Atabani 2019
Atabani 2019
Atabani 2019
Fuel
journal homepage: www.elsevier.com/locate/fuel
Integrated valorization of waste cooking oil and spent coffee grounds for T
biodiesel production: Blending with higher alcohols, FT–IR, TGA, DSC and
NMR characterizations
A.E. Atabania, , Sutha Shobanab, M.N. Mohammedc, Gediz Uğuzd, Gopalakrishnan Kumare, ,
⁎ ⁎
GRAPHICAL ABSTRACT
Keywords: This work deals for the first time with recycling a binary blend of waste cooking oil (WCO) and spent coffee
Waste cooking oil grounds oil (SCGO) (50/50% by volume) into biodiesel. Additionally, this work aimed at valorizing both wastes
Spent coffee grounds oil while improving fatty acids composition to obtain a stabilized product due to low degree of unsaturation of
Biodiesel SCGO besides, reducing its high acid value. Furthermore, blending with Euro diesel and higher alcohols of
Higher alcohols
butanol, pentanol and hexanol to improve some biodiesel properties was investigated. Furthermore, FT–IR,
FT–IR
UV–vis., TGA, DSC and NMR analyses to assess the blends quality was examined. Results have proved the ad-
TGA
DSC vantage of pre-blending and higher alcohols addition to improve the fuel properties. The biodiesel exhibits
NMR excellent cetane number of 56.64, oxidation stability of about 6.51 h and iodine value of 83.71 (g iodine/100 g)
due to the low degree of unsaturation (89.02). Blending with higher alcohols improves density and cold flow
properties while keeping acceptable level of cetane number, flash point and higher heating value (HHV) rather
than lower alcohols. Consequently, this opens new gates for biodiesel research area. In conclusion, this work
reinforces that blending of WCO and SCGO improves biodiesel properties and add to the economy of this
⁎
Corresponding authors.
E-mail addresses: [email protected] (A.E. Atabani), [email protected] (G. Kumar).
https://doi.org/10.1016/j.fuel.2019.01.169
Received 5 December 2018; Received in revised form 25 January 2019; Accepted 30 January 2019
Available online 13 February 2019
0016-2361/ © 2019 Elsevier Ltd. All rights reserved.
A.E. Atabani, et al. Fuel 244 (2019) 419–430
industry. Therefore, further examination on the blends effect in compression ignition diesel engines through
engine emissions and combustion characteristics is recommended.
420
A.E. Atabani, et al. Fuel 244 (2019) 419–430
Ultraviolet–visible absorption spectra (UV–vis.), TGA, DSC and Nuclear heating to 60 °C (hold for 2 min), then to 200 °C with heating rate
magnetic resonance (NMR) analyses was extensively studied and pre- 10 °C/minute and finally to 240 °C with heating rate 5 °C/minute (hold
sented. Moreover, some important properties of the blends were also for 7 min). Analyses were made as per the standard of fat and oil de-
investigated as qualitative and quantitative analyses as well that can rivatives (TS EN 14103:2011).
valorize both WCO and SCGO in one pot. Consequently, such a study DU was calculated from the proportion of monounsaturated and
opens new directions for the biodiesel research area in the future. In polyunsaturated fatty acids using the following equations [4,52,54,55]:
conclusion, authors have recommended to further examining the effects
of these blends in CI ignition diesel engines through engine emissions DU = (monosaturated Cn: 1, wt %) + 2(polunsaturated Cn: 2, wt %)
and combustion characteristics. + 3(polyunsaturated Cn: 3, wt %) (1)
421
A.E. Atabani, et al. Fuel 244 (2019) 419–430
Table 2 was employed to purify the sample crystal before the background col-
Summary of biodiesel production process. lection. The scan results were processed with a computer software
Esterification Parameter Transesterification Parameter program Spectrum.
Hydrochloric acid 1% (v/v oil) Potassium hydroxide 1% (m/m oil) 3.4.2. UV–vis. Spectra
(HCL) (KOH)
UV–vis. spectra provide information about the electronic properties
Methanol (CH3OH) 100% (v/v oil) Methanol (CH3OH) 25% (v/v oil)
Reaction time 3h Reaction time 1.5 h of the oil molecules. All organic compounds absorb light in the visible
Reaction 60 °C Reaction temperature 60 °C and ultraviolet region. It is possible to see various aromatic compounds
temperature in diesel fuel sources [58]. In this study, all the samples were analyzed
by UV–vis. to obtain the spectra and the quartz cuvettes were filled with
the blends. The UV absorption spectra of the biodiesel blends were
obtained using Agilent Technologies Cary 60 UV–vis. Spectro-
MWi represents the molecular weight of the fatty acids and N re- photometer in the wavelength range of 500–200 nm at 1 nm intervals of
presents the number of double bonds in monounsaturated and poly- using standard 1 cm quartz cuvettes at room temperature.
unsaturated fatty acids.
3.4.3. Differential Scanning Calorimetric (DSC) and thermogravimetric
3.3. Fuel modification and selection of binary and ternary blends analysis (TGA) studies
DSC is used to determine the crystallization onset temperature of
Blending of WCOME/SCGOME with But, Pen, Hex and D100 was the samples and can be plotted on the graph against heat flow. The DSC
adopted in this study to modify some properties of WCOME/SCGOME. curves were obtained in a calorimeter model TA Q-2000 with RCS90
Thus, two binary blends of (D80B70) and (D70B30) and three ternary coupled with a cooling system. 3 ± 0.5 mg of each sample was loaded
blends of (D80B10But10), (D80B10Pen10) and (D80B10Hex10) were in an aluminum pan and heat flow was measured differentially by
prepared and kept in reagent bottles. All the blends show good mis- comparing the heat flow of an empty reference pan as a reference. The
cibility and stability over a long period. Properties such as density, rate of heating/cooling was 10 °C min−1 between −80 °C and 20 °C
kinematic viscosity, pour point, cloud point and cold filter plugging under inert nitrogen atmosphere (N2) with a flow rate of 50 mL min−1.
point were examined and the equipment list can be found in our pre- In this study, all the samples and blends were analyzed by TA Q-2000
vious work [52]. The densities of all the samples are predicted math- DSC under the flow of nitrogen. Each DSC test takes about 17 mins to
ematically with a polynomial curve-fitting method, a polynomial of complete. Results were processed with the help of TA Orchestrator
order k in X is expressed in the following form [57]. Version V7.2.2.1.
The thermogravimetric (TG) thermograms of WCO/SCGO,
Y = C + C1 X + C2 X 2 +Ck X k (9)
WCOME/SCGOME, D100, But100, Hex100, Pen100 and their blends
were recorded in a Shimadzu DTG 60H thermogravimetric analyzer
3.4. Spectral studies using platinum pans. 10 mg of each sample was loaded into pans. After
each analysis, platinum pans were cleaned well in order to make the
3.4.1. FT–IR spectra pans ready for the next run. In this analysis, the temperature ranges
The infrared spectra of all the samples were obtained in a Fourier from 25 to 700 °C with a heating rate of 10 °C/min under dry air at-
Transform Infrared Spectrometer by Perkin Elmer Spectrum Two. The mosphere of 100 mL /min to determine the oxidative stability. Each
supporting software was spectrum software. Each spectrum was taken sample took a period of time, almost an hour. A maximum temperature
within few seconds at the resolution of 4 cm−1 and 4 scans, in the re- of 700 °C was used as the mass remains constant until the end of the
gion of 4000–650 cm−1 and 0.5 mL per sample. Cleaning with ethanol analysis and all the chemical reactions were completed at this
Table 3
FAC, DU, LCSF, KV, CN, OS, HHV, IV and SN of WCOME/SCGOME.
FAC Analyses
Fatty acid MWi Structure N Systematic name Formula (%) DU MPn LCSF SN HHV
Myristic 228.37 14:00 0 Tetradecanoic C14H28O2 0.32 54.4 0.174 0.78 0.123
Palmitic 256.40 16:00 0 Hexadecanoic C16H32O2 32.56 62.9 20.480 71.11 12.761
Stearic 284.48 18:00 0 Octadecanoic C18H36O2 6.0 69.3 4.1580 11.81 2.393
Oleic 282.47 18:01 1 Cis-9-Octadecenoic C18H34O2 29.69 29.69 58.86 11.766
Linoleic 280.45 18:02 2 Cis-9-cis-12 Octadecadienoic C18H32O2 28.02 56.04 55.95 11.032
Linolenic 278.43 18:03 3 α-Linolenic acid C18H30O2 1.0 3.0 2.01 0.391
Arachidic 312.53 20:00 0 Eicosanoic C20H40O2 1.86 75.5 1.404 3.33 0.752
Gendoic 310.55 20:01 1 Cis-11-eicosenoic acid C20H38O2 0.29 0.29 0.52 0.117
Behenic 340.58 22:00 0 1-Docosanoic acid C22H44O2 0.26 80 0.208 0.43 0.106
Saturated 41.00 DU, LCSF, KV, CN, OS, HHV, IV and SN
Monounsaturated 29.98 DU 89.02
Polyunsaturated 29.02 LCSF 26.42
Total 100 KV at 40 °C (mm2/s) 3.88
CN 56.64
OS (h) 6.51
HHV (MJ/kg) 39.44
IV (g Iodine/100 g) 83.71
SN 204.8
422
A.E. Atabani, et al. Fuel 244 (2019) 419–430
Table 4
Summary of biodiesel, biodiesel-diesel and biodiesel-diesel-higher alcohols blends properties.
Property WCOME/SCGOME B20 B30 D80B10But10 D80B10Pen10 D80Hex10B10 D100
3
Density at 15 °C (kg/m ) 885 850.1 854.1 842.5 843.3 843.5 842.5
Cloud point (°C) 9 −3 −2 −3 −3 −2 −5
Pour point (°C) 9 −6 −3 −7 −6 −9 −14
Cold filter plugging point (°C) 2 −14 −11 −22 −20 −19 −26
B ≡ WCOME/SCGOME.
Fig. 2a. Infrared spectra of D100, But100, Pen100, Hex100, WCO/SCGO and
WCOME/SCGOME.
Table 5
Prediction of density (kg/m3) at temperature range (15–40 °C). Fig. 2b. Infrared spectra of B20D80, B30D70, B10But10D80, B10Pen10D80
and B10Hex10D80.
Sample Equation R2
423
A.E. Atabani, et al. Fuel 244 (2019) 419–430
Table 6
Functional group frequencies of the fuel samples [63].
Wavenumber, cm−1 Types of vibration Functional Groups
D100 −8.02
WCOME/SCGOME −1.43
But −15.13
Pen −18.09
Hex −22.52
B10But10D80 −5.20
B10Pen10D80 −5.48
B10Hex10D80 −5.76
B20D80 −3.42
B30D70 −2.54
WCO/SCGO −1.99
4.1. FAC, DU, LCSF, KV, CN, OS, HHV, IV and SN of WCOME/SCGOME
The results of FAC (Table 3) confirmed the existence of fatty acid Fig. 4b. Derivative Thermogravimetric analyzes (DTGA) of fuel samples.
methyl ester with saturated fatty acids having 41% of the total fatty
acid methyl ester content compared to 88.5% of SCGOME [2]. This is
respectively compared to 71.2, 10.4 h and 12.4 g iodine/100 g of
followed by a reasonable amount of both mono and polyunsaturated
SCGOME [2] all of which are satisfactory according to ASTM D6751
fatty acids (29.98% and 29.02% respectively) compared to 10.3% and
standards.
1.2% of SCGOME [2]. DU and LCSF of WCOME/SCGOME were 41.25
and 49.27 compared to 13.9 and 61.76 of SCGOME [2] respectively.
4.2. Blends properties
This excellent fatty acid profile positively affects the final fuel proper-
ties and proves the advantage of feedstock pre-blending strategy. This
Additionally, the results indicate that the advantage of higher al-
profile resulted in a cetane number of 56.64, oxidation stability of
cohols addition to the final fuel properties (Table 4). Compared to pure
6.51 h, HHV of 39.44 MJ/kg and iodine value of 83.71 g iodine/100 g
biodiesel and biodiesel-diesel blends, the three biodiesel-diesel-higher
424
A.E. Atabani, et al. Fuel 244 (2019) 419–430
Table 8 alcohols blends have closer densities (842.5–843.5 kg/m3) and cold
TGA analysis of fuel samples. filter plugging points (−19 to −22 °C) to D100 (842.5 kg/m3 and
Sample name Heating Temperature Onset Mass loss −26 °C respectively). Furthermore, Fig. 1 shows the effects of tem-
rate (°C/ Ranges (°C) temperature for under perature on densities of all samples. Table 5 summarizes the formulae
min) thermal dry air to predict the density of each sample at a temperature range of about
degradation (°C) (%) 15–40 °C.
WCO/SCGO 10 30–400 240.2 99.7
WCOME/SCGOME 10 95–350 185.3 97.0 4.3. FT–IR spectral studies
But100 10 40–115 78.4 98.1
Hex100 10 50–165 102.4 99.0 FT–IR has been employed as an easy modern analytical technique to
Pen100 10 45–165 87.9 98.0
identify the existing functionals and type of binding linkages, corre-
D100 10 75–290 127.3 97.8
B30D70 10 30–345 206.5 96.4 sponding to various stretching and bending vibrations in the biodiesel
B20D80 10 75–280 167.7 97.3 samples. Oils and esters are noted as fairly strong absorbers in the in-
B10But10D80 10 30–300 166.4 96.3 frared region of the electromagnetic spectrum [59]. Biodiesel is often
B10Hex10D80 10 30–280 170.2 92.4
blended with Euro diesel to make the biodiesel blends [60]. In recent
B10Pen10D80 10 70–280 162.4 96.7
years, the addition of alcohols has been investigated to improve the
B ≡ WCOME/SCGOME. properties of the fuel used in diesel engines. Alcohols contain alkyl and
hydroxyl functional groups. As the number of carbons in the carbon
chain of alcohols increases, it is easier to blend with organic compo-
nents. Thus, biodiesel can be mixed with higher alcohols more easily
than with lower alcohols. Fuel with alcohol additive is called oxyge-
nated fuel, and this type of fuel is thought to contribute positively to-
wards the environment through engine emissions. The previous studies
focused on lower alcohols such as methanol (CH4O) and ethanol
(C2H6O). However, some issues contribute to poor fuel properties and
performance such as lower cetane number, flash point, and heat of
evaporation besides corrosion and miscibility issues [61]. Recently, the
higher alcohols such as propanol (C3H8O), butanol (C4H10O), and
pentanol (C5H12O) are getting more attention in diesel engines [62]. In
this study, WCOME/SCGOME and its blends with D100, But, Pen and
Hex was analyzed to determine the existence of methyl ester and al-
cohols functionals by FT–IR spectra. The peaks of these spectrum were
identified and interpreted and shown in Figs. 2a & 2b. The lists of
functional groups identified were shown in Table 6 [63]. This led to the
identification of functional groups in the WCOME/SCGOME and its
blends with the aid of the structure correlation chart. It can be seen
from Figs. 2a & 2b that the absorbance peak heights of each sample
were taken at nearly 1745 cm−1 where strong absorption of fatty acid
methyl esters exist. The strong characteristic absorption bands for
ν(eOeH) occurs nearly at 3324 cm−1 in But, Pen and Hex and blends
with higher alcohols (B10But10D80, B10Pen10D80, B10Hex10D80) in
different intensities of absorbance. When the biodiesel-diesel samples
were blended with different higher alcohols, the intensities of the ab-
sorbance decreased as the number of carbons increased.
The results confirmed the existence of fatty acid methyl esters
(FAME) and alcohols. The FT–IR spectrum for the samples revealed the
functional groups with their characteristics bands especially for ν (eC]
O), ν (eCH2)n, ν(eCeO), ν(C]C) and ν(CeH) in the spectrum. The
peaks for the blends show variations at nearly about 1745 and between
1240 and 1030 cm−1 which all are responsible to ν (eC]O) and ν
(eCeO) functionals respectively. The peak intensities vary according to
the biodiesel ratio. These peaks increase with the concentration of
biodiesel in each blended sample. Variations in the intensities from 722
to 688 cm−1indicate the existence of (]CeH) bending; cis-dis-
ubstituted alkenes and aromatic functionals. The intensities of peaks at
2949 cm−1, 2920 cm−1 and 2858 cm−1 are attributed to ν (CeH)alkyl &
alkenes and are found to remain unchanged in each of the blend sample,
which is an indication of the existence of D100 [64].
425
A.E. Atabani, et al. Fuel 244 (2019) 419–430
Fig. 5b. 1H NMR spectra of D100, WCOME/SCGOME, WCO/SCGO, But, Pen, Hex, D80B10But10, D80B10Pen10 and D8010Hex10.
600–200 nm on account of INCT (Intra Charge Transfer) (n − π*) well as to study the mode of decomposition of the fuel materials. In
transitions. When the spectra were plotted, the change in suction den- addition, the DSC studies provide the cold flow properties of the fuel
sity can be ignored. The absorbance for all samples also produced very materials as a function of temperature and time; however, these studies
similar absorbance, with no statistically significant differences. It is provide the crystallization phenomenon of the fuel materials.
concluded that the absorbance in this band is not affected by the type of The crystallization onset temperatures of all the samples under N2
alcohol, as the blended biodiesel does not have chemical components atmosphere were studied using DSC and presented in Table 7. The onset
that contribute to significant absorption in this range. temperatures for WCOME/SCGOME, D100, But, Pen and Hex are
−1.43, −8.02, −15.13, −18.09 and −22.52 °C respectively and
4.5. DSC and thermal studies hence biodiesel crystallizes faster than diesel and higher alcohols. The
crystallization onset temperatures of B10But10D80, B10Pen10D80 and
Both the thermal (TGA and DSC) analysis techniques are quick, B10Hex10D80 are −5.20, −5.48 and −5.76 °C respectively, below
sensitive and relatively inexpensive that can be utilized to determine WCOME/SCGOME which signifies the addition of higher alcohols
the boiling point of methyl ester formed and also to check the chemical lowers the crystallization temperature of biodiesel. Addition of diesel
changes that occurs during transesterification reaction. From TGA and and higher alcohols reduces the crystallization of biodiesel [65]. The
DTA studies, the thermograms clearly implies that the composition crystallization onset temperature of B30D70 and B20D80 are −2.54 °C
present in the fuel materials and the formed intermediate compounds or and −3.42 °C respectively, below WCOME/SCGOME which represents
oxidative stabilized compounds during the decomposition pathways as that addition of biodiesel increases the crystallization temperature of
426
A.E. Atabani, et al. Fuel 244 (2019) 419–430
13
Fig. 6a. C NMR spectra of D100, WCOME/SCGOME, B20D80 and B30D70.
diesel. The oxidative stability of all the samples was quantified by between the curves, with small variations of Tonset between the blends.
Thermogravimetric analysis (TGA), which is a technique for char-
acterizing thermal and oxidative stability of a material (compound or 4.6. 1H and 13
C NMR spectral studies
mixture) by measuring changes in its physicochemical properties. These
properties were expressed in terms of weight change as a function of In this study, all the samples were characterized by 1H NMR spec-
increasing temperature. The oxidative stability is a quality indicator troscopy and their spectra are shown in Figs. 5a and 5b. The char-
parameter for esters. Thermogravimetric curves (TGA) show the weight acteristic peaks of methoxy protons δ(eOeCH3) were observed as a
loss in different temperatures and the first derivative of the weight loss singlet at around 3.65, 3.67 and 3.68 ppm for WCOME/SCGOME,
in relation to the temperature curves. B20D80 and B30D70, respectively. These the chemical shifts indicate
TGA/DTGA curves show that oxidative degradation occurred in a the presence of methyl esters in biodiesel. Other observed peaks were
single continuous step in the temperature range 75–240 °C for WCO/ around between 0.87 and 0.90 ppm of the terminal methyl δ (eCH3)
SCGO ester and its blends, shown in Figs. 4a and 4b respectively. In protons [65]. Certain charateristic signals obtained were ascribed to
these temperature ranges, the mass losses occur at levels of 92.4–99.7% protons α- to the methylene carbonyl groups δ (eCH2eC]Oe) at
for WCO/SCGO biodiesel and its blends. The oxidative degradation 1.88–1.96 ppm, at 2.57–2.69 ppm responsible for diallylic methylene
occurred at temperature of 78.4, 87.9, 102.4 and 127.3 °C, for But100, proton δ (eC]CeCH2eC]Ce) and at 4.17 ppm for the protons which
Pen100, Hex100 and D100 respectively. In this temperature range, the are in α- and β-position in glyceryl functionals in WCOME/SCGOME,
mass losses occur at levels of 98.198, 99and 97.8% for But100, Pen100, B20D80 samples.
Hex100 and D100 respectively. The oxidation onset temperatures were A representative spectrum of 13CNMR of all the samples are shown
represented in Table 8. It is possible to observe the great similarity in Fig. 6a and 6b, which shows the characteristic peaks of ester and
427
A.E. Atabani, et al. Fuel 244 (2019) 419–430
13
Fig. 6b. C NMR spectra of D100, WCOME/SCGOME, WCO/SCGO, But, Pen, Hex, D80B10But10, D80B10Pen10 and D8010Hex10.
carbonyl δ (eCOOe) and δ (eC]Oe) at around 170 and 50 ppm, re- improves the density and cold flow properties of biodiesel while
spectively. The peaks around 130 and 125 ppm indicate the unsatura- keeping acceptable level for cetane number, flash point and HHV rather
tion in methyl esters. Other peaks at 14 ppm are due to the terminal than lower alcohols. Therefore, this work further shall be further car-
carbon of methyl groups δ (eCH3) and signals at 10–35 ppm are related ried on assessing the performance in compression ignition engines in
to methylene carbons δ (eCH2e) of long carbon chain in FAMEs [66]. the future study.
428
A.E. Atabani, et al. Fuel 244 (2019) 419–430
429
A.E. Atabani, et al. Fuel 244 (2019) 419–430
430