Original papers

INFLUENCE OF ACTIVATED BIOMASS FLY ASH ON

PORTLAND CEMENT HYDRATION
#
RIMVYDAS KAMINSKAS, VYTAUTAS CESNAUSKAS

Department of Silicate Technology, Faculty of Chemical Technology,

Kaunas University of Technology, Kaunas, Lithuania

#
E-mail: [email protected]

Submitted June 20, 2014; accepted November 23, 2014

Keywords: Portland cement, Pozzolana, Microfiller, Biomass fly ash, Hydration

This study investigated the possibility of using biomass fly ash as an additive for ordinary Portland cement (OPC), where
5 %, 10 %, 15 %, and 25 % (by weight) of the Portland cement was replaced with biomass fly ash (BFA) and tribochemically
activated biomass fly ash (AFA). The mixture was then hardened for 28 days in water at 20°C. It is estimated that both
additives are distinguished for very low pozzolanic activity. Under normal conditions, the BFA additive had a negative
effect on the strength properties of cement. After 28 days of hydration, the compressive strength of samples with 15 wt. %
BFA additive was more than 10 % lower than that of OPC (54 MPa), whereas the compressive strength of the sample with
25 wt. % additive (35 MPa) was 30 % lower than that of the OPC. Additional milling of biomass fly ash samples changes the
particle size distribution and composition of the compounds. Tribochemically activated biomass fly ash additive significantly
accelerates the hydration process of calcium silicates and had a positive effect on the strength properties of the samples,
because up to 15 wt. % of the cement can be replaced with this additive without impairing the strength properties of Portland
cement. Also, activated biomass fly ash impedes the alkali-silica reaction in cement mortars.

INTRODUCTION Many kinds of biomass fly ash have pozzolanic

properties similar to coal fly ash, such as those from rice
In recent years, pressures on the global environment husk, wood, wheat straw, and sugar cane straw [5-7],
and energy security have led to an increasing demand some of which have been added in concrete as mineral
for renewable energy sources and diversification of admixtures, improving the performance of concrete.
Europe’s energy supply. Among these resources, biomass Further study of pozzolanic reaction kinetics has found
could play an important role because it is considered that at one month of curing, coal fly ash (Class C and F)
a renewable and CO2-neutral energy resource once the and biomass fly ash (either cofired or wood fly ash by
consumption rate is lower than the growth rate and can itself) have undergone significant pozzolanic reactions at
potentially provide energy for heat, power, and transports ambient temperatures [8]. On the other hand, some data
from the same installation [1]. published in the literature [9] shows that the chemical
The characteristics of ashes from biomass com- composition of wood ash indicates very high LOI, high
bustion vary widely and depend on the cleanliness of alkali content, and very low SiO2 values (and hence do
the biomass fuel, combustion technology, the collection not satisfy the requirements to be classified as Class
location, and other factors [2, 3]. Therefore, the biomass N pozzolans). The authors also state that a high SiO2
composition can be highly variable depending on the content in the biomass ash does not automatically imply
geographic locations and industrial processes. The che- an effective pozzolanic material (high efficiency factor).
mical composition of biomass fly ash is an important It is known that a higher quantity of water is used
property to consider as a supplementary cementing for waste wood ash concrete [10, 11]. The compressive
material in blended cement. The general term “the strength increases with duration, but it decreases with
activity of additives” defines the chemical activity the use of more additive. It was determined that the
(usually pozzolanic activity) and microfiller effect. The compressive strength of waste wood ash concrete is
pozzolanic activity is the ability to react with portlandite lower than that of concrete without additives [12-14].
(Ca(OH)2) in the presence of an excess of water, which This can be explained by the different behavior of
is expressed in the quantity of bound CaO mg per 1 g waste wood ash, because it acts more as a microfiller
of additive. Moreover, the microfiller effect depends in concrete matrix than as a binder. Tkaczewska at al.
mainly on the shape and size of the particles, the particle [15] found that, during hydration, the cement samples,
size distribution, and specific surface area [4]. which contain coal and biomass fly ash, show lower heat
260 Ceramics – Silikáty 58 (4) 260-268 (2014)
 Influence of activated biomass fly ash on portland cement hydration

flow values, a higher quantity of Ca(OH)2, and slower the samples for instrumental analysis were prepared
C3S hydration in comparison with samples containing without the sand that is usually used. The storage con-
bituminous coal fly ash as an additive. ditions of the samples were analogous to those used
The aim of this work is to investigate the influence during the strength test. The hydration of the samples
of biomass fly ash (wood chips and waste wood) on the was stopped using acetone.
hydration and hardening processes of Portland cement. The pozzolanic activity was assessed using the
modified Chapelle method. This test consists of placing
1.000 g of a mineral admixture into 500 ml of a lime
EXPERIMENTAL solution (1.200 g∙l-1 CaO). The solutions were kept
for the first 48 h in a thermostat at 45°C. At the end of
The fly ashes used in the experiment were from this period, 50 ml of the solution was taken, and the
a biomass (wood chips and waste wood)-fired power CaO content was determined by titration with 0.05 N
plant in Lithuania. The main chemical composition hydrochloric acid (HCl) solution using methyl orange as
and characteristics of the biomass fly ash are shown the indicator. The results were expressed in milligrams
in Table 1. Ordinary Portland cement CEM I 42.5 R of fixed CaO per gram of pozzolanic additive. The rest
was used in this work. The chemical composition of the solution (450 ml) was again kept for 24 h at 45°C.
of the cement is shown in Table 1. The mineralogical The process was repeated until the estimated value of the
composition of the clinker was as follows: 3CaO·SiO2 pozzolanic activity was insignificantly low (7 days).
57.80 wt. %; 2CaO·SiO2 22.15 wt. %; 3CaO·Al2O3 6.65 For the alkali-silica reaction test, reactive quartzitic
wt. %; and 4CaO·Al2O3·Fe2O3 13.40 wt. %. fine gravel is used as the aggregate. The aggregate is
ground down to 0.125 mm, and the comminuted material
Table 1. Chemical composition and characteristics of raw is used to produce a mortar (4 × 4 × 16 cm mortar prisms).
materials. The binder-to-aggregate ratio was 1:3, and the water-
to-binder ratio was 0.5. Samples were formed of pure
Component Biomass Portland
ordinary Portland cement and Portland cement with 5 %,
(wt. %) fly ash cement
15 %, and 25 % (by weight) replacement with biomass
SiO2 45.08 20.8 fly ash. During the first day, the samples were kept in
Al2O3 2.98 5.7 moulds at 20 ± 1°C and 100 % humidity. After removal
Fe2O3 1.37 4.4 from the moulds, the bars were transferred to storage
CaO 16.58 61.9 containers filled with 1 N NaOH solution maintained at
MgO 1.90 1.0 80 ± 1°C. The lengths of the mortar bars were periodically
K 2O 4.47 – measured over a 14-day period. The expansion value was
Na2O 0.37 0.9 calculated as the average percentage length change of
MnO 0.56 – three mortar bar samples based on length change since
SO3 2.26 2.5 initial immersion in NaOH.
P 2O 5 2.07 – The XRD analysis was performed using the D8 Ad-
TiO2 0.2 – vance diffractometer (Bruker AXS, Karlsruhe, Germany)
ZnO 0.12 – operating at the tube voltage of 40 kV and tube current
BaO 0.10 – of 40 mA. The X-ray beam was filtered with a Ni 0.02
Loss on ignition 8.89 0.4 mm filter to select the CuKα wavelength. Diffraction
Specific surface area (m2∙kg-1) 310 350 patterns were recorded in a Bragg-Brentano geometry
using a fast counting detector Bruker LynxEye based on
silicon strip technology. The specimens of samples were
Samples were formed of pure ordinary Portland scanned over the range 2θ = 3 − 70° at a scanning speed
cement and Portland cement with 5 %, 10 %, 15 %, of 6°∙min-1 using a coupled two theta/theta scan type.
and 25 % (by weight) replacement with fly ash. The Simultaneous thermal analysis (STA) (differential
consistency of the cement paste and the initial and final scanning calorimetry and thermogravimetry) was carried
setting times of the cement were estimated according to out on a Netzsch STA 409 PC Luxx instrument with
European Standard EN 196-3. Samples for compressive ceramic sample handlers and crucibles of Pt-Rh. At a
strength analysis (prisms 4 × 4 × 16 cm) were formed heating rate of 15°C∙min-1, the temperature ranged from
according to European Standard EN 196-1 (the cement- 30°C to 1000°C under the ambient atmosphere.
to-sand ratio was 1:3, and the water-to-cement ratio Infrared spectra were measured using a Perkin-Elmer
was 0.5). During the first day, the samples were kept in Fourier transform infrared (FT-IR) system Spectrum X
moulds at 20 ± 1°C and 100 % humidity. After 24 hours spectrometer. Samples were prepared by mixing 1 mg of
of formation, the samples were transferred to deionized the sample in 200 mg of KBr. The spectral analysis was
water and stored there for 27 days at 20 ± 1°C. In order to performed in the range of 4000 − 400 cm-1 with a spectral
make a more exact estimation of the hydration process, resolution of 1 cm-1.
Ceramics – Silikáty 58 (4) 260-268 (2014) 261
 Kaminskas R., Cesnauskas V.

The particle size distribution of fly ash was measured particle size of BFA estimated by laser diffraction is
by a CILAS 1090 LD laser scattering particle size inaccurate because a significant portion of the particles
distribution analyzer. An amount of 0.1 g of powdered are larger than 500 μm.
sample was put in 100 ml of ethanol and underwent The results of the sieve analysis are shown in
dispersion treatment by an ultrasonic dispersion unit for Table 2. The data of Figure 2 and Table 2 show that
70 s. BFA particle diameter varies over very wide limits, and
The chemical composition analysis of samples were particles larger than 80 μm constitute 78.34 wt. % of the
performed by X-ray fluorescence spectroscopy (XRF) on total amount.
a Bruker X-ray S8 Tiger WD spectrometer equipped with
a Rh tube with energy of up to 60 keV. Powder samples
Table 2. Sieve analysis results of biomass fly ash.
were measured in a helium atmosphere, and the data
were analyzed with SPECTRA Plus QUANT EXPRESS Sieve size Mass Percent Cumulative
standardless software. (μm) residue (g) residue percent
The calorimetric analysis data were gathered using 600 0.33 11 11
a TAM AIR III calorimeter. The range of measurement 425 0.23 7.69 18.69
was ± 600 mW, the sensitivity of the signal was 4 μW, 212 0.97 32.33 51.02
the time constant was < 500 s, the temperature of the 106 0.65 21.66 72.68
experiment was 25 ± 0.1°C, and the water-to-solid ratio 80 0.17 5.66 78.34
was 0.5. < 80 0.65 21.66 100
BFA Mass (g) 3.00 100 –

RESULTS AND DISCUSSION

In the next stage of the research the pozzolanic
In order to monitor the trend of variation of the activity investigation was carried out. It was established
elements in the biomass fly ashes, a number of samples that the pozzolanic activity of BFA was very low and
were collected over several months and were analysed reached only 22 mg CaO/g. The data confirm that
for chemical composition using the XRF technique. The the biomass fly ash properties can be highly variable
results are shown in Figure 1. A variation of the absolute depending on geographic location and industrial pro-
maximum of around 20 % is observed in the distribution cesses because many authors [16-19] note high biomass
of major elements. The variations can be attributed to fly ash pozzolanic activity.
the variation in the operating conditions as well as the Investigation of the compressive strength of the
biomass fuels used for the combustion. samples, in which the content of BFA additive varied
For further research, the sample for which the from 5 to 25 wt. %, showed that this additive had
chemical composition is given in Table 1 was chosen a negative effect on the strength properties of cement. It
(Sample 1). The loss on ignition was 8.89 wt. %. The was estimated (Figure 3) that after 7 days of hydration,
organic content as well as the decomposition of calcite is all samples with BFA additive had lower compressive
the reason for the loss on ignition. strength than OPC samples. When hydration time was
The particle size distribution of biomass fly ash increased to 28 days, the compressive strength of the
(BFA) measured by laser diffraction is shown in Figure 2. samples with BFA additive remained lower than that of
The Blaine value of the BFA powder is 310 m2∙kg-1. The the OPC samples. It should be noted that the compressive

50 100 10

45 Sample 1 90 9
Sample 2 80 8
40
Cumulative value (%)

Sample 3 70 7
Histogram (×20)

35
Content of oxide (%)

Sample 4 60 6
30 1
Sample 5 50 5
25 40 4
20 30 2 3
15 20 2
10 10 1
5 0 0
0.1
0.9
1.5
2.4
3.6
5.3
8
13
19
28
40
60
85
5.3
130
190
280
450

0
SiO2 CaO K2O Al2O3 MgO Fe2O3 SO3 P 2O 5 Diameter (µm)
Figure 1. Variation of chemical composition of the biomass fly Figure 2. Particle size distribution of biomass fly ash (BFA).
ash samples.

262 Ceramics – Silikáty 58 (4) 260-268 (2014)

 Influence of activated biomass fly ash on portland cement hydration

strength of samples with 15 wt. % BFA additive was more Portlandite

than 10 % lower than that of OPC (53 MPa), whereas Quartz
the compressive strength of the sample with 25 wt. % Calcium oxide
Manganese sulfide
additive after 28 days of hydration (35 MPa) was 30 % Aluminum phosphate
lower than that of the OPC. Thus, the investigated Calcium carbonate
biomass fly ash (as-received) is an unsuitable additive K 2Mn 2(SO 4) 3
for Portland cement because it has a negative influence
2
on the strength properties of the OPC.

Intensity (a.u.)
70
Compressive strength (MPa)

60 0% 5% 10 % 15 % 25 %

50

40 1
30

20 3 8 13 18 23 28 33 38 43 48 53 58 63
2θ (°)
10
Figure 5. X-ray diffraction pattern of biomass fly ash; 1 – fly
0
7 days 28 days ash as-received (BFA), 2 – ground fly ash (AFA).
Figure 3. Compressive strength of cement samples with diffe-
rent amounts of BFA additive after 7 and 28 days of hydration. According to the XRD analysis data (Figure 5, cur-
ve 1), quartz (d-spacing: 0.425; 0.334; 0.181 nm; JCPDS
77-1060), CaO (d-spacing: 0.277; 0.240; 0.169 nm;
According to the obtained experimental results, it JCPDS 82-1690), MnS (d-spacing: 0.324; 0.198 nm;
was decided in the next stage of research to grind BFA, JCPDS 40-1288), Al3PO4 (d-spacing: 0.423; 0.334 nm;
which can then be used as microfiller. The BFA was JCPDS 76-228), and calcite (d-spacing: 0.302; 0.228;
ground using a vibrating disc mill to a specific surface 0.209; 0.189; 0.186 nm; JCPDS 72-1650) were found
area of 430 m2∙kg-1. in BFA. After milling of BFA samples, the intensity of
the quartz and Al3PO4 peaks considerably decrease, and
new peaks characteristic of portlandite (d-spacing: 0.489;
100 10
0.262; 0.192 nm; JCPDS 44-1481) were found in AFA
90 9
samples (Figure 5, curve 2). Portlandite formation can be
80 8
explained by the fact that during the grinding process CaO
Cumulative value (%)

70 7
becomes very active, joins the environmental humidity,
Histogram (×20)

60 6
1 and becomes portlandite. Thus, grinding of BFA results
50 5
not only in mechanical fly ash particle milling, but it also
40 4
changes the composition of the compounds; therefore
30 2 3 this process can be called a tribochemical activation of
20 2 biomass fly ash.
10 1 The results of the XRD data were confirmed by STA
0 0 and FT-IR data (Figures 6 and 7). In the DSC curve of
0.1
0.9
1.5
2.4
3.6
5.3
8
13
19
28
40
60
85
5.3
130
190
280
450

the AFA sample, three significant endothermic effects

Diameter (µm) at 440°C, 575°C, and 705°C were found, which are
Figure 4. Particle size distribution of additionally milled attributed to the decomposition of portlandite, α-mo-
biomass fly ash (AFA). dification quartz converting to β-modification, and de-
composition of calcite, respectively.
The particle size distribution of additionally milled In the FT-IR curve of the AFA sample the absorption
biomass fly ash (AFA) is shown in Figure 4. After maximum characteristic of portlandite at 3644 cm-1
milling, 100 wt. % of the particles are lower than 450 μm, frequency was observed [20]. The absorption bands of
and the uniformly distributed particles are dominated by calcium carbonate that are described by the stretching
diameters less than 63 μm, constituting 70 wt. % of the ν3–CO32- (713 cm-1), ν2–CO32- (875 cm-1), and ν3–CO32-
total amount. Thus, these prepared biomass fly ashes (1430 cm-1) vibrations were also visible in the spectrum
can be used as a microfiller (EN 12620:2003+A1:2008). [21]. The infrared spectra of silicate in the range
It was determined that the pozzolanic activity of 1200 − 400 cm-1 are classified into four characteristic
additionally milled biomass fly ash was higher than BFA, bands around 1000 (overlapped), 780, 695, and 450 cm-1
but it remained very low and reached only 30 mg CaO/g. in the standard quartz spectra. Among these four

Ceramics – Silikáty 58 (4) 260-268 (2014) 263

 Kaminskas R., Cesnauskas V.

100 0.3 100

99 0.2 90
98 0.1 80

Transmittance T (%)

695
Heat flow (mW mg-1)

780
DSC

568
70

1194
97 0
Mass (%)

60

1797
96 -0.1

713
601
2514
Exo

50

450
3644
95 -0.2

1122
1055
40

875
1105
Endo

94 -0.3 30
TG

1430
93 -0.4 20
92 -0.5 10
91 -0.6 0
30 180 330 480 630 780 930 4000 3600 3200 2800 2400 2000 1600 1200 800 400
Temperature (°C) Wavenumber ν (cm-1)

Figure 6. STA curves of activated fly ash (AFA) sample. Figure 7. FT-IR curve of AFA sample.

characteristic peak regions, the peak at 695 cm-1 is unique samples. When hydration time was increased to 28 days,
to the crystalline materials [22]. The band observed in the compressive strength of the samples with 5 − 15 wt. %
the FT-IR spectrum around 1122 cm-1 corresponding to AFA additive is higher than that of the OPC samples.
the symmetric stretching mode of [PO4]3- is specific for It should be noted that even for samples with 25 wt. %
Al3PO4 [23]; the presence of two characteristic bands AFA additive the compressive strength was less than
around 568 and 601 cm-1 correspond to the ν4 (O–P–O) 10 % lower (48.3 MPa) than that of the OPC samples
bending mode [24]. The frequencies ν1, ν2, ν3, and ν4 (53 MPa). Thus, the tribochemically activated biomass
of the SO42- ion occur at 983 (overlapped), 450, 1105, fly ash is an available additive for Portland cement
and 611 cm-1, respectively [25]. An absorption band because up to 15 wt. % of the cement can be replaced with
characteristic of MnS can’t be identified, because metal- this additive without impairing the strength properties of
sulphide stretching vibrations occur below 400 cm-1 [26]. Portland cement.
In the next stage of research the influence of AFA
additive on the cement hydration process was investi- 70
gated. The results of physical tests–setting time and 0% 5% 10 % 15 % 25 %
Compressive strength (MPa)

60
normal consistency of different cement compositions–
are summarised in Table 3. Taking these data into 50
account, a negligible increase of the setting time and a 40
continuous increase of the normal consistency according
30
to the increasing percentage of added AFA additive were
determined. This is explained by the partial substitution 20
of cement by adding a character with fly ash, causing 10
excessive consumption of water and absorption of part
0
of the water for hydration. 7 days 28 days
Investigation of the compressive strength of the
Figure 8. Compressive strength of cement samples with diffe-
samples, in which the content of AFA additive varied
rent amounts of AFA additive after 7 and 28 days of hydration.
from 5 to 25 wt. %, showed that this additive had a
positive effect on the strength properties of cement. It
was estimated (Figure 8) that after 7 days of hydration, In order to assess the influence of fly ash additive
samples with 5 and 10 wt. % AFA additive had higher on the initial hydration of cement, a measurement of heat
strength, whereas samples with 15 wt. % AFA additive evolution during the hydration process (calorimetric
had the same compressive strength (46.8 MPa) as OPC analysis) was performed (Figure 9). As shown in Figure 9,
two sharp peaks of heat evolution on the calorimetric
Table 3. Influence of AFA on normal consistency and setting curve of cement were noted: the first indicates the active
time of portland cement. evolution of heat due to wetting a powder of cement and
the initial kinetic reaction, during which Ca2+, OH-, SiO44-
Amount of Normal consistency Setting time (min) , and SO42- ions pass into the solution, and the second
additive (wt. %) (W/C*) (%) initial final has to do with the reaction between the deeper layers of
0 0.255 89 139 C3S particles and water [27, 28]. During hydration of the
5 0.265 95 153 cement samples with AFA (except sample with 5 wt. %
15 0.275 105 155 additive), on the released heat flow measurement curve
25 0.285 104 160 (Figure 9a, curves 3, 4), an earlier flow of heat evolution

264 Ceramics – Silikáty 58 (4) 260-268 (2014)

 Influence of activated biomass fly ash on portland cement hydration

was identified at the second period of hydration than centers, around which portlandite and calcium silicate
in the pure cement sample (Figure 9a, 1 curve). The hydrate begin to crystallize. In both cases, the Ca2+
maximum heat flow value of the sample with 15 wt. % of concentration reduces in the solution, and due to this the
AFA additive was reached 5 h 50 min after the beginning cement hydration accelerates [27, 28]. All of the samples
of hydration. For the sample with 25 wt. % of AFA with AFA additive up to 70 h of hydration emit less heat
additive it was reached 5 h 30 min after the beginning energy (Figure 9b, curves 2-4). This can be explained by
of hydration, whereas for pure cement and the sample the fact that the AFA additive doesn’t have binding or
with 5 wt. % of AFA additive (Figure 9a, curves 3, 4) pozzolanic properties; therefore the hydration with less
it was reached after 6 h 20 min. Therefore, it can be cement emits less heat energy.
assumed that the AFA additive significantly accelerates The results of XRD analysis showed (Figure 10a)
the initial cement hydration. It is possible that during the that after 7 days of hydration, in all X-ray patterns of samp-
induction period Ca2+ ions adsorbed on the surface of the les, the diffraction peaks of unhydrated tricalcium silicate
fly ash or finely ground fly ash can act as crystallization (d-spacing: 0.304; 0.277; 0.260; 0.218 nm; JCPDS 42-551)

0.006 350

0% 300
0.005
5%
15 % 250
Heat flow (W g-1)

0.004 25 %
Heat (J g-1)

200
0.003
150
0.002 0%
100
5%
0.001 15 %
50
25 %
0 0
0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70
Time (h) Time (h)
a) thermal flow b) heat

Figure 9. Calorimetric curves of thermal flow a) and heat b) in samples with different amounts of AFA additive.

Portlandite
Portlandite
Quartz
Quartz
Calcium silicate hydrate
Calcium silicate hydrate
Brownmillerite
Brownmillerite
Alite
Alite
Calcite
Calcite
4 4
Intensity (a.u.)

Intensity (a.u.)

3 3

2 2

1 1

3 8 13 18 23 28 33 38 43 48 53 58 63 3 8 13 18 23 28 33 38 43 48 53 58 63
2θ (°) 2θ (°)
a) 7 days b) 28 days

Figure 10. X-ray diffraction patterns of cement samples with AFA additive cured for 7 a) and 28 b) days under normal conditions:
1 – pure Portland cement; 2 – Portland cement with 5 wt. % additive; 3 – Portland cement with 15 wt. % additive; 4 – Portland
cement with 25 wt. % additive.

Ceramics – Silikáty 58 (4) 260-268 (2014) 265

 Kaminskas R., Cesnauskas V.

and unhydrated brownmillerite (d-spacing: 0.725 nm; with 5 and 15 wt. % of AFA additive the reduction
JCPDS 30-226) were identified. Also, the ordinary pro- was greater, and in the samples with 25 wt. % of AFA
duct of cement hydration, portlandite (d-spacing: 0.489; additive, only a trace of unhydrated C3S was observed. It
0.262 nm; JCPDS 44-1481), was identified in all curves. should be noted that the intensity of peaks characteristic
It was determined that, after 7 days of hydration, peaks of CSH and portlandite grows with an increasing amount
of silica dioxide (d-spacing: 0.425; 0.334 nm; JCPDS of AFA additive in the samples. The higher intensity of
77–1060) were visible only in samples with fly ash the diffraction peaks of portlandite may be associated
(Figure 10 a, curves 2-4), because quartz is an integral with a more intense hydration process of calcium silicate
part of this additive. In addition to the main products and with portlandite as one of the AFA components;
in XRD curves, the characteristic peaks of CaCO3 however, the higher intensity of CSH peaks can be
(d-spacing: 0.302; 0.189; 0.186 nm) were observed. attributed only to the more intense hydration process of
The peak of calcium carbonate in the samples is related calcium silicates.
to the carbonization process of them, and a negligible In all DSC curves of hydrated samples (Figure 11),
content of CaCO3 is an integral part of cement (up to three significant endothermic peaks in the temperature
5 wt. %) and biomass fly ash. It should be noted that ranges of 100 − 200°C, 440 − 460°C, and 725 − 760°C
a low-intensity characteristic peak of calcium silicate were observed. The endothermic peak at 100 − 200°C
hydrate (d-spacing: 0.305; 0.271; 0.190 nm; JCPDS is due to the dehydration of most cement hydration
3-588) was identified in the XRD curves of the samples products (calcium silicate hydrates, calcium aluminates,
with AFA additive (Figure 10 a curves 2-4). ettringite, etc.), whereas those at 440 − 460°C and
After 28 days of hydration the nature of the XRD 725 − 760°C indicate the decomposition of portlandite
curves remains similar to the curves after 7 days of and calcium carbonate, respectively.
hydration: peaks characteristic of portlandite, unhydrated The results of TG analysis (Table 4) showed that
C3S, CSH, CaCO3, and quartz were identified in the when the amount of AFA additive was increased, after 7
curves. The intensity of unhydrated tricalcium silicate days of hydration the mass loss due to the presence of
diffraction peaks was slightly reduced in the cement portlandite (~ 455°C) was also increased. The higher
sample (Figure 10b, curve 1). Meanwhile, in the samples mass loss was also prominent on the samples with AFA

4 4

126 725
Exo

Exo

3 3 142
451 462 752

140 730
2 2 138
∆Q

∆Q

452 457 747

137 137
1 735 1 459
Endo

Endo

753
462

137 729 137

451 737
456

40 140 240 340 440 540 640 740 40 140 240 340 440 540 640 740
Temperature (°C) Temperature (°C)
a) 7 days b) 28 days

Figure 11. Differential scanning calorimetry patterns of cement samples with AFA additive cured for 7 a) and 28 b) days under
normal conditions: 1 – pure Portland cement; 2 – Portland cement with 5 wt. % additive; 3 – Portland cement with 15 wt. %
additive; 4 – Portland cement with 25 wt. % additive.

Table 4. Thermogravimetry analysis results of samples cured for 7 and 28 days under normal conditions.
Amount Mass loss (wt. %)
of AFA additive After 7 days of hydration After 28 days of hydration
(wt. %) 100 − 140°C ~ 455°C 680 − 760°C 100 − 140°C ~ 455°C 680 − 760°C
0 1.58 1.83 3.66 1.68 2.0 4.69
5 1.61 1.85 4.09 1.72 2.16 5.05
15 1.70 1.91 4.61 1.81 2.24 5.12
25 1.74 2.15 4.65 2.08 2.67 5.25

266 Ceramics – Silikáty 58 (4) 260-268 (2014)

 Influence of activated biomass fly ash on portland cement hydration

additive in the area of calcium silicate hydrate dehydration higher compressive strength of the samples with additive
(90 − 140°C). A clear tendency was observed: when the (up to 15 wt. %) after 7 and 28 days of hydration. It
amount of AFA additive increases in the samples, the should be noted that in samples with AFA additive
mass loss in decomposition areas of portlandite and greater amounts of both hydration products (CSH and
calcium silicate hydrates also increases. This confirmed portlandite) were formed, which indicates that in these
the assumption that AFA additive promoted the hydration samples a pozzolanic reaction doesn’t occur–otherwise
of calcium silicates. It should be noted that this tendency part of the CH would react with active SiO2 and the
remains for the samples cured for 28 days under normal amount of CH should decrease.
conditions. Even in the sample with 25 wt. % of additive After 28 days of hydration, in the FT-IR curves
a larger amount of portlandite and CSH were formed (Figure 12) of all samples the absorption band charac-
than in the OPC sample. Considering that 25 % less teristic of portlandite at 3644 cm-1 frequency was obser-
calcium silicates were hydrated in this sample (cement ved. The wide absorption band at ~ 3440 cm-1 frequency
was replaced by AFA), the TG analysis results showed a is assigned to valence ν(H2O) vibrations. The absorption
significant acceleration of the calcium silicate hydration band at 975 cm-1 frequency is characteristic of ν(Si-O)
process. This influence of AFA additive explained a vibrations, which identifies calcium silicate hydrates,
whereas ~ 465 cm-1 is characteristic of the internal
δ[SiO4]4- tetrahedral deformations. The absorption bands
characteristic of calcium carbonate at 713 cm-1, 875 cm-1,
4 and 1420 cm-1 also were observed in FT-IR curves of all
samples.
3 In FT-IR curves of the samples with AFA additive
Transmittance T (%)

the intensity of the absorption band of portlandite

2
(3644 cm-1) was higher than in the pure cement sample.
Also, the intensity of absorption bands at 973 cm-1 and
especially at 465 cm-1 in the samples with AFA additive
1110

1
was higher than in the pure cement sample. This confirms
500
450

the previous statement that portlandite and calcium

3644

silicate hydrate were formed more completely in the

713
3440

1120

465
975
875

1420 samples with AFA additive.

4000 3600 3200 2800 2400 2000 1600 1200 800 400 The absorption band shoulder at ~ 1120 cm-1, cha-
Wavenumber ν (cm-1) racteristic of δ(S–O) vibrations in [SO4]2- groups and
assigned to ettringite [29], was observed in the pure
Figure 12. FT-IR curves of cement samples with AFA additive
cement sample. Meanwhile, in the samples with AFA
cured for 28 days under normal conditions: 1 – pure Portland
additive, the intensity of absorption bands characteristic
cement; 2 – Portland cement with 5 wt. % additive; 3 –
Portland cement with 15 wt. % additive; 4 – Portland cement of S–O vibrations in sulfates were reduced and dis-
with 25 wt. % additive. placed to a lower wavelength range (~ 1110 cm-1) typi-

0.150 0.40
CEM+0 %
0.35
0.125 CEM+5 %
CEM+15 % 0.30
Length expansion (%)

Mass expansion (%)

0.100 CEM+25 %
0.25

0.075 0.20

0.15
0.050 CEM+0 %
0.10 CEM+5 %
0.025 CEM+15 %
0.05
CEM+25 %
0 0
1 5 8 12 14 1 5 8 12 14
Time (h) Time (h)

a) length expansion b) mass expansion

Figure 13. Length expansion a) and mass expansion b) of cement samples with AFA additive after 14 days of storage in 1 molar
NaOH solution at 80°C.

Ceramics – Silikáty 58 (4) 260-268 (2014) 267

 Kaminskas R., Cesnauskas V.

cal of monosulfoaluminate [30]. This shows that AFA 4. Activated biomass fly ash impedes the alkali-silica
additive promotes recrystallization of ettringite to mono- reaction in cement mortars.
sulfoaluminate. It is possible that the SO42- ions, like Ca2+
ions, were adsorbed on the surface of ground biomass fly
ash particles. REFERENCES
Because biomass fly ash contains a number of 1. Rajamma R., Ball R.J., Tarelho L.A.C., Allen G.C., Labrincha J.A.,
alkalis, in the next stage of research the influence of Ferreira V.M.: J. Hazard. Mater. 172, 1049 (2009).
2. Yin C., Rosendahl L.A., Kaer S.K.: Prog. Energy Combust. Sci. 34,
AFA additive on the course of the alkali-silica reaction
725 (2008).
in cement mortars was determined. It was found that the 3. Obernberger I., Biedermann F., Widmann W., Riedl R.: Biomass
mass of all investigated samples during maintenance Bioenergy 12, 211 (1997).
in 1 M NaOH solution at 80°C for up to 14 days 4. Shvarzman A., Kovler K., Schamban I., Grader G.S., Shter G.E.:
Adv. Cem. Res. 13, 1 (2001).
(Figure 13b) increased exponentially; however, in 5. Martirena F., Middendorf B., Day R.L., Gehrke M., Roque P.,
samples with an AFA additive the mass gain was less Martínez L., Betancourt S.: Cem. Concr. Res. 36, 1056 (2006).
than in the pure OPC sample. The reduced mass increase 6. Yu Q., Sawayama K., Sugita S., Shoya M., Isojima Y.: Cem Concr.
Res. 29, 37 (1999).
is probably due to the fact that the addition of AFA 7. Naik T.R., Kraus R.N.: Concr. Int. 25, 55 (2003).
additive provides a better grain-size composition in the 8. Wang S., Dalton R., Bragonje S., Tullis J., Baxter L. in: Kinetics of
mixture and a denser cement stone structure. The biggest pozzolanic reaction and strength build-up in the mortar of biomass/
coal fly ash with calcium hydroxide, p. 99-110, Ed. Malhotra V. M.,
change in length (Figure 13a) is also observed in the pure American Concrete Institute, Miami 2007.
cement sample (0.095 %), while in samples with AFA 9. Demis S., Rodrigues C.S., Santos L.A., Papadakis V.G. in: Proc.
additive the length expansion is less and varies from XIII DBMC - 13th International Conference on Durability of
Building Materials and Components, Sao Paulo, Brazil; 2013
0.088 % (sample with 25 wt. % additive) to 0.023 % (available online: www.researchgate.net/publication/260287514_
(sample with 15 wt. % additive). It was observed that An_investigation_of_the_effectiveness_and_durability_
the expansion rate of all mortars was high in the initial characteristic_of_biomass_ashes_as_pozzolanic_materials).
10. Udoeyo F.F., Inyang H., Young D.T., Oparadu E.E.: J. Mater. Civ.
stages, and after that it seemed to decrease. According
Eng. 18, 605 (2006).
to the applied methodology, the critical value of length 11. Fehrs J.E. in: Ash from the combustion of treated wood:
expansion is 0.1 %. Thus, the expansion of the reference characterization and management options. The National Bioash
mortar prepared with the OPC was only slightly less than Utilization Conference, Portland, 1996.
12. Etiegni L. in: Wood ash recycling and land disposal, p. 174, Thesis
the critical value; meanwhile, the expansion results for Ph.D., Department of Forest Products. University of Idaho at
mortar mixtures containing activated biomass fly ash Moscow, Idaho, 1990.
showed a reduction in the expansion compared to the 13. Abdullahi M.: Leonardo El. J. Pract. Technol. 89, (2006).
14. Wang S., Miller A., Llamazos E., Fonseca F, Baxter L.: Fuel. 87,
plain cement mortars. 365 (2008).
15. Tkaczewska E., Małolepszy J.: Constr. Build. Mater. 23, 2694
(2009).
CONCLUSIONS 16. Wang S., Baxter L.: Fuel Process. Technol. 88, 1165 (2007).
17. Elinwa A.U., Mahmood Y.A.: Cem. Concr. Compos. 24, 219
1. The investigated biomass fly ash (as-received) is (2002).
distinguished by an uneven grain-size composition 18. Cheah C.B., Ramli M.: Resour. Conserv. Recycl. 55, 669 (2011).
19. Naik T.R.: Tests of wood ash as a potential source for construction
and very low pozzolanic activity and had a negative materials (Report No. CBU-1999-09), p. 1-61, UWM Center for
effect on the strength properties of Portland cement By-Products Utilization, Department of Civil Engineering and
samples. After 28 days of hydration, the compressive Mechanics, University of Wisconsin-Milwaukee, Milwaukee,
strength of samples with 15 wt. % BFA additive was 1999.
20. Ylmen R., Jaglid U., Steenari B.M., Panas I.: Cem. Concr. Res. 39,
more than 10 % lower than that of OPC (54 MPa), 433 (2009).
whereas compressive strength of the sample with 21. Stepkowska E.T.: J. Therm. Anal. Calorim. 84, 175 (2006).
25 wt. % additive (35 MPa) was 30 % lower than that 22. Saikia B.J., Parthasarathy G., Sarmah N.C.: Bull. Mater. Sci. 31,
775 (2008).
of the OPC. 23. Hepzi Pramila Devamani R., Alagar M.: Int. J. Appl. Sci. Eng. Res.
2. After additional milling of biomass fly ash samples, 1, 769 (2012).
24. Paz A., Guadarrama D., López M., Gonzalez J.E., Brizuela N.,
the particles were distributed uniformly, particle Aragón J.: Quim. Nova. 35, (2012).
diameters less than 63 μm constituted 70 wt. % of the 25. Bensted J.: Naturwiss. 63, 193 (1976).
total amount, and it changes the composition of the 26. Dhanam M., Kavitha B., Shanmugapriya M.: Chalcogenide
Letters. 6, 541 (2009).
compounds. 27. Langan B.W., Weng K., Ward M.A.: Cem. Concr. Res. 32, 1045
3. Tribochemically activated biomass fly ash additive (2002).
28. Snelson D.G., Wild S., O‘Farrell M.: Cem. Concr. Res. 38, 832
significantly accelerates the hydration process of
(2008).
calcium silicates and has a positive effect on the 29. Fernández-Carrasco L., Torrens-Martín D., Morales L.M.,
strength properties of samples, because up to 15 wt. % Martínez-Ramírez S.: Infrared Spectroscopy in the Analysis of
of the cement can be replaced with this additive Building and Construction Materials, p. 369-382, InTech, Croatia
2012.
without impairing the strength properties of Portland 30. Taylor H.F.W.: Cement Chemistry, 2nd edition, p. 361, Thomas
cement. Telford Publishing, London 1997.

268 Ceramics – Silikáty 58 (4) 260-268 (2014)