Journal of Sustainable Cement-Based Materials

ISSN: 2165-0373 (Print) 2165-0381 (Online) Journal homepage: http://www.tandfonline.com/loi/tscm20

Binding gel characterization of alkali-activated

binders based on palm oil fuel ash (POFA) and fly
ash

Nurulhuda Nadziri, Idawati Ismail & Sinin Hamdan

To cite this article: Nurulhuda Nadziri, Idawati Ismail & Sinin Hamdan (2017): Binding gel
characterization of alkali-activated binders based on palm oil fuel ash (POFA) and fly ash, Journal
of Sustainable Cement-Based Materials, DOI: 10.1080/21650373.2017.1299054

To link to this article: http://dx.doi.org/10.1080/21650373.2017.1299054

Published online: 07 Mar 2017.

Submit your article to this journal

Article views: 105

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=tscm20

Download by: [The UC San Diego Library] Date: 06 April 2017, At: 13:56
Journal of Sustainable Cement-Based Materials, 2017
http://dx.doi.org/10.1080/21650373.2017.1299054

Binding gel characterization of alkali-activated binders based on

palm oil fuel ash (POFA) and y ash
Nurulhuda Nadziri, Idawati Ismail* and Sinin Hamdan

Faculty of Engineering, Universiti Malaysia Sarawak, Kota Samarahan 94300, Sarawak,

Malaysia

Pastes of palm oil fuel ash (POFA) and y ash geopolymer activated with sodium
hydroxide (NaOH) and combination of NaOH with sodium silicate (Na2SiO3) were
prepared and cured at 80 C for a duration of 7, 28, and 90 days. The hydrated bind-
ing gel products were studied on phase, morphology, and bonding characterization.
From these analyses, calciumsodiumsilicatehydrate (C/NSH) gel type appear to
form in sole POFA binder but at a very low Ca/Si ratio, while sole y ash geopolymer
binder consists of sodiumaluminasilicatehydrate (NASH) binding gel product
with a signicant Na/Si ratio, regardless of activator types. When both precursors were
used, a combination of both gel types was formed, however, Fourier transform infra-
red spectroscopy and energy-dispersive X-ray analyses show that the NASH gel is
a more dominant product in the binder compared to the C/NSH gel.
Keywords: geopolymer; alkali-activated binder; palm oil fuel ash; y ash

1. Introduction blast furnace slags (GBBS), rice husk

It is a well-known fact that the manufac-
turing of ordinary Portland cement
(OPC), the worlds most common con-
struction material releases approximately
one tonne of carbon dioxide (CO2) for
every tonne of cement produced [1]. It is
stipulated that up to 16% of this green-
house gas will be released from the man-
ufacturing of cement by 2050 [2] even
when its usage is utilized by incorporat-
ing supplementary cementitious materials
(SCM). Therefore, there is an urgent need
to reduce this dependency by adopting
alternative binders using sole or blended
precursors. Numerous researchers have
reported the potential of industrial
byproducts or waste materials as cement
substitute or as partial replacement. In
many studies, alkali-activated binders
derived from y ash, ground granulated
ash, palm oil fuel ash (POFA), or natural
minerals such as kaolin have been
successfully identied [37].
Geopolymer binders were introduced
by Joseph Davidovits in the 1971, pro-
duced from the reaction of earth minerals
which are rich in silica (Si) and alumina
(Al) with alkaline and proposed that the
structure of the binder was a three-dimen-
sional network of SiO4 and AlO4 tetrahe-
dra sharing oxygen atoms, with negative
charges on the AlO4 tetrahedra balanced
by alkali cations (Na+, K+) [8,9]. Later
ndings showed that precursors of similar
compositions rich in Si and Al such as
y ash can dissolve in alkaline solutions
and resulted in a distinctive NASH
(Na2OAl2O3SiO2H2O) or KASH
(K2OAl2O3SiO2H2O) gel binder [10],
depending on the type of activator used.

*Corresponding author. Email: iidawati@unimas.my

2017 Informa UK Limited, trading as Taylor & Francis Group

2 N. Nadziri et al.
This gel type has a typical Na/Si or Al/Si
ratio ranging 0.51.0 and 1.152.15,
respectively. Raw materials that are rich
in calcium such as slag can also react
with alkaline solution to form a hardened
binder, producing a structure similar to
hydration of OPC whose composition is
a CSH (CaOSiOH2O) with a Ca/Si
ratio of near 2.0. However, an alkali-acti-
vated binder from a calcium-rich precur-
sor composed of tetrahedral substitution
of Al for Si, forming a CASH (CaO
Al2O3SiO2H2O) gel [11] with a com-
parable Ca/Si ratio of 1.01.5. Therefore,
activator choice and compositions of the
raw precursors has signicant inuence
on the type of gel formation in binder,
which contribute to mechanical and
microstructural properties.
POFA are wastes from palm oil
industry which has potentials as partial
replacement materials in concrete owing
to its pozzolanic properties [12,13].
POFA is a highly pozzolanic material
which has been tested as cement replace-
ment materials with improvement in
strength and durability [7,1416]. POFA
chemical composition varies with average
of 60% silica and 7% carbon depending
on manufacturing conditions [1720]. In
the hydration process, the silica content
in POFA will react with calcium hydrox-
ide Ca(OH)2 while pozzolanic reactions
produce more calcium silicate hydrate
(CSH) which is a gel compound as
well as reducing the amount of calcium
hydroxide. This contributes to the
strength of the concrete and produces
stronger and denser concrete as well as
enhanced durability. Fly ash is an indus-
trial byproduct during the combustion of
coal for energy production. Due to its
high silica and alumina, y ash is a good
source for geopolymerization, a process
where upon interaction with alkali solu-
tion releases Al and Si species for the
formation of a dense binder matrix [10].
The use of POFA as alkali-activated
binder is limited. In most studies, POFA
has shown good combination with cement
via partial replacement, but not as sole
precursor derived from the geopolymeriza-
tion process. However, there is no detailed
study done to analyze the microstructural
properties of alkali-activated POFA added
with y ash. The objective of this study
are to understand the compositional char-
acteristics of the binding gel based on sole
POFA and y ash, and a combination of
both precursors are analyzed using X-ray
diffraction (XRD) for phase characteriza-
tion, Fourier transform infrared (FTIR)
spectroscopy for structure analysis and
scanning electron microscopy with
energy-dispersive X-ray (SEM-EDX) for
morphology analysis and types of gel after
geopolymerization.
2. Experimental program
2.1. Materials
The raw precursors used in this investiga-
tion are POFA obtained from palm oil mill
in Lundu, Sarawak and Class F y ash
was received from Sejingkat Power Sta-
tion, Kuching, Sarawak, Malaysia. POFA
was dried at 100 C for 24 h and sieved in
45 micron sieve to remove large particles
and to ensure good pozzolanic reactivity
[18,21]. Access carbon on POFA was
removed by calcination at 500 C in a fur-
nace for approximately one hour. Access
carbon and other unburnt organic materi-
als contained in POFA were removed by
calcination at 500 C. This is important to
avoid their negative effect on nished
product. High amount of carbon will
reduce the pozzolanic reaction of POFA.
Particle size of the raw materials was ana-
lyzed using CILAS laser particle size ana-
lyzer where d50 obtained for POFA and y
ash are at 16.7 and 9.67 m, respectively.
The chemical compositions of the raw
materials are shown in Table 1 which
 Journal of Sustainable Cement-Based Materials
indicate signicant amount of silica oxide
(SiO2) and aluminum oxide (Al2O3) in
POFA and y ash. POFA also shows some
traces of calcium oxide (CaO) and magne-
sium oxide (MgO) but low in alumina
oxide (Al2O3). Meanwhile the composi-
tion of y ash is according to ASTM C
618, Standard Specication for Coal Fly
Ash and Raw or Calcined Natural Poz-
zolan for Use in Concrete [22], and is
rich in silica and alumina. These elements
are present in both precursors which are
important in development of binding prop-
erties. Two types of alkaline solutions are
prepared as activator. Solid anhydrous
sodium hydroxide (NaOH) was dosed at
12 wt.%, by total mass of solid precursor.
The dry solid was prepared by dissolving
NaOH pellets in tap water and allowed to
cool to room temperature before adding to
the dry precursors. A combination of
NaOH with sodium silicate (Na2SiO3) at
1:1 mass ratio (modulus ratio SiO2/
Na2O = 0.6) was also prepared as second
activator in this study. Water to solid pre-
cursor ratio is adjusted according to work-
ability factor described in Table 2. Table 2
shows the mix proportion of the alkali-ac-
tivated pastes prepared for this study. The
samples were measured by weight per-
centage (wt.%). A higher water binder
ratio is required for sole alkali-activated
POFA due to surface characteristics and
chemical content that leads to low worka-
bility of POFA-based samples. POFA is
highly hygroscopic material and therefore
more water is required to achieve satisfac-
tory workability [14].
2.2. Sample preparation and methods
The fresh pastes were mixed manually and
poured in 25 mm diameter 50 mm
Table 1. Chemical composition of raw materials by XRF (%).
Precursor SiO2 Al2O3 Fe2O3 CaO
POFA 44.10 1.20 2.10 10.30
Fly Ash 55.90 21.80 6.62 4.91
3
height cylindrical molds. The samples
were sealed to prevent moisture loss and
cured at 80 C until age of testing, i.e. 7,
28 and 90 days. The inuence of using
higher temperature may have some effect
to the mechanical strength [23,24]. XRD
analysis was conducted using a Bruker D8
Advance instrument, with Cu K radiation
scanning from 5 to 65 2, with a 0.02
step size and 2 s/step count time. FTIR
spectroscopy was carried out using Nico-
let iS10 instrument to characterize chemi-
cal functional groups of the pastes using
KBr pellet method, where approximately
1 mg of crushed sample and 100 mg KBr
were mixed and scanned 32 times from
4000 to 400 at 4 cm1 resolution. SEM-
EDX allows the observation of morphol-
ogy and the EXD data were taken from
fractured surface. Overall composition of
the material was carried out using JEOL
JSM6390La instrument at an accelerating
voltage of 10 kV in high vacuum environ-
ment, and working distance of 10 mm.
3. Results and discussion
3.1. Phase analysis using X-ray
diffraction
The X-ray diffractograms of the unre-
acted POFA and y ash are shown in
Figure 1. The unreacted POFA consists
of main crystalline phases of kyanite
(Al2SiO5, PDF# 01-083-1567). Kyanite is
a polymorph of andalusite and silliman-
ite. At 1100 C, kyanite will decompose
into mullite and vitreous silica [25]. The
unreacted y ash shows high intensity of
quartz (SiO2, PDF# 00-005-0490) and
mullite (Al6Si2O13, PDF# 00-015-0776).
After alkali activation, XRD diffrac-
tograms of the pastes shows no signicant
MgO SO3 Na2O K2O Others
9.13 5.80 0.55 12.40 14.95
2.00 0.32 0.32 2.20 5.94
4 N. Nadziri et al.
Table 2. Mix proportions and the water to solid precursor ratio of the alkali-activated POFA/y
ash binder.
Binder ratio (wt.% POFA/wt.% y ash)
100/0
0/100
50/50
Figure 1. X-Ray diffractograms of the unre-
acted precursors used to synthesize the alkali-
activated POFA y ash pastes.
formation of new peaks for POFA pastes
as observed in both activators in Fig-
ure 2(A) (NaOH) and (B) (NaOH with sil-
icate). However, kyanite crystalline shows
increment in intensity providing evidence
of reaction between activator and POFA
precursor. By comparing diffractograms
derived from both activators at region 2
angle between 25 and 35, higher
intensity is observed by NaOH-activated
POFA at early age (7 days) compared to
Figure 2. XRD diffractograms of alkali-activated POFA pastes, as a function of the curing ages
and activator (A) NaOH, (B) NaOH + Na2SiO3.
Water/solid precursor ratio
0.55
0.35
0.35
diffractograms of POFA in NaOH with sil-
icate activator as shown in Figure 3(A).
The comparison in intensity between these
diffractograms became less remarkable at
older age (90 days) as the reaction became
stable in both activators as depicted in
Figure 3(B). This indicates that both types
of activator have different mechanism of
reaction where higher reaction occurred
during earlier time of curing in NaOH
activator compared to NaOH with silicate.
This is observable in most NaOH-
activated precursors such as in kaolinite
and y ash as the supplied OH reduced in
time leading to a constant reaction at later
age of curing [26].
Meanwhile there are no signicant
formation of new peaks observed in y
ash geopolymer in both activators as
shown in Figure 4(A) and (B). Fly ash
geopolymer using both activators shows
similar crystalline phases as identied in
unreacted y ash which mainly consists
of quartz. However, a closer look at
region 2 angle between 10 and 30
 Journal of Sustainable Cement-Based Materials 5

Figure 3. Intensity comparisons of XRD diffractograms at selected 2 angle of alkali-activated

POFA as a function of activator obtained at (A) 7 days and (B) 90 days.

show slight variation by the formation of
new peak in NaOH with silicate activator.
Figure 5(A) shows the appearance of
low-intensity zeolite Na-P1 is observed in
NaOH with silicate at 28 days and simi-
larly shown at 90 days (Figure 5(B)).
This phase was not detected at 7 day cur-
ing age. This zeolitic phase is common in
activated y ash cured in high tempera-
tures, also reported in [27]. Zeolite Na-P1
(Na6Al6Si10O32, PDF# 01-074-1787) is
usually formed during hydrothermal
preparation of analcime from pure silica
and alumina sources.
The structural evolution of pastes with
50 wt.% POFA/50 wt.% y ash are shown
in Figure 6 as a function of activator. In
all diffractograms, quartz from y ash
shows a dominant prole indicating the
strong inuence of y ash compared to
POFA precursor in blended paste. Low
intensity of kyanite is also observable in
all blends for both types of activator. The
intensity of kyanite is observable at 2
degree angle between 15 and 35 as
shown in Figure 7 as a function of activa-
tor. At 7 days (Figure 7(A)), kyanite is
more intense under NaOH activation com-
pared to NaOH with silicate during early
curing time. However, there is an increase
amount of detection of kyanite in the
blends activated using NaOH with silicate
activator at 90 days curing (Figure 7(B))
suggesting the reaction of POFA in form-
ing gel binder, which occur at later times
when mixed with y ash.

Figure 4. XRD diffractograms of y ash geopolymer pastes, as a function of the curing ages
and activator (A) NaOH, (B) NaOH + Na2SiO3.
6 N. Nadziri et al.

Figure 5. Intensity comparisons of XRD diffractograms at selected 2 angle of y ash geopoly-

mer as a function of activator obtained at (A) 28 days and (B) 90 days (Note: No distinct peaks
observable at 7 days curing age).

Figure 6. XRD diffractograms of alkali-activated POFA/y ash pastes, as a function of the cur-
ing ages and activator (A) NaOH, (B) NaOH + Na2SiO3.

Figure 7. Intensity comparisons of XRD diffractograms at selected 2 angle of alkali-activated

50wt.% POFA/50wt.%y ash as a function of activator obtained at (A) 7 days and (B) 90 days.

3.2. Structure analysis using Fourier study with marked wavenumbers. The
transform infrared band observed at 3423 and 3415 cm1
Figure 8 shows the infrared spectra of the corresponds to the stretching vibration
unreacted precursors material used in this modes of hydrogen in HOH group in
 Journal of Sustainable Cement-Based Materials
y ash and POFA, respectively. A broad
peak at 3415 cm1 in POFA is due to
loss of water during drying process in
500 C temperature. The asymmetric
stretching vibration mode of SiOT
bonds (T = tetrahedral Si or Al) is
between 1200 and 950 cm1 band [28].
This band is shown at 1035 cm1 in the
POFA and 1077 cm1 in the y ash, con-
sistent with the differences in the chemi-
cal composition (Table 1). The small
band at 791 and 782 cm1 for unreacted
y ash and POFA, respectively, attribute
to the presence of quartz. Absorption
bands at 459 and 455 cm1 are assigned
to SiOSi bonding, while bands at
579 cm1 in the unreacted POFA spec-
trum are assigned to SiO bending [29].
The FTIR spectra for activated POFA
activated with NaOH, and NaOH with sili-
cate are shown in Figures 9(A) and (B),
respectively. In general, the spectrums
show similar prole in both types of acti-
vator. Overall, the intensity of the broad
band assigned to the vibration modes of
HOH group which is between 3200 and
3600 cm1 appears to increase over time
due to the hydration process of the bin-
ders. Absorption bands at 1620
1640 cm1 are assigned to the bending
vibration of the OH groups of the
hydrated reaction products. Small decre-
ment in peak intensity from 1640 cm1 at
early age and at 1630 cm1 at advanced
age indicates some removal of chemically
bonded water during the curing period for
Figure 8. FTIR spectra of the unreacted
material used to synthesize alkali-activated
POFA/y ash binders.
7
these activated POFA. The main binding
gel of NaOH-activated POFA pastes (Fig-
ure 9(A)) is characterized by the broad
peak in the wavenumber range between
1000 and 1100 cm1, a shift from
wavenumber 1030 cm1 from the unre-
acted POFA. The nal position at
advanced age is around 1006 cm1 indi-
cating a progressive reaction that is rich in
silica gel. This is consistent with the for-
mation of a sodium aluminosilicate-type
gel [30]. Meanwhile, the wavenumber
located between 1050 to 1095 cm1 refers
to the assignment of kyanite, which is con-
sistent with XRD [31,32]. There is also a
slight shift in the symmetric stretching
vibrations of SiOSi near 788 cm1 from
772 cm1, and bending vibrations of Si
OSi and OSiO in the 440450 cm1
region from 460 cm1. These results are
consistent with the formation of a Ca2+ in
aluminosilicate gel or C/NASH gel
[33]. Calcium is available in POFA
(Table 1) and therefore could participate in
formation of the gel binder at a low rate.
This phenomenon was also observed in y
ash geopolymer with partial addition of
calcium-rich slag [28]. In NaOH with sili-
cate-activated POFA paste (Figure 9(B)),
the major binding gel is centered at the
wavenumber near 1000 cm1 at early age
and continues to shift slightly towards
1009 cm1 as the nal position at later age
corresponding to also a silica-rich-type
gel. These major peaks are also appeared
broader when compared with NaOH-acti-
vated POFA whose main gel shows an
intense and sharp peak which is also con-
sistent with the observation via XRD (Fig-
ure 3). There is also a slight shift at
wavenumber 777 cm1 and the wavenum-
ber between 443 and 447 cm1 which
show the appearance of SiOSi bond in
the products. A weak bond centered at
wavenumber 562 cm1 corresponds to Si
O composition.
The FTIR spectra for alkali-activated
y ash with NaOH and NaOH with silicate
are shown in Figure 10(A) and (B),
8 N. Nadziri et al.

Figure 9. FTIR spectra of alkali-activated POFA pastes, as a function the time of curing and
activator (A) NaOH, (B) NaOH + Na2SiO3.

respectively. Similar peak assignment of
the stretching and bending vibration of the
water compound is characterized between
3200 and 3600 cm1 and 1620 and
1640 cm1, respectively. No signicant
changes occurred in the intensity of the
water structure of this type of gel, contrary
to what was observed in POFA pastes in
Figure 10(A). It shows that the curing con-
dition used in this assessment did not dis-
turb the water structure in the gel, which
gives an indication of a NASH-type
gel compared to POFA whose gel could
have some calcium silicate hydrate [34],
although not apparent via this assessment
method. This also shows that aluminosili-
cate-type gel is more dominant in both y
ash and POFA pastes compared to a cal-
cium silicate hydrate. Small band were
observed between 1400 and 1420 cm1,
associated with the stretching mode of O
CO bonds in carbonate groups. The main
binding gel shows that NaOH-activated
paste (Figure 10(A)) was centered at
1100 cm1, consistent with sodium alumi-
nosilicate gel in the y ash pastes
[26,28,32]. Comparing this major peak
location to the y ash pastes activated
using NaOH with silicate (Figure 10(B)),
it appears that NaOH activator gives a
sharper and intense peak, similar to what
observed in sole POFA pastes. These pre-
sented the high silica gels in the reaction
under this type of activator [35]. The week
bands occur at 518 cm1 are related to
vibrations in the unreacted solid aluminate
in the products [36].
Meanwhile, FTIR spectra of alkali-ac-
tivated POFA with y ash activated with
NaOH and the combination of NaOH and
silicate are shown in Figures 11(A) and
(B), respectively. Both FTIR spectra
show similar prole. Major binding gel
in NaOH-activated POFA/y ash is char-
acterized between 1002 and 1007 cm1.
By comparing Figures 10(A) and 11(A),
the reduction of the wavenumber in these
band is due to the lower degree of cross-

Figure 10. FTIR spectra of alkali-activated y ash pastes, as a function the time of curing and
activator (A) with NaOH, (B) NaOH + Na2SiO3.
 Journal of Sustainable Cement-Based Materials
linking of the amorphous phase on the
unreacted material, owing to the calcium
contents in POFA [28]. The small vibra-
tion located 1088 cm1 refers to the pres-
ence of kyanite, which is also consistent
with XRD. The simultaneous activation
of y ash is promoting the formation of a
more cross-linking of aluminosilicate-type
gel which can be seen from the vibration
on the band between 544 and 775 cm1
where the y ash spectra prole is domi-
nant in the product. Symmetric stretching
vibrations of SiOSi and AlOSi can
be identied near the 755 cm1 until
755 cm1 band, while OSiO bonding
occurs near 450 cm1 band [28,37]. It
can be clearly seen that there are no sig-
nicant differences in FTIR using both
types of activator in blended precursors.
3.3. Scanning electron microscope
energy-dispersive X-ray
3.3.1. Morphology analysis using
scanning electron microscope
The SEM image of unreacted precursor
can be observed in Figure 12. The unre-
acted POFA (Figure 12(A)) appears coar-
ser compare to y ash particle with oval
or round edges and irregular in shape.
The unreacted y ash (Figure 12(B))
shows spherical shape of the particle.
This spherical shape of the y ash
improves the consolidation and reduced
the permeability of concrete binder [38].
Figure 11. FTIR spectra of alkali-activated 50wt.%POFA/50wt.%y ash pastes, as a function of
the time of curing and activator (A) NaOH, (B) NaOH + Na2SiO3.
9
Figure 13 shows the appearance of
POFA paste after activation with NaOH
(Figure 13(A)) and NaOH with silicate
(Figure 13(B)) taken at 28-day of curing.
It seems that POFA activated using
NaOH is denser compared to NaOH with
silicate whose appearance is more porous.
There appear to be some formation of
crystalline microstructure within the bind-
ing gel in NaOH activator. Calcium,
aluminum, and silica contents of the sam-
ples were reported, renormalized to 100%
on an oxide basis when plotting the tern-
ary diagram. Calcium, aluminum and sil-
ica contents of the samples were
reported, renormalized to 100% on an
oxide basis when plotting the ternary dia-
gram. It shows the variation of gel com-
position across the binders as a function
of activator (Figure 16(A)). POFA acti-
vated using NaOH with silicate shows a
heterogeneous gel formation within a
porous microstructure.
Figure 14 shows the appearance of
y ash paste after activation with NaOH
(Figure 14(A)) and NaOH with silicate
(Figure 14(B)) taken at 28-day of curing.
Fly ash paste using NaOH appears denser
compared to NaOH with silicate with
partially reacted y ash. Fly ash with
NaOH with silicate shows a heteroge-
neous gel formation within a porous
microstructure. Fly ash particle was par-
tially reacted after 28 days curing ages
compare to raw y ash in Figure 12.
10 N. Nadziri et al.

Figure 12. SEM of unreacted precursors (A) POFA and (B) Fly Ash.

Figure 13. SEM images of alkali-activated POFA pastes, as a function of the time of curing and
activator (A) NaOH, (B) NaOH + Na2SiO3.

Figure 14. SEM images of alkali-activated y ash pastes, as a function of the time of curing
and activator (A) NaOH, (B) NaOH + Na2SiO3.
 Journal of Sustainable Cement-Based Materials
Meanwhile, in binary precursors,
POFA/y ash activated using NaOH (Fig-
ure 15(A)) is denser compared to NaOH
with silicate (Figure 15(B)) whose
appearance is more porous. There appear
to be some formation of crystalline
microstructure within the binding gel in
NaOH activator. The crystalline
microstructure also represents the silica.
POFA/y ash with NaOH with silicate
shows a heterogeneous gel formation
within a porous microstructure.
3.3.2. Energy-dispersive X-ray
spectroscopy analysis
EDX results obtained from multiple
points from the binder regions of the
alkali-activated POFA/y ash paste are
analyzed in this section. Calcium oxide
(CaO), aluminum oxide (Al2O3) and sil-
ica oxide (SiO2) contents of pastes plot-
ted as ternary diagrams. These plots
clearly show the variations in binding gel
composition across the different POFA/
y ash binders as a function of activator,
and also conrm the approximate chemi-
cal compositions of the types of gel
forming with different binder proportions.
Figure 16(A) shows the clustered regions
in POFA binder activated with NaOH
and NaOH with silicate. There is no
signicant difference in the binding gel
Figure 15. SEM images of alkali-activated 50 wt.% POFA/ 50 wt.% y ash pastes, as a function
of the time of curing and activator (A) NaOH, (B) NaOH + Na2SiO3.
11
of POFA pastes activated by both types
of activator. The average Ca/Si ratios are
distinctively low for both sets of activa-
tor. This is a large different when com-
pared to the normal cement-based pastes
whose Ca/Si is around 1.52.0 [39]. This
indicates that in alkali-activated POFA,
calcium is not largely dominating in the
formation of the binding gel, contradicts
to some previous ndings [40]. This is
also consistent with the XRD and FTIR
data. The binding phase observed in
POFA pastes can be described as an C/
NSH regardless of the activator used.
Aluminum is very low (Table 1) to be
considered in the formation of alumi-
nosilicate (NASH) gel with EDX cal-
culation of Al/Si are 0.03 and 0.06 for
both types of activating solution used,
which is in contrast to a typical value of
0.350.4 derived from aluminosilicate
precursors such as y ash and slag [41].
Figure 16(B) shows the clustered regions
in y ash binder as a function of activa-
tor. The points of the clustered region
clearly show the composition N(A)S
H (dashed circle) indicating a completely
different gel composition in y ash com-
pared to POFA pastes. There is also no
obvious difference in the gel composition
with respect to the types of activator
used, although some Ca from y ash
(Table 1) may participate when using
12 N. Nadziri et al.
Figure 16. Ternary representation of EDX
for alkali-activated POFA/y ash pastes as a
function of activator and POFA/y ash ratio:
(A) 100wt.% POFA, (B) 100wt.% y ash and
(C) 50wt.%POFA/50wt.% y ash. Composi-
tion in black circle represents a C/NSH type
gel, dashed circle represents an NASH gel
and gray lled circled represent an N(C)A
SH type gel.
NaOH with silicate as activator, produc-
ing a NC(A)SH gel (shaded circle).
This type of gel was discovered previ-
ously in binders with low calcium con-
tent [28]. Meanwhile Figure 16(C)
marked the clustered regions in blended
POFA/y ash paste as a function of acti-
vator. The incorporation of y ash in the
pastes promotes the formation of two dis-
tinct gels associated with the activation
reaction of both precursors. The C/NS
H product seems to show a reduction,
and consequently the binder becomes
enriched in aluminum and silicon, consis-
tent with the formation of a N(C)AS
H (shaded gray circle)-type product.
These gels are evident from the FTIR
analysis where the inclusion of y ash
slight shifting the chemistry of the bind-
ing paste. Thus, the inclusion of y ash
in POFA precursor give the coexistence
of N(A)SH (dashed circle) gel which
dominant in paste.
4. Conclusion
The nature of the gel in alkali-activated
pastes using POFA and y ash is strongly
dependent on the precursors and choice
of activator, as the differences in
microstructural and mineralogy of these
precursors promote the formation of dif-
ferent binder products. A calcium
sodiumsilicatehydrate (C/NSH)-type
gel is formed in the sole POFA binder
while sodiumaluminasilicatehydrate
(NASH) binding gel dominant in y
ash binder. When a combination of POFA
and y ash were used, NASH gel
strongly inuences the binder matrix. The
types of alkaline activator do not inu-
ence signicantly the type of gel forming
in the binder.
Disclosure statement
No potential conict of interest was reported
by the authors.
 Journal of Sustainable Cement-Based Materials
Funding
This work was supported by the Ministry of
Higher Education Malaysia under MyMaster,
MyBrain15 and Universiti Malaysia Sarawak
under Small [grant number [SGS 02 (S138)/
1113/2014(3)].
ORCID
Idawati Ismail http://orcid.org/0000-0003-
2960-4230
References
[1] Damtoft JS, Lukasik J, Herfort D, et al.
Sustainable development and climate
change initiatives. Cem Concr Res.
2008;38:115127.
[2] Provis JL. Green concrete or red herring?
future of alkali-activated materials.
Adv Appl Ceram. 2014;113:472477.
[3] Diaz EII, Allouche ENN, Eklund S.
Factors affecting the suitability of y ash
as source material for geopolymers. Fuel.
2010;89:992996.
[4] Fernndez-Jimnez AM, Palomo A,
Lpez-Hombrados C. Engineering prop-
erties of alkali-activated y ash concrete.
ACI Mater J. 2006;103:106112.
[5] Guo X, Shi H, Dick WA. Compressive
strength and microstructural characteris-
tics of class C y ash geopolymer. Cem
Concr Compos. 2010;32:142147.
[6] van Deventer JSJ, Provis JL, Duxson P,
et al. Reaction mechanisms in the
geopolymeric conversion of inorganic
waste to useful products. J Hazard Mater.
2007;139:506513.
[7] Chindaprasirt P, Homwuttiwong S, Jatu-
rapitakkul C. Strength and water perme-
ability of concrete containing palm oil
fuel ash and rice husk-bark ash. Constr
Build Mater. 2007;21:14921499.
[8] Davidovits J. Synthesis of new high-tem-
perature geo-polymers for reinforced
plastic/composites. SPE PACTEC 79 Soc
Plast Eng. 1979;4151.
[9] Davidovits J. Geopolymers J Therm
Anal. 1991;37:16331656.
[10] Fernndez-Jimnez A, Palomo A, Criado
M. Microstructure development of
alkali-activated y ash cement: a
descriptive model. Cem Concr Res.
2005;35:12041209.
[11] Brough A, Atkinson A. Sodium silicate-
based, alkali-activated slag mortars. Cem
Concr Res. 2002;32:865879.
13
[12] Tangchirapat W, Jaturapitakkul C, Chin-
daprasirt P. Use of palm oil fuel ash as
a supplementary cementitious material
for producing high-strength concrete.
Constr Build Mater. 2009;23:26412646.
[13] Hassan IO, Ismail M, Forouzani P, et al.
Flow characteristics of ternary blended
self-consolidating cement mortars incor-
porating palm oil fuel ash and pulverised
burnt clay. Constr Build Mater.
2014;64:253260.
[14] Sata V, Jaturapitakkul C, Kiattikomol K.
Utilization of palm oil fuel ash in high-
strength concrete. J Mater Civ Eng.
2004;16:623628.
[15] Foo KY, Hameed BH. Value-added uti-
lization of oil palm ash: a superior recy-
cling of the industrial agricultural waste.
J Hazard Mater. 2009;172:523531.
[16] Tangchirapat W, Jaturapitakkul C.
Strength, drying shrinkage, and water
permeability of concrete incorporating
ground palm oil fuel ash. Cem Concr
Compos. 2010;32:767774.
[17] Megat Johari MA, Zeyad AM, Muha-
mad Bunnori N, et al. Engineering and
transport properties of high-strength
green concrete containing high volume
of ultrane palm oil fuel ash. Constr
Build Mater. 2012;30:281288.
[18] Altwair NM, Megat Johari MA, Saiyid
Hashim SF. Flexural performance of
green engineered cementitious compos-
ites containing high volume of palm oil
fuel ash. Constr Build Mater. 2012:
518525.
[19] Al-mulali MZ, Awang H, Abdul Khalil
HPS, et al. The incorporation of oil palm
ash in concrete as a means of recycling:
A review. Cem Concr Compos.
2015;55:129138.
[20] Hussin MW, Bhutta MAR, Azreen M,
et al. Performance of blended ash
geopolymer concrete at elevated temper-
atures. Mater Struct. 2014;48:709720.
[21] Tay JH, Show KY. Use of ash derived
from oil-palm waste incineration as a
cement replacement material. Res Con-
serv Recycl. 1995;13:2736.
[22] ASTM. Standard specication for coal
y ash and raw or calcined natural poz-
zolan for use. Annu B ASTM Stand.
2003:36.
[23] Puertas F, Martnez-Ramrez S, Alonso
S, et al. Alkali-activated y ash/slag
cements: strength behaviour and hydra-
tion products. Cem Concr Res.
2000;30:16251632.
14 N. Nadziri et al.
[24] Bakharev T. Geopolymeric materials
prepared using Class F y ash and
elevated temperature curing. Cem Concr
Res. 2005;35:12241232.
[25] Speyer R. Thermal Analysis of materi-
als: dilatometry and interferometry. New
York (NY): CRC Press; 1993.
[26] de Vargas AS, Dal Molin DCC, Masuero
B, et al. Strength development of alkali-
activated y ash produced with combined
NaOH and Ca(OH)2 activators. Cem
Concr Compos. 2014;53:341349.
[27] Oh JE, Monteiro PJM, Jun SS, et al. The
evolution of strength and crystalline
phases for alkali-activated ground blast
furnace slag and y ash-based geopoly-
mers. Cem Concr Res. 2010;40:189196.
[28] Ismail I, Bernal SA, Provis JL, et al.
Modication of phase evolution in
alkali-activated blast furnace slag by the
incorporation of y ash. Cem Concr
Compos. 2014;45:125135.
[29] Duxson P, Mallicoat SW, Lukey GC,
et al. The effect of alkali and Si/Al ratio
on the development of mechanical prop-
erties of metakaolin-based geopolymers.
Colloids Surf, A. 2007;292:820.
[30] Garca-Lodeiro I, Fernndez-Jimnez A,
Blanco M, et al. FTIR study of the sol-
gel synthesis of cementitious gels: CS
H and NASH. J Sol-Gel Sci Technol.
2008;45:6372.
[31] Farmer VC. The infrared spectra of min-
erals. London: Mineralogical Society;
1974.
[32] Zhang Z, Wang H, Provis JL. Quantita-
tive study of the reactivity of y ash in
geopolymerization by FTIR. J Sustain
Cem Mater. 2012;1:154166.
[33] Garcia-Lodeiro I, Palomo A, Fernndez-
Jimnez A, et al. Compatibility studies
between NASH and CASH gels.
Study in the ternary diagram Na2O
CaOAl2O3SiO2H2O. Cem Concr
Res. 2011;41:923931.
[34] Ismail I, Bernal SA, Provis JL, et al.
Drying-induced changes in the structure
of alkali-activated pastes. J Mater Sci.
2013;48:35663577.
[35] Hanjitsuwan S, Hunpratub S, Thongbai
P, et al. Effects of NaOH concentrations
on physical and electrical properties of
high calcium y ash geopolymer paste.
Cem Concr Compos. 2014;45:914.
[36] Barrer RM, Mainwaring DE. Chemistry
of soil minerals. Part XIII. Reactions of
metakaolinite with single and mixed
bases. J Chem Soc Dalton Trans.
1972:25342546.
[37] Zhang Z, Wang H, Provis JL. The reac-
tivity of y ash in geopolymer synthesis:
a qualication study by FTIR. J Sustain
Cem Mater. 2012;1:68.
[38] Pacheco-Torgal F, Castro-Gomes J, Jalali
S. Alkali-activated binders: a review.
Constr Build Mater. 2008;22(7):1305
1314.
[39] Richardson IG. The nature of CSH in
hardened cements. Cem Concr Res.
1999;29:11311147.
[40] Salih MA, Abang Ali AA, Farzadnia N.
Characterization of mechanical and
microstructural properties of palm oil
fuel ash geopolymer cement paste. Con-
str Build Mater. 2014;65:592603.
[41] Escalante Garca JI, Campos-Venegas K,
Gorokhovsky A, et al. Cementitious
composites of pulverised fuel ash and
blast furnace slag activated by sodium
silicate: effect of Na2O concentration
and modulus. Adv Appl Ceram.
2006;105:201208.