Journal of Cleaner Production 112 (2016) 816e822

Contents lists available at ScienceDirect

Journal of Cleaner Production

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

Strength and microstructural properties of fly ash based geopolymer

concrete containing high-calcium and water-absorptive aggregate
Rahimah Embong a, b, Andri Kusbiantoro a, *, Nasir Shafiq b, Muhd Fadhil Nuruddin b
a
Faculty of Engineering Technology, Universiti Malaysia Pahang, Pahang 26300, Malaysia
b
Civil and Environmental Engineering Department, Universiti Teknologi PETRONAS, Perak 31750, Malaysia

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

Article history: In this research, performance of limestone and oil palm shell (OPS) as the alternative aggregates in fly ash
Received 20 March 2015 based geopolymer concrete was evaluated based on the fresh and hardened concrete test. Fly ash and oil
Received in revised form palm shell were studied due to their abundant availability as industrial and agricultural waste. Experi-
4 June 2015
mental works were conducted by replacing the portion of coarse and fine aggregate in concrete with
Accepted 11 June 2015
Available online 20 June 2015
limestone and oil palm shell in various percentages. Evaluation of its performance was done based on
slump, compressive strength, and porosity test. Microstructure analysis was added via Scanning Electron
Microscope (SEM) and Energy Dispersive X-Ray (EDX) test to support the generated hypothesis. From the
Keywords:
Limestone
fresh properties test, the increasing amount of limestone and OPS have shortened the setting time of
Oil palm shell fresh geopolymer concrete, even though they presented small variations in the slump's height and
Fly ash diameter results. The availability of Calcium Aluminate Silicate Hydrate (CeAeSeH) structure in lime-
Geopolymer stone based specimen and higher polycondensation rate due to optimum water content in OPS based
Microstructure specimens have stimulated the rapid production of geopolymer gel and improved the strength perfor-
mance of geopolymer concrete in ambient curing condition. Moderate water absorption from OPS fibers
and additional dissolution of Al and Si elements due to formation of CeAeSeH gels from limestone
inclusion have presented 10% OPS and 25% limestone as the optimum replacement to river sand and
coarse granite aggregate, respectively. In ambient curing, 25% limestone presented a 35.23% higher
strength while 10% OPS had 11.01% higher strength than control. An escalation to the curing temperature
reduced the efficiency of both aggregates, where only 25% limestone was able to produce specimen with
2.34% higher strength than control A supportive analysis from SEM and EDX has also revealed the ex-
istence of poly(sialate), poly(sialate-siloxo), poly(sialate-disiloxo) and sialate link in the geopolymer
matrix. From this research, the applications of limestone and oil palm shell in geopolymer concrete have
presented a nurturing result, particularly as the alternative aggregates in ambient-cured application.
© 2015 Elsevier Ltd. All rights reserved.

1. Introduction popularity as an environmentally friendly binder and invited

several researches to investigate further into its feasibility as ag-
Environmental concerns due to CO2 emissions from the gregates binder in concrete.
manufacturing process of Ordinary Portland Cement (OPC) has McLellan et al. (2011) have studied that based on numerous
encouraged former researchers to investigate the suitability of geopolymer concrete researches conducted in Australia region,
alternative material to replace OPC roles in concrete. Geopolymer geopolymer concrete indicates a potential 44e64% reduction in
was first introduced by Joseph Davidovits in 1972 by reacting greenhouse gas emissions, yet implicate financial costs 7% lower to
alumina-silica rich material with sodium or potassium hydroxide 39% higher compared to OPC. Another recent analysis on the
solution. Geopolymer's zero-OPC approach has increased its environmental evaluation of geopolymer using Life Cycle Assess-
ment method has also revealed that due to the heavy impacts of
sodium silicate solution production, geopolymer concrete presents
Abbreviations: OPS, oil palm shell; ITZ, interfacial transition zone; SEM, scan- a similar environmental impact to conventional OPC concrete
ning electron microscope; EDX, energy dispersive X-ray. (Habert et al., 2011). Utilization of waste material, including source
* Corresponding author. Tel.: þ60 95492337. material containing a suitable Si/Al molar ratio, is suggested to
E-mail address: [email protected] (A. Kusbiantoro).

http://dx.doi.org/10.1016/j.jclepro.2015.06.058
0959-6526/© 2015 Elsevier Ltd. All rights reserved.
 R. Embong et al. / Journal of Cleaner Production 112 (2016) 816e822 817

reduce the negative impact of this excessive sodium silicate usage degradation hence makes the disposal process become a critical
 and Hanzlí
(Habert et al., 2011). In 2014, Perna cek have studied the issue. Currently, Malaysia and Indonesia contribute to almost 80% of
solidification of aluminum production waste by using geopolymer total palm oil fruit world production (FAO, 2015). Large concen-
technology. The solid product was found to be feasible for imme- tration of palm oil industry in these regions will result in an urgent
diate use as building material, particularly as thermal-insulation demand for an environmentally friendly solution that able to
material. Other research carried by Topçu et al. (2014) by using consume this by-product in large quantity.
bottom ash also presented positive results where a durable geo- Most of the studies in geopolymer have described the funda-
polymer concrete only loss 6.77% of its strength after freezeethaw mental difference between aluminosilicate framework in geo-
cycle. Phosphogypsum is also feasible to be applied in geopolymer polymer and calcium silicate hydrate in hydraulic Portland cement
technology as the alternative source material (Rashad, 2015). A system, yet the role of water and free calcium in the strength
combination of 90% fly ash: 10% phosphogypsum has presented a development of geopolymer concrete has not yet fully understood.
positive strength improvement, even after exposure to 1000  C Therefore this study was conducted to acquire the complete ca-
temperature for 2 h. Nevertheless, the abundant availability of fly pacity of limestone and oil palm shell as a replacement material for
ash as an industrial waste and its suitable Si/Al ratio has been the aggregate, considering their interesting feature that contradict to
focus of many researchers to reuse and keep it as a feedstock for the well-known basic features of geopolymer, which are alumino-
geopolymer (McLellan et al., 2011). silicate based framework in non-hydraulic binder system. The
Class-F fly ash is a common source material in the study also focused on the area of ITZ between large particle of
manufacturing of geopolymer concrete. Low calcium oxide content aggregate and geopolymer paste, since it affects the microstructural
in this class is the important feature to introduce a moderate loading mechanism in generating compressive strength
environment for the geopolymerization of alumina and silica pre- performance.
cursors in geopolymer system. It has been studied that the exces-
sive content of calcium in the geopolymer system is presenting a 2. Experimental investigation
large number of nucleation centers in the initial paste solution
hence increasing its solidification rate (Lee and Van Deventer, 2.1. Materials
2002; Van Deventer et al., 2007). Shorter setting time developed
from the high solidification rate is proportionally related to the Independent approaches were designed for both coarse and fine
lower workability of paste produced. In the manufacturing of aggregate analysis in this research. Due to certain workability issue
conventional concrete, workability is a decisive factor since it will in the combination of limestone and OPS as aggregate in geo-
affect the quality of subsequent works i.e. transporting, casting, polymer concrete, analyses were conducted separately where
compacting, which in turn will contribute to the mechanical limestone replaced granite as the coarse aggregate and OPS
strength and durability of concrete produced. replaced sand in fine aggregate. OPS was taken as a by-product of
In geopolymer concrete, immoderate amounts of calcium are palm oil industry in Pahang, Malaysia. As an organic material, oil
mostly derived from the source material and scarcely from impu- palm shell typically consists of carbon and other volatile materials
rities in other geopolymer components. Even though excessive (>90%) with few portions of absorbed moisture. It also shares a
amount of calcium will decrease geopolymer paste workability, yet similar ratio of holocellulose, cellulose, and lignin contents with
its existence may improve some mechanical properties of low- coconut shell (Daud and Ali, 2004). Nevertheless, normally an oil
calcium fly ash based geopolymers, particularly in the ambient palm shell possesses larger lignin and cellulose portion
temperature. Temuujin et al. (2009) have observed that the sub- (lignin > 50% and cellulose > 25%) as compared to coconut shell,
stitution of fly ash with calcium compounds (CaO and Ca(OH)2) therefore, a proper disposal method has always become a major
improved mechanical properties of geopolymer paste cured at environmental issue since oil palm shell is able to withstand a
ambient temperature (20  C) yet decreased the properties for the natural degradation. Its utilization as alternative aggregate in
70  C specimens. Measurement on the geopolymer paste speci- concrete is expected to provide a clear and more sustainable
mens after seven days of ambient curing recorded a strength disposal method. Fig. 1 presents the particle size distribution of
increment from 11.8 MPa to 22.8 MPa and 29.2 MPa for 3% CaO and both aggregates used in this research.
3% Ca(OH)2 additions, respectively. This feature is important to Geopolymer binder used in this research was produced by
diminish the restriction in the application of geopolymer concrete activating class F fly ash with alkali solutions. Fly ash is an industrial
for massive in-situ structure. waste collected from coal fired power plant Manjung, Malaysia
Aggregate as the main occupant in a concrete, contributes up to while the alkali solution is a mixture of 8 M NaOH solution and
70e80% of concrete volume which in turn will be significantly Na2SiO3 solution. 8 M NaOH was prepared by diluting 297 g of
affecting concrete strength as a composite material. Individually, NaOH pellet (99% purities) with 703 g distilled water, while Na2SiO3
hydrated cement paste and aggregate will perform an elastic solution is a mixture of 14.73% Na2O; 29.75% SiO2; and 55.52% H2O.
behavior before its fracture in a uniaxial compression test, while a Table 1 shows the chemical composition of fly ash used in this
concrete as the combination of them will present an inelastic research.
behavior (Mehta and Monteiro, 2006). This enigmatic behavior is
mainly originating from the existence of interfacial transition zone 2.2. Specimen preparations
(ITZ) between hydrated cement paste and large particle of aggre-
gate (Mehta and Monteiro, 2006). The existence of reactive sub- Scopes of geopolymer concrete mixtures in this research were
stance on aggregate surface is also responsible in the phenomenon divided into coarse and fine aggregate based. In coarse aggregate
of expansion and cracking of concrete structure, which is referred based mixture, limestone replaced 25%, 50%, 75%, and 100% (by
as alkali silica reaction (ASR) (Neville, 2006). weight) of granite content, while OPS replaced 10%, 20%, 30%, and
This study focused on the effect of reactive calcium element 40% (by weight) of sand content in fine aggregate based mixture.
from limestone and free-water availability in the presence of The amounts of binding paste constituents were kept constant in
pervious oil palm shell to the performance of geopolymer concrete. every mixture where 350 kg/m3 of fly ash were reacted with
Oil palm shell (OPS) is a waste product from palm oil production 103 kg/m3 Na2SiO3 solutions and 41 kg/m3 NaOH (8 M) solutions.
that is abundantly available and also highly resistant to natural Table 2 shows the detail of mixture proportions used in this
818 R. Embong et al. / Journal of Cleaner Production 112 (2016) 816e822

Fig. 1. Particle size distribution of aggregates.

research, where batching of materials was conducted using by- based on the study conducted by Swanepoel and Strydom in
weight method. 2002. Even though a prolonged curing duration from 4 to 96 h
Conventional concrete mixing method was adopted to prepare is beneficial to the improvement of fly ash based geopolymer
geopolymer concrete in this study. Divided into dry and wet mixing concrete, yet the strength increment is not significant beyond
phase, all dry materials i.e. fly ash, coarse aggregate, and fine 24 h curing period (Hardjito and Rangan, 2005). Therefore 24 h
aggregate were initially dry-mixed to the constant homogeneity. A was chosen as the main curing duration in oven curing regime. In
combination of NaOH and Na2SiO3 was then included into the dry ambient curing, specimens were kept at the average temperature
mixture and continued with wet-mixing until a consistent fresh of 28 ± 2  C until the testing day. In oven curing, the specimens
concrete mixture was obtained. were oven-cured in the temperature of 60  C for 24 h. After 24 h,
Curing methods were divided into room-temperature and the specimens were de-molded and placed in room-temperature
60  C oven curing to provide more comprehensive analysis on until the testing days. Slump test was conducted on fresh geo-
the effect of aggregate substance to the geopolymerization pro- polymer concrete with reference to BS EN 12350-2:2009.
cess. Oven curing was applied since elevated temperature could Compression test on the hardened cube was conducted at 1, 7,
accelerate the condensation and polymerization process of the and 28 days with porosity, Scanning Electron Microscope (SEM),
supersaturated aluminosilicate from the dissolution process. and Energy Dispersive X-Ray (EDX) observations were carried out
60  C was selected as the optimum oven curing temperature at 28 days. BS EN 12390-3:2009 was used as a reference standard
for the compression test while ASTM C 642 saturation technique
Table 1 was used to measure the volume of permeable voids (%) in
Chemical composition of fly ash. hardened concrete. Analysis on the microstructural observation
Oxide Al2O3 SiO2 CaO Fe2O3 MgO K2O SO3 TiO2 LOI and elemental compositions were conducted using Hitachi S3400
SEM & Horiba EMAX EDX.
Percentage (%) 36.37 17.57 10.58 12.43 3.05 1.77 1.39 0.88 1.19

Table 2
Detail of mixture proportions.

Mix code Fly ash (kg/m3) NaOH (kg/m3) Na2SiO3 (kg/m3) Coarse Aggregate Fine Aggregate Curing method

Granite (kg/m3) Limestone (kg/m3) Sand (kg/m3) OPS (kg/m3)

GA 350 41 103 1250 0 650 0 Ambient

C1A 350 41 103 937.5 312.5 650 0 Ambient
C2A 350 41 103 625 625 650 0 Ambient
C3A 350 41 103 312.5 937.5 650 0 Ambient
C4A 350 41 103 0 1250 650 0 Ambient
F1A 350 41 103 1250 0 585 65 Ambient
F2A 350 41 103 1250 0 520 130 Ambient
F3A 350 41 103 1250 0 455 195 Ambient
F4A 350 41 103 1250 0 390 260 Ambient
GO 350 41 103 1250 0 650 0 Oven
C1O 350 41 103 937.5 312.5 650 0 Oven
C2O 350 41 103 625 625 650 0 Oven
C3O 350 41 103 312.5 937.5 650 0 Oven
C4O 350 41 103 0 1250 650 0 Oven
F1O 350 41 103 1250 0 585 65 Oven
F2O 350 41 103 1250 0 520 130 Oven
F3O 350 41 103 1250 0 455 195 Oven
F4O 350 41 103 1250 0 390 260 Oven
 R. Embong et al. / Journal of Cleaner Production 112 (2016) 816e822 819

3. Result and discussion

3.1. Workability of fresh concrete

Observations on the workability performances indicated insig-

nificant different in terms of slump height and diameter, however
manual observation through continuous mixing had shown a
decreasing setting time profile when limestone and OPS were
introduced into the mixture. Consistent slump height and diameter
figures were caused by the high fluidity and small cohesion of fresh
geopolymer paste, therefore setting time analysis was added to
provide more realistic information in the workability characteristic
of fresh geopolymer concrete. Fig. 2 shows the workability results
of fresh concrete specimens with limestone while Fig. 3 is for
specimen with OPS inclusion. It is evident in here that the hard- Fig. 3. Workability of geopolymer concrete containing OPS.
ening mechanism of geopolymer concrete is significantly affected
by the dissolution of free calcium ion from limestone surface or the
natural water absorption characteristic of organic substances like calcium silicate hydrate gels produced by Portland cement. Geo-
OPS. polymerization involves simultaneous reaction of dis-
As stated by Puligilla and Mondal (2013), calcium element solutionehydrolysisepolycondensation that generally uses
dissolved in the solution tends to react with silicon ions and form elevated temperature to accelerate its polycondensation process
calcium silicate hydrate formations which define the hardening (Khale and Chaudhary, 2007; Xu et al., 2001). Polycondensation
profile of fresh geopolymer concrete. The hypothesis is also (releasing water) process, which is a simultaneous continuity from
supported by Garcia-Lodeiro et al. (2011) where the formation of hydrolysis (consuming water) process, is accelerated rapidly during
aluminate-substituted calcium silicate hydrate (CeAeSeH) gel is elevated temperature curing to form geopolymer gels (Duxson
preferred over sodium aluminosilicate hydrate in the presence of et al., 2007). As presented in Figs. 4(b) and 5(b), the compressive
Ca element in a high pH environment. In this research, with pH in strengths of oven-cured specimens can achieve up to 65.91% higher
the range of 11.5e12.5, precipitation of CeAeSeH framework has than ambient-cured specimens (Figs. 4(a) and 5(a)). However, the
accelerated the hardening process of geopolymer concrete increasing proportion of limestone has presented a strength
mixture through the solidification of structural gel in the early reduction trend, even though the average values are still higher
stage. It explains the rapid loss of plasticity in fresh geopolymer than the ambient-cured specimens.
concrete when limestone is included as the aggregate An interesting phenomenon occurred when limestone was
replacement. added into the concrete and cured in the ambient condition. As
A similar hardening profile was also observed in OPS based compared to control, 25%, 50% and 75% of limestone presented
mixture. Slump heights were measured in the range of strength improvements up to 35.23%, 19.53%, and 14.15%,
190e210 mm with slump diameters were within 350e410 mm.
Nevertheless, the setting time has decreased from 45 min to 30 min
when 10% OPS was included as the aggregate replacement. This
setting time depreciation was largely believed due to water ab-
sorption characteristic of OPS that has limited the lubrication sys-
tem within geopolymer concrete components. It can be seen from
the setting time results where 20%, 30%, and 40% of OPS inclusion
have reduced the setting time to 26, 23, and 21 min, respectively.

3.2. Compressive strength

Structural framework of geopolymer concrete is supported by

aluminosilicate formations, which is completely different from

Fig. 4. Compressive strength of geopolymer concrete containing limestone in: (a)

Fig. 2. Workability of geopolymer concrete containing limestone. ambient curing; (b) oven curing.
820 R. Embong et al. / Journal of Cleaner Production 112 (2016) 816e822

commence, which in turn will also release the aluminate com-

pound (Puligilla and Mondal, 2013). This AleSi rich layer will
absorb Hþ to increase the concentration of OH, hence the high
alkalinity of geopolymer system will stimulate further dissolution
of Al and Si elements from fly ash particles (Shi et al., 2006;
Rajaokarivony-Andriambololona et al., 1990). Nevertheless, a full
replacement of granite with limestone in geopolymer will
contribute nil improvement to the concrete strength. Lower ag-
gregate's packing density provided by uniformly graded limestone
in the concrete system has reduced the loading path connectivity
where not all voids are filled. Semi-rigid mixture having low plas-
ticity due to rapid solidification CeAeSeH gels has created
numerous unconsolidated voids in geopolymer concrete frame-
work. Therefore the existence of these pores/voids will facilitate
premature failure by providing generous crack paths in the hard-
ened concrete specimens.
Meanwhile in oven curing regime, a consistent strength
improvement was obtained at 25% limestone inclusion. Further
additions of limestone proportions up to 50%, 75%, and 100%
gradually decreased the compressive strength performance. The
availability of elevated temperature in this curing regime is sup-
porting the rapid hydrolysis and polycondensation to diminish the
higher activation energy barrier in C1O mixture, yet poor packing
density is showing more dominant influence in the strength per-
formance of other limestone based mixtures.
Fig. 5(a) illustrates the absorptive characteristic of organic ma-
terials in OPS that assisted the initial strength development of
Fig. 5. Compressive strength of geopolymer concrete containing OPS in: (a) ambient
curing; (b) oven curing. ambient cured specimens, particularly in the first 24 h. It is a
notorious fact that hydroxide activity is considerably affected by
the excessive water content in geopolymer concrete system. The
respectively. Meanwhile 100% of limestone reduced the strength up absorptive characteristic of OPS has decreased the water content in
to 10.21%. Poor gradation of limestone and possible dissolution of geopolymer system through free-water absorption, which will
additional calcium ion into the geopolymer system are believed as subsequently result in a rapid transportation of hydroxyl ion on to
the main causes to the divergent trend of compressive strength fly ash particles (Kusbiantoro et al., 2012). With additional disso-
development in this research. lution of aluminosilicate compound, it will also be followed by the
The escalating strength of limestone based specimens in gelation of supersaturated aluminosilicate solution (Duxson et al.,
ambient curing (Fig. 4(a)) was believed as a result of CeAeSeH gels 2007).
formation that provided solid structural framework in geopolymer Based on compressive strength performance, 10% OPS presented
system. The effect of low reactivity due to lack of activation energy the highest compressive strength among others OPS based mix-
in ambient curing was reduced via additional dissolution of Si tures. Even though its strength was inferior to control specimen in
element from fly ash to form CeAeSeH gels structure (Puligilla and oven curing condition, yet this result has illustrated the importance
Mondal, 2013). Consequently, as the concentration of Si ion needs of water component in geopolymer mixture during early age pe-
to be preserved in the system, further dissolution of fly ash will riods. The presence of excessive water in geopolymer system will

Fig. 6. SEM images of geopolymer concrete containing various types of aggregates: (a) granite, (b) limestone, (c) oil palm shell.
 R. Embong et al. / Journal of Cleaner Production 112 (2016) 816e822 821

subside the polycondensation rate hence producing geopolymer the surface. Imperfect geopolymerization product was also
frameworks with lower strength. Meanwhile, insufficient water observed in Fig. 6(c) where porous gels condensate on fly ash sur-
amount will reduce the condensation/polymerization rate, which face and formed a mesh layer covering fly ash surface.
will further decrease geopolymer concrete compressive strength
(Provis et al., 2005; Blum and Lasaga, 1988). It explains the higher
compressive strength of 10% and 20% OPS mixtures than control 3.3.2. Energy dispersive X-ray spectroscopy
specimen. The amount of water available for geopolymerization EDX measurement was conducted along with SEM test to pro-
process appears to be optimum at 20% OPS, further addition of OPS vide a quantitative analysis on the elemental composition of geo-
will create insufficient water for dissolution of fly ash particles polymer microstructure. Table 3 shows the elemental weight
hence reducing the condensation of aluminosilicate compound in percentage of specimens tested in SEM. As observed in this table,
geopolymer system. Nevertheless, rapid polycondensation of con- the ratio of Si/Al decreased from 5.1 to 3.7 and 3.4 in the addition of
trol specimen in oven curing was still superior to all OPS based limestone and OPS, respectively. It agrees the hypothesis on the
mixtures. The availability of elevated temperature in oven curing CeAeSeH formations in limestone based specimen that initiate
has accelerated every process in geopolymer system. Rapid disso- further dissolution of Al and Si elements from fly ash particles.
lution of fly ash particles to produce supersaturated aluminosilicate Since the available Si has affiliated with CeAeSeH networks,
solution for further condensation/polymerization process has pro- additional dissolution of Si element was needed to maintain the
duced solid geopolymer framework that contributes to the superior concentration of Si in geopolymer system; hence initiated further
compressive strength. Nevertheless, slower dissolution due to dissolution of Al and Si elements from fly ash particles.
insufficient water in geopolymer system has also resulted in a Meanwhile in OPS based specimen, the absorption of excessive
lower quality of geopolymer framework produced. It is clearly water has promoted slight increment in the dissolution of Al and Si
illustrated by the decreasing compressive strength along with the elements. However excessive water absorption has reduced poly-
increasing OPS amounts in Fig. 5(b). condensation rate of these geopolymer precursors, hence they
remained unreacted and deposited on geopolymer matrix. Based
on Si/Al and Na/Al ratio provided in Table 3, three types of geo-
3.3. Microstructure analysis
polymer structure, namely poly(sialate), poly(sialate-siloxo), and
poly(sialate-disiloxo) were detected in this geopolymer system.
3.3.1. Scanning electron microscopy
Additional structure of sialate link or poly(sialate-multisiloxo) was
SEM and EDX analysis were used to evaluate the characteristic of
also detected in control specimens. Sialate link is generally con-
geopolymer microstructure. Three samples were collected from
necting poly(siloxonate), poly(silanol), or poly(sialate) chains via
each mixture group that represented control, limestone based
SieOeAl bridge and having Si/Al > 3 (Davidovits, 2008). Meanwhile
mixture, and OPS based mixture. Fig. 6 shows SEM images of the
poly(sialate), poly(sialate-siloxo), and poly(sialate-disiloxo) are
respective mixtures in oven curing condition. The main feature in
geopolymer structure with Si/Al ¼ 1, 2, and, 3 respectively
these images is the appearance of ITZ between aggregate and
(Davidovits, 2008).
geopolymer matrix. An enhancement in the dissolution of fly ash
particles and polycondensation of aluminosilicate compound has
created a good bonding between these two elements. With proper
formation of aluminosilicate gels to fill the ITZ in Fig. 6(a) and (b),
compressive strength of control and limestone based specimens
were significantly higher than OPS based specimens.
The existence of large ITZ gap in OPS based mixture (Fig. 6(c))
has expedited the crack path connectivity in geopolymer matrix
(average ITZ gap in OPS based mixture was 80.07 mm), which will
further trigger the premature failure of geopolymer concrete.
Defective load transfers from aggregate-to-matrix and vice versa
has facilitated cracking path to be easily connected with micro
pores in geopolymer matrix. An interesting part of OPS based
mixture in Fig. 6(c) is the existence of fiber straw covering geo-
polymer matrices. It is believed the lower dissolution of fly ash
particle was also due to substantial water absorption by OPS fibers.
As observed in this image, unreacted fly ash particles were
dispersed in geopolymer matrices with several fiber straws covered

Table 3
Elemental weight percentage from EDX analysis (wt%).

Element Control Limestone based OPS based

C 4.3 6.3 15.6

O 39.3 44.8 49.2
Na 2.5 5.6 2.7
Al 2.6 5.5 5.5
Si 14.3 20.7 16.0
K 1.0 5.3 1.4
Ca 14.2 5.6 3.9
Mg 1.2 0.7 0.8

Si/Al 5.5 3.8 2.9

Fig. 7. Correlation between porosity and compressive strength of geopolymer concrete
Na/Al 1.0 1.0 0.5
containing: (a) oil palm shell; (b) limestone.
822 R. Embong et al. / Journal of Cleaner Production 112 (2016) 816e822

3.3.3. Porosity References

Fig. 7 shows the correlation between compressive strength and
porosity for both Oil Palm Shell and Limestone mixture. As a Blum, A.E., Lasaga, A.C., 1988. Role of surface speciation in the low temperature
dissolution of minerals. Nature 331, 431e433.
composite material, geopolymer concrete displayed a common Daud, W.M.A.W., Ali, W.S.W., 2004. Comparison on pore development of activated
correlation trend between porosity and compressive strength carbon produced from palm shell and coconut shell. Bioresour. Technol. 93,
properties, yet higher reliability of the trend line is obtained when 63e69.
Davidovits, J., 2008. Geopolymer Chemistry and Application. Institut Geopolymere,
geopolymer concrete is exposed to the elevated temperature dur- France.
ing curing process. A consistent polymerization process in elevated Duxson, P., Fernandez-Jimenez, A., Provis, J.L., Lukey, G.C., Palomo, A., van
temperature curing was able to provide more binding gel to fill the Deventer, J.S.J., 2007. Geopolymer technology: the current state of the art.
J. Mater. Sci. 42, 2917e2933.
previously water-filled pores. This curing regime has provided less Food and Agriculture Organization of the United Nations, 2005. http://faostat3.fao.
data variations, hence a decent correlation between porosity, and org (accessed 06.05.15.).
compressive strength can be established. Garcia-Lodeiro, I., Palomo, A., Fernandez-Jimenez, A., Macphee, D.E., 2011.
Compatibility studies between NeAeSeH and CeAeSeH gels: study in the
The presence of excessive pores formation in geopolymer
ternary diagram Na2OeCaOeAl2O3eSiO2eH2O. Cem. Concr. Res. 41 (2),
framework is directly proportional to the declining compressive 923e931.
strength of geopolymer concrete specimens. As presented in Habert, G., d'Espinose de Lacaillerie, J.B., Roussel, N., 2011. An environmental
evaluation of geopolymer based concrete production: reviewing current
Fig. 7, OPS based specimens has yielded a stronger prediction in
research trends. J. Clean. Prod. 19, 1229e1238.
the correlation between compressive strength and porosity than Hardjito, D., Rangan, B.V., 2005. Development and Properties of Low-Calcium Fly
limestone based mixture. Anomalies in the Ca reactivity Ash Based Geopolymer Concrete. Curtin University of Technology, Perth,
appeared to contribute more on to the fluctuating results; hence Australia.
Khale, D., Chaudhary, R., 2007. Mechanism of geopolymerization and factors
generated difficulties to provide a reliable prediction particularly influencing its development: a review. J. Mater. Sci. 42, 729e746.
in ambient curing condition. It was also aggravated by poor Kusbiantoro, A., Nuruddin, M.F., Shafiq, N., Qazi, S., 2012. The effect of microwave
distribution of limestone particle in geopolymer concrete system. incinerated rice husk ash on the compressive and bond strength of fly ash based
geopolymer concrete. Constr. Build. Mater. 36, 695e703.
Meanwhile, the porous nature of OPS fragments has presented a Lee, W.K.W., Van Deventer, J.S.J., 2002. The effect of ionic contaminants on the early-
consistent water absorption rate which affected poly- age properties of alkali-activated fly ash-based cements. Cem. Concr. Res. 32,
condensation process in producing geopolymer gel. A consistent 577e584.
McLellan, B.C., Williams, R.P., Lay, J., van Riessen, A., Corder, G.D., 2011. Costs and
reduction in the amount of available solution has made OPS carbon emissions for geopolymer pastes in comparison to ordinary Portland
based specimens presented better fitness in predicting the cor- cement. J. Clean. Prod. 19, 1080e1090.
relation between porosity-compressive strength than limestone Mehta, P.K., Monteiro, P.J.M., 2006. Concrete Microstructure, Properties, and Ma-
terials. McGraw-Hill, United States of America.
based specimens.
Neville, A.M., 2006. Properties of Concrete. Pearson Education Limited, Harlow,
England.
4. Conclusion Perna, I., Hanzlí
cek, T., 2014. The solidification of aluminium production waste in
geopolymer matrix. J. Clean. Prod. 84, 657e662.
Provis, J.L., Duxson, P., van Deventer, J.S.J., Lukey, G.C., 2005. The role of mathe-
Based on the results, it can be concluded that the inclusion of matical modeling and gel chemistry in advancing geopolymer technology.
limestone and OPS as alternative aggregates have produced geo- Chem. Eng. Res. Des. 83, 853e860.
polymer concrete with different characteristic, particularly at the Puligilla, S., Mondal, P., 2013. Role of slag in microstructural development and
hardening of fly ash-slag geopolymer. Cem. Concr. Res. 43, 70e80.
microstructural level. Large interfacial gap between geopolymer Rajaokarivony-Andriambololona, Z., Thomassin, J.H., Baillif, P., Touray, J.C., 1990.
matrix and aggregate in OPS based specimens significantly influ- Experimental hydration of two synthetic glassy blast furnace slags in water and
enced the compressive strength performance of hardened con- alkaline solutions (NaOH and KOH 0.1 N) at 40  C: structure, composition, and
origin of the hydrated layer. J. Mater. Sci. 25, 2399e3410.
crete specimens. Excessive water absorption from OPS fibers has Rashad, A.M., 2015. Potential use of phosphogypsum in alkali-activated fly ash
created unfavorable environment for the polymerization of under the effects of elevated temperatures and thermal shock cycles. J. Clean.
aluminosilicate precursors. However, at its optimum amount, OPS Prod. 87, 717e725.
Shi, C., Krivenko, P.V., Roy, D., 2006. Alkali-Activated Cements and Concrete. Taylor
was able to assist the initial strength development via water ab- and Francis, Abingdon, UK.
sorption mechanism to increase the hydroxide activity and Swanepoel, J.C., Strydom, C.A., 2002. Utilisation of fly ash in a geopolymeric ma-
transportation of hydroxyl ion on to fly ash particles. In ambient terial. Appl. Geochem. 17, 1143e1148.
Temuujin, J., Van Riessen, A., Williams, R., 2009. Influence of calcium compounds on
curing, 10% OPS and 25% limestone were the optimum replace-
the mechanical properties of fly ash geopolymer pastes. J. Hazard. Mater. 167,
ment levels for fine and coarse aggregate proportion in geo- 82e88.
polymer concrete. While in oven curing, only 10% of limestone Topçu, I.B., Toprak, M.U., Uygunog lu, T., 2014. Durability and microstructure char-
acteristic of alkali activate bottom ash geopolymer cement. J. Clean. Prod. 81,
proportion that was able to produce geopolymer concrete with
211e217.
higher strength than control. Van Deventer, J.S.J., Provis, J.L., Duxson, P., Lukey, G.C., 2007. Reaction mechanisms
in the geopolymeric conversion of inorganic waste to useful products. J. Hazard.
Acknowledgment Mater. A139, 506e513.
Xu, H., van Deventer, J.S.J., Lukey, G.C., 2001. Effect of alkali metals on the prefer-
ential geopolymerization of stilbite/kaolinite mixtures. Ind. Eng. Chem. Res. 40,
The authors would like to thank the financial support from 3749e3756.
Ministry of Education Malaysia and Universiti Malaysia Pahang
under research grant RDU 130614.