Construction and Building Materials 98 (2015) 417–424

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier.com/locate/conbuildmat

Performance of slag blended alkaline activated palm oil fuel ash mortar
in sulfate environments
Moruf Olalekan Yusuf
Construction Engineering Department, University of Dammam, P.O. Box 1982, Dammam 31451, Saudi Arabia

h i g h l i g h t s

 Combination of slag and POFA in AAGU mortar provide good resistance to sulfate attack.
 Sulfate attack of AAGU mortar depends on the cations present.
 Slag content provides better resistance to sulfate attack than activators ratio.
 MgSO4aq attack causes decalcification of C–S–H in AAGU mortar to form M–S–H.
 AAGU mortar has a higher resistance to MgSO4aq than Na2SO4aq attack.

a r t i c l e i n f o a b s t r a c t

Article history: This study investigated the performance of alkaline activated ground steel slag-ultrafine palm oil fuel ash
Received 27 March 2015 (AAGU) mortar exposed to 5% Na2SO4 and 5% MgSO4aq for 6 months. The samples cured at 60 °C for 24 h
Received in revised form 3 July 2015 were prepared with 10 M-NaOHaq and Na2SiO3aq (Ms-SiO2/Na2O = 3.3) activators. Microstructural and
Accepted 7 July 2015
characterization tools were employed to examine the type and nature of the resulting products. The find-
ings showed that there was a better strength retention in the MgSO4aq-exposed samples compared to the
Na2SO4aq exposed due to the formation of surfacial deposits and intra-microstructurally crystallized anhy-
Keywords:
drite (CaSO4). The loss of weight/strength and skeletal disintegrations in Na2SO4aq exposure resulted from
Sulfate-attack
Geopolymer
the leaching of active elements (Ca, Al and Mg). Slag significantly contributed to the sulfate resistance of
Alkaline activated binder AAGU mortars, while there was an insignificant difference in strength retention when the activator ratios
Slag (Na2SiO3aq:NaOHaq) varied within 1.0–2.5 for the samples prepared with a workable mixture.
Palm oil fuel ash Ó 2015 Elsevier Ltd. All rights reserved.
Durability

1. Introduction way of studying its durability performances with a view to comple-

The release of excess greenhouse gases (GHGs) and the need for
the reduction in the environmental agro-industrial solid wastes
have recently prompted the synthesis of alkaline activated ground
slag/ultrafine palm oil fuel ash (AAGU) binding products [1,2].
Ultrafine palm oil fuel ash (UPOFA) is the post treated palm oil fuel
ash (POFA) solid wastes obtained from burnt palm kernel shell and
bunches used in electricity generation in an oil mill plant.
Further, POFA contributes to the solid wastes whose disposal
poses an environmental challenge. For instance, 3 million tons of
POFA was generated in Malaysia in 2007, while the annual POFA
generation in Thailand amounted to about 0.1 million ton [3,4].
The strength of AAGU binder was reported to be better than that
of ordinary Portland cement (OPC) [2,5]. This paper addresses the
response of AAGU mortars to the sulfate solutions (Mg, Na) as a
E-mail address: moruff@gmail.com
menting the available studies on the response of alkaline activated
binders (AAB) prepared from solid waste products such as slag and
fly-ash to sulfate environment [6–8]. Other studies previously con-
ducted on the durability of AAB or geopolymer include their
responses to sulfuric acid [9–11] and the elevated temperature
[12,13].
Sulfate attack on concrete products could cause both internal
and external defects due to the contamination of the constituent
materials and their exposure to contaminated environments,
respectively [14,15]. Various factors affect the sulfate resistance
of mortar or concrete. These include the type of the cations (Mg,
Na) carried by the sulfate ions (SO2
4 ) in the salt solutions, the con-
centration of the sulfate solution, and the permeability coefficient
of the exposed mortar samples [16]. The chemistry of the solid
skeleton of AAB, which in turn depends on the type of activators,
mineralogy of the aggregates and base materials used in the syn-
thesis are also among the factors controlling the density of their

http://dx.doi.org/10.1016/j.conbuildmat.2015.07.012
0950-0618/Ó 2015 Elsevier Ltd. All rights reserved.
418 M.O. Yusuf / Construction and Building Materials 98 (2015) 417–424
microstructures [17]. Some of the failures that accompany the sul-
fate attack on mortar or concrete include the loss of strength,
microstructural expansion, formation of products such as gypsum
(calcium sulfate-dihydrate, (CaSO42H2O)) and ettringite (Ca6Al2
(SO4)3(OH)1226H2O) [14]. The more aggressive nature of MgSO4aq
compared to Na2SO4 solution has been reported [18,19], especially
for the OPC products exposed to the sulfate solutions, while the
products like brucite (Mg(OH)2) and gypsum are also identified
in the aftermath of the attack.
The possibility of Ca–Mg ionic exchange that resulted in the for-
mation of serpentine or magnesium silicate hydrate (M–S–H) from
calcium–silicate–hydrate (C–S–H) has been adduced to more
aggressive nature of MgSO4aq compared to Na2SO4aq [20]. In fact,
the M–S–H (Mg8Si8O20(OH)812H2O) formed could be the combi-
nations of poorly crystallized sepiolite (Mg2H2Si3O9xH2O) and ser-
pentine (Mg3(OH)4(Si3O5)) [21]. Besides, Aye and Oguchi
established that MgSO4aq deleterious effect emanated from chem-
ical attack or ionic exchange, while Na2SO4aq exposure of OPC
products was observed to be more harmful compared to MgSO4aq
from physical attack standpoint [20].
Further, Thokchom et al. [22] studied the performance of fly-ash
based geopolymer mortar in 10 wt.% MgSO4 solution within
24 weeks. The effects of such exposure were found to be a deposi-
tion of white crystal on the sample surface with additional gain in
weight. Besides, changes in pH and reduction in the residual com-
pressive strength were also identified among other consequences
of sulfate attacks. The deterioration of fly ash-based geopolymer
exposed to 5% of Na2SO4 and MgSO4 salts was also reported to be
caused by leaching of alkali and penetration of sulfate ions through
diffusion whose degree was more in the former than the latter [23].
The least deterioration, however, was observed when samples
were exposed to the combined sulfate solutions (Na2SO4aq +
MgSO4aq) [23].
In addition, [7] observed that durability of the alkaline activated
water cooled slag exposed to 5% MgSO4 displayed a better
microstructural matrix and high resistivity such that, the perfor-
mance of the product prepared with the quantities of NaOHaq
and Na2SiO3aq of 3:3 (wt.%), surpassed that prepared with only 2
or 6% of NaOHaq and 4 or 0% of Na2SiO3aq, respectively. Moreover,
the lesser susceptibility of bottom ash (BA) geopolymer to sulfate
and sulfuric acid attacks compared to OPC was also reported
[10]. Investigation on fly-ash/slag geopolymer binder performance
in the sulfate (5% MgSO4 and Na2SO4 solutions) environments
revealed more extensive physical deteriorations in MgSO4aq due
to the formation of an expansive gypsum product compared to
Na2SO4aq exposure [16]. They also identified that low water/binder
ratio was one of the factors that mitigated against the sulfate
attack of geopolymer or alkaline activated binder [16].
Be that as it may, this study provides a better understanding of
durability performance of AAGU mortar in the sulfate environ-
ments of different cations (Mg, Na), and also enhances the practical
application of AAGU mortar as an alternative material of construc-
tion to OPC. It also promotes the utilization of solid wastes that
constitute landfill and environmental pollution as construction
materials for building structures or other civil engineering
applications.
2. Materials and methods
2.1. Materials
2.1.1. Ultrafine palm fuel ash and ground steel slag
Slag and palm oil fuel ash (POFA) were collected from Malaysian Southern Steel
Company and United Palm Oil Industry, respectively. These pozzolanic materials
(PMs) – original raw POFA and slag – were subjected to oven drying at the temper-
ature of 105 °C ± 5 °C for 24 h to get rid of their inherent moistures. The samples
were sieved through 300-lm sieve to remove stones and debris. Slag was then
ground for 16 h to become ground slag (GS). POFA was first ground for 8 h before
subjecting it to heat treatment in a gas-powered clay furnace at 500 °C ± 50 °C for
90 min to remove carbon content, and any volatile substances [17]. The treated
POFA was further ground for 8 h to become ultrafine palm oil fuel ash (UPOFA).
The milling of the raw materials was done by using a ball-milling machine rotating
at a speed of 180 rev/min.
The particle size distributions (PSDs) of the PMs were determined with Micro-
trac (Turbotrac) model S3500 particle size analyzer. The oxide compositions of the
PMs were obtained by using X-ray florescence (XRF) technique while the surface
areas (BET) of the PMs were determined with Micromeritics ASAP2020 by using
nitrogen gas adsorption.
2.1.2. Preparation of alkaline activators
The commercially available Na2SiO3aq was used together with 10 M-NaOHaq as
the alkaline activators. Na2SiO3aq comprised H2O: 62.11%, SiO2: 29.13% and Na2O:
8.76% such that its initial silica modulus (Ms = SiO2/Na2O) was 3.3. The
10 M-NaOHaq was prepared by dissolving 404.4 g of NaOH pellet (99% assays) into
distilled water to form 1 L of alkali solution (10 M-NaOHaq).
2.1.3. Preparation of sulfate solutions
The sulfate salts Na2SO4 (CAS Number 7757-82-6) and MgSO4 (CAS Number
7487-88-9) of percentage purity > 99% were used. Each solution was prepared by
dissolving 50 g of the solutes in distilled water to obtain 1 L (50 g/dm3) of the sul-
fate solution.
2.1.4. Aggregates
Mortars prepared comprised fine aggregates (FAs), which were dune sand with
the specific gravity of 2.62 in the saturated and surface dry condition (SSD) and the
fineness modulus of 1.85. The percentage water absorption for fine aggregates was
0.57% [24].
2.2. Analytical methods
The 14-day pulverized mortars were used for the X-ray diffraction (XRD) and
Fourier transform infrared (FTIR) spectroscopy analyses. The samples were col-
lected at an approximate depth of 10 mm from the sample surface. FTIR of the sam-
ples was conducted using Perking Elmer 880 spectrometer (KBr pellet technique),
while X-ray diffraction analysis (XRD) was obtained with the semi-quantitative
X-ray diffraction (XRD) instrument. Copper (Cu) Ka monochromatic radiation
(k = 1.5406 Å), 2h from 10° to 90°; with step interval 0.0341°, while a commercial
software was used for the XRD mineral compound or phase analyses.
X-ray florescence analyses (XRF) was obtained by using XRF-spectrometer with
a fully automated programmable operation of Rh/W dual X-ray tube, which is cap-
able of analyzing liquid, solids and powder. The fragmented portions of the samples
obtained from compressive strength test were used to determine the morphologies
of the matrix by using field emission scanning electron microscope couple with
energy dispersive spectroscopy (FESEM + EDS) equipment.
2.3. Experimental design
2.3.1. Mixture proportion for the constituents in AAGU mortar
Alkaline activated mortar has an approximate unit weight of 2400 kg/m3 [25].
The mortars were prepared with GS/PMs of 0 and 0.2. Na2SiO3aq/NaOHaq (NS/NH)
varied as 1.0 and 2.5 as shown in Table 3. Free water content to pozzolanic materi-
als ratio (FWC/PMs) was kept at 2 wt.% with the adjustment for the water absorp-
tion of the aggregates. The fine aggregates to pozzolanic materials (FA/PMs) and
(NS + NH)/PMs were 1.8 and 0.5, respectively.
2.3.2. Mixing, placement and testing of AAGU mortar
The constituent materials were mixed in a 4.73 L planetary bench mixer. The
PMs were first mixed for 2 min to remove the air pockets and this was followed
by adding free (portable) water content (FWC), NaOHaq, Na2SiO3aq and FAs, sequen-
tially. Each material was mixed for 2 min such that the total mixing time for the
mortar was 12 min. The mixtures were cast in oil smeared prismatic steel mould
of 50  50  50 mm3 in two layers and each layer was vibrated for 15 s. The cast
samples were then covered with vinyl bag to prevent moisture loss, and then later
kept in the laboratory at temperature of 25 °C for 12 h for easy demoulding.
Subsequently, the samples were then cured in an oven at 60 °C for 24 h. The
cured samples were exposed to the sulfate solution 30 min after removing them
from the oven. The unexposed (control) samples were used for the determination
of 3, 7 and 28-day compressive strength. The strength was tested with a crushing
machine at a loading rate of 0.9 kN/s. The average failure stress (MPa) of three spec-
imens was recorded as the compressive strength value. The loss in the sulfate solu-
tions in which the samples were submerged were made up by freshly prepared
solution. The weights of the samples were initially determined initially on weekly,
and subsequently on monthly basis for 6 months.
 M.O. Yusuf / Construction and Building Materials 98 (2015) 417–424 419

2.3.3. Sample designations anhydrite (CaSO4), portlandite (Ca(OH)2), diopside (MgCaSi2O6)

The samples were designated as MD1, MD2 and MD3. Sample MD1 (AAGU0R2.5)
represented the slag free mortar prepared with the alkaline activator ratio, (Na2-
SiO3aq/10 M-NaOHaq) ratio of 2.5 at the total quantity of activators (Na2SiO3aq + 10-
MNaOHaq)/PMs of 0.5. Similarly, MD2 (AAGU0.2R2.5) samples were produced with
the similar alkaline activator ratio as MD1 but with slag content (GS/(GS + UPOFA))
of 0.2. MD3 (AAGU0.2R1) samples had the same slag content as the MD2 but with
alkaline activator ratio of 1.0. Therefore, MD1 and MD2 were used to study the con-
tributions of slag to the sulfate resistance of AAGU mortar, while MD2 and MD3
samples were employed for the monitoring of the effects of variation in the alkaline
activator ratios (AAR) of the activators on the sulfate resistance of the mortar.
3. Results and discussion
3.1. Physical properties and characterization of pozzolanic base
materials (PMs)
The grinding of both pozzolanic materials (PMs) led to an
increase in their fineness and surface areas. The median particle
sizes (d50) of GSS and UPOFA were approximately equal to
1.1 lm. The d10 and d90 particle sizes of UPOFA were 0.566 lm
and 10.12 lm, respectively while it was 0.591 lm and 0.984 lm,
respectively for GSS as shown in Table 1. Calcination of POFA
reduced its loss on ignition (LOI) value and percentage of its carbon
as shown in Table 2. From XRF results in Table 2, the basicity coef-
ficient of GS (kb = CaO + MgO + Al2O3/SiO2) was 2.10, (>1.0) hence,
classified as basic. The results also indicated that UPOFA had higher
composition of SiO2 than GSS, which in turn had higher CaO, SO3
and MgO contents than UPOFA. The Al2O3 of both materials was
relatively low and almost the same (4%). The hydration modulus
of the slag (HM = CaO + MgO + Al2O3/SiO2), was 2.62, (>1.4)
implied that it had a good hydration tendency as reported by [26].
XRD diffractograms of GS and UPOFA showed that the two base
materials have crystalline and amorphous phases (Fig. 1). The
amorphous phase identified by the diffractive halo of 2-h angle
between 20° and 40° was caused by the presence of short-range
order of CaO–MgO–Al2O3–SiO2 structure within the materials.
The peaks in UPOFA sample include calcite (CaCO3), quartz
(SiO2), potassium aluminate phosphate (k2Al2(PO4)3) and cristo-
balite (SiO2), while the GS contained calcium–silicate based com-
pounds like bregidite (Ca2SiO4), belite (c-C2S or c-Ca2SiO4), alite
(Ca3SiO5 or C3S), and larnite (C2S or bCa2SiO4). Others included
Table 1
Physical properties of ultrafine palm oil fuel ash and ground steel slag.
Base Specific Particle Median Particle Specific
materials gravity size, d10 particle size, size, d90 surface (BET)
(lm) d50 (lm) (lm) area (m2/g)
UPOFA 2.6 0.566 1.069 10.120 13.40
GSS 2.9 0.591 1.098 9.840 14.92
and spinel (Al2O4Mg), which is an alumina-based compound.
3.2. Physical examination of the exposed samples
After two weeks of sample exposure, the Na2SO4aq-exposed
samples exhibited no physical damage, and no extract was seen
on their peripheries. However, MgSO4aq-exposed samples showed
deposition of white flaky substance as shown in Fig. 2. In other
words, the behavior or response of the mortars in the sulfate envi-
ronment depended on the type of cations (Mg2+ or Na+) carried by
the sulfate salts as previously observed in the geopolymer product
[16].
From Fig. 2, the fewer white flaky deposition on the
slag-blended MD2 (AAGU0.2R2.5) compared to slag-free MD1
(AAGU0R2.5) specimens suggested that Ca2+ present in slag played
the role of pore filling [27]. This is due to the formation of
calcium–(alumino)silicate–hydrate (C–(A)–S–H) resulted from the
hydration of alite (C3S) and belite (C2S) compounds present in
the slag. This consequently reduced the permeability or ion
exchanges between the sample and the MgSO4aq solution. The sur-
facial depositions also reduced as the Na-silicate to Na-hydroxide
(Na2SiO3aq/NaOHaq) ratio (AAR) increased from 1.0 to 2.5, which
is evident by comparing MD2 and MD3 (AAGU0.2R1). Thus, slag
inclusion and preponderance of Na-silicate over NaOHaq in the syn-
thesis of AAGU mortar improved the tortuosity of the capillary
pores thereby limiting the mobility of ions (Na+, Mg2+, OH,
SO2
4 ) in and out of the specimens.
3.3. Compressive and residual strength of sulfate exposed AAGU
mortars
From Table 4, the compressive strength increased from 3 to
28 days in all the samples. The 28-day strengths were 34.0 MPa,
44.4 MPa and 40.1 MPa for MD1, MD2 and MD3, respectively. At
the end of 6 months, MD1 sample was found to substantially lose
its strengths such that only 16% of the 28-d strength (unexposed)
was retained as a consequence of Na2SO4aq attack, while the
strength retention increased to 76.8% in MgSO4aq exposure. The
strength loss could be as a result of formation of M–S–H or serpen-
tine (Mg3[Si2O5](OH)4) as identified in Fig. 5 and Eq. (1), due to the
replacement of C–S–H with the infiltrated Mg2+ ions [21].
MgSO4 þ C  S  H ! CaSO4  2H2 O þ M  S  H ð1Þ
Similarly, the hydroxylation of base materials (slag and POFA)
by the activators generated Ca(OH)2 that reacted with MgSO4 to
form Mg(OH)2 or brucite as identified in Fig. 5. However, upon
increasing the alkaline activator ratio (AAR) from 1.0 (MD3) to
2.5 (MD2), the 6-month strength retentions became 23.6% and

Table 2
Oxide compositions of the base materials.

Materials/Oxides SiO2 Al2O3 Fe2O3 CaO Na2O TiO2 MgO K2O P2O5 SO3 C LOI
POFA (%) 46.00 3.10 2.45 8.46 0.13 0.13 4.40 4.08 3.95 0.30 26.40 21.60
UPOFA (%) 60.42 4.26 3.34 11.00 0.18 0.19 5.31 5.03 4.48 0.45 4.85 2.55
GSS (%) 25.78 4.42 1.94 53.54 0.03 0.26 9.50 0.05 0.04 3.85 0.00 2.05

Table 3
Material proportion in the sulfate exposure AAGU mortars.

ID UPOFA (kg/m3) Slag (kg/m3) Sand (kg/m3) Water (FWC) (kg/m3) Na2SiO3 (kg/m3) 10 M NaOH (kg/m3)
MD1AAGU0R2.5 722.89 0.00 1301.20 14.46 258.18 103.27
MD2AAGU0.2R2.5 578.31 144.58 1301.20 14.46 258.18 103.27
MD3AAGU0.2R1.0 578.31 144.58 1301.20 14.46 180.72 180.72
420 M.O. Yusuf / Construction and Building Materials 98 (2015) 417–424

21.4% in the former and latter, respectively in the mortar exposure

to Na2SO4aq. The strength retentions increased to 95.5% and 84.45%
in MD2 and MD3, respectively in the MgSO4aq exposure.
Unlike OPC based mortar [19], no visible crack was found on the
specimens as a result of formation of MgSO4nH2O. The more
strength retention in MgSO4aq exposed sample could be due to for-
mation of M–S–H as a result of the decalcification of C–S–H that led
to the formation of M–S–H (Mg8Si8O20(OH)812H2O) as shown in
Eq. (1) [19,21]. This implies that the strength retention of AAGU
mortar in the sulfate environments depends largely on the cations
(Na, Mg) carried by sulfate solution, and the type of oxide compo-
sitions of base materials (CaO and SiO2) but less dependent on the
AAR in the range of 1.0–2.5 as used in this study.

3.4. Effect of sulfate attacks on the change in weight of the specimens

From Fig. 3, the weight of all the samples exposed to Na2SO4aq

Fig. 1. XRD diffractogram of the original ultrafine palm oil fuel ash and ground slag.
was found to be decreasing with time. The initial weight loss of
the slag free sample, MD1 exposed to Na2SO4aq was found to be
more than in MD2. In fact, about 6.3% of the initial weight was lost
in MD1 after 6-month of exposure. The weight further reduced to
3.7% and 4.8% in MD2 and MD3, respectively. This indicated that
when the AAR changed from 1.0 to 2.5, there was also a reduction
in the weight loss of the exposed samples.
However, the response of the samples to MgSO4aq attack in
terms of weight loss displayed a somewhat different performance
as shown in Fig. 4. There was an initial gain in weight in the first
3 weeks due to the deposition of surfacial white crystalline depos-
its on the sample, and probably due to absorption of sulfate solu-
tion after 1 week of exposure (Fig. 2). This observation was in the
contrary to the performance of OPC mortar in sulfate exposures,
where Aye and Oguchi [20] reported that Na2SO4aq lost less weight
than MgSO4aq exposed samples due to surface corrosion that led to
weight loss of cohesionless particle resulted from gypsum forma-
tion and decalcification process.

Fig. 2. Samples physical appearance upon their exposure to Na2SO4 (top) and MgSO4 (bottom) attacks.
 M.O. Yusuf / Construction and Building Materials 98 (2015) 417–424 421

Table 4
Initial activators oxide molar ratio and the strength results of AAGU mortars.

Designation SiO2/Na2O H2O/Na2O H2O/SiO2 3-d 14-d 28-d 6-Month residual 6-Month residual
(MR) (MR) (MR) (MPa) (MPa) (MPa) strength strength
MPa in Na2SO4 MPa in MgSO4
MD1AAGU0R2.5 1.64 18.47 11.23 28.0 30 34 5.6 26.1
MD2AAGU0.2R2.5 1.64 18.47 11.23 41.7 42.2 44.4 10.5 37.5
MD3AAGU0.2R1 0.92 15.44 16.73 37.5 39.2 40.1 8.6 38.3

MR-molar ratio.

Fig. 3. Weight loss of AAGU samples in Na2SO4aq solution.

Fig. 4. Weight loss of AAGU samples in MgSO4aq solution.

Further, the slag blended MD2 continued gaining weight than edge of the localized ruptured region of the Na2SO4aq exposed
MD1 for the first 3 weeks. Addition of pozzolanic materials such samples. The erosion of the solid skeleton framework was caused
as slag has also been reported to reduce the deterioration of OPC by the infiltration of SO2
4 from sulfate solution, which combined

mortar [19,28]. At the end of 6 months, MD1, MD2 and MD3 sam- with Na+ from Na2SiO3 thereby causing the replacement of
ples lost 3.43%, 1.23% and 1.11%, respectively of their initial weight SiO2
3 that stabilized the skeletal framework. This caused the disin-

upon exposing them to Mg-sulfate solution. Therefore, the slag
composition played major roles in the resistance of the materials
to sulfate attack but alkaline activated ratio (AAR) insignificantly
affected the resistance of AAGU mortars to MgSO4aq attack.
3.5. Mineral compounds and microstructure of sulfate exposed AAGU
mortars
Fig. 5 shows the presence of crystalline phases such as
serpentine, anhydrite and aragonite, which had resemblance with
M–S–H, gypsum and calcite, respectively were found in the XRD
diffraction of MgSO4aq. These peaks were conspicuously absent in
the Na2SO4aq exposed samples diffractogram with no peaks other
than calcite and brucite (Mg(OH)2) [18]. This suggested that the
appearance of these crystalline phases in MgSO4aq contributed to
the samples strength retention (Fig. 2). The formation of
Mg-based compounds (brucite) was due to the infiltration of
Mg2+ that resulted from the dissociation of its salt (MgSO4) as
shown in Eq. (2), and the reactions of Mg-salt with alkali as shown
in Eq. (3).
MgSO4aq ! Mg2þ þ SO2
4 ð2Þ
MgSO4aq þ NaOHaq ! MgðOHÞ2aq þ Na2 SO4aq ð3Þ
By observing the monograph of MgSO4aq exposed sample in
Fig. 6, there exist microcracks, partially reacted UPOFA particles,
and fragmented heterogeneous phases of crystalline solids. How-
ever, Fig. 7 shows the leaching effect that resulted in the defrag-
mentation of microstructure with the accompanied cracks at the
tegration of the microstructure as shown in Fig. 7 and Eq. (4)
2
Na2 SiO3 þ SO2
4 ! Na2 SO4 þ SiO3 ð4Þ
The perforated skin layers shown in Fig. 7 further corroborated
the microstructural erosion of the active minerals in form of leaching
from the samples subjected to Na2SO4aq attack. Fig. 5 shows the pres-
ence of anhydrite (CaSO4), which has a similarity with gypsum
(CaSO42H2O) but the latter could not be identified in the
XRD diffractogram (Fig. 5) of MgSO4aq exposed sample. Further,
anhydrite disappeared completely in the Na2SO4aq exposed samples.
Therefore, white deposit crystalline substance on the surface of the
MgSO4aq-exposed sample could have emanated from the reaction
of Ca(OH)2 with MgSO4 to form anhydrite as shown in Eq. (5).
CaðOHÞ2aq þ MgSO4aq ! CaSO4aq þ MgðOHÞ2aq ð5Þ
However, further investigation in terms of chemical analysis is
required to ascertain the presence of anhydrite or gypsum in AAGU
mortar as observed in the exposure of OPC binder to MgSO4 attack
[18].
From Fig. 8, the FTIR spectra show the asymmetric stretching of
Si–O–Si(Al) of MD2 in the Na2SO4aq and MgSO4aq exposed samples
at the bands 1027 cm1 and 1033 cm1, respectively. The fre-
quency of vibration of Si–O–Si(Al) bonds increased as the Al lea-
ched out from the framework due to the weaker bond of Al–O–Si
as compared to Si–O–Si. The bending vibrations of Si–O–Si in the
exposed samples (457–459 cm1) were found to be higher than
the control (unexposed) sample (449 cm1) [29,30]. Therefore,
more leaching of Al that occurred in Na2SO4aq compared to the
MgSO4aq exposed sample and stronger Si–O bond vibrations
implied that sulfate exposure caused the reduction in the bond
422 M.O. Yusuf / Construction and Building Materials 98 (2015) 417–424

Fig. 5. XRD diffractogram of AAGU0.2R2.5 (MD2) samples exposed to MgSO4aq and Na2SO4aq.

Fig. 6. Micrographs of 180-d AAGU0.2R2.5 (MD2) mortar (control-top) MgSO4aq exposures.

 M.O. Yusuf / Construction and Building Materials 98 (2015) 417–424 423

Fig. 7. Micrographs of 180-d AAGU0.2R2.5 mortar (control-top) Na2SO4aq exposures.

Fig. 8. FTIR spectra of the unexposed and exposed AAGU mortars in Na2SO4aq and MgSO4aq after 6-month exposure.
424 M.O. Yusuf / Construction and Building Materials 98 (2015) 417–424
angle of the silicate bonds. This caused the less amorphous nature
of the exposed sample compared to the control.
Besides, the O–C–O vibrations in the FTIR spectra at the bands
1443 and 1428 cm1 in the samples exposed to MgSO4aq and
Na2SO4aq, respectively indicated O–C–O vibration from carbonate
source (CO23 ). However, the presence of calcite shown by the
bands 2962 and 2925 cm1 in the spectrum band of the control
sample that disappeared in the MgSO4aq and Na2SO4aq samples
suggested that sulfate (Mg, Na) solutions regardless of the attached
cations disintegrated the calcite phase of the solid skeleton of the
products.
From Fig. 8, the elimination of cations such as (Ca2+, Mg2+, Na+)
caused the clustering of –O–H bonds as shown in the FTIR spectra
with widened trough at wave numbers 3564 cm1 in Na2SO4aq
exposed sample. This absorption reduced to 3476 cm1 in the
MgSO4aq-exposed samples. The asymmetric vibration of –OH was
also noted in the control at 3459 cm1. The shallow trough of this
peak in the Na2SO4aq exposed sample compared to MgSO4aq could
further support the fact that more leaching occurred in the former
than the latter. This also supported the sample microstructural
characteristics of the samples exposed to both sulfate solutions
as identified in Figs. 6 and 7.
4. Conclusions
The performances of the synthesized alkaline activated ground
steel slag and ultrafine palm oil fuel ash mortar (AAGU) due to
exposure to sulfate environments of 5% Na2SO4aq and 5% MgSO4aq
were investigated and the followings are the conclusions:
 The response of AAGU binder to sulfate attack was found to be
dependent on the type of cation carried by the sulfate solutions
or the environment of the exposure.
 There were surfacial white depositions of anhydrite (CaSO4) on
the MgSO4aq-exposed mortar due to cation exchanges, this phe-
nomenon decreased with the increase in the slag composition
and alkaline activator ratio (Na2SiO3aq/NaOHaq) but no such
deposit was observed in the Na2SO4aq-exposed mortars.
 Exposure of AAGU mortars to Na2SO4aq solution resulted in the
leaching of active mineral elements thereby causing the disinte-
gration of microstructural framework
 There was higher strength retention and initial weight gain for
the samples exposed to MgSO4aq in comparison with Na2SO4aq
exposed samples, which had a comparably significant weight
loss.
 There was a better microstructural stability for the AAGU
samples exposed to MgSO4aq than Na2SO4aq environments.
 Finally, slag inclusion (up to 20%) into the AAGU product
contributed to its better sulfate resistance while change in the
activator ratio (Na2SiO3aq/NaOHaq) from 1.0 to 2.5 resulted in
an insignificant difference in strength retention of the products
especially for a workable mixture.
Acknowledgements
Universiti Sains Malaysia is gratefully appreciated for her
supports in the testing and characterization of materials. United
Palm Oil Industries and Southern Steel Company are also thanked
for the provisions of palm oil fuel ash and waste steel slag,
respectively.
References
[1] M.J.A. Mijarsh, M.A. Megat Johari, Z.A. Ahmad, Synthesis of geopolymer from
large amounts of treated palm oil fuel ash: application of the Taguchi method
in investigating the main parameters affecting compressive strength, Constr.
Build. Mater. 52 (2014) 473–481.
[2] M.O. Yusuf, M.A. Megat Johari, Z.A. Ahmad, M. Maslehuddin, Evolution of
alkaline activated ground blast furnace slag–ultrafine palm oil fuel ash based
concrete, Mater. Des. 55 (2014) 387–393.
[3] M.A. Megat Johari, A.M. Zeyad, N.M. Bunnori, K.S. Ariffin, Engineering and
transport properties of high-strength green concrete containing high volume
of ultrafine palm oil fuel ash, Constr. Build. Mater. 30 (2012) 281–288.
[4] P. Chindaprasirt, S. Homwuttiwong, C. Jaturapitakkul, Strength and water
permeability of concrete containing palm oil fuel ash and rice husk-bark ash,
Constr. Build. Mater. 21 (2007) 1492–1499.
[5] M.O. Yusuf, M.A. Megat Johari, Z.A. Ahmad, M. Maslehuddin, Strength
and microstructure of alkali-activated binary blended binder containing
palm oil fuel ash and ground blast-furnace slag, Constr. Build. Mater. 52
(2014) 504–510.
[6] N.M. Al-Akhras, Durability of metakaolin to sulfate attack, Cem. Concr. Res.
(2006) 36.
[7] H.A. El-Sayed, S.A.A. El-Enein, H.M. Khate, S.A. Hasanein, Resistance of alkali
activated water-cooled slag geopolymer to sulphate attack, Ceram. Silik. 55
(2011) 153–160.
[8] N.P. Rajamane, M.C. Nataraja, J.K. Dattatreya, N. Lakshmanan, D. Sabitha,
Sulphate resistance and eco-friendliness of geopolymer concretes, Indian
Concr. J. 86 (2012) 13–22.
[9] T. Bakharev, Resistance of geopolymer materials to acid attack, Cem. Concr.
Res. 35 (2005) 658–670.
[10] V. Sata, A. Sathonsaowaphak, P. Chindaprasirt, Resistance of lignite bottom ash
geopolymer mortar to sulfate and sulfuric acid attack, Cement Concr. Compos.
34 (2012) 700–708.
[11] X. Song, M. Marosszeky, M. Brungs, Z. Chang, Response of geopolymer concrete
to sulfuric acid attack, in: J. Davidovits (Ed.), Geopolymer, Green Chemistry
and Sustainable Development Solutions: Proceedings of the World Congress
Geopolymer, 2005, p. 4.
[12] D.L.Y. Kong, J.G. Sanjayan, Effect of elevated temperatures on geopolymer
paste, mortar and concrete, Cem. Concr. Res. 40 (2010) 334–339.
[13] A.M. Rashad, Potential use of phosphogypsum in alkali-activated fly ash under
the effects of elevated temperatures and thermal shock cycles, J. Cleaner Prod.
87 (2015) 717–725.
[14] A. Neville, The confused world of sulfate attack on concrete, Cem. Concr. Res.
34 (2004) 1275–1296.
[15] M.D. Cohen, B. Mather, Sulfate attack on concrete Research needs, ACI Mater. J.
88 (1991) 62–69.
[16] I. Ismail, S.A. Bernal, J.L. Provis, S. Hamdan, J.S.J. Deventer, Microstructural
changes in alkali activated fly ash/slag geopolymers with sulfate exposure,
Mater. Struct. 46 (2012) 361–373.
[17] M.O. Yusuf, M.A. Megat Johari, Z.A. Ahmad, M. Maslehuddin, Performance of
different grades of palm oil fuel ash and ground slag in the synthesis of
alkaline activated mortar, J. Adv. Concr. Technol. 12 (2014) 378–387.
[18] R.S. Gollop, H.F.W. Taylor, Microstructural and microanalytical studies of
sulfate attack. I. Ordinary Portland cement paste, Cem. Concr. Res. 22 (1992)
1027–1038.
[19] E.E. Hekal, E. Kishar, H. Mostafa, Magnesium sulfate attack on hardened
blended cement pastes under different circumstances, Cem. Concr. Res. 32
(2002) 1421–1427.
[20] T. Aye, C.T. Oguchi, Resistance of plain and blended cement mortars exposed to
severe sulfate attacks, Constr. Build. Mater. 25 (2011) 2988–2996.
[21] T. Zhang, L.J. Vandeperre, C.R. Cheeseman, Formation of magnesium silicate
hydrate (M-S-H) cement pastes using sodium hexametaphosphate, Cem.
Concr. Res. 65 (2014) 8–14.
[22] S. Thokchom, P. Ghosh, S. Ghosh, Performance of fly ash based geopolymer
mortars in sulphate solution, J. Eng. Sci. Technol. Rev. 3 (2010) 36–40.
[23] T. Bakharev, Durability of geopolymer materials in sodium and magnesium
sulfate solutions, Cem. Concr. Res. 35 (2005) 1233–1246.
[24] M.O. Yusuf, Towards Optimal Design of Reinforced Concrete Structures in
Corrosive Environment, VDM Verlag Dr. Muller Aktiengesellschaft & Co. KG,
Germany, 2009.
[25] B.V. Rangan, Low-Calcium, Fly-Ash-Based Geopolymer Concrete, in: E.G. Nawy
(Ed.), Concrete Construction Engineering Handbook, CRC Press, 2008, pp. 1–20.
[26] J.J. Chang, A study on the setting characteristics of sodium silicate-activated
slag pastes, Cem. Concr. Res. 33 (2003) 1005–1011.
[27] C.K. Yip, G.C. Lukey, J.L. Provis, J.S.J. van Deventer, Effect of calcium silicate
sources on geopolymerisation, Cem. Concr. Res. 38 (2008) 554–564.
[28] S.U. Al-Dulaijan, Sulfate resistance of plain and blended cements exposed to
magnesium sulfate solutions, Constr. Build. Mater. 21 (2007) 1792–1802.
[29] I. Lecomte, C. Henrist, M. Liégeois, F. Maseri, A. Rulmont, R. Cloots, (Micro)-
structural comparison between geopolymers, alkali-activated slag cement and
Portland cement, J. Eur. Ceram. Soc. 26 (2006) 3789–3797.
[30] W.K.W. Lee, J.S.J. van Deventer, The effects of inorganic salt contamination on
the strength and durability of geopolymers, Colloids Surf., A 211 (2002) 115–
126.