerian still water
PET bottles that were allowed to stay with their contents for 13 months were used.
Pearsons correlation for the data showed strong statistically significant positive
correlation (r(8) = 0.778; p = 0.005. r2 = 0.60). The scatter plot for the relationship is
shown in Figure 9.12. The results obtained here proved that acetaldehyde
concentration in bottle wall increases as the thickness of the wall increases.
360

Bottle thickness (m)

340
320
300
280
260
240
220
200
0.4

0.6

0.8

1.2

1.4

1.6

1.8

Residual acetaldehyde (g/g)

Figure 9.12 Relationship between bottle wall thickness and residual acetaldehyde
(Nigerian PET)

165

9.9

Effect of carbonation on acetaldehyde quantitation

As explained in Subsection 5.9.7 experiments were carried out to ascertain whether

the carbon dioxide in sparkling water and soft drink has any effect on acetaldehyde
quantitation using headspace GC-FID. The results in Table 9.6 revealed that the
quantitation of acetaldehyde in sparkling water and soft drinks is not affected by
carbon dioxide content as the concentrations of acetaldehyde in both spiked still water
and sparkling water are approximately equal.
Table 9.6 Effect of carbonation on acetaldehyde quantitation
Replicate samples

Peak area
Still water

sparkling water

3.45

4.73

4.33

3.45

4.64

4.12

Mean

4.14

4.1

Standard deviation

0.62

0.64

9.10 Acetaldehyde in water and soft drinks

Acetaldehyde has been documented to migrate from the PET bottle wall into the
water and soft drinks (Nijssen et al 1996, zlem 2008), yet acetaldehyde is also added
to some soft drinks as a flavouring agent (Miyake and Shibamoto 1993, Food Safety
Commission

2005,

National

Toxicology

Program

2010).

Concentration

of

acetaldehyde found in fresh British and Nigerian bottled water and soft drinks in this
study are given in Table 9.7. In all cases the concentrations were within the range
reported in the literature (Miyake and Shibamoto 1993, Sugaya et al 2001, Nawrocki et
al 2002, Mutsuga et al 2006). The average concentrations found in this study in British
fruit juices, carbonated soft drinks, sparkling water and still water are shown in Figure
9.13. The average concentrations based on packaging types are given in Figure 9.14.
From Figure 9.13 it can be seen that acetaldehyde concentration is highest in fruit
juices followed by soft drinks, sparkling water and still water. Acetaldehyde was found
in soft drinks packaged in PET bottles as in soft drinks packaged in bottles made from
glass and carton (Figure 9.14). This confirmed the origin of the acetaldehyde in soft
drinks as the soft drinks themselves rather than migration from PET bottle. A juice
166

carton is usually made up of layers of paper, polyethylene, and aluminium foil. Again as
can be seen in Figure 9.14 and Table 9.7 acetaldehyde has not been detected in all
bottled water in glass bottles. Similar results were reported in the work of Sugaya et al
(2001). This provides evidence that the acetaldehyde detected in water bottled in PET

Acetadehyde concentration
(g/L)

originated from the PET materials.

6000
5113

5000
4000
3000
2000

1458

1000
22

0
Fruit juice

carbonated drinks sparkling water

still water

Figure 9.13 Average acetaldehyde concentrations by bottle content type (dotted line
shows EU SML)

Acetadehyde concentration (g/L)

7000

6450

6408

6000
5000
4000
3000
2095

2000

1798

1341

1033

1000
31

nd

nd

Figure 9.14 Average acetaldehyde concentrations by bottle content and packaging type
(dotted line shows EU SML)

167

Table 9.7 Acetaldehyde concentration in British and Nigerian bottled water and soft
drinks
Acetaldehyde concentration (g/L)
British bottled
British soft drinks
water
UPB5
72.25
UPC37
3116.48
UPB3
29.81
UPC38
9902.19
UPB6
29.33
UPC39
395.05
UPB4
54.65
UPC30
4952.19
UPB10
12.13
UPC33
330.76
UPB1
10.06
UPC40
14247.43
UGB9
nd
UPC32
7145.05
UGB3
nd
UPC41
3330.76
UGB10
nd
UCC42
383.14
UPA17
7.51
UCC43
12516.48
UPA5
15.32
UGC34
187.9
UPA7
nd
UGC36
1683.14
UPA15
12.29
UGC31
2152.19
UPA1
10.86
UOC35
2095.05
UPA19
33.79
UPC40
14249.81
UPA12
nd
UPC22
311.71
UPA6
2.74
UPC24
4752.19
UPA17
7.99
UPC25
1397.43
UPA3
2.58
UPC26
597.43
UGA8
nd
UPC23
1930.76
UGA3
nd
UGC22
109.33
UGC23
3142.67
UGC26
599.81
UGC27
278.38
nd not detected

Nigerian bottled
water
NPA5
nd
NPA8
6.4
NPA6
16.43
NPA2
14.04
NPA4
4.17

Nigerian soft
drinks
NPC16
178.38
NGC16
173.62

As mentioned earlier, guidelines and standards for acetaldehyde in drinking-water have

not been established. Consequently acetaldehyde in bottled water and soft drinks can
only be evaluated based on the EU specific migration limit (SML) for acetaldehyde from
packaging into foods (6mg/kg), tolerable daily intake of acetaldehyde (0.1mg/kg body
weight per day) and odour and taste threshold limits for acetaldehyde in drinking
water. The EU specific migration limit (SML) for acetaldehyde from packaging into
foods/water is 6mg/kg. In the case of water and soft drinks this is approximately
equivalent to 6mg/L considering the fact that one litre of water at 4C weighs 1Kg. It is
worth mentioning that SML is meant only for assessing the safety of food packaging
material in terms of contaminants migration. It is thus not a standard for assessing the

168

concentration of contaminants in food or water. It will be used here for comparative

purpose because it still refers to the concentration of contaminant that can be
achieved in foods and water. Additionally, the SML was designed to ensure that
exposure to acetaldehyde, as a result of intake of bottled water and soft drinks in PET
bottles, is below the tolerable daily intake (TDI) for acetaldehyde. In this study the
concentration of acetaldehyde found in 5 fruit juice samples was beyond the EU SML.
An important question worth asking is on the significance of the acetaldehyde EU SML
considering the fact that acetaldehyde can be added as flavouring agent in soft drinks at
concentration greater that the SML.
The Panel on Dietary Reference Intakes for Electrolytes and Water (2011) set the daily
recommended adequate intake for water for an adult man at 3.7 litres from all
beverages and foods. According to their estimate about 80 percent of people's total
water intake comes from drinking water and beverages with the other 20 percent
coming from food. Thus 2.96 litre of the recommended water intake comes from
drinking water and beverages. According to Ogden et al (2004) the average weight of
an adult American is 86.1kg. If the 2.96 litre of water taken by an adult on daily basis is
assumed to be totally in the form of soft drinks then the tolerable daily intake of
acetaldehyde will be exceeded as a result of intake of 2 out of 9 carbonated soft drinks
and 8 out of 15 fruit juices investigated in this study (Appendix 2). If only half of the
daily intake comes in the form of the soft drinks the tolerable daily intake of
acetaldehyde will be exceeded as a result of the intake of only 5 out of the 15 fruit
juices investigated in this study (Appendix 2). The odour and taste threshold limit for
acetaldehyde in bottled water is reported to be 2040 g /L (Nijssen et al, 1996;
Schrder, 2001). From this it can be seen that the lower level of the threshold value
was exceeded in 5 out of 26 bottled water brands analysed.
The average concentrations of acetaldehyde in Nigerian and British still water are
basically similar as can be seen in Figure 9.15. However the concentration of
acetaldehyde in a cola drink brand from Nigeria and Britain bottled in PET and glass
bottles look different (Figure 9.16). The concentrations for samples from the same
country are similar regardless of bottling material. The concentration in British samples
is about three times the concentration in Nigerian samples. This most likely implies the

169

use of different amounts of acetaldehyde in the production of this drink in the two

Acetadehyde concentration (g/L)

countries.
10

9.31
8.21

8
6
4
2
0

Nigerian
British
Still water in PET bottles

Acetadehyde concentration (g/L)

Figure 9.15 Acetaldehyde concentration in Nigerian and British still water

700
600
500

PET
Glass

400
300
200
100
0

NC16

UC26

Figure 9.16 Acetaldehyde content in a cola drink brand from Nigeria and Britain
bottled in PET and glass bottles.

Figure 9.17 shows the concentration of acetaldehyde in 5 British bottled water brands
marketed as still and sparkling water. The concentration of acetaldehyde in all but one
brand is higher in sparkling water than in the corresponding still water. As mentioned
earlier, carbonation, which lowers the pH of sparkling water by raising the hydrogen
170

ion concentration through formation of carbonic acid, enhances migration of

acetaldehyde from PET bottles into water. The mechanism through which carbonation
enhances the migration of acetaldehyde has not been determined. From these results it
follows that bottled water pH will be positively correlated with acetaldehyde

Acetadehyde concentration (g/L)

concentration.
80
70
60

Sparkling

50

Still

40
30
20
10
0
UP5

UP3

UP6

UP1

Figure 9.17 Acetaldehyde in 5 British brands of bottled water marketed as still and
sparkling water

9.11 Conductivity and acetaldehyde concentration

High solute concentrations in aqueous samples decrease the solubility of polar organic
volatiles in the sample matrix and promote their transfer out of the sample. This is the
basis for addition of sodium chloride to samples in vials in the determination of
acetaldehyde in water using headspace GC. As a result of this it was hypothesised that
conductivity, which approximate the measure of total concentration of inorganic
substances in water, will be inversely associated with acetaldehyde concentration in
bottled water. Pearsons correlation showed strong statistically significant negative
correlation between conductivity and acetaldehyde correlation in bottled water (r(9) =
-0.722; p = 0.028. r2 = 0.52). The scatter plot for the relationship is shown in Figure
9.18. The results obtained here proved that acetaldehyde concentration in bottled
water decreases as water conductivity increases.

171

Water conductivity (S/cm)

800
700
600
500
400
300
200
100
0
0

10

15

Acetaldehyde concentration (g/L)

Figure 9.18 Relationship between conductivity and acetaldehyde concentration in
bottled water

9.12 Storage and acetaldehyde migration into bottle contents

As stated in Chapter 6, purchased bottled water and soft drinks are mostly consumed
within 30 days of purchase in both Nigeria and Britain. In both countries only a very
small fraction of respondents reported consuming contents in periods greater than 3
months (Britain <1%, Nigeria 6%). Nevertheless storage period of up to one year has
been reported. In this study bottled water and soft drinks were stored for up to 25
months at room temperature and the change in acetaldehyde concentration with time
was studied. Table 9.8 and Figure 9.19 show the acetaldehyde content in fresh British
bottled water samples and samples stored for 20 months. Table 9.9 shows the
acetaldehyde content in fresh Nigerian bottled water samples, samples stored for 12
months and samples stored for 25 months. Figure 9.20 shows change in acetaldehyde
concentration in a Nigerian soft drink stored for 12 months. From Table 9.8 and
Figure 9.19 it can be seen that acetaldehyde concentration in 5 out of 8 samples
dropped after 20 months storage. Acetaldehyde content increase in sample UPA3 but
remain constant in sample UPB6. No acetaldehyde could be detected in both fresh and
stored samples of brand UPA12. Unlike antimony which accumulates continuously in
bottled content acetaldehyde migrating into content can be degraded by bacterial flora
in the content as reported by Matsuga et al (2006). Heterotrophic bacteria may exist in
British bottled water because the European Union Directive 2009/54/EC prohibited
disinfection of natural mineral water. Acetaldehyde is not likely to be regenerated in

172

the bottle material as regeneration usually occurs only at high temperatures. Nawrocki
et al (2002) reported observing gradual increase in the content of acetaldehyde and
formaldehyde in carbonated water over a period of 89 months followed by a decline.
Additionally in this study acetaldehyde concentration has been shown to fall if
acetaldehyde solutions are stored in polypropylene bottles (to be discussed later).
Nawrocki et al (2002) associated the gradual decrease of aldehydes concentration they
observed with the gradual loss of dissolved CO2. The decrease in most of the British
samples in this study is most likely related to depletion of the acetaldehyde content in
the bottle material. The absence of acetaldehyde in both fresh and stored samples of
brand UPA12 is probably related to bacterial content in the water as this PET material
has residual acetaldehyde content comparable to other still water PET materials.
Table 9.8 Change in acetaldehyde content in stored British bottled water
Bottled
water sample
UPB6
UPB4
UPB5
UPA19
UPA12
UPA3
UPA15
UPA17
nd not detected

Acetaldehyde
concentration (g/L)
fresh
20 months
29.33
32.67
54.65
26.3
72.25
14.04
33.79
17.7
nd
nd
2.58
6.24
12.29
5.76
7.99

3.21

Acetadehyde concentration
(g/L)

80
70
60
50

fresh

40

20 months

30
20
10
0

Figure 9.19 Change in acetaldehyde content in British bottled water stored for 20
months
173

Most Nigerian samples are table bottled water and even for spring water and natural
mineral water a regulation similar to the European Union Directive 2009/54/EC may
not be in existence. Consequently most Nigerian bottled water is disinfected. Increase
in acetaldehyde concentration was therefore expected. However other factors that
may determine the stability of acetaldehyde in the water include oxygen content and
temperature. From Table 9.9 an increase in acetaldehyde concentration is observed
with most samples. Importantly while the concentrations generally lie within the odour
and taste threshold limit for British sparkling water the concentration failed to reach
the lower border in British still water and in all Nigerian water samples.
Table 9.9 Change in acetaldehyde content in stored Nigerian bottled water
Bottled water sample
NPA1
NPA2
NPA3
NPA4
NPA5
NPA6
NPA7
NPA8
NPA9
NPA10
NPA11
nd not detected

Acetaldehyde concentration (g/L)

fresh
12 months
25 months
7.35
5.92
14.04
15.79
18.98
3.69
4.17
7.19
nd
4.49
12.13
16.43
3.85
4.81
6.4
10.38
nd
8.95
nd

From Figure 9.20 the concentration of acetaldehyde increased by 586% in the fresh
Nigerian cola drink (NPC13) after storage period of 12 months. zlem (2008) has
reported similar observation after storing carbonated drink sample for 6 months.
However the actual amount of acetaldehyde in the 28.59g bottle of 500ml capacity
with a residual acetaldehyde concentration of 3.09g/g is 88.4g. Even if all the residual
acetaldehyde migrate into the bottle content the concentration in the bottle content
will only increase by 178.6 g/L (about 100%). Unless acetaldehyde was being
regenerated in the PET bottle material or from degradation of the soft drink the
source of the increase could not be ascertained. The method used in acetaldehyde
determination could not have been responsible for the increase, for it were it would
have affected both fresh and stored samples to the same extent.

174

Acetadehyde concentration (g/L)

1400
1223.62
1200
1000
800
600
400
200

178.38

fresh

12 months

Figure 9.20 Change in acetaldehyde content in a fresh Nigerian cola drink brand
(NPC13)

9.13 Acetaldehyde in PET materials and in bottle contents

For British water bottles as bottle wall thickness increases the acetaldehyde
concentration in the wall matrix decreases (Section 9.8). This observation was in part
attributed to sparkling waters greater ability to extract acetaldehyde from PET bottle
wall than still water. Sparkling water is contained in bottles which are thicker than still
water bottles. The results in Figure 9.21 provide further evidence in support of this
argument. The Figure revealed that even though the acetaldehyde content is lower in
the thicker sparkling water bottles than in the thinner still water bottles the
acetaldehyde content is higher in sparkling water than in still water. From Figure 9.21
the mean acetaldehyde concentration in sparkling water is higher than the odour and
taste threshold level for acetaldehyde in water To avoid the acetaldehyde
concentration going above the odour and taste threshold in bottling of water, bottles
with lower acetaldehyde contents should be used for sparkling water since sparkling
water has greater ability to extract acetaldehyde than still water.

175

Acetaldehyde concentration

Residual AA in PET
AA in bottle content

43.8g/L

6.9g/L
3.7g/g
1.9g/g

Still

sparkling

Figure 9. 21 Relationship between acetaldehyde concentration in PET and its presence

in bottled water

9.14 Acetaldehyde outgassing from PET bottles

New plastics materials including those used in car interiors, shower curtains, and
flooring initially outgas chemicals into the air which accounts for the distinctive smell
associated with some of these products when new. Except for flavoured water
chemicals other than carbon dioxide are usually not added to bottled drinking water.
Consequently no additive-associated smell is expected to come from used PET bottles.
In this study some used PET bottles were found to be releasing a gas with fruity smell
which was suspected to be acetaldehyde vapour. To characterise and quantify the gas,
bottles from 3 brands of bottled water were studied as explained in Subsection 5.9.11.
The gas was confirmed to be acetaldehyde and the results obtained are shown in
Figure 9.22. From the Figure it can be seen that sample UPB5 which is a green bottle
for sparkling water releases acetaldehyde with the concentration increasing with time.
A similar bottle aged for 266 days was not releasing acetaldehyde. Also 2 bottle
samples used for still water (UPA12 and UPA15) were not releasing acetaldehyde. The
residual acetaldehyde content for UPB5 and UPA12 is shown in Table 9.4. Even though
UPB5 has higher residual PET than UPA12 the observation is believed to be more
related to the usage of the bottle for sparkling water than with the residual
acetaldehyde concentration. Since only 3 bottles were studied it cannot be said with
certainty whether this phenomenon is more associated with sparkling water bottles
than with still water bottles

176

Acetadehyde concentration (g/L)

14

11.43

12
10

8.66

6.23

6.84

6
4
2

nd

nd

nd

0
UPB5 10
days

UPB5 20
days

UPB5 30
days

UPB5 240
days

Aged UPB5 UPA12 30

10 days
days

UPA15 30
days

Figure 9.22 Outgassing of acetaldehyde from empty PET bottles (nd not detected)

9.15 Acetaldehyde migration from water medium into PET

From the literature and results discussed earlier acetaldehyde has been established to
migrate from PET bottle wall into the surrounding environment. To verify whether
PET material could absorb acetaldehyde PET materials from samples UPB5 and UPA12
were soaked in acetaldehyde solutions of varying concentrations for 24 hours as

Acetadehyde concentration (g/L)

explained in 5.9.12.
100000
10000

UPA12
UPB5

1000
100
10
1
None

0.1mg/ml

10mg/ml

Neat

Acetaldehyde solution strength

Figure 9.23 Absorption of acetaldehyde by PET

The results for these experiments are shown in Figure 9.23. From the results it can be
seen that PET material can absorb acetaldehyde in solution and the amount absorbed

177

increases with the concentration of the acetaldehyde solution. The results also
revealed that at low concentrations the quantity of acetaldehyde absorbed varies
directly with the concentration of the residual acetaldehyde. However at higher
concentrations absorption pattern becomes similar. In the case of the two PET
materials here at acetaldehyde concentration of 0.1mg/ml the absorption was about
100% of the concentration of residual acetaldehyde in each of the 2 materials. But the
amount absorbed became roughly equal at acetaldehyde concentration of 10mg/ml and
for neat acetaldehyde.
The lowest concentration used in this study (0.1mg/ml = 100mg/L) has been reported
to be obtainable in some fruit juices. According to Food Safety Commission of Japan
(2005) acetaldehyde is contained in fruit juices at concentrations as high as 230mg/L. In
this study the highest concentration found was 14.25mg/L (Table 9.7). The implication
of this finding is that PET materials from soft drinks and fruit juices cannot be used to
assess actual residual acetaldehyde in PET since the materials could absorbed as much
acetaldehyde from their contents as the residual acetaldehyde contained in them. As
observed in this study the residual acetaldehyde concentrations found in soft drink and
fruit juices bottles are higher than the residual concentrations in water PET bottles.
Acetaldehyde in soft drinks (g/L)

5000
4500
4000
3500
3000
2500
2000
1500
1000
500
0
3

11

13

Residual acetaldehyde in PET (g/g)

Figure 9.24 Relationship between residual acetaldehyde in carbonated drinks PET and
the acetaldehyde concentration in the carbonated drinks
This is most likely as a result of acetaldehyde migration from soft drinks and fruit juices
into the PET matrix. Further evidence giving some indication of acetaldehyde migration
from content into PET is shown the scatter plot in Figure 9.24. In the Figure residual

178

acetaldehyde in carbonated drinks PET has been shown to strongly correlate with the
acetaldehyde concentration in the contents (r(6) = 0.895; p = 0.016. r2 = 0.802).
However for unknown reason no statistically significant correlation could be observed
between the residual acetaldehyde in fruit juice PET and the acetaldehyde
concentration in the fruit juices.

9.16 Stability of acetaldehyde with storage

Even though acetaldehyde is generated from PET bottle materials into bottle content it
has also been reported to degrade in water (Nijssen et al, 1996). To study the stability
of acetaldehyde in water three acetaldehyde solutions (50, 500 and 5000g/L) in boiled
and unboiled water in polypropylene bottles were stored in a refrigerator (about 5C)
and at room temperature (Subsection 5.9.13). The expectation was that acetaldehyde
loss due to degradation and evaporation will be higher for acetaldehyde solution in
unboiled water and water stored at room temperature. This is because the volatility of
acetaldehyde dissolved in water increases as temperature increases and acetaldehyde
stability in water increases in the absence of oxygen which is flushed out during boiling.
Also according to Matsuga et al (2006) microorganisms in unsterilized water can
remove acetaldehyde from the water by degradation. The expected pattern in terms of
acetaldehyde reduction with time is given below:
BF < BO < NBF < NBO
Decreasing acetaldehyde concentration with time
B stands for boiled, F stands for stored in refrigerator, NB stands for not boiled
and O stands for stored at room temperature.
For acetaldehyde at 50g/L the concentration reduced to below detection limit for all
storage conditions after 4 days. This finding agrees with the findings of Nijssen et al
(1996) where acetaldehyde concentration was reduced from 100g/L to less than
1g/L within 8 days of storage. The results for solutions with higher acetaldehyde
concentrations are given in Figures 9.25 and 9.26. For all unboiled solutions and boiled
solution stored in refrigerator acetaldehyde concentration reduced gradually with
time. Additionally for these samples the observation agreed to some extent with the
expected results. For 500g/L solutions lowest concentration was recorded in

179

unboiled samples stored at room temperature, followed by unboiled sample stored in

refrigerator and then boiled sample stored in refrigerator. For 5000g/L solutions
lowest concentration was recorded in unboiled samples stored at room temperature.
However the concentration is similar for unboiled sample stored in refrigerator and

Aceatldehyde concentration (g/L)

boiled sample stored in refrigerator.

700

Acetaldehyde in boiled water stored in

fridge

600

Acetaldehyde in boiled water stored at

room temperature

500

Acetaldehyde in unboiled water stored in

fridge

400

Acetaldehyde in unboiled water stored at

room temperature

300
200
100
0
0

10

15

20

25

30

35

Days of storage
Figure 9.25 stability of 500g/L aqueous acetaldehyde solution
At both concentrations (500 and 5000g/L) boiled solutions stored at room
temperature showed abrupt decrease in acetaldehyde. For these solutions the
acetaldehyde concentrations reduced to levels below detection limit and 977.9g/L
after 8 and 12 days respectively. The reason why boiled samples stored at room
temperature behaved in this manner remains a mystery. Nijssen et al (1996) earlier
reported similar observation with glass bottles. According to these authors the rapid
decrease in acetaldehyde in boiled water stored in glass bottles may be caused by the
higher reactivity of the silicium oxides of the bottle wall. This is however not applicable
to the observation in this research as polypropylene rather than glass bottles were
used for the study. The observation may be related to greater solubility of oxygen in
water than nitrogen. Due to the higher solubility of oxygen in water than nitrogen the
ratio of oxygen to nitrogen is greater in water than the same ratio in air. As a result of
boiling both oxygen and nitrogen are expelled from water. However as water cools
the oxygen concentration rises faster than the less soluble nitrogen resulting in its
transient higher concentration in proportion to nitrogen. Even though this should also

180

happen in boiled samples stored in refrigerator, the rate of degradation may still be
lower due to lower temperatures. The results of these experiments established that

Aceatldehyde concentration (g/L)

acetaldehyde concentration in water reduces if it is not been replenished.

Acetaldehyde in boiled water stored in
fridge
Acetaldehyde in boiled water stored at
room temperature
Acetaldehyde in unboiled water stored in
fridge
Acetaldehyde in unboiled water stored at
room temperature

6000
5000
4000
3000
2000
1000
0
0

10

15

20

25

30

35

Days of storage
Figure 9.26 stability of 5000g/L aqueous acetaldehyde solution

9.17 Summary
The concentration of residual acetaldehyde in PET bottle materials and the
concentration of dissolved acetaldehyde in bottled water and soft drinks from Nigeria
and Britain were determined. The influence of some parameters on the acetaldehyde
concentration in PET materials and in bottle contents was also assessed. Accuracy and
precision were good as recoveries were around 100% and coefficients of variation
were less than 15% for all analysis types.
All Nigerian and British PET bottle materials contained residual acetaldehyde within
the concentration range reported in the literature. The total average and the average
for still water, sparkling water and soft drinks PET materials for British samples were
4.76, 4.10, 1.55 and 6.70 g/g respectively. The average concentration for Nigerian soft
drink PET materials was 2.17g/g. Acetaldehyde concentration in PET material was
found to decrease as the bottle material ages. The concentration was also established
to be directly proportional to thickness of the bottle wall. Higher acetaldehyde
concentration was observed in the thinner British still water bottles compared to the
thicker sparkling water bottles. This observation was attributed to the sparkling
181

waters greater ability to extract acetaldehyde from the bottle wall matrix than still
water.
The average concentrations of acetaldehyde in fresh British fruit juices, carbonated soft
drinks, sparkling water and still water were 5112.5, 1457.75, 21.8 and 7.84g/L
respectively. While acetaldehyde was detected in all fruit juices and carbonated drinks
regardless of packaging, it was only detected in bottled water packaged in PET bottles.
The average concentration in Nigerian still water and the concentration in a Nigerian
carbonated cola drink are 8.21 and 176 g/L respectively.
Acetaldehyde content of some soft drinks was found to be beyond the EU specific
migration limit of acetaldehyde from PET bottles. The study also found that tolerable
daily intake of acetaldehyde in human could be exceeded as a result of exclusive
consumption of some fruit juices and carbonated soft drinks analysed. Additionally, the
odour and taste threshold limit for acetaldehyde in bottled water has been exceeded
in 5 out of 26 bottled water brands analysed. The presence of carbon dioxide in
bottled water and soft drinks have no effect on acetaldehyde determination in those
samples. Acetaldehyde concentration in bottled water

showed an inverse

proportionality to water conductivity. The concentration of acetaldehyde decreased in

most British bottled water sample after storage for 20 months. Conversely an increase
was observed in a Nigerian carbonated drink sample and most Nigerian still water
sample after storage for 12 and 25 months.
Some empty bottles were established to have the ability to release acetaldehyde
vapour continuously as was observed with a green sparkling water bottle. However a
similar bottle aged for 266 days was not releasing acetaldehyde. Also two new still
water bottles observed were not releasing acetaldehyde. It has also been ascertained
that PET material do absorbs acetaldehyde when soaked in solution containing
acetaldehyde at concentrations obtainable in soft drinks. Stability of acetaldehyde in
water studies revealed that acetaldehyde dissolved in water diminishes with time even
at low temperatures, so far as it is not been replenished.

9.18 Conclusion
In this Chapter acetaldehyde concentration in PET bottle materials and in bottle
contents were analysed. From the results acetaldehyde concentration in bottled water

182

is low and may only pose taste and odour problem to bottled water rather than been a
safety concern. However acetaldehyde in soft drinks could pose safety problems as in
some cases intake of acetaldehyde as a result of consumption of the soft drinks can
exceed the tolerable daily intake of acetaldehyde in humans. Acetaldehyde
concentration in water and in PET material has been shown to reduce with time. The
decrease in acetaldehyde concentration in PET implies that aged PET bottles are safe
for reuse.

183

CHAPTER 10: CONCLUSION AND FURTHER WORK

10.1 Introduction
This research aimed to assess the pattern and extent of antimony and acetaldehyde
migration from British and Nigerian polyethylene terephthalate bottles into bottle
contents under typical use and reuse conditions and to relate the migration patterns to
current regulations and controls. This research aim was further subdivided into four
objectives which were;
1. To examine the pattern and extent of bottled water and soft drink use and PET
bottle reuse in Britain and Nigeria.
2. To assess the extent of antimony and acetaldehyde migration from PET into
water and soft drinks under conditions representative of real use and reuse in
Britain and Nigeria
3. Drawing on findings from 1 and 2, as well as existing regulations and controls
to assess whether acceptable limits of chemicals in foods and water are being
exceeded as a result of the chemical migrations
4. To generate recommendations about the extent to which existing regulations
and controls might merit re-examination
The first objective was further divided into 3 sub-objectives namely;
i.

To establish the pattern and extent of typical bottled water and soft drink use
and PET bottle reuse including, the approximate proportion of individuals/
households using and reusing plastic bottles, bottled water and soft drinks
storage periods prior to use, bottle sizes most often used and reused, bottle
reuse periods, the number of bottles being used and reused by an individual
or a household at any one time, etc.

ii.

To determine public perceptions of the safety of reusing plastic containers to

store drinking water, beverages, etc

iii.

To find out factors influencing peoples preferences with respect to reuse of

plastic water containers

This Chapter reviews the steps taken through the entire research process, states how
the approach in this research differed from the approaches in previous works,
identifies the strengths and limitations associated with the research, and specifies the
184

main findings of the research in relation to the objectives and the recommendations
for action. The Chapter concludes by exploring potential areas for further research.

10.2 Main findings in relation to objectives

10.2.1 Survey on bottled water and soft drink use and PET bottle reuse
10.2.1.1

Pattern and extent of bottled water and soft drinks use and PET
bottle reuse

British respondents use more bottled water and soft drinks than Nigerian respondents,
however Nigerian respondents store unopened bottles for longer durations before
use. An important implication of this observation is that attributes associated with
consumption of bottled water and soft drinks will manifest differently in the two
countries. For example risk of dental erosion, which has been associated with the
consumption of soft drinks, may manifest more in Britain than in Nigeria. Conversely,
the likelihood of exposure to chemicals migrating from bottle wall into contents may
be more in Nigeria as a result of longer storage. Importantly, this study has established
that storage of bottle contents prior to use, to periods beyond their best before
dates is small in both countries. This also implies that for both countries the
possibility of accumulation of chemicals beyond international standards is likely in only
few cases where storage periods are extended.
Bottled water and soft drinks storage pattern in terms of storage places are similar in
both countries. Unrefrigerated storage in the summer will not have much implication
in Britain and other temperate countries in terms of temperature elevation of bottle
content. In Nigeria where ambient temperature can approach 40C, storage outside
the refrigerator can result in the elevation of temperature of bottle contents. This
issue can even be more interesting if North African and some Middle Eastern
countries, which are usually hotter than Nigeria, are considered. As mentioned earlier
elevation of temperature enhances leaching of chemicals from PET bottle wall into the
contents.
In spite of the lack of information on bottle reuse in the literature the results in this
study together with the earlier results from the US shows that PET bottle reuse is high
and practised to the same extent in both developing and developed countries. These
results are interesting in that reuse was initially hypothesised to be much higher in
developing countries than in developed countries. This study showed that PET bottles

185

are reused for longer periods in Nigeria than in Britain. A similar extent of reuse in
both countries with longer reuse duration in Nigeria than in Britain is believed to be
due to greater bottle availability in Britain. Reuse of a larger number of bottles for
storage of drinking water in Nigerian places of residence than in British places of
residence probably results from the lower availability of drinking tap water and lower
consumption of bottled water in Nigeria.

10.2.1.2

Bottle reuse, safety perception and safety debate

Perception of bottle reuse as an unsafe practice by respondents from both countries

gives some indication about the extent of this concern in both developing and
developed countries. A small proportion of British respondents believe that cancer
causing chemicals migrate from PET bottles into contents. The extent to which
chemical migration risk is overstated by the media and internet is believed to have
some influence on the bottle reuse safety perception observed in this study. The
greater concern about chemical migration and risk of cancer from bottle reuse in
Britain than in Nigeria could be connected to greater health/environmental awareness
in Britain than in Nigeria. In particular greater access to internet is believed to play an
important role.
The level of safety concern from both reusing and non-reusing respondents in both
countries was found to be similar. The similar levels of safety concern imply that safety
concern is not an important determinant of bottle reuse. It cannot be ascertained why
the existence of this perception is not affecting the extent of bottle reuse.

10.2.1.3

Factors influencing bottle reuse

The most important factors this study identified as affecting bottle reuse are
convenience associated with PET bottle reuse, saving money, concern for
environment, bottle age and original use of bottle. Economic status and climate for a
country did not show significant relationship with the proportion of people reusing
PET bottles; however these two factors may have some influence on reuse duration
and also the number of bottles being reused in households. For British respondents
the single most important motivation for reusing bottle is the desire to preserve the
integrity of the environment followed by the need to save money. For the Nigerian
respondents convenience associated with reuse of PET bottle was the most important
motivating factor followed by the need to save money. The emergence of

186

environmental concern as the most important factor influencing reuse in Britain is a

reflection of greater global environmental awareness in developed countries than in
developing countries.

10.2.2 Antimony and acetaldehyde in PET and their migration into bottle
content
The second objective assesses the extent of antimony and acetaldehyde migration from
PET into water and soft drinks under typical use and reuse conditions. The
concentration of antimony and acetaldehyde in PET and other materials related to
bottling was initially assessed. This is because antimony and acetaldehyde can only
migrate into contents if they are present in the PET materials. This was followed by
assessment of their concentrations in freshly purchased bottled water and soft drinks
samples to establish baseline concentrations for fresh samples. Migration was assessed
under typical use and reuse conditions and also under some extreme conditions.
In this study only the bottle material in the form of PET was found to contain antimony
in quantities that could leach into the bottle contents. Plastic bottle caps and cap liners
were eliminated as source of antimony that can migrate into bottle content. Bottle
caps and cap liners were not expected to contain antimony as they are not
manufactured using antimony catalyst. All Nigerian and British PET bottle materials
contained antimony within the concentration range reported by industry sources
confirming the use of antimony catalyst in their synthesis rather than germanium,
titanium, cobalt, manganese, magnesium or zinc-based catalysts. Antimony was found in
some soft drinks bottled in glass; however, antimony content in glass materials was not
quantified.
Antimony was detected in all fresh soft drinks and bottled water samples. The
antimony concentration in fresh soft drinks samples was higher than in bottled water.
This difference is in large part believed to be due to presence of antimony in the
constituent materials used for making soft drinks. Bottled water and soft drinks with
higher conductivity tend to have higher concentrations of antimony, however little or
no relationship exists between pH and antimony concentration in freshly purchased
samples. But lower pH values (acidic) appeared to be associated with higher antimony
concentration in storage experiments. Concentrations of the trace elements
investigated (Cd, Ge, Zn, Al, Be, Ti, Co and Pb) were higher in soft drinks than in

187

bottled water and tap water. Similarly concentrations of all the elements except
antimony were higher in glass bottled contents than in PET bottled contents. High
levels of titanium were detected in soft drinks from both countries. Titanium is
believed to be in the soft drinks as a result of its usage as food additive.
An increase in antimony concentration with time was observed in storage experiments
in PET bottles. Similarly, increase in antimony concentration was observed in
temperature exposure experiments in PET bottles. Glass bottles demonstrated much
lower release of antimony in comparison to PET bottles. Release of lead was observed
from a Nigerian glass bottle. Antimony migration into water was found to be directly
proportional to the antimony concentration in PET and to bottle thickness for some
exposure conditions. Aged bottles leach lower amounts of antimony than new.
Similarly larger bottles leach lower amount of antimony than smaller bottles.
Glass materials and plastics other than PET identified in this study were not expected
to release acetaldehyde. However, none of these materials were analysed for
acetaldehyde. All Nigerian and British PET bottle materials contained residual
acetaldehyde within the concentration range reported in the literature. Acetaldehyde
concentration in PET material was found to decrease as the bottle material ages. A
phenomenon explaining this was observed in the form of acetaldehyde outgassing from
some PET bottles. The concentration in PET materials was also established to be
directly proportional to thickness of the bottle wall. Higher acetaldehyde
concentrations were observed in the thinner British still water bottles compared to
the thicker sparkling water bottles. This observation was attributed to the sparkling
waters greater ability to extract acetaldehyde from the bottle wall matrix than still
water. Higher acetaldehyde concentrations were found in soft drinks PET materials
than in other PET materials. It was suspected and confirmed that soft drink PET
materials absorbs acetaldehyde from the soft drinks. The implication of this finding is
that PET materials from soft drinks and fruit juices cannot be used to assess actual
residual acetaldehyde in PET since the materials could absorbed as much acetaldehyde
from their contents as the residual acetaldehyde contained in them.
Acetaldehyde was detected in all fruit juices and carbonated drinks regardless of
packaging type. However acetaldehyde was detected only in bottled water packaged in
PET bottles. Concentration of acetaldehyde in soft drinks was so high that it could only
188

be attributed to the use of acetaldehyde as flavouring agents in the soft drinks.

Acetaldehyde concentration in bottled water decreases with increase of water
conductivity.
The concentration of acetaldehyde decreased in most British bottled water sample
after storage for 20 months. Conversely an increase was observed in a Nigerian
carbonated drink sample and most Nigerian still water sample after storage for 12 and
25 months. Studies of the stability acetaldehyde in water revealed that the
concentration of acetaldehyde dissolved in water diminishes with time even at low
temperatures, if acetaldehyde is not been added from another source.
As mentioned earlier the most important concern in bottle reuse is the safety of the
bottles with regard to release of chemicals into contents. This study has confirmed
that age of bottle decreases the concentration of antimony and acetaldehyde in bottle
wall and their leachability from the bottle wall into the content. This is attributed to
depletion of the migrants in the bottle wall matrix as a result of migration.

10.2.3 Migration in relationship to usage pattern and existing regulations

This component of the objective seeks to draw on from findings from the first and
second objective, existing regulations on chemical migration and literature to assess
whether acceptable limits of chemicals in foods and water are being exceeded as a
result of the chemical migrations.

10.2.3.1

Antimony and other trace elements in fresh samples

The EU maximum admissible concentration for antimony is 5g/L (European

Commission, 2003; EPA, 2010). The EU specific migration limit (SML) for antimony
from PET into foods is 40g/kg of food (EFSA, 2004). Since one litre of water weighs
approximately one kilogram (Lide, 1990), the SML for water can be presented as
40g/L. Detectable levels of antimony were found in tap water and all the 47 freshly
purchased British bottled water and soft drinks samples analysed. However the
concentration exceeded the EU maximum admissible concentration only in one fruit
juice sample from a PET bottle. The concentration found in the fruit juice was 6.6g/L.
Concentrations were below the EU SML in all samples. The concentrations of lead,
cadmium and beryllium are also all within the EU MAC and/or US MCL. Guidelines and
standards for cobalt, titanium, germanium and zinc in drinking-water have not been

189

established. In the US secondary non-enforceable guidelines regulating contaminants

that may cause cosmetic or aesthetic effects in drinking water covers zinc and
aluminium. The concentration of zinc in all the samples falls within the US MCL. The
concentration of aluminium in two freshly purchased British soft drinks is however
greater than the British MAC.

10.2.3.2

Antimony in samples exposed to different conditions

Concentration of antimony in some Nigerian bottled water and soft drinks was above
the EU MAC after 11 months of storage at room temperature. For 10 British bottled
water samples the concentration remained below the EU MAC even after 19 months
of storage. A Nigerian glass bottle for soft drink leached both antimony and lead above
EU MAC after 11 months. Lead concentration in the contents of the same glass bottle
was above EU MAC after 2 months of storage. However 2 British glass bottles
subjected to antimony migration test at elevated temperatures demonstrated low level
of antimony and lead migration. Antimony concentrations in water exposed at 40, 60
and 80C for up to 48 hours in PET and glass bottles remained below the EU specific
migration limit for antimony from plastic materials and other articles intended to come
in contact with food. At realistic temperatures of 40 and 60C antimony concentration
in the water remained below the EU MAC even after 48 hours of exposure but the
concentration exceeded the EU MAC for most exposures at 80C.

10.2.3.3

Acetaldehyde in fresh samples

In the absence of guidelines and standards for acetaldehyde in drinking-water,

acetaldehyde in bottled water and soft drinks was evaluated based on the EU specific
migration limit (SML) for acetaldehyde from packaging into foods (6000g/kg),
tolerable daily intake of acetaldehyde (0.1mg/kg body weight per day) and odour and
taste threshold limits for acetaldehyde in drinking water. For water and soft drinks the
SML for acetaldehyde can be approximated to 6000g/L given that one litre of water at
4C weighs 1kg. In this study the concentration of acetaldehyde found in 5 soft drinks
samples was beyond the EU SML. An important question worth asking is on the
significance of the acetaldehyde EU SML considering the fact that acetaldehyde can be
added as flavouring agent in soft drinks at concentration greater than the SML. If the
2.96 litre of water taken by an adult on daily basis is assumed to be totally in the form
of soft drinks then the tolerable daily intake of acetaldehyde will be exceeded as a

190

result of intake of 2 out of 9 carbonated soft drinks and 8 out of 15 fruit juices
investigated in this study. If only half of the daily intake comes in the form of the soft
drinks the tolerable daily intake of acetaldehyde will be exceeded as a result of intake
of only 5 out of the 15 fruit juices investigated in this study. The odour and taste
threshold limit for acetaldehyde in bottled water is reported to be 2040 g/L (Nijssen
et al, 1996; Schrder, 2001). From this it can be seen that the lower level of the
threshold value was exceeded in 5 out of 26 bottled water brands analysed.

10.2.3.4

Acetaldehyde in samples exposed to different conditions

The concentration acetaldehyde decreased in most British bottled water sample after
storage for 20 months. An increase was observed in a Nigerian carbonated drink
sample and most Nigerian still water sample after storage for 12 and 25 months.
Importantly while the concentrations generally lie within the odour and taste threshold
limit for British sparkling water the concentration failed to reach the odour and taste
threshold limit in British still water and in all Nigerian water samples. Acetaldehyde
concentration increased by 586% in the fresh Nigerian cola drink after storage period
of 12 months. However the concentration still failed to reach the EU SML for
acetaldehyde. Acetaldehyde could increase in bottled water and soft drinks with
storage. Nevertheless, based on the residual acetaldehyde found in PET in this study
and in the literature and also the level of migration observed, acetaldehyde increase in
bottled contents will not reach concentrations at which tolerable daily intake of
acetaldehyde can be exceeded as a result of consumption of the bottled contents. For
bottled water the observed increase is only of importance in terms of the acetaldehyde
odour and taste threshold limit in water.

10.2.4 Re-examination of existing regulations and controls

The fourth objective explores the extent to which existing regulations and controls
might merit re-examination. Acetaldehyde associated with alcoholic beverages (derived
from the alcoholic beverages and from their metabolism in the body) has been
concluded to be carcinogenic to humans (Group 1) in a recent IARC review of human
carcinogens. In this study acetaldehyde has been identified in soft drinks at
concentrations higher than the European Union specific migration limit for
acetaldehyde from food packaging materials into foods and that the tolerable daily
intake of acetaldehyde could be exceeded as a result of intake of some soft drinks.

191

Consequently the absence of international guidelines for acetaldehyde in water and

foods is in need of reassessment.

10.3 Other findings

In addition to addressing the primary objectives of the research, the research has
other findings which were not anticipated.

10.3.1 Use of PVC as cap liner in Nigeria

The characterisation of materials associated with bottling is a secondary issue not
initially targeted by the research. However as the research involves working with all
bottle materials having contact with bottle contents it became worthwhile to
characterize them. The characterisation exercise has revealed useful information
relating to use of different materials in bottling in the two countries of interest. The
characterisation has revealed the use of EVA/PP copolymer and plasticised PVC as
bottle cap lining material for glass bottles by the same multinational bottling company
in Britain and Nigeria respectively. While EVA/PP has not been associated with any
health risk, plasticised PVC is a plastic material associated with health risk issue.

10.3.2 New PET digestion method

In the course of determination of antimony in PET and other bottling materials a new
simpler method of PET material digestion was developed. The developed method has
the potential to make the procedure for determination of antimony simpler and
cheaper because the procedure requires a microwave oven for household use rather
than purpose-built laboratory microwave digestion system.

10.3.3 Material minimization in PET bottle manufacture

The research has revealed the use of bottles with thinner walls in Britain than in
Nigeria. The minimization of bottle wall thickness translate into utilisation of fewer raw
materials to make bottles and generation of less waste associated with bottled water
and soft drinks. Thicker bottle walls in Nigeria may also facilitate the longer re-use
periods associated with this country.

10.3.4 Elevation of water temperature on sunny summer day

The research has shown the temperature elevation pattern in bottle contents exposed
to sunlight and the maximum temperatures that could be attained by bottle contents
on exposing bottles with the contents to brilliant sunlight on a British summer day. As
192

mentioned earlier migration of chemicals from bottles into contents is directly

associated with temperature.

10.4 Implication of results

10.4.1 Plasticized PVC cap linings in Nigeria
The ban on the use of 6 phthalates including DEHP at concentrations greater than
0.1% by mass in the manufacture of plastic toys and childcare article, the consideration
of a bill in the State of California in the US for banning of use of PVC in consumer
packaging and the efforts to eliminate PVC from products and packaging by major
corporations is a clear indication of potential for harm associated with use of PVC
(usually made up of 30 40% plasticizers), especially in consumer packaging. As said
earlier phthalates were reported to be associated with allergies in children, decrease in
anogenital distance among male infants exposed before birth, inducement of less male
typical play behaviour in boys and other manifestations related to mimicry of human
hormones. In developing countries like Nigeria regulations guiding the use of materials
in packaging may either be non-existent or where they exist enforcement may be
poor. In many situations local and multinational companies, which are profit-oriented
organisations, exploit the situation to use materials that have potential to cause harm
in consumers. The use of PVC cap lining in Nigeria but not in Britain by the same
multinational company is believed to be one of such situations where the poor
enforcement of safety regulations is being exploited. The National Agency for Food
and Drug Administration and Control (NAFDAC) is the Nigerian agency vested with
the task of safeguarding public health by ensuring that only the right quality drugs, food
and other regulated products are manufactured, imported, exported, advertised,
distributed, sold and used. The PVC cap lining issue in Nigeria will need to be
investigated by NAFDAC and appropriate actions need to be taken.

10.4.2 Antimony and other trace elements in bottled water and soft
drinks
The detection of antimony and the other trace elements in all freshly purchased British
bottled water samples at concentrations below the regulatory limits corroborate on
the safety of the water brands for use as drinking water with regard to antimony and
the other trace elements investigated. The same cannot be said of freshly purchased
Nigerian samples as these were not analysed. Antimony in freshly purchased fruit juice

193

has been found to exceed the regulatory standards in this works and in earlier works.
The worrisome aspect of this finding is that the high concentrations were found in
freshly purchased samples. By implications if these samples are to be allowed to stay
longer in the bottles, especially in the tropical countries like Nigeria where ambient
temperatures are generally high, the antimony concentrations will be even higher as a
result of migration. Findings like these raise questions about the effectiveness of
monitoring activities by the agencies in charge of food safety. Actions need to be taken
to ensure that antimony in fruit juices and other bottled products remain within the
regulatory standard from bottling to consumption for the purpose of safeguarding the
health of consumers.
The detection of titanium at a concentration approaching two parts per million in soft
drinks is not necessarily alarming considering the fact titanium dioxide is an approved
substance for use as food additive in water-based flavoured drinks and other foods.
However as titanium dioxide dust, taken into the body by inhalation, has recently been
reclassified by the International Agency for Research on Cancer (IARC) from Group 3
(not classifiable as to its carcinogenicity to humans) to Group 2B (possibly carcinogenic
to humans), the use of titanium dioxide as food additive will need greater scrutiny in
future.
Elevation of antimony and lead content to levels above the regulatory limit after long
term storage of bottled water and soft drinks in PET and glass is not a serious issue
since most of these bottled water and soft drinks are not typically stored for long
periods before consumption in both Britain and Nigeria. However the elevation of lead
to concentration above the EU MAC in a Nigerian soft drink brand after 2 months of
storage in refillable glass bottle (most likely reused several times for bottling) will need
greater scrutiny from the regulatory authorities. Reuse of refillable glass bottle for soft
drink bottling is an environment-friendly behaviour. So the most important issue here
is not the reuse of the glass bottle but the chemical composition of the bottle.
NAFDAC in Nigeria needs to act to ensure that refillable glass bottles used in bottling
do not contain high levels of lead or other chemical substances that can cause harm to
consumers through migration into contents.
The antimony migration behaviour of the new and aged PET bottles studied at the
realistic temperatures of 40 and 60C demonstrated the safety of using the bottles
194

with regard to antimony migration at the said temperatures. This is good news
considering the fact that drinking water being purified by SODIS in tropical developing
countries has not been reported to heat up above 60C. Consequently treating
drinking water by SODIS in bottles will not elevate the antimony concentration to
levels above the regulatory limits even for bottles reused over a long period of time.
As explained earlier SODIS is a low-cost drinking water disinfection technology with a
great potential to improve the health of people without access to safe drinking water.
Even though exposure of water or soft drinks in PET bottle at temperature of 80C is
associated with release of antimony several times above the accepted limit, this
exposure situation is not typically encountered. The likelihood of encountering this
situation become even lower considering the fact that at such temperatures
deformation in bottle shape can start to manifest as the result of the elevated
temperature. However in developing countries like Nigeria where PET bottle is reused
in very many ways exposure at 80C can still not be completely ruled out. So for
situations like this the regulatory authority needs to inform the people reusing bottles
to be aware of the risk associated with reusing the bottles at such high temperatures.

10.4.3 Acetaldehyde in bottled water and soft drinks

Acetaldehyde was found in freshly purchased bottled water in PET bottles but not in
glass bottles. However the concentration was so low that it is only relevant to odour
and taste of the bottled water rather than to the safety of consumption of the water.
Acetaldehyde concentration in some freshly purchased soft drinks exceeded the EU
specific migration limit for acetaldehyde from food packaging material into packaged
food (6mg/kg). The SML was established based on the tolerable daily intake level of
0.1mg/kg. Maintaining the concentration of acetaldehyde below the SML is meant to
assure that exposure remains under the Tolerable Daily Intake. For the sake of safety
this study recommends that the amount of acetaldehyde that can be added to soft
drinks as flavouring agent should be below the specific migration limit (SML) for
migration of acetaldehyde from PET bottle into bottle contents. For the purpose of
preventing the acetaldehyde concentration from going above the odour and taste
threshold limit in sparkling water in PET bottles, bottles with low acetaldehyde
contents should be used for bottling sparkling water since sparkling water has greater
ability to extract acetaldehyde than still water.

195

10.5 Research strengths and limitations

Unlike previous studies in the field of chemical leaching from water and soft drinks
containers this research attempted to approach the problem by exploring and linking
the two different but complementary aspects of the problem namely, the pattern and
extent of bottle use and reuse and the chemical migration, which to appreciable extent
depends on the former. Fusion of the two aspects conferred some uniqueness to the
research in allowing the problem of migration to be visualised from the perspective of
the actual bottle handling. For example the research has found that single PET bottles
are reused for as long as one year but that bottles subjected to actual reuse condition
for as long as a year release less antimony and acetaldehyde than new bottles. In other
words reused bottles are safer to use than new bottles. As advantageous as this
approach may be it is still associated with some difficulties. For example the two
aspects of the research belong to two different fields (social science and physical
science) with different skills requirement on the side of the researcher and also
different supervision requirements.
As in all research works some issues were encountered that affected the achievement
of some of the objectives in this research. Some of these limitations are discussed
below
The survey, which was meant to reveal the pattern and extent of bottled water
and soft drink use and PET bottle reuse in the general population, considered
university environments only, as that was the most feasible strategy to adopt.
Surveying the general population, even though more challenging would have
provided truer information on the general population.
Due to practical issues, the survey was only piloted in a UK context. Had a
pilot been conducted in Nigeria, apparently inaccurate reporting of bottle sizes
might have been avoided (see section 6.4 and 6.11).
In spite of the importance of determining the concentration of the analytes of
interest in freshly purchased samples, the concentration of antimony could not
be determined from freshly purchased Nigerian samples due to sample storage
and instrument availability limitations.
The research aimed to determine the concentration of antimony in both
bottling materials and in bottle contents. However antimony content in glass

196

bottle materials was not measured as the methods used in this research were
not designed to quantify antimony in glass. Quantifying antimony in glass bottle
materials would have been useful in that antimony has been reported to be
used as fining agent in the manufacture of bottle material and in this research it
has been found in similar quantities in some soft drink brands bottled in both
glass and PET.
The best way to assess the migration of both antimony and acetaldehyde with
time is to measure them in a freshly purchased sample and then measure them
in the same sample after desired storage period. This research assessed
migration with time by concurrently measuring the migrating chemicals in a
sample stored for a desired storage period and in a fresh sample. The method
involving one-sample could not be used due to instrument availability
limitations. While the two-sample method can reveal chemical migration, it is
not as reliable as the method using single samples as some differences could
arise in samples of different batch.

10.6 Further work

While this research attempted to address the questions in the research objectives,
other question have arisen as a result of the research work. The most important of
these question are listed here as a basis for future work
In Nigeria the water packaged in plastic pouches accounts for 68 percent of total
commercial water and is consumed by the low-income group as stated in the
literature review. This type of packaged water was not fully investigated in this
study because the plastic packaging for the water is not PET plastic.
Notwithstanding studies will be needed to see how the consumption of this type
of water affects the extent of use of bottled water in Nigeria.
Antimony and acetaldehyde in freshly purchased soft drink samples evidently
originate from both migration from bottle walls and from the soft drinks
themselves, their ingredients or the processes associated with their of production.
A study will be needed to ascertain the origin of these chemicals in soft drinks to
see whether the manufacturing processes used and/or the ingredients will need to
be reviewed.

197

 In this study outgassing of acetaldehyde was observed in some empty bottled

water PET bottles but not in others. This observation was believed to be related
to the presence of sparkling water in bottles. However, since only three bottles
were observed, the reason behind the observed outgassing could not be
ascertained with certainty. It will be worthwhile to investigate this previously
unreported phenomenon.
In this study an increase of 586% in the acetaldehyde concentration was observed
after storage of a cola drink for a period of 12 months. zlem (2008) has
reported a similar observation. However both in this work and in the work of
zlem (2008) the acetaldehyde increase could not be accounted for from the
migration of the residual acetaldehyde in PET material. In other words migration
of all the residual acetaldehyde in the PET material could not elevate the level of
acetaldehyde to the concentrations observed. Studies are needed to ascertain the
source of this increase, i.e. whether it is as a result of degradation of some
components in the soft drinks or from the degradation of the PET material.
For the sake of standardisation studies comparing the different methods available
for determination of antimony and residual acetaldehyde in PET will need to be
carried out. These studies will reveal whether the methods have similar
effectiveness. For determination of antimony the microwave digestion-ICP-MS
carried out in this study will need to be compared with laser ablation-ICP-MS,
Instrumental

Neutron

Activation

Analysis

and

energy

dispersive

X-ray

fluorescence spectroscopy. For determination of acetaldehyde the headspace-GCFID used in this study can be compared with the methods described by Bashir et al
(2002) and the method described by Matsuga et al (2005) which Involved
simultaneous dissolution of PET in trifluoroacetic acid and derivatization of the
acetaldehyde content with 2,4-dinitrophenylhydrazine followed by precipitation,
solvent

extraction

with

dichloromethane,

evaporation,

reconstitution

in

acetonitrile and analysis by HPLC. The method of Bashir (2002) is similar to the
method used in this study. The two differ in that rather than introducing the PET
material directly into the headspace vial, in the method of Bashir (2002) the PET is
cooled in liquid nitrogen and ground.
Because glass bottles were not the main interest in this study, only few were
assessed for chemical migration. Considering the finding made in this study with

198

regards to migration of antimony and lead from some glass bottles, and the fact
that antimony and arsenic are in some cases used as fining agent in glass
manufactures, glass bottles from different countries will need to be investigated
for migration of antimony, lead and arsenic. A further justification for a research
like this lies in the fact that in Nigeria use of refillable glass bottles is very common
in the soft drink bottling industry. Among other issue a research like this will need
to investigate is the relationship between the age and frequency of refill of a bottle
and migration of chemicals.
The phthalate plasticizer in the Nigerian PVC cap lining was not characterised in
this study. Chromatographic studies to characterise the plasticizer are worth
carrying out. Studies assessing the extent to which the plasticizer can migrate into
the soft drinks are also worthwhile.

Overall, this investigation has pushed forward knowledge about how PET bottles are
used in practice and the implications this has for exposure to migrant chemicals
leaching from bottle walls.

199

REFERENCES
ACGIH (2001) Documentation of the Threshold Limit Values for Chemical Substances, 7th
ed. Cincinnati: American Conference of Governmental Industrial Hygienists. Available
from: www.acgih.org [Accessed 29th December 2010]
Ahmad, M., Bajahlan, A. S. (2007) Leaching of styrene and other aromatic compounds
in drinking water from PS bottles. Journal of Environmental Sciences, 19, 421426
Ainsworth, N., Cooke, J. A., Johnson, M. S. (1991) Behaviour and toxicity of antimony
in the short-tailed field vole (Microtus agrestis). Ecotoxicology and Environmental Safety,
21(2), 165-170.
Al-Malack, M. H. (2001) Migration of lead from unplasticized polyvinyl chloride pipes.
Journal of Hazardous Materials, B82, 263274
Al-Malack, M. H., Sheikheldin, S. Y. (2001) Effect of solar radiation on the migration of
vinyl chloride monomer from unplasticized PVC pipes. Water research, 35(14), 32833290
Alt, A., Haldimann, M. (2008) Diffusion coefficient of antimony catalysts in polyethylene
terephthalate (PET) materials. [online]. Available from: http://www.akts.ch/dwlds/SMLSoftware-Diffusion-of-antimony-catalysts-in-PET.pdf [Accessed 23rd July 2008]
Appleman, L. M., Woutersen, R. A., Feron, V. J. (1982) Inhalation toxicity of
acetaldehyde in rats. I. Acute and subacute studies. Toxicology, 23, 293-307
Arizona Solar Center (2008) Passive solar heating and cooling. [online] Phoenix: Arizona
Solar Center. Available from: www.azsolarcenter.com/design/documents/passive.DOC
[Accessed 23rd July 2008]
Arnold, E., Larson, J. (2006) Bottled Water: Pouring resources down the drain. [online]
Washington DC: Earth Policy Institute. Available from: http://www.earthpolicy.org/Updates/2006/Update51.htm [Accessed 30th June 2008]
ATSDR (2007) ToxFAQs for Antimony and Compounds. [online] Atlanta: Agency for
Toxic Substances and Disease Registry. Available from:
http://www.atsdr.cdc.gov/tfacts23.html [Accessed 2nd May 2008]
Australian Department of the Environment, Water, Heritage and the Arts (2006)
Antimony & compounds fact sheet.[online] Department of the Environment, Water,
Heritage and the Arts. Available from:
http://www.npi.gov.au/substances/antimony/index.html [Accessed 5th May 2008]
Azapagic, A., Emsley, A. Hemerton, I. (2003) Polymers, the environment and sustainable
development. Chichester: John Wiley and sons.
Bashir, Z., Al-Uraini, A-A., Jamjoom M., Al-Khalid, A., Al-Hafez ,M., Ali, S. (2002)
Acetaldehyde generation in poly(ethylene terephthalate) resins for water bottles.
Journal of Macromolecular Science, Part A: Pure and Applied Chemistry, 39(12), 1407 1433

200

Begley, T.H., Biles, J.E., Cunningham, C., Piringer, O. (2004) Migration of a UV stabilizer
from polyethylene terephthalate (PET) into food simulants. Food Additives &
Contaminants, 21(10), 1007-1014
Benfenati, E., Natangelo, M., Davoli E., Fanelli R. (1991) Migration of vinyl chloride into
PVC- bottled drinking- water assessed by gas chromatography-mass spectrometry.
Food and Chemical Toxicology, 29(2), 1991, 131-134
Biedermann-Brem, S., Grob, K., Fjeldal, P. (2008) Release of bisphenol A from
polycarbonate baby bottles: mechanisms of formation and investigation of worst case
scenarios. European Food Research and Technology. 227(4), 1053-1060
Birke, M., Rauch, U., Harazim, B., Lorenz, H., Glatte, W. (2010) Major and trace
elements in German bottled water, their regional distribution, and accordance with
national and international standards. Journal of Geochemical Exploration, 107, 245271
Bornehag, C., Sundell, J., Weschler, C. J., Sigsgaard, T, Lundgren, B., Hasselgren, M.,
Hgerhed-Engman, L., (2004) The Association Between Asthma and Allergic Symptoms
in Children and Phthalates in House Dust: A Nested Case-Control Study. Environmental
Health Perspectives, 112(14),1393-1397.
Borucke, M., Elice, S. J., Parsons, J., Varghese, A. (2001) Point-of-use Household Water
Treatment in Haiti [online] Cambridge (US): Massachusetts Institute of Technology
Department of Civil and Environmental Engineering Available from:
http://ceeserver3.mit.edu/~Haiti/docs/Proposal-all.htm [Accessed: 23rd July 2008]
Bottled Water Information Office (2008) Facts and Figures. [online] London: Bottled
Water Information Office. Available from:
http://www.bottledwaterinformation.co.uk/default.asp?section=2&subsection=24 [Accessed
30th June 2008]
British Soft Drinks Association (no date) General information about the soft drinks
industry. [online] London: British Soft Drinks Association. Available from:
http://www.britishsoftdrinks.com/default.aspx?page=329 [Accessed 26th January 2011]
Butterman, W.C., Carlin Jr., J.F. (2004) Mineral commodity profiles: Antimony. U.S.
Department of The Interior, U.S. Geological Survey Open-File Report 03-019.
Available from: http://pubs.usgs.gov/of/2003/of03-019/of03-019.pdf [Accessed 29th
December 2010]
Californians Against Waste (2010) AB 2505 (Brownley) PVC packaging phase out. [online]
Sacramento: Californians Against Waste. Available from:
http://www.cawrecycles.org/issues/current_legislation/ab2505_08 [Accessed 23rd January
2011]
Center for Health, Environment and Justice (2009) PVC Policies across the world. [online]
Falls Church: Center for Health, Environment and Justice. Available from:
http://www.chej.org/BESAFE/pvc/documents/2009/FactSheets/110909%20PVC%20Policies%20Across%20the%20World.pdf [Accessed 23rd
January 2011]

201

Cicchella, D., Albanese, S., De Vivo, B., Dinelli, E., Giaccio, L., Lima, A. and Valera, P.
(2010) Trace elements and ions in Italian bottled mineral waters: Identification of
anomalous values and human health related effects. Journal of Geochemical Exploration,
107(3), 336-349
Clayton, G. D., Clayton, F. E. (1993) Pattys Industrial Hygiene and Toxicology, vol. II, Part
A Toxicology, 4th ed.), New York: John Wiley & Sons.
Cohen, J. (1988) Statistical power analysis for the behavioural sciences. 2nd edn., Hillsdale,
NJ: Lawrence Erlbaum Associates
Corradini, P., Guerra, G., Cavallo, I. (2004) Do new century catalysts unravel the
mechanism of stereocontrol of old ziegler-natta catalysts? Accounts of Chemical
Research, 37, 231-241
Dental abstracts (2007) Soft drink acidity and enamel dissolution. Dental Abstracts,
52(5), 300-301
Dobney, A. M., Wiarda, W., de Joode, P., van der Peijl, G. J. Q. (2002) Sector field ICPMS applied to the forensic analysis of commercially available adhesive packaging tapes.
Journal of Analytical Atomic Spectrometry, 17, 478484
EAWAG (2008) The application of Solar Water Disinfection. [online] Dbendorf:
EAWAG/SANDEC. Available from: http://www.sodis.ch/Text2002/T-Application.htm
[Accessed 21st June 2008]
EFSA (2004) Opinion of the Scientific Panel on food additives, flavourings, processing
aids and materials in contact with food (AFC) on a request from the Commission
related to a 2nd list of substances for food contact materials. EFSA Journal, 24, 1-13
Egle Jr., J. L. (1972) Effects of inhaled acetaldehyde and propionaldehyde on blood
pressure and heart rate. Toxicology and Applied Pharmacology, 23, 131-135
EHSO (2006) Shelf life of bottled water. [online] Atlanta: Environment, Health and Safety
Online. Available from:
http://www.ehso.com/ehshome/DrWater/drinkingwaterfaqs.php#shelflife [Accessed 28th
January 2011]
El-Toufaili, F. (2006) Catalytic and Mechanistic Studies of Polyethylene Terephthalate
Synthesis. Dr. rer. nat. Dissertation, The Technical University of Berlin.
EPA (1994) Chemical summary for acetaldehyde. [online] Washington, DC: United States
Environmental Protection Agency. Available from:
http://www.epa.gov/chemfact/s_acetal.txt [Accessed: 29th December 2010
EPA (2007) Acetaldehyde. [online] Washington, DC: United States Environmental
Protection Agency. Available from: http://www.epa.gov/ttnatw01/hlthef/acetalde.html
[Accessed: 29th December 2010
EPA (2010) National Primary Drinking Water Regulations. [online] Washington, DC:
United States Environmental Protection Agency. Available from:
http://water.epa.gov/drink/contaminants/index.cfm [Accessed: 29th December 2010]

202

European Commission (1998) Opinion of the Scientific Committee on Food on an additional

list of monomers and additives for food contact materials (adopted the 18 September 1998).
[online] Brussels: European Commission. Available from:
http://ec.europa.eu/food/fs/sc/scf/out16_en.html [Accessed 3rd January 2011]
European Commission (2002) COMMISSION DIRECTIVE 2002/72/EC of 6 August
2002 relating to plastic materials and articles intended to come into contact with
foodstuff. Official Journal of the European Union. [online] L220, 18-57 Available from:
http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2002:220:0018:0058:EN:PDF
[Accessed: 22 December 2010]
European Commission (2003) COMMISSION DIRECTIVE 2003/40/EC of 16 May 2003
establishing the list, concentration limits and labelling requirements for the constituents
of natural mineral waters and the conditions for using ozone-enriched air for the
treatment of natural mineral waters and spring waters. Official Journal of the European
Union, [online] L126, 34-39. Available from:
http://eurlex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2003:126:0034:0039:EN:PDF
[Accessed: 22 December 2010]
FACS (2009) PET bottle reuse. [online] Sunnyside: Food Advisory Consumer Service.
Available from: http://www.foodfacts.org.za/Articles/PETbottlereuse.asp [Accessed 5th
January 2011]
FAO/WHO (2010) Food additive details: Titanium dioxide. [online] Rome: FAO. Available
from: http://www.codexalimentarius.net/gsfaonline/additives/details.html?id=184 [Accessed
27th January 2011]
FDA (2001) Guidance for Industry: Bioanalytical Method Validation. [online]. Silver Spring:
United States Food and Drug Administration. Available from:
http://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/uc
m070107.pdf
FDA (2009) Bottled water regulation and the FDA. [online]. Silver Spring: United States
Food and Drug Administration. Available from:
http://www.fda.gov/Food/FoodSafety/ProductSpecificInformation/BottledWaterCarbonatedSoftDrinks/ucm077079.htm [Accessed: 26th
December 2010]
FDA (2011) Everything Added to Food in the United States (EAFUS). [online]. Silver Spring:
United States Food and Drug Administration. Available from:
http://www.accessdata.fda.gov/scripts/fcn/fcnNavigation.cfm?rpt=eafusListing [Accessed: 8th
June 2011]
Federal Institute for risk assessment (2007) Frequently Asked Questions about PET bottles.
[online] Berlin: Federal Institute for risk assessment. Available from:
http://www.bfr.bund.de/cm/279/frequently_asked_questions_about_pet_bottles.pdf
[Accessed: 22 November 2009]
Feron, V. J., Kruysse, A., Woutersen, R. A. (1982) Respiratory tract tumours in
hamsters exposed to acetaldehyde vapour alone or simultaneously to benzo(a)pyrene
or diethylnitrosamine. European Journal of Cancer and Clinical Oncology, 18(1), 13-21, 2331

203

Filella, M., Belzile, N., Chen, Y. (2002) Antimony in the environment: A review focused
on natural waters. I Occurrence. Earth-Science Reviews 57, 125176
Finewaters (2005) Bottled water - Clear Growth in UK. [online] London: Food Standards
Agency. Available from: http://www.eatwell.gov.uk/foodlabels/labellingterms/bestbefore/
[Accessed 27th January 2011]
Food navigator (2004) Soft drinks boast hard sales. [online] Decision News Media.
Available from: http://www.foodnavigator.com/news/ng.asp?id=50252-soft-drinks-boast
[Accessed 30th June 2008]
Food Safety Commission (2005) Evaluation Report of Food Additives Acetaldehyde.
[online]. Tokyo: Food Safety Commission. Available from:
http://www.fsc.go.jp/english/evaluationreports/foodadditive/acetaldehyde_report.pdf [Accessed
21st April 2011]
Food Standards Agency (no date) Best before. [online] Finewaters. Available from:
http://www.finewaters.com/Newsletter/November_2005/Bottled_water__Clear_Growth_in_UK.asp [Accessed 30th June 2008]
Fu, Z., Wu, F., Amarasiriwardena, D., Mo, C., Liu, B., Zhu, J., Deng, Q., Liao, H. (2010)
Antimony, arsenic and mercury in the aquatic environment and sh in a large antimony
mining area in Hunan, China. Science of the Total Environment, 408, 34033410
Gaffa, T., Jideani, I. A., Nkama, I. (2002) Nutrient composition of different types of
kunu produced in Bauchi and Gombe states of Nigeria. International Journal of Food
Properties, 5, 351-357.
Garba, I., Yakubu, S.E., Olonitola, O.S. (2007) Antibiotic susceptibility studies on some
enteropathogenic bacteria isolated from public water supplies in Gusau municipal,
Zamfara State. Journal of pharmaceutical and allied sciences 4 (1), 403 412
Gebel, T., (1997) Arsenic and antimony: comparative approach on mechanistic
toxicology. Chemico-Biological Interactions, 107(3), 131144.
Gleick, H. (2008) The World's Water 2008-2009: The Biennial Report on Freshwater
Resources. Washington, D.C.: Island Press.
Grimes, D. J., Ficklin, W. H., Meier, A. L., McHugh J. B. (1995) Anomalous gold,
antimony, arsenic, and tungsten in ground water and alluvium around disseminated
gold deposits along the Getchell Trend, Humboldt County, Nevada. Journal of
Geochemical Exploration, 52, 351-371.
Gler, C. (2007) Evaluation of maximum contaminant levels in Turkish bottled drinking
waters utilizing parameters reported on manufacturers labelling and governmentissued production licenses. Journal of Food Composition and Analysis, 20, 262272
Hansen, H. R., Pergantis, S. A. (2006) Detection of antimony species in citrus juices and
drinking water stored in PET containers. J. Analytical Atomic Spectrometry, 21, 731733
Health Canada (2009) Frequently asked questions about bottled water.[online]s.l.: Health
Canada. Available from: http://www.hc-sc.gc.ca/fn-an/securit/facts-faits/faqs_bottle_watereau_embouteillee-eng.php#a2 [Accessed 28th January 2011]

204

Hendra, P. J., Agbenyega, J.K. (1993) The raman spectra of polymers. Chichester: John
Wiley and sons.
Howdeshell, K. L., Peterman, P. H., Judy, B. M., Taylor, J. A., Orazio, C. E., Ruhlen, R.
L., vom Saal, F. S., Welshons, W. V. (2003) Bisphenol a is released from used
polycarbonate animal cages into water at room temperature. Environmental Health
Perspectives, 111(9), 1180-187
Howell, E.D. and Ponasik, J.A. (2006) Acetaldehyde scavenging by addition of active
scavengers to bottle closures. European Patent EP 1 442 993 B1, Appl.19th January 2004,
Acc. 8th November 2006
IARC (1989) Some organic solvents, resin monomers and related compounds,
pigments and occupational exposures in paint manufacture and painting: antimony
trioxide and antimony trisulphide. IARC Monographs on Evaluation of Carcinogenic Risks to
Humans, [online]. 47, p.291. Available from:
http://monographs.iarc.fr/ENG/Monographs/vol47/volume47.pdf [Accessed 29th December
2010]
IARC (1999) Re-evaluation of some organic chemicals, hydrazine and hydrogen
peroxide: Acetaldehyde. IARC Monographs on Evaluation of Carcinogenic Risks to Humans,
[online]. 71, p.331. Available from:
http://monographs.iarc.fr/ENG/Monographs/vol71/mono71-11.pdf [Accessed 29th
December 2010]
IARC (2010) Carbon Black, Titanium Dioxide, and Talc. IARC Monographs on the
Evaluation of Carcinogenic Risks to Humans[online]. 93, p.275. Available from:
http://monographs.iarc.fr/ENG/Monographs/vol93/mono93.pdf [Accessed 13th May
2011]
IARC (2009) A review of human carcinogensPart E: tobacco, areca nut, alcohol, coal
smoke, and salted fish. The Lancet Oncology, 10(11), 1033-1034
IDES (2011) Raman spectroscopy 101 for plastics identification. [online] s.l.: IDES.
Available from: http://www.ides.com/articles/materials/2010/Raman-Spectroscopy-PlasticsIdentification.asp [Accessed 8th February 2011]
IMF (2010) World economic database 2010. [online] Washington DC: International
Monetary Fund Available from:
http://www.imf.org/external/pubs/ft/weo/2010/02/weodata/weorept.aspx?sy=2004&ey=2004
&scsm=1&ssd=1&sort=country&ds=.&br=1&c=449%2C466&s=NGDPDPC&grp=0&a=&pr
.x=46&pr.y=8 [Accessed 28th January 2011]
IPCS (1995) Acetaldehyde. International Programme on Chemical Safety, Environmental
Health Criteria No. 167.[online]. Available from:
http://www.inchem.org/documents/ehc/ehc/ehc167.htm [Accessed 29th December 2010]
International Antimony Oxide Industry Association (2006) Antimony Trioxide Frequently
Asked Questions. [online] s.l.: IAOIA. Available from:
http://www.nihonseiko.co.jp/english/environment/060418faq_e.pdf [Accessed: 1st February
2009]

205

Joint Monitoring Programme (2006) The Joint Monitoring Programme for water supply.
[online] Geneva: WHO. Available from: http://www.wssinfo.org/en/ [Accessed 19th June,
2006]
Just-Drinks (2008) UK: Bottled water consumption continues stellar growth. [online]
Bromsgrove: Just-Drinks. Available from: http://www.justdrinks.com/article.aspx?ID=94140 [Accessed 30th June 2008]
Kenplas (2008) PET Preform/Bottle Project. [online]. Hangzhou: Kenplas Industry Limited.
Available from: http://www.kenplas.com/project/pet/ [Accessed: 21 July 2008]
Keresztes, S., Tatr, E., Mihucz, V. G., Virg, I., Majdik, C., Zray, G. (2009) Leaching of
antimony from polyethylene terephthalate (PET) bottles into mineral water.
Science of the Total Environment, 407, 47314735
Kirkland D., Whitwell J., Deyo J., Serex T. (2007) Failure of antimony trioxide to
induce micronuclei or chromosomal aberrations in rat bone-marrow after sub-chronic
oral dosing. Mutation Research. 627(2), pp 119-28.
Kolawole, J., A., Maduenyi A. (2004) Effect of zobo drink (Hibiscus sabdariffa water
extract) on the pharmacokinetics of acetaminophen in human volunteers. European
Journal of Drug Metabolism and Pharmacokinetics, 29(1), 25-29.
Kuisma-Kursula, P. (2000) Accuracy, Precision and Detection Limits of SEMWDS,
SEMEDS and PIXE in the Multi-Elemental Analysis of Medieval Glass. X-Ray
Spectrometry, 29, 111118
Kuznesof, P. M. (2006) Titanium dioxide : Chemical and Technical Assessment.
[online]Rome: Food and Agricultural Organisation. Available from:
ftp://ftp.fao.org/ag/agn/jecfa/cta_tio2.pdf [Accessed 30th June 2008]
Le, H. H., Carlson, E. M., Chua, J. P., Belcher, S. M. (2008) Bisphenol A is released from
polycarbonate drinking bottles and mimics the neurotoxic actions of estrogen in
developing cerebellar neurons. Toxicology Letters, 176, 149156
Li, L. (2008) Bottled Water Consumption Jumps. [online] Washington DC: Worldwatch
Institute. Available from: http://www.worldwatch.org/node/5475 [Accessed 30th June
2008]
Lichter, S. R. (2009) Are chemicals killing us? [online] Washington DC: Statistical
Assessment Service. Available from:
http://stats.org/stories/2009/are_chemicals_killing_us.html [Accessed 5th January 2011]
Lide, D.R. (1990) CRC Handbook of Chemistry and Physics. 70th Edn., Boca Raton, FL:
CRC Press.
Lilya, D. (2001) Analysis and risk assessment of organic chemical migration from
reused PET plastic bottles. Society for Risk Analysis 2001 Annual Meeting. December 2-5,
2001. Available at: http://www.riskworld.com/abstract/2001/SRAam01/ab01aa189.htm
[Accessed 2nd July 2008]

206

Linssen, J., Reitsma, H., Cozijnsen, J. (1995) Static headspace gas chromatography of
acetaldehyde in aqueous foods and polythene terephthalate. Zeitschrift fr
Lebensmitteluntersuchung und -Forschung A, 201(3), 253 255.
Lopez-Molinero, A., Calatayud, P., Sipiera, D., l Falcon, R., Lian, D., Castillo J.R. (2007)
Determination of antimony in poly(ethylene terephthalate) by volatile bromide
generation flame atomic absorption spectrometry. Microchim Acta, 158, 247-253
Matsuga, M., Tojima, T., Kawamura, Y., Tanamoto, K. (2005) Survey of formaldehyde,
acetaldehyde and oligomers in polyethylene terephthalate food-packaging materials.
Food Additives & Contaminants, 22(8), 783 - 789
Matsuga, M., Kawamura, Y., Sugita-Konishi, Y., Hara-Kudo, Y., Takatori, K., Tanamoto,
K. (2006) Migration of formaldehyde and acetaldehyde into mineral water in
polyethylene terephthalate (PET) bottles. Food Additive and Contaminants, 23(20), 212218
Matthews, V. (2000) Packaging materials: 1. Polyethylene terephthalate (PET) for food
packaging applications. International Life Sciences Institute: ILSI Europe Report Series.
McCarthy, M. (2007) Understanding plastic. Sea stars, 17(7), 1-2.
Miranda, E. S., Miekeley, N., De-Carvalho, R. R., Paumgartten F. J .R., (2006)
Developmental toxicity of meglumine antimoniate and transplacental transfer of
antimony in the rat. Reproductive Toxicology. 21, 292300
Miyake, T., Shibamoto, T. (1993) Quantitative Analysis of Acetaldehyde in Foods and
Beverages. Journal of Agricultural and Food Chemistry, 41, 19681970
Morris J. B. (1997) Dosimetry, toxicity and carcinogenicity of inspired acetaldehyde in
the rat. Mutation Research, 380, 113124
Nasser, A.L.M., Lopes, L.M.X., Eberlin, M.N., Monteiro, M. (2005) Identification of
oligomers in polyethylene terephthalate bottles for mineral water and fruit
juice/Development and validation of a high-performance liquid chromatographic
method for the determination of first series cyclic trimer. Journal of Chromatography A,
1097, 130137
Nathan, J.S., Philip, L. (2009) Leaching of DEHA and DEHP from PET bottles to water.
Department of Civil Engineering Indian Institute of Technology
National Environmental Service Centre (2006) UNICEF Report Says One Billion Lack Safe
Drinking Water. [online] Morgantown: West Virginia University. Available from:
http://www.nesc.wvu.edu/ndwc/articles/OT/FA06/OT_Fl_06_NNweb.pdf [Accessed 19th
June, 2008]
National Toxicology Program (2010) Acetaldehyde (substance profile): Report on
Carcinogens, Eleventh Edition; U.S. Department of Health and Human Services, Public Health
Service, National Toxicology Program. Available from:
http://ntp.niehs.nih.gov/ntp/roc/eleventh/profiles/s001acet.pdf [Accessed 1st January 2011]
Nawrocki, J., Dbrowska, A., Borcz, A. (2002) Investigation of carbonyl compounds in
bottled waters from Poland. Water Research 36, 48934901

207

Nexant (2008) Polyethylene Terephthalate (PET). [online] San Fransisco: Nexant.

Available from: http://nexant.ecnext.com/coms2/gi_0255-144/Polyethylene-TerephthalatePET.html [Accessed 19th June 2008
New Jersey Department of Health and Senior Services (2010) Hazardous substance fact
sheet: Acetaldehyde. [online] Trenton: New Jersey Department of Health and Senior
Services. Available from: http://nj.gov/health/eoh/rtkweb/documents/fs/0001.pdf. [Accessed
29th December 2010]
Newton, P. E., Bolte, H. F., Daly, I. W., Pillsbury, B. D., Terrill, J. B., Drew, R. T., BenDyke, R., Sheldon, A. W., Rubin, L. F. (1994) Subchronic and Chronic Inhalation
Toxicity of Antimony Trioxide in the Rat. Fundamental and Applied Toxicology, 22, 561576.
Niedzielski, P. Siepak, J. Siepak, M. (2001) Total content of arsenic, antimony and
selenium in groundwater samples from western Poland. Polish Journal of Environmental
Studies, 10(5), pp.347-350
Niedzielski, P., Siepak, M. (2005) The occurrence and speciation of arsenic, antimony,
and selenium in ground water of Pozna city (Poland). Chemistry and Ecology, 21(4), 241
253
Nijssen, B., Kamperman, T., Jetten, J. (1996) Acetaldehyde in mineral water stored in
polyethylene terephtalate (PET) bottles: odour threshold and quantification. Packaging
technology & science, 9, 17585.
Ogden, C. L., Fryar, C. D., Carroll, M. D., Flegal, K. M. (2004) Mean Body Weight,
Height, and Body Mass Index, United States 19602002. Advance Data from Vital and
Health Statistics, 347, 1 17
Olajubu, F. A., Mope, D. A. (2007) Bacteriological Assessment of Pure Water
Samples in Ogun State of Nigeria. Nigerian Journal of Health and Biomedical Sciences,
6(2), 45-48
Oliphant, J.A., Ryan, M.C., Chu, A. (2002). Bacterial water quality in the personal water
bottles of elementary students. Canadian Journal of Public Health, 93(5), 366-367.
zlem, K.E.(2008) Acetaldehyde migration from polyethylene terephthalate bottles
into carbonated beverages in Trkiye. International Journal of Food Science and
Technology, 43, 333338
Panel on Dietary Reference Intakes for Electrolytes and Water, Standing Committee
on the Scientific Evaluation of Dietary Reference Intakes. (2004) Dietary Reference
Intakes: Water, Potassium, Sodium, Chloride, and Sulfate. Institute of Medicine of the
National Academies.
PET Planet Insider (2001) Pet Bottle Production Preforms. [online] Heidelberg:
heidelberg business media GmbH. Available from:
http://www.hbmedia.net/petplanet/petplanet/insider/2001/0405/articles/preformproduction1.shtml [Accessed: 21 July 2008]

208

PTCL (2005) Safety data for antimony trioxide. [online] Oxford: University of Oxford.
Available at: http://msds.chem.ox.ac.uk/AN/antimony_trioxide.html [Accessed 4th July 2008)
Radojevi, M., Bashkin, V.N. (1999) Practical environmental analysis. Cambridge: The
Royal Society of Chemistry.
Reimann, C., Birke, M., Filzmoser, P. (2010) Bottled drinking water: Water
contamination from bottle materials (glass, hard PET, soft PET), the influence of colour
and acidification. Applied Geochemistry, 25, 10301046
Resano, M., Garca-Ruiz, E., Vanhaecke, F. (2005) Laser ablationinductively coupled
plasmadynamic reaction cellmass spectrometry for the multi-element analysis of
polymers. Spectrochimica Acta Part B, 60, 1472 1481
Rothschild, J., Nzeka, U. (2005) Nigeria Food Processing Ingredients Sector Nigeria's Food
Processing Ingredients Market2005. [online] USDA Foreign Agricultural Service GAIN
Report/ Global Agriculture Information Network. GAIN Report Number: NI5011.
Available from: http://www.fas.usda.gov/gainfiles/200507/146130221.pdf [Accessed 30th
June 2008]
Royte, E. (2008) Bottlemania: How water went on sale and why we bought it. New York:
Bloomsbury.
Sakurai, H., Noro, J., Kawase, A., Fujinami, M., Oguma, K. (2006) Digestion of plastic
material for the determination of metals with a microwave oven for household use.
Analytical Sciences. 22, 225-228
Schrder, T. (2001) Preforms for the production of containers holding goods sensitive to
taste - Acetaldehyde, a controllable aromatic?! -. [online] Regensburg: PETnology GmbH.
Available from: http://www.petnology.com/zine/zine.php?c_ID=1603&ilang=e [Accessed:
21st July 2008]
Seitz, H. K., Meier, P. (2007) The role of acetaldehyde in upper digestive tract cancer
in alcoholics. Translational Research, 149(6), 293-297
Seward, T. P., Vascott, T. (2005) High temperature glass melt property database for
process modelling. Westerville: The American Ceramic Society.
Shelby, J.E. (2005) Introduction to glass science and technology, 2nd edition. Cambridge:
The Royal Society of Chemistry.
Shimoyama, M., Maeda, H., Matsukawa, K., Inoue, H., Ninomiya, T., Ozaki, Y. (1997)
Discrimination of ethylene/vinyl acetate copolymers with different composition and
prediction of the vinyl acetate content in the copolymers using Fourier-transform
Raman spectroscopy and multivariate data analysis. Vibrational spectroscopy, 14, 253-259
Shotyk, W., Krachler, M. (2007) Contamination of Bottled Waters with Antimony
Leaching from Polyethylene Terephthalate (PET) Increases upon Storage. Environmental
Science & Technology., 41, 1560-1563
Shotyk, W., Krachler, M., Chen, B. (2006) Contamination of Canadian and European
bottled waters with antimony from PET containers. Journal of Environmental Monitoring,
8, 288292
209

Singh, N. P., Khan, A. (1995) Acetaldehyde: genotoxicity and cytotoxicity in human

lymphocytes. Mutation Research, 337, 9-17
Skoog, D. A., Holler, F. J.,Nieman, T. A. (1998) Principles of Instrumental analysis. 5th
edn., London: Thomson Learning
Sugaya, N., Nakagawa, T., Sakurai, K., Morita, M. and Onodera, S. (2001) Analysis of
Aldehydes in Water by Head Space-GC/MS. Journal of Health Science, 47(1), 2127
Swan, S.H., Main, K.M., Liu, F., Stewart, S.L., Kruse, R.L., Calafat, A.M., Mao, C.S.,
Redmon, J.B., Ternand, C.L., Sullivan, S., Teague, J.L., Study for Future Families
Research Team (2005) Decrease in anogenital distance among male infants with
prenatal phthalate exposure. Environ Health Perspectives., 113(8), 1056-61
Swan, S. H., Liu, F., Hines, M., Kruse, R. L., Wang, C., Redmon, J. B., Sparks A., Weiss,
B. (2010) Prenatal phthalate exposure and reduced masculine play in boys. International
Journal of Andrology, 33, 259269
Thiele, U.K. (2004) Quo vadis polyester catalyst? Chemical Fibers International, 54, p.
162.
Tickner, J. A., Rossi, M., Haiama, N., Lappe, M., Hunt, P. (1999) The use of di(2ethylhexyl) phthalate in PVC medical devices: exposure, toxicity and alternatives. [online]
Lowell: Center for Sustainable Production. Available from:
http://www.sustainableproduction.org/downloads/DEHP%20Full%20Text.pdf [Accessed 8th
February 2011]
Tukur, A., Yakubu, S.E., Galadima, M. (2006) Evaluation of inexpensive, small-scale
wellwater disinfection methods. 3rd Annual National Conference. Awka, 8-11th November
2006, s.l.: Society for Occupational Safety and Environmental Health (SOSEH), 124-129
USGS (2010) Antimony: Statistics and Information. [online]. Reston: United States
Geological Survey. Available from:
http://minerals.usgs.gov/minerals/pubs/commodity/antimony/ [Accessed: 29th December
2010]
VISY (2008) VISY Beverage Packaging: Quick Reference guide. [online] Melbourne: VISY.
Available from: http://www.visy.com.au/uploads/BeverageManufacturingProcesses.pdf
[Accessed: 21 July 2008]
Wakayama, H. (2005) Revision of Drinking Water Quality Standards and QA/QC for
Drinking Water Quality Analysis in Japan. Technical Note of National Institute for Land
and Infrastructure Management, 264, 74-88
Walther-Meiner-Institute for Low Temperature Research (2007) SEM / EDX. [online]
Garching:WMI. Available from: http://www.wmi.badw-muenchen.de/methods/sem_edx.htm
[Accessed 1st February 2009]
Wegelin, M. (2006) Training Manual for SODIS promotion (SANDEC Report No.13/06 ,
2006). [online] Duebendorf: SANDEC/EAWAG. Available from:
http://www.sodis.ch/files/TrainingManual_sm.pdf [Accessed 2nd July 2008]

210

Welle F., Franz R. (2011) Migration of antimony from PET bottles into beverages:
determination of the activation energy of diffusion and migration modelling compared
with literature data. Food Additives & Contaminants: Part A, Chemistry, Analysis, Control,
Exposure & Risk Assessment, 28(1), 115-26.
Westerhoff, P., Prapaipong, P., Shock, E., Hillaireau, A. (2008) Antimony leaching from
polyethylene terephthalate (PET) plastic used for bottled drinking water. Water
Research 42, 551 556
WHO (2003) Antimony in Drinking-water: Background document for development of WHO
Guidelines for Drinking-water Quality. World Health Organisation.
WHO/SDE/WSH/03.04/74
Wienaah, M., M. (2007) Sustainable plastic waste management a case of Accra, Ghana.
MSc Thesis, Royal Institute of Technology (KTH)
Woutersen, R. A., Appleman, L. M., Feron, V. J., Van der Heijden, C.A. (1984)
Inhalation toxicity of acetaldehyde in rats. II. Carcinogenicity study: interim results after
15 months. Toxicology, 31, 123-133
Woutersen, R. A., Appleman, L. M., Van Garderen-Hoetmer, A., Feron, V. J. (1986)
Inhalation toxicity of acetaldehyde in rats. III. Carcinogenicity study. Toxicology, 41, 213231
WRAP (2007) Packaging optimisation: the impact to date. [online] Banbury: Waste &
Resource Action Programme. Available from:
http://www.shropshire.gov.uk/waste.nsf/viewAttachments/JTHN-85ZFWP/$file/Packaging-workby-WRAP.pdf [Accessed 21st June 2008]
WRAP (2008) Case study: Minimising packaging: Coca-Colas lightest ever PET 500ml bottle
in GB. [online]Banbury: Waste & Resource Action Programme. Available from:
http://www.wrap.org.uk/downloads/1562402_PET_Case_Study_HiRes_PPV.76ffad0c.3975.pdf [Accessed 21st June 2008]
Yoshioka, T., Okayama, N., Okuwaki, A. (1998) Kinetics of Hydrolysis of PET Powder
in Nitric Acid by a Modified Shrinking-Core Model. Industrial & Engineering Chemistry
Research, 37, 336-340

211

Appendix 1 Survey questions

212

Appendix 1 Survey questions (continued from page 210)

213

Appendix 2 Acetaldehyde intake from soft drinks

Sample

Acetaldehyde Acetaldehyde Mean

daily
daily
concentration in 2.96 litres weight of
intake
intake
(g/L)
of sample
male adult
(mg/kg)* (mg/kg)**
UPC37
3116.48
9.22
86.1
0.11
0.05
UPC38
9902.19
29.31
86.1
0.34
0.17
UPC39
395.05
1.17
86.1
0.01
0.01
UPC30
4952.19
14.66
86.1
0.17
0.09
UPC33
330.76
0.98
86.1
0.01
0.01
UPC40
14247.43
42.17
86.1
0.49
0.24
UPC32
7145.05
21.15
86.1
0.25
0.12
UPC41
3330.76
9.86
86.1
0.11
0.06
UCC42
383.14
1.13
86.1
0.01
0.01
UCC43
12516.48
37.05
86.1
0.43
0.22
UGC34
187.9
0.56
86.1
0.01
0.00
UGC36
1683.14
4.98
86.1
0.06
0.03
UGC31
2152.19
6.37
86.1
0.07
0.04
UOC35
2095.05
6.20
86.1
0.07
0.04
UPC40
14249.81
42.18
86.1
0.49
0.24
UPC22
311.71
0.92
86.1
0.01
0.01
UPC24
4752.19
14.07
86.1
0.16
0.08
UPC25
1397.43
4.14
86.1
0.05
0.02
UPC26
597.43
1.77
86.1
0.02
0.01
UPC23
1930.76
5.72
86.1
0.07
0.03
UGC22
109.33
0.32
86.1
0.00
0.00
UGC23
3142.67
9.30
86.1
0.11
0.05
UGC26
599.81
1.78
86.1
0.02
0.01
UGC27
278.38
0.82
86.1
0.01
0.00
*if all water intake is in the form of soft drinks
**if half of water intake is in the form of soft drink the other half as water

214