Materials and Design 56 (2014) 833–841

Contents lists available at ScienceDirect

Materials and Design

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

The development of compressive strength of ground granulated blastfurnace slag-palm oil fuel ash-fly ash based geopolymer mortar

0261-3069/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.matdes.2013.11.080

⇑ Corresponding author. Tel.: +60 3 7967 7632; fax: +60 3 7967 5318.E-mail addresses: johnson@um.edu.my, ujohnrose@yahoo.com (U.J. Alengaram).

Azizul Islam, U. Johnson Alengaram ⇑, Mohd Zamin Jumaat, Iftekhair Ibnul BasharDepartment of Civil Engineering, Faculty of Engineering, University of Malaya, 50603 Kuala Lumpur, Malaysia

a r t i c l e i n f o

Article history:Received 15 July 2013Accepted 30 November 2013Available online 7 December 2013

Keywords:Palm oil fuel ashGround granulated blast furnace slagFly ashManufactured sandGeopolymer mortarCompressive strength

a b s t r a c t

The paper presents the report on the use of optimum level of palm oil fuel ash (POFA), ground granulatedblast furnace slag (GGBS) and low calcium fly-ash (FA) with manufactured sand (M-sand) to producegeopolymer mortar. Eleven mixtures were prepared with varying binder contents with the POFA contentvarying between 25% and 100%; the other constituent materials such as fine aggregate and water werekept constant. All the specimens were cured in oven for 24 h at 65 �C and thereafter kept in room tem-perature (about 26–29 �C) before testing for the compressive strength. The highest compressive strengthof about 66 MPa was achieved for the mortar containing 30% of POFA and 70% of GGBS with a total bindercontent of 460 kg/m3. The increase in the POFA content beyond 30% reduces the compressive strength.The density reduction after 3 days was found negligible.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction One of the main goals in achieving sustainable construction

The huge demand for concrete using Ordinary Portland Cement(OPC) has resulted in high volume of CO2 emission, and lead to eco-logical imbalance due to continuous depletion of natural resources.The reality of air pollution through carbon dioxide (CO2) emissioninto the atmosphere from the production of cement is well known.The contribution of OPC production worldwide to greenhouse gasemission is estimated to be about 6% of the total greenhouse gasemissions [1]. In addition, the depletion of natural sand due toquarrying activities has already caused flooding in many parts ofthe world; the need for alternative materials to reduce naturalsand through the use of recycling of old mortar [2–4] has also beeninvestigated; however there have been efforts to utilize the manu-factured sand, commonly known as M-sand from the waste ofcrushed granite aggregates.

The process of formation of CO2 by calcining can be expressedby the following equation:

CaCO3ð1 kgÞ

! CaOð0:56 kgÞ

þ CO2ð0:44 kgÞ

ð1Þ

The share of CaO in clinker amounts to 64–67%. The remainderconsists of silicon oxides, iron oxides, and aluminium oxides.Therefore, CO2 emission from clinker production amounts to about0.5 kg/kg. The CO2 emission per tonne of cement depends on theratio of clinker to cement. This ratio varies normally from 0.5 to0.95 [5].

materials is to reduce the overuse of virgin materials used toproduce cement, coarse and fine aggregates. The utilization ofindustrial by-products such as fly ash (FA), silica fume, groundgranulated blast furnace slag (GGBS), and rice husk ash, as the ce-ment replacement or as the additional cementitious materials hashad a constructive effect in minimizing greenhouse gas emissions.Every year millions of tons of industrial wastes are generated andmost of these wastes are unutilized or underutilized; these wastescause environmental issues due to storage problem and pollutionto the surrounding field. In recent years, there is an increasingawareness on the quantity and diversity of hazardous solid wastegeneration and its impact on the human health. Increasing concernabout the environmental consequences of waste disposal has ledresearchers to investigate the utilization of the wastes as potentialconstruction materials [6].

In order to achieve an environmentally friendly concrete,several studies are on-going on the utilization of waste materialsto produce green concrete. Among the researches, the successfulone was through the development of geopolymer concrete to elim-inate the use of cement. FA based geopolymer concrete was firstintroduced by Davidovits [7] in 1979 to reduce the use of OPC inconcrete. Geopolymer concrete is well-suited to manufactureprecast concrete products that can be used in infrastructure devel-opments [8]. A number of researchers [8–13] have published arti-cles on the use of FA as source material in the development ofgeopolymer concrete. The significant research in geopolymerincludes thermal behaviour [9], durability in sodium and magne-sium sulfate solutions [10], and resistance to acid attack [11] ofgeopolymeric materials.

Fig. 1. Structural models of geopolymer concrete – Davidovits [7].

Table 1CO2 emissions for OPC and blast furnace slag [1].

CO2 emission/t OPC (kg) CO2 emission/t GGBS (kg)

Calcination of CaCO3 540 0Fossil fuel (coal) 340 20 (drying)Electricity generation 90 50

Total 970 70

834 A. Islam et al. / Materials and Design 56 (2014) 833–841

Geopolymer is an inorganic alumino-hydroxide polymer syn-thesized from predominantly silicon and aluminium materials ofgeological origin and industrial by-product material such as FA(with low calcium) as shown in Fig. 1.

FA is a fine powder of mainly spherical glass particles havingpozzolanic properties which shall consist essentially of reactive sil-icon dioxide (SiO2) and aluminium oxide (Al2O3), the remainderbeing iron (III) oxide (Fe2O3) and other oxides. It can be obtainedby electrostatic or mechanical precipitation of dust-like particlesfrom the flue gases of power station furnaces fired with pulverizedbituminous or other hard coal [14]. The government of Malaysiadecided that by 2010 the share of coal in the fuel mix for electricitygeneration would rise to about 40% [15]. The increased use of coalburning in thermal power plants has increased the production ofFA to an estimated 3 million tons per annum. The abundance ofFA in Malaysia could pave way for the development of geopolymerconcrete.

The other waste material that is abundant in Malaysia is GGBS,a by-product of the production of iron in a blast furnace and it iscomposed chiefly of calcium and magnesium silicates and alumi-nosilicates. The history of slags used for cement is not new. A quitenumber of investigations have been performed on the use of GGBSas a cementitious material in cement production since 1939 and toevaluate its performance [16]. GGBS can be used for producinghigh quality self-compacting concrete (100 MPa) [17]. Researchersexpress the reactivity of GGBS in terms of slag activity index (SAI)[17]. The use of GGBS as cement replacement material in geopoly-mer concrete reduces the CO2 emission. Bakharev et al. [18,19]reported that alkali-activated slag concrete had lower resistanceto carbonation and alkali-aggregate than that of OPC concrete ofsimilar grade. Table 1 shows the comparison of CO2 emission be-tween the OPC and GGBS and as seen from the results, one tonneof GGBS releases only about 70 kg of CO2, that is only 7% of CO2

of cement for the same quantity of material produced. Table 1shows the comparison of CO2 emission from the production ofOPC and the GGBS.

Another locally abundant waste material from palm oil industryis palm oil fuel ash (POFA). Malaysia is currently producing morethan half of the world’s total output of palm oil, planted over 5 mil-lion hectares of land, yielding about 18.89 tonnes/hectare of freshfruit bunch (FEB) [20]. The empty fruit bunches (EFB), have tradi-tionally been burnt and their ash recycled into the plantation asfertilizer. Another waste from palm oil industries is Oil Palm Shell(OPS) and it is obtained during the extraction of palm oil by crush-ing of the palm nut in the palm oil mills. After the extraction ofpalm oil from the palm oil fruit, both palm oil husk and OPS areburnt as fuel in the boiler to generate electricity. Palm oil fuelash is commonly known as POFA, which is about 5% of solid wasteproduct, have the potentiality to be used as pozzolanic materials inconcrete industry [21]. A large area is required to dispose thesewaste materials; the disposal of these wastes in the vicinity ofthe factories results in environmental concerns as both the landand ground water are likely to be affected due to pollution. Theproduction of these wastes is on the rise every year due to high de-mand for palm oil. There has been a number of research works thatare being carried out to utilize these waste material i.e., OPS usedfor lightweight concrete [22–26] and POFA used for high strengthconcrete [27,28]. Alengaram et al. [24] reported that the increasein sand content coupled with reduction in OPS content enhancedthe compressive strength of concrete. POFA is a cementitiousmaterial, rich in silica that could be used with recycled aggregateconcrete that could result in a higher compressive strength thanthat of recycled aggregate concrete without ground POFA [29].

(P) - Passing, (R) – Retained

0

10

20

30

40

50

0

20

40

60

80

100

0.01 0.1 1 10 100 1000

% R

etai

ned

% P

assi

ng

Particle size (µm) (log scale)

POFA (P) GGBS (P) Fly-ash (P)

Fly-ash (R) POFA (R) GGBS (R)

Fig. 2. Particle size distribution of GGBS, POFA and fly ash.

A. Islam et al. / Materials and Design 56 (2014) 833–841 835

The POFA produced sometimes varies in colour from whitish greyto darker shade based on the carbon content in it. High-strengthconcrete can be produced using POFA as a pozzolanic materialand it also improves the durability, reduces cost due to reductionin the use of cement [28,30]. It will also be beneficial for the envi-ronment with respect to reducing the waste disposal volume oflandfills [28].

Quarrying of natural sand has a great irreversible environmentimpact [31] as it causes reduction in the ground water that affectsthe moisture content of the soil. Due to the drop in river water le-vel, the drinking water will be affected badly specially during thedry season and salt water intrusion is a potential problem. Sandmining causes erosion of nearby land leading to instability in theecosystem [32]. Globally, natural sand and gravel extraction isbecoming less of an option and the cost of river sand increasesdue to the raising demand in construction sector. During the year2010, Malaysia consumed 2.76 billion metric tons of natural aggre-gate of this amount 1.17 billion metric tons or 42.4%, was sand[32]. In many regions of the world, the extraction of sand and grav-el is heavily taxed or banned completely to try to preserve remain-ing deposits [33]. The increasing difficulty in extraction has had anegative effect on the bottom line for many producers. Thus, it isimperative for the construction industries should find alternativesto meet the growing demand for fine aggregates. One of the op-tions is to utilize the waste materials from the crushing of graniteaggregates [34], commonly known as manufactured sand (M-sand)as the fine aggregates. A growing number of quarry operators havefound that processed, high-quality manufactured sand can im-prove their bottom line and significantly reduce the percentageof waste and low-value by-products.

The main objective of this research was to investigate the devel-opment of the compressive strength of geopolymer mortar usingfour locally available waste materials such as FA, POFA and GGBSas binders and M-sand as fine aggregate. The effect of varyingthe percentages of these three binders on the compressive strengthwas also investigated and reported. The optimum compressivestrength of cube specimen was determined with various binderdosages, but by keeping other parameters such as sand, waterand activator contents constant.

2. Experimental programme

2.1. Materials

2.1.1. Ground granulated blast furnace slagGround granulated blast furnace slag (GGBS) was obtained from

YTL Cement Marketing Sdn Bhd, Malaysia. The slag activity indexof GGBS was 62% and 108% for 7 and 28 days respectively. The spe-cific gravity was 2.89 g/cm3, specific surface area was 405 m2/kgand the soundness was 1 mm. The particle size distribution ofGGBS is shown in Fig. 2. It is off-white in colour and substantiallylighter than Portland cement. The chemical composition of GGBS isshown in Table 2; while its physical properties are given in Table 3.GGBS shall contain at least two-thirds by mass of glassy slag. Theslag shall consist of at least two-thirds by mass of the sum ofCaO, MgO and SiO2. The remainder contains Al2O3 together withsmall amounts of other oxides. The ratio by mass (CaO + MgO)/(SiO2) shall exceed 1.0 [35].

2.1.2. Palm oil fuel ashPalm oil fuel ash (POFA) was obtained from Jugra Palm Oil Mill

Sdn Bhd, Malaysia. It was then dried in an oven for at least 24 h at100 �C to remove the moisture and then it was sieved through300 lm sieve. Forty mild steel rods of 10 mm diameter and400 mm length were placed in the rotating drum to grind

approximately 10 kg of POFA that was sieved through 300 lm.According to ASTM:C618-12a, the mass of FA and natural pozzolanpassing 45-lm wet sieve shall be at least 66%. This criterion wasfollowed for POFA. The grinding of POFA was carried out for30,000 cycle in 16 h to obtain the desired level of fineness (>66%)and its particle size distribution is shown in Fig. 2. Tables 2 and3 show the chemical composition and physical properties of POFA,respectively. The fineness of POFA was checked at every 4 h ofgrinding interval using a 45-lm sieve according to ASTM: C430-08. The sieve fineness of POFA for different grinding period isshown in Fig. 3. The fineness of POFA was found 88.4%. It was dar-ker in colour. It was found that after processing of raw POFA, theprocessed POFA obtained was about 57%.

2.1.3. Fly ashFly ash (FA) was obtained from Lafarge Malayan Cement Bhd,

Malaysia. According to ASTM:C618-12a, FA is divided into two dis-tinct categories i.e., low-calcium FA (Class F, CaO < 10%) and high-calcium FA (Class C, CaO > 10%). In this study low-calcium FA wasused. The chemical composition and physical properties are shownin Tables 2 and 3, respectively.

2.1.4. Manufactured-sandManufactured sand (M-sand) was obtained from Batu Tiga

Quarry Sdn Bhd (YTL), Malaysia. Generally, the quarry dust (QD)obtained during the crushing of granite aggregate is consideredwaste and sometime used in land filling; however, recently thereis a renewed interest to reuse the QD. Thus, the QD is processedthrough centrifuge action to smoothen the angular edges and theresulting particles are rounded and it is used to replace for naturalsand (Fig. 4). The processed QD is christened as M-sand and widelyused in Singapore, India and some other countries to replace con-ventional sand where the demand for sand is high/not available inabundance. The M-sand has a wide range of particles as shown inthe distribution curve (Fig. 5).

From the particle-size distribution curve shown in Fig. 5.

D10 = 0.17 mmD60 = 1.60 mmD30 = 0.60 mmCu = (D60/D10) = 9.41Cc = (D30)2/(D10 � D60) = 1.32

It is observed that Cu (uniformity co-efficient) is greater than 6and Cc (co-efficient of gradation) is between 1 and 3. Hence, theM-sand is well graded and it is under zone-C [BS 882:1992].

Table 2Chemical composition (wt%) of the raw materials, X-ray Fluorescence (XRF) analysis.

Chemical compounds CaO SiO2 Al2O3 MgO Na2O SO3 P2O5 K2O TiO2 MnO Fe2O3 SrO Cl CuO LOI

GGBS 45.83 32.52 13.71 3.27 0.25 1.80 0.04 0.48 0.73 0.35 0.76 0.08 0.02 – 0.60Fly-ash 5.31 54.72 27.28 1.10 0.43 1.01 1.12 1.00 1.82 0.10 5.15 0.36 0.01 0.01 6.80POFA 4.34 63.41 5.55 3.74 0.16 0.91 3.78 6.33 0.33 0.17 4.19 0.02 0.45 6.54 6.20

Table 3Physical properties of constituent materials.

Materials Properties

GGBS Specific gravity: 2.89Specific surface area: 405 m2/kg (min 275 m2/kg, [36])Soundness: 1 mmColour: off-white

POFA Specific gravity: 2.2% Passing 45-lm sieve: 88.4Specific surface area: 172 m2/kgColour: dark

FA Specific gravity: 2.4Specific surface area: 341 m2/kgColour: grey

M-sand Specific gravity: 2.78Fineness modulus (FM): 3.19Grading zone: Zone C

0

20

40

60

80

100

0 4 8 12 16 20

% o

f pa

ssin

g th

ru 4

5 µm

w

et s

ieve

Grinding period (h)

Fig. 3. Fineness of POFA with different grinding periods.

Fig. 4. Processing of M-sand.

0

20

40

60

80

100

1010

% F

iner

Particle size (mm) (log scale)

M-sand(zone-C)

Grading Lower Limit Upper Limit

Fig. 5. Particle size distribution of M-sand (BS 882:1992).

836 A. Islam et al. / Materials and Design 56 (2014) 833–841

2.2. Specimen preparation and curing

2.2.1. Activator solutionThe alkaline activator used was from the combination of so-

dium silicate and sodium hydroxide solution. The activator from

the sodium silicate solution (Na2O = 12%, SiO2 = 30%, andwater = 57% by mass) and sodium hydroxide (NaOH) in flakes orpellets form with 99% purity was prepared according to the refer-ence [37]. The concentration of the sodium hydroxide solutionused was 12 molarity (M) and the mixture contained additionalwater.

2.2.2. Preparation of fresh mortar and castingA total 11 mixtures were prepared by varying the POFA, FA and

GGBS contents. The sand and activators contents were kept con-stant to investigate the effect of the binders. The proportion of bin-der to sand ratio was 1:4. The mixture proportions for mortar aregiven in Table 4. The binder content, solution to binder ratio,molarity of sodium hydroxide solution and curing temperatureare given in Table 5.

The binder and the M-sand were first mixed together in a rotarymixer for about 3 min. The alkaline liquid was then added to thedry materials followed by water and the mixing was continuedfor further 4 min to produce the fresh mortar as shown in Fig. 6.The fresh mortar was compacted and the excess mortar removed.The moulds were covered by plastic film to avoid evaporation ofwater. For each mortar mixture, twelve (12) 50 mm cube specimenwere cast to determine the compressive strength.

2.2.3. CuringImmediately after casting, the test specimens were covered

with plastic film to minimize the water evaporation during curingat an elevated temperature as shown in Fig. 7. The test specimenswere cured in an oven at 65 �C for 24 h. After the curing period, thetest specimens were left in the moulds for at least six hours anddemoulded. After demoulding, the specimens were left to air-drycondition in the laboratory with the temperature and humidityof 27 �C and 70%, respectively until the day of test [37,38].

2.3. Compressive strength

The cubes were tested in compression in accordance with thetest procedures given in ASTM: C109/C109M-13. The compressive

Table 4Mixture proportion (kg/m3).

Mix No. Binding raw materials

GGBSa POFAb FAc

(%) Weight (%) Weight (%) Weight

M1 100 460 0 0 0 0M2 0 0 100 460 0 0M3 0 0 0 0 100 460M4 50 230 50 230 0 0M5 0 0 50 230 50 230M6 50 230 0 0 50 230M7 50 230 25 115 25 115M8 40 184 60 276 0 0M9 40 184 30 138 30 138M10 60 276 40 184 0 0M11 70 322 30 138 0 0

a Ground granulated blast furnace slag.b Palm oil fuel ash.c Class F fly ash.

A. Islam et al. / Materials and Design 56 (2014) 833–841 837

strength value was determined as the average of three specimens.The testing machine and the failure mode of the specimens areshown in Fig. 8.

Compressive strength, fc = failure load (P)/loaded area (a).

3. Results and discussion

3.1. Effect of specific gravity and fineness on the density

Table 6 shows the 3-day oven-dry density (ODD) of the speci-mens. The ODD depends on the specific gravity and fineness ofthe materials. The mix with 100% GGBS that has higher specificgravity and fineness produced the highest density of 2163 kg/m3

(mix M1); on the contrary the mix M2 with 100% POFA producedthe lowest density of 2014 kg/m3. Another factor that influencesthe density is the ability of finer particle to fill the voids withinthe mortar. Fig. 2 shows that POFA has relatively coarser particleswithin a narrow range compared to that of GGBS and FA. ThusGGBS with finer particles enhanced its density of about 7.5% ascompared to mortar with POFA. It was observed that the densityof mortar varies between 2014 kg/m3 and 2163 kg/m3. As indicatedearlier, in the previous works [27–29] investigations on the effectsof ash particle size on properties of geopolymer showed that the fi-ner the particle size, the better the properties.

3.2. Development of compressive strength

The development of compressive strength at the age of 3-, 7-,14- and 28-days are shown in Fig. 9. It can be observed from

Fig. 6. Preparation of mortar.

Fig. 9 that the mixture M3 that contains 100% FA and cured at65 �C for 24-h produced the lowest compressive strength. ThoughBakharev [9] reported that alkali activated cementitious pastesprepared using class F FA produced higher initial compressivestrength for specimens cured at 100 �C compared to specimenscured at 80 �C, the mixes in this investigation were cured at65 �C that contained class F FA, GGBS and POFA. Hence, the reduc-tion in the curing temperature allowed the mixtures with high Cacontent able to achieve the desired strength. It is also reported thatlong pre-curing at room temperature is beneficial for strengthdevelopment of geopolymeric materials utilizing FA; curing at ele-vated temperature allows shortening the time of heat treatment toachieve high strength [12]. For materials utilizing FA activated bysodium silicate, 6-h heat curing is more beneficial for the strengthdevelopment than 24-h heat treatment [12]. The curing at 65 �C for24-h was chosen in this investigation for practical reasons even-though the effect of short curing period is beneficial for FA basedgeopolymer mortars. The mixture M2 contained 100% POFA andwhen it was mixed with GGBS, the strength was increased signifi-cantly. This might be attributed to the packing ability of finer par-ticles. On the contrary the mixture M1 and M11 produced highercompressive strength which contained higher Ca and Al2O3. POFAcontained very less Al2O3 and Ca but when it was mixed withGGBS, the compressive strength increased. So it is observed thatCa and Al2O3 influenced the compressive strength of the mortar[6,16]. The average compressive strength and standard deviationare given in Table 7. Wongpa et al. [39] reported that highersolution to binder ratios (s/b) and higher paste to aggregate (P/Agg) ratios result in lower compressive strength and higher waterpermeability.

Table 8 and Fig. 10 show the increase in the compressivestrength between 3 and 28 days expressed as a percentage. The28-day compressive strength was taken as the reference and the3-, 7-, and 14-days and the ratio of increase in the strengths wascalculated. Most of the specimen achieved 86% of the 28-daystrength at the age of 3-day. Similarly the 7-day and 14-daystrengths were 90% and 94%, respectively of the 28-day strength.

3.3. Effect of GGBS on the compressive strength of the mortar

The ground blast furnace slag employed is a latent hydraulicproduct, which can be activated by suitable activators. Withoutan activation, the development of the strength of the GGBS is ex-tremely slow and the development of the slag necessitates apH P 12 [7]. GGBS plays an important role in the development ofthe compressive strength. Higher concentrations of G.G.B.S (slag)result in higher compressive strength of geopolymer concrete[40]. Fig. 11 shows that the compressive strength of mortar with70% of GGBS (mix No. M11) produced the highest strength whilefurther increase in the GGBS content (mix No. M1) reduces thecompressive strength. The mixture M1 contains 100% GGBS whilethe mix M11 contains 70% GGBS and 30% POFA. A comparison be-tween the mixes M1 and M11 shows that the former with 100%GGBS produces 3% lower compressive strength compared to themix with 70% GGBS and 30% POFA. The mixes M10, M4 and M8show that the reduction of GGBS contents of 10%, 20% and 30%,respectively and the remainder is replaced by POFA. Thus, the ef-fect of GGBS replacement with POFA shows that the mixes M10,M4 and M8 with high content of POFA produced lower strengthof about 9%, 15%, 35%, respectively compared to the mix M11 (with70% GGBS and 30% POFA). The effect of POFA in enhancing thecompressive strength can be seen from Fig. 11 as the mix M4(50% GGBS and 50% POFA) produced 22% higher strength thanthe mix M6 (50% GGBS and 50% FA). Further comparison betweenthe mixes M6 and M7 (50% GGBS, 25% POFA and 25% FA) shows asincrease of about 17% for the latter. As explained earlier that GGBS

Table 5Experimental parameters.

Binder: M-sand Binder(kg/m3)

M-sand(kg/m3)

Activators (1:2.5) (kg/m3) Added water(kg/m3)

s/b(wt/wt)

w/b(wt/wt)

Curing temp.�C

NaOH solution (12 M) Na2SiO3

1:4 460 1840 53 131 184 0.4 0.4 65

s/b: solution to binder weight ratio, w/b: water to binder weight ratio.

Fig. 7. Test specimens covered with plastic film and specimens in air drying condition.

Fig. 8. Compression testing machine and failure mode of cubes.

0

10

20

30

40

50

60

70

0 7 14 21 28

Com

pres

sive

str

engt

h (M

Pa)

Test age (days)

M1 M2 M3 M4

M5 M6 M7 M8

M9 M10 M11

Fig. 9. Development of compressive strength of mortar with varying binder contentratio.

838 A. Islam et al. / Materials and Design 56 (2014) 833–841

has finer particles compared to POFA and its contribution in thestrength development cannot be ignored; however further testsare required to validate the compactness of the structure withinthe mortar.

Naidu et al. [40] reported that the compressive strength ofgeopolymer concrete increases with increase in percentage ofreplacement of fly-ash with GGBS and 90% of the compressivestrength was achieved in 14 days and from this investigation mostof the mortar specimens achieved it in 7 days.

Shafigh et al. [41] used lightweight concrete with high volumeof GGBS and reported that 30% of cement replacement by GGBS in-creased the workability of OPS concrete; however they reported

Table 6Average oven-dry density (ODD) (kg/m3) of mortar at the age of 3-day.

Mix no. M1 M2 M3 M4 M5

(kg/m3) 2163 2014 2020 2116 2021

further increase in GGBS content decreased the workability. Theyalso reported that by introducing initial heating at 60 �C for 20 hafter demoulding, it is possible to improve the compressivestrength of GGBS OPS concrete at early ages.

3.4. Effect of POFA on the compressive strength of the mortar

Fig. 12 shows the effect of POFA content on the compressivestrength at the age of 28-day. The compressive strengths of themixes M2 and M3 that contain 100% POFA and 100% FA, respec-tively shows that the later produced about 50% lower compressivestrength than the mix M2. As seen from Fig. 12 any increase

M6 M7 M8 M9 M10 M11

2135 2121 2107 2112 2157 2159

Table 7Development of the compressive strength (MPa) and standard deviation of mortar at different ages.

Label M1 M2 M3 M4 M5 M6 M7 M8 M9 M10 M11

3-day 60 13 7 53 6 45 47 41 47 55 54(1.45) (0.125) (1.03) (2.4) (0.38) (1.18) (0.05) (1.76) (4.03) (8.7) (1.62)

7-day 61 13 7 53 7 45 53 42 47 57 60(1.3) (2.05) (0.95) (0.56) (0.43) (1.3) (1.76) (1.3) (0.73) (1.26) (0.39)

14-day 62 14 8 54 9 45 54 42 48 58 64(1.26) (0.73) (0.26) (0.5) (0.29) (1.26) (1.46) (1.83) (2.19) (2.19) (0.12)

28-day 64 18 9 56 10 46 54 43 50 60 66(0.39) (0.05) (0.54) (1.3) (0.48 (0.4) (0.4) (0.73) (0.58) (1.45) (0.5)

Note: The standard deviation of the corresponding compressive strength is shown in the brackets ().

Table 8The comparison of increase in the compressive strength (%) among 3-, 7-, 14- and 28-days.

Test age (day) M1 M2 M3 M4 M5 M6 M7 M8 M9 M10 M11

3 94 72 78 95 60 98 87 95 94 92 827 95 72 78 95 70 98 98 98 94 95 91

14 97 78 89 96 90 98 100 98 96 97 97

50

55

60

65

70

75

80

85

90

95

100

1470

% A

chie

ved

in c

ompr

essi

ve s

tren

gth

Test age (days)

M1 M2 M3 M4

M5 M6 M7 M8

M9 M10 M11

Fig. 10. The increase in compressive strength (%).

Fig. 11. The effect of GGBS on the compressive strength of mortar mixed with POFAand FA at the age of 28-day.

A. Islam et al. / Materials and Design 56 (2014) 833–841 839

beyond 30% increase in POFA content decreases the strength. Saf-iuddin et al. [42] reported that a POFA content higher than 40%may adversely affect the properties and durability of concretewhich was reflected in the geopolymer mortar as well. The mix

M11 with 30% of POFA and 70% of GGBS produced the higheststrength of 66 MPa. The coarser particles of the POFA with cohesivecharacteristic could not be mixed properly and hence the strengthdevelopment was poor. The ground POFA with high fineness(d50 = 10.1 lm) is a reactive pozzolanic material and can be usedto produce high-strength concrete. The suggested level of POFAcontent as cement replacement in normal concrete was 20% to pro-duce high-strength concrete [27,30].

3.5. Density reduction

Fig. 13 shows the change in density of mortar specimens left inthe laboratory at temperature of 26–29 �C and relative humidity of75–80%. The ODD of the mortar decreased slightly in the order ofabout 2% in the first few weeks but remained almost constantthereafter. Similar finding was reported by Wallah and Rangan[37].

3.6. Analysis of chemical composition

Khale and Chaudhary [6] reported in their review that certainsynthesis limits existed in the formation of strong products(Table 9) but the ratio changes while working with the waste.

The rate of polymerization is influenced by parameters such ascuring temperature, water content, alkali concentration, initial sol-ids content, silicate and aluminate ratio, pH and the type of activa-tors used. Table 10 and Fig. 14 represent the comparison of themajor oxide composition of the three materials i.e., GGBS, POFAand FA. The mixes M1 and M11 achieved higher strength and themix M3 achieved the lowest strength compared to other mixes;however the mix M11 produced slightly higher strength than themix M1. Table 10 and Fig. 14 show that the mix M1 contains morelime than the mix M11; nevertheless, the SiO2 content in the mixM11 is slightly higher than the M1 and the ratio of SiO2/Al2O3 ofM11 is 3.7 (Table 11) which compiled with Table 9. On the con-trary, this ratio (SiO2/Al2O3) for M1 is 2.37 and for M3 is 2.01.The lowest compressive strength of mix M3 that contains 100%FA might be attributed to the lowest SiO2/Al2O3 ratio of 2.01 asseen from Table 11. Lime (CaO) plays a very important role. It con-trols strength and soundness but any excess in the lime contentmakes the material unsound and causes expansion and disintegra-tion. Excessive quantity of lime (CaO) is the essence of the harden-ing mechanism of mortar [7]. It has been reported that theformation of Ca compounds in geopolymers is greatly dependent

Legend: S – Slag (GGBS), P – POFA, F – Fly ash and mix compositions

are shown in bracket in percentage (%)

0

10

20

30

40

50

60

70

M2 M3 M4 M5 M7 M8 M9 M10 M11

Com

pres

sive

str

engt

h(M

Pa)

Mix no.

M2 (P=100,S=0.0,F=0.0) M3 (P=0.0,S=0.0,F=100) M4 (P=50,S=50,F=0.0) M5 (P=50,S=0.0,F=50) M7 (P=25,S=50,F=25) M8 (P=60,S=40,F=0.0) M9 (P=30,S=40,F=30) M10 (P=40,S=60,F=0.0) M11 (P=30,S=70,F=0.0)

Fig. 12. The effect of POFA on the compressive strength of mortar mixed with GGBS and FA at the age of 28-day.

90

92

94

96

98

100

0 2 4 6 8 10

Rat

io o

f de

nsit

y (%

)

Test age (weeks)

Fig. 13. Reduction in density of mortar specimen.

Table 9Oxide-mole ratios of the reactant mixture [6,7].

Composition M2O/SiO2 SiO2/Al2O3 H2O/M2O M2O/Al2O3

Range 0.2–0.48 3.3–4.5 10–25 0.8–1.6

0

10

20

30

40

50

60

70

0

10

20

30

40

50

60

70

M1 M2 M3 M4 M5 M6 M7 M8 M9 M10 M11

Com

pres

sive

str

engt

h (M

Pa)

Che

mic

al c

ompo

siti

on(w

t %

)

Mixture no.

CaO SiO2 Al2O3

MgO Fe2O3 Na2O

K2O 28-d Comp. Strength

Fig. 14. The comparison between major chemical composition and compressivestrength of mortar.

840 A. Islam et al. / Materials and Design 56 (2014) 833–841

on the pH and Si/Al ratio [43]. The SiO2 content provides greaterstrength but at the same time it prolongs its setting time though

Table 10Major chemical composition of mortar and 28-day compressive strength.

Mix No. CaO SiO2 Al2O3 MgO

M1 45.83 32.52 13.71 3.27M2 4.34 63.41 5.55 3.74M3 5.31 54.72 27.28 1.10M4 25.09 47.97 9.63 3.51M5 4.83 59.07 16.42 2.42M6 25.57 43.62 20.50 2.19M7 25.33 45.79 15.06 2.85M8 20.94 51.05 8.81 3.55M9 21.23 48.447 15.333 2.76M10 29.23 44.88 10.45 3.46M11 33.38 41.79 11.26 3.41

Table 11The SiO2/Al2O3 ratio for the mixes.

Mix No. M1 M2 M3 M4 M5

SiO2/Al2O3 2.37 11.43 2.01 4.98 3.60

mix M2 contains the maximum percentages (%) of SiO2 amongall the mixes but the silicate and aluminate ratio is very high(11.4). Also M2 contains higher percentages of K2O and MgO whichare harmful ingredients in cement. If the amount of Na2O and K2Oexceeds 1%, it leads to the failure of concrete and if the content ofMgO exceeds 5%, it causes cracks in the hardened concrete.

Fe2O3 Na2O K2O 28-d Comp. strength (MPa)

0.76 0.25 0.48 644.19 0.16 6.33 185.15 0.43 1.00 92.48 0.21 3.41 564.67 0.30 3.67 102.96 0.34 0.74 462.72 0.27 2.07 542.82 0.20 3.99 433.11 0.28 2.39 502.13 0.21 2.82 601.79 0.22 2.24 66

M6 M7 M8 M9 M10 M11

2.13 3.04 5.79 3.16 4.29 3.71

A. Islam et al. / Materials and Design 56 (2014) 833–841 841

4. Conclusions

The following conclusions can be drawn from this experimentalstudy:

(1) The compressive strength of geopolymer mortar increases asthe GGBS content is increased up to 70%. Further increase inGGBS content did not produce desired effect.

(2) The addition of POFA up to 30% with GGBS produced thehighest strength and hence it is recommended for strengthbeyond 60 MPa.

(3) In most of the specimens 90% of compressive strength ofgeopolymer mortar was achieved at the age of 7 days.

(4) The coarser POFA produces cohesive mix and additionalwater has to be added in the mortar to achieve desiredworkability.

(5) The finer particles of GGBS produce dense mix and hence thedensity of mortar produced using GGBS resulted in about 8%of density increase.

(6) The use of locally available waste materials such as GGBS,POFA, FA and M-sand could be used for development of sus-tainable construction material.

Acknowledgement

The authors are grateful to University of Malaya for the financialsupport through the University of Malaya Research Project ‘‘RP018/2012B: Development of Geopolymer Concrete for StructuralApplication’’.

References

[1] Robert McCaffrey. Climate Change and the Cement Industry, Environmentaloverview. GCL Magazine, environmental special issue. London, UK: CementTrends; 2002.

[2] Yi H, Xu G, Cheng H, Wang J, Wan Y, Chen H. An overview of utilization of steelslag. Procedia Environ Sci 2012;16:791–801.

[3] Meyer C. The greening of the concrete industry. Cem Concr Compos2009;31:601–5.

[4] Kou S-C, Poon C-S. Effects of different kinds of recycled fine aggregate onproperties of rendering mortar. J Sust Cem-Based Mater 2013;2:43–57.

[5] Ernst Worrell, Lynn Price, Nathan Martin, Chris Hendriks, Leticia Ozawa Meida.Carbondioxide emissions from the global cement industry. Annu. Rev. EnergyEnviron., US: Lawrence Berkeley National Laboratory, Berkeley, California94720; 2001. p. 303–29.

[6] Khale D, Chaudhary R. Mechanism of geopolymerization and factorsinfluencing its development: a review. J Mater Sci 2007;42:729–46.

[7] Davidovits J. Geopolymer chemistry and applications. 3rded. France: Geopolymer Institute; 2011.

[8] Lloyd N, Rangan BV. Geopolymer concrete with fly ash. In: Zachar JaC P, Naik T,Ganjian G, editor. Second international conference on sustainable constructionmaterials and technologies. Ancona, Italy: UWM Center for By-ProductsUtilization; 2010. p. 1493–504.

[9] Bakharev T. Thermal behaviour of geopolymers prepared using class F fly ashand elevated temperature curing. Cem Concr Res 2006;36:1134–47.

[10] Bakharev T. Durability of geopolymer materials in sodium and magnesiumsulfate solutions. Cem Concr Res 2005;35:1233–46.

[11] Bakharev T. Resistance of geopolymer materials to acid attack. Cem Concr Res2005;35:658–70.

[12] Bakharev T. Geopolymeric materials prepared using Class F fly ash andelevated temperature curing. Cem Concr Res 2005;35:1224–32.

[13] Hardjito D, Wallah SE, Sumajouw DM, Rangan BV. On the development of flyash-based geopolymer concrete. ACI Mater J-Am Concr Inst 2004;101:467–72.

[14] BS 3892: Part 1. Pulverized-fuel ash. Specification for pulverized-fuel ash foruse with Portland cement. London: BSI; 1997.

[15] Kupaei RH, Alengaram UJ, Jumaat MZ, Nikraz H. Mix design for fly ash based oilpalm shell geopolymer lightweight concrete. Constr Build Mater2013;43:490–6.

[16] Li C, Sun H, Li L. A review: the comparison between alkali-activated slag(Si + Ca) and metakaolin (Si + Al) cements. Cem Concr Res 2010;40:1341–9.

[17] Dinakar P, Sethy KP, Sahoo UC. Design of self-compacting concrete withground granulated blast furnace slag. Mater Des 2013;43:161–9.

[18] Bakharev T, Sanjayan JG, Cheng YB. Resistance of alkali-activated slag concreteto alkali–aggregate reaction. Cem Concr Res 2001;31:331–4.

[19] Bakharev T, Sanjayan JG, Cheng YB. Resistance of alkali-activated slag concreteto carbonation. Cem Concr Res 2001;31:1277–83.

[20] MPOB (Malaysian Palm Oil Board). Overview of the Malaysian Oil PalmIndustry; 2012.

[21] Sata V, Jaturapitakkul C, Rattanashotinunt C. Compressive strength and heatevolution of concretes containing palm oil fuel ash. J Mater Civ Eng2010;22:1033–8.

[22] Alengaram UJ, Al-Muhit BA, Jumaat MZ. Utilization of oil palm kernel shell aslightweight aggregate in concrete – a review. Constr Build Mater2013;38:161–72.

[23] Jumaat MZ, Alengaram UJ, Mahmud H. Shear strength of oil palm shellfoamed concrete beams. Mater Des 2009;30:2227–36.

[24] Alengaram UJ, Mahmud H, Jumaat MZ. Enhancement and prediction ofmodulus of elasticity of palm kernel shell concrete. Mater Des 2011;32:2143–8.

[25] Alengaram UJ, Jumaat MZ, Mahmud H, Fayyadh MM. Shear behaviour ofreinforced palm kernel shell concrete beams. Constr Build Mater2011;25:2918–27.

[26] Shafigh P, Alengaram UJ, Mahmud H, Jumaat MZ. Engineering properties of oilpalm shell lightweight concrete containing fly ash. Mater Des2013;49:613–21.

[27] Sata Vanchai, Jaturapitakkul Chai, Kiattikomol Kraiwood. Utilization of palmoil fuel ash in high-strength concrete. J Mater Civ Eng 2004;16:623–8.

[28] Tangchirapat Weerachart, Jaturapitakkul Chai, Chindaprasirt Prinya. Use ofpalm oil fuel ash as a supplementary cementitious material for producinghigh-strength concrete. Constr Build Mater 2009;23:2641–6.

[29] Tangchirapat Weerachart, Khamklai Supat, Jaturapitakkul Chai. Use of groundpalm oil fuel ash to improve strength, sulfate resistance, and waterpermeability of concrete containing high amount of recycled concreteaggregates. Mater Des 2012;41:150–7.

[30] Kroehong Wunchock, Sinsiri Theerawat, Jaturapitakkul Chai, ChindaprasirtPrinya. Effect of palm oil fuel ash fineness on the microstructure of blendedcement paste. Constr Build Mater 2011;25:4095–104.

[31] Short Andrew, Kinniburgh William. Lightweight concrete. 3rd ed. UK: AppliedScience Publishers; 1978.

[32] Ashraf MA, Maah MJ, Yusoff I, Wajid A, Mahmood K. Sand mining effects,causes and concerns: a case study from Bestari Jaya, Selangor, PeninsularMalaysia. Sci Res Essays 2011;6:1216–31.

[33] Sreenivasa G. Use of manufactured sand in concrete and construction analternate to river sand. India: NBMCW; 2012.

[34] Celik T, Marar K. Effects of crushed stone dust on some properties of concrete.Cem Concr Res 1996;26:1121–30.

[35] BS 146: Specification for Portland blastfurnace cements. 9th ed. London: BSI;1996.

[36] BS 6699: Ground granulated blastfurnace slag for use with Portland cement.2nd ed. London: BSI; 1992.

[37] Wallah SE, Rangan BV. Low-calcium fly ash-based geopolymer concrete: long-term properties. Perth, Australia: Faculty of Engineering, Curtin University;2006.

[38] Hardjito D, Rangan BV. Development and properties of low-calcium fly ash-based geopolymer concrete. Perth, Australia: Faculty of Engineering, CurtinUniversity; 2005.

[39] Wongpa J, Kiattikomol K, Jaturapitakkul C, Chindaprasirt P. Compressivestrength, modulus of elasticity, and water permeability of inorganic polymerconcrete. Mater Des 2010;31:4748–54.

[40] Ganapati Naidu P, Prasad ASSN, Adiseshu S, Satayanarayana PVV. A study onstrength properties of geopolymer concrete with addition of G.G.B.S. Int J EngRes Develop 2012;2:19–28.

[41] Shafigh P, Jumaat MZ, Mahmud H, Alengaram UJ. Oil palm shell lightweightconcrete containing high volume ground granulated blast furnace slag. ConstrBuild Mater 2013;40:231–8.

[42] Safiuddin M, Abdus Salam M, Jumaat MZ. Utilization of palm oil fuel ash inconcrete: a review. J Civ Eng Manage 2011;17:234–47.

[43] Yip CK, Lukey GC, van Deventer JSJ. The coexistence of geopolymeric gel andcalcium silicate hydrate at the early stage of alkaline activation. Cem Concr Res2005;35:1688–97.