dt

jun3007ivers

3007

Article history:Received 28 November 2014Revised 5 March 2015Accepted 7 March 2015Available online 9 March 2015

This study aims to investigate durability and microstructure of uidized bed y ash and metakaolin based

of less than 20 lm and low bulk density (0.540.86 g/cm3), highsurface area (300500 m2/kg) and light texture with spherical inshape and consist of solid spheres, cenospheres, irregular-shapeddebris and porous unburnt carbon), blast furnace slag (producedwhen iron ore is reduced by coke at about 13501550 C in a blast

2

t of the cement

fforts [29f potentia

cations of these by-product materials, which has gained incattention due to the energy conservation, economic and enmental considerations.

By-products of industry are some of the most complex andabundant of anthropogenic materials. They cause water and soilpollution, disrupt ecological cycles and environmental hazards[10]. The current worldwide annual production of y ash, one ofsuch by-products, is estimated around 750 million tones [11]. Atpresent, only a minor part of this material is utilized (2030%) on

Corresponding author at: Faculty of Materials Science and Chemistry, ChinaUniversity of Geosciences, Wuhan 430074, China.

E-mail address: dp19851128@sina.com (P. Duan).

Materials and Design 74 (2015) 125137

Contents lists availab

Materials an

elsing industrial by-products is one of existing strategies.Geopolymer is such an emerging alternative binder, which is

prepared using by-product materials such as y ash (by-productof coal combustion in thermal power plants with an average size

emission to reduce the environmental impacproduction.

During the last decade, increased research ebeen directed to this area due to the wide range ohttp://dx.doi.org/10.1016/j.matdes.2015.03.0090261-3069/Crown Copyright 2015 Published by Elsevier Ltd. All rights reserved.] havel appli-reasingviron-1. Introduction

Cement production is associated with the emission of consider-able amount of greenhouse gas [1]. The amount of carbon dioxidereleased in the manufacture of cement is about one ton for one toncement clinker. Therefore, it is considered vital to search for alter-native low CO2 emission binders for concrete in order to reduce itscarbon footprint and the development of alternative binders utiliz-

furnace. It normally contains more than 95% of glass. Generally,they are ground to ne powder, called ground granulated blastfurnace slag), metakaolin (produced by calcining kaolin at650800 C. The main components are amorphous Al2O3 andSiO2 with high pozzolanic activity. Besides the lling effect, meta-kaolin reacts with calcium hydroxide, which is one of the hydrationproducts of Portland cement, to form calcium silicate hydrate gels)or a combination of them instead of cement and results in less COKeywords:GeopolymerFly ashMetakaolinDurabilityOPCgeopolymer exposed to elevated temperatures and acid attack. Geopolymer specimens were prepared bycombination of y ash and metakaolin activated by sodium silicate and sodium hydroxide solutions andwere cured in microwave radiation environment plus a heat curing period. Compressive strength andseveral key durability parameters for geopolymer and ordinary Portland cement (OPC) were assessedand compared. Microstructure formation and development was characterized in terms of morphologyand pore structure as well as simulation.The experimental results reveal a dense microstructure of geopolymer compared to OPC. In terms of

resistance to the acidic solution and elevated temperatures, geopolymer is superior to OPC as indicatedby the relatively lower strength loss and lower mass change. Compressive strength shows a dramaticdrop in OPC while geopolymer shows a strength increase after 400 C. The mass loss curves of geopoly-mer are similar to OPC, but it shows relatively lower mass loss compared to OPC. The result of saturatedwater absorption after 28 days curing indicates less water absorption in geopolymer before and afterthermal and acid exposure. Durability of geopolymer is demonstrated by monitoring the pore structure.

Crown Copyright 2015 Published by Elsevier Ltd. All rights reserved.a r t i c l e i n f o a b s t r a c tAn investigation of the microstructure any ashmetakaolin geopolymer after hea

Ping Duan a,b,c,, Chunjie Yan a,b,c, Wei Zhou a,b, Wena Faculty of Materials Science and Chemistry, China University of Geosciences, Wuhan 4b Engineering Research Center of Nano-Geomaterials of Ministry of Education, China Unc Zhejiang Research Institute, China University of Geosciences, Hangzhou 311305, ChinadCenter of Materials Research and Analysis, Wuhan University of Technology, Wuhan 4

journal homepage: www.durability of a uidized bedand acid exposure

Luo a,b, Chunhua Shen d

4, Chinaity of Geosciences, Wuhan 430074, China

0, China

le at ScienceDirect

d Design

evier .com/locate /matdes

nd Dworldwide basis, while the rest is still disposed of in landlls, thuscontributing to the pollution of soil, water and air.

The environmental impact of coal y ash is well known for itsmassive generation, large usage of land for disposal and shortand long term impact on surrounding areas. The principal environ-mental concern stems from the possible leaching of heavy metalsand organic compounds and their migration into ground water ornearby surface water. In addition, y ash could also affect humanhealth through direct inhalation or ingestion of airborne or settledash. The unproductive use of land and its maintenance results inlong-term nancial burden [12].

Aggressive efforts have been undertaken recently to recycle yash [1317] in concrete production as mineral admixture, soilamendment, zeolite synthesis, and as ller in polymers. However,these applications are not sufcient for complete utilization of yash, thereby it is imperative to develop new recycling techniquesfor y ash.

The concept of sustainable solutions for y ash is closely linkedwith technologies with aligned vision for environment, economyand societal goals. Fly ash was widely used as the source materialto manufacture geopolymer products owing to its aluminosilicatecomposition, ne size, signicant amount of glassy content andavailability across the world [14,1820]. This is one of the existingimportant strategies and it is believed to be the sustainable solu-tion for utilization of y ash. It will consume large part of y ashand relieve the pollution of soil, water and air.

The term geopolymer was introduced in 1970 by Davidovits,who made a signicant breakthrough in understanding anddevelopment of binders from metakaolin and alkaline metalsolutions. Geopolymer was originally applied to three-dimensionalaluminosilicate materials formed by condensation of a solid alumi-nosilicate source such as dehydroxylated kaolinite (metakaolin)with an alkali silicate solution under highly alkaline conditions [21].

This new material was likely to have enormous potential tobecome an alternative to Portland cement and it is receivingincreased attention due to the need of new binders with enhanceddurability performance [2228].

Although different source materials can be used to preparegeopolymer binders, y ash, which provides the greatest opportu-nity for commercial utilization and has the potential to reduce thecarbon footprint, has been extensively used and found to be themost practical source material suitable for concrete applicationsdue to the plentiful worldwide raw material supply, which isderived from coal-red electricity generation [29].

Excellent properties of y ash-based geopolymer concrete havebeen reported in the last decades [3038]. Fly ash-based geopoly-mer concrete has properties favourable for its potential use as acementitious material due to excellent durability aspects. Someauthors [3943] have reported good engineering properties ofgeopolymer concrete that were favourable for its use as a construc-tion material.

Geopolymer can be composed from metakaolin or wastes, suchas y ash, slag, and tailing [4448]. Geopolymer can be used asbuilding materials and llers due to their ameproof characteris-tics [4951]. In addition, geopolymer has several other advantagessuch as high strength, acid/alkaline resistance and heat resistance[44].

Having more outstanding mechanical properties and envi-ronmental friendliness, geopolymer, is considered to be a newcementing material with widely potential application value in con-struction. With the aim of reducing consumption of non-renewableraw materials whilst increasing the use of industrial by-products(residue), research has recently focused on the alkali activation of

126 P. Duan et al. /Materials ametakaolin and y ash.Some authors [5257] explored some basic aspects of metakao-

lin or y ash-based geopolymer activated using sodium alkali forhigh-temperature applications. They estimated that the structurewas stable enough to resist to high temperatures.

On balance, ash-based geopolymer was prepared based on aprecursor derived from y ash, generally consisting of sphericalparticles, glassy or amorphous as well as crystalline phases.

Recently, increased utilization of uidized bed technology led tothe production of a great amount of uidized bed y ash includinglittle glassy spherical particles, which forms by burning coalgangue in a low temperature. Circulation uidized bed combustionis an advanced, clean and reliable coal ring technology for powergeneration. There are many unreacted CaO, desulphurized prod-ucts CaSO4 and a little CaCO3 remaining in the uidized bed y ash.

Coal y ash is typically found in the form of coarse bottom ashand ne y ash, which represent 515 and 8595 wt% of the totalash, respectively. Bottom ash refers to the ash that falls downthrough the airow to the bottom of the boiler and is mechanicallyremoved. The term coal y ash is often used to refer to ne y ash,particles of which are captured from ue gas and collected by elec-trostatic or mechanical precipitation.

Bottom ash is a coarse, granular, incombustible by-product thatis collected from the bottom of furnaces that burn coal for the gen-eration of steam, the production of electric power, or both. Bottomash is coarser than y ash, with grain sizes spanning from ne sandto ne gravel.

In general, uidized bed y ash differs from coal y ash in termsof particle shape, chemical composition and amorphous phase con-tent. Fluidized bed y ash particles are approximately 1300 lm insize, with irregular shape, while coal y ash particles are normallyslightly ner at approximately 1200 lm in size and the content ofamorphous phase is usually higher than that of uidized bed yash. Coal y ash is widely used as pozzolanic material for partialreplacement of Portland cement due to its spherical shape and highreactivity. Comparing with coal y ash, uidized bed y ash hashigher contents of lime (CaO), gypsum (CaSO4) and crystallinephase thus its usage as pozzolanic material is limited. The uniquethermal history, featuring low combustion temperatures of800950 C, makes uidized bed y ash differ greatly in physicaland chemical characteristics from coal y ash, whose typical ringtemperatures are 12001400 C.

It is well acknowledged that compositions and structure ofprecursors have signicant effects on performance of geopolymer.Previous research efforts about the relevant topic mainly focusedon cement paste [5870], or preparation and properties of geopoly-mer but below 1000 C, and acid attack was not mentioned[7176]. Existing references on the utilization of uidized bed yash to prepare geopolymer and the durability aspects investigationare scant, and the ash used belonged to high calcium y ash, therehave been very few published references on uidized bed y ashbased geopolymer exposed to elevated temperature at 1000 Cfrom ambient temperature and acid attack.

In the present study, geopolymer specimens were prepared bycombination of uidized bed y ash and metakaolin activated byalkaline activator and were cured in microwave radiation environ-ment. Compressive strength was investigated and several keydurability parameters for geopolymer and ordinary Portlandcement (OPC) were assessed and compared. Microstructure forma-tion and development was characterized in terms of morphology,pore structure and simulation.

2. Experimental

2.1. Materials

esign 74 (2015) 125137Fluidized bed y ash was provided by Shenhua Junggar EnergyCorporation in Junggar, Inner Mongolia, China. Metakaolin wasobtained from Yunnan, China. The chemical analysis of those raw

C6

nd DTable 1Chemical compositions of raw materials by XRF analysis (mass, %).

SiO2 Al2O3 Fe2O3 MgO

Cement 21.35 7.67 3.31 3.08Fly ash 29.47 51.72 2.25 0.15Metakaolin 53.32 42.09 2.33 0.21

a LOI: mass loss of dried sample after calcined between 1000 and 1100 C.

P. Duan et al. /Materials amaterials mentioned above was listed in Table 1 and their micro-graphs were shown in Figs. 1 and 2, respectively. Alkali activatorwas a combination of sodium silicate and sodium hydroxide(99.2% NaOH) in analytical reagent degree. The liquid portions inthe mixture were 10 M sodium hydroxide (NaOH) and sodiumsilicate (Na2SiO3) with 14.51% Na2O, 33.39% SiO2, and 48.53% H2O.

Portland cement (CEM I 42.5) (relative density 3100 kg/m3,specic surface area 369.6 m2/kg) was used as binder and wascompared to geopolymer. The properties of cement were shownin Table 2.

2.2. Test procedure

Geopolymer specimens were synthesized by alkaline-activationof combination of uidized bed y ash and metakaolin with massratios of 1.0 in alkali silicate solutions (MR = 1.6, MR meansmodulus of alkaline activator). The liquid/solid (L/S) mass ratiowas kept constant and equaled to 0.4 (the liquid consists of waterin alkaline solutions and the extrawater, the solidmaterials consistsof y ash, metakaolin or a combination of them), which dependedon an acceptable workability for each paste sample. Fresh geopoly-mer pastes and cement paste were cast in triplet steel cubes moldswith size of 40 40 40 mm3 and 20 20 20 mm3, respectively,and vibrated to remove entrained air bubbles. The molds were thensealed with polyethylene lm and set into a standard curing box inmicrowave radiation environment for 5 min. After initial curing at40 C for 1 day, the samples were released from themolds andweresubjected to further curing at 40 C for acquired ages.

Fig. 1. The micrographs (SEMaO Na2O K2O MnO TiO2 LOIa

2.60 0.35 0.39 0.05 0.25 0.955.21 0.05 0.35 0.03 1.83 8.580.09 0.49 0.64 0.02 0.63 0.08

esign 74 (2015) 125137 127All samples were calcined from ambient temperature to 1000 Cin an electric mufe furnace with a heating rate of 2 C/min andthen cooled to room temperature. Sample characterization wasconducted on geopolymer and OPC specimens to evaluate theeffects of the elevated temperature exposure on properties ofthem.

2.2.1. Compressive strengthAfter casting, the molds containing the specimens were covered

with a plastic sheet and stored in the laboratory environment(20 C and 90% RH) for 24 h. The paste specimens with size of40 mm 40 mm 40 mm were then removed from the molds.The compressive strength was determined using a universaltesting device with a loading capacity of 3000 kN. The loading ratesapplied during the compression tests was 0.6 MPa/s. For eachmixture, the compressive strength was measured at the age ofacquired days.

2.2.2. Drying shrinkagePaste specimens of the size 20 20 200 mm3 were prepared

for the drying shrinkage test. The specimens were removed fromthe molds after being cured for 24 h. The prism specimens wereinstalled onto the setup for the length change tests and cured ina room with constant temperature and relative humidity(20 3 C, RH = 90 5%). Length changes of the prism before andafter high temperature exposure were recorded by reading the dialgauge regularly [77].

) of uidized bed y ash.

nd D128 P. Duan et al. /Materials a2.2.3. Acid attackCement and geopolymer pastes were placed in a cylinder mold

with size of 100 mm 50 mm. Test pieces were kept in a moistroom. After 24 2 h curing, the test pieces were removed from themolds. The test pieces were immerged in the sulfuric acid (2%) plushydrochloric acid (2%) solutions for 28 days. After the immersiontest, compressive strength loss compared to the original strengthwas measured and appearance of the exposure surfaces wasobserved to evaluate the resistance to acid attack.

2.2.4. Water sorptivityWater sorptivity test was carried out on specimens with size of

100 mm 50 mm for both cement and geopolymer paste. Thespecimens were oven dried for 3 days at 50 C followed by 28 dayssealed storage. After this curing stage, the specimens wereimmerged into deionized water for several days and the masschanges before and after immerged into water were measured toevaluate the resistance to water permeability.

2.2.5. Mercury intrusion porosimetryThe porosity and pore size distribution of paste were measured

by mercury intrusion porosimetry (MIP, AutoPore IV 9500 type)with a maximum pressure of 207 MPa. The contact angle was140 and the measurable pore size ranged from about 6 nm to360 lm. The samples in the shape of pellets of 5 5 5 mm3 insize for pore structure testing were separated from the crushedspecimens. The samples were immerged in ethanol to avoid rehy-dration immediately after being crushed and dried at about 105 Cfor 24 h before pore structure test.

Fig. 2. The micrographs (

Table 2The properties of cement.

Compressivestrength (MPa)

Flexuralstrength (MPa)

Initial setting time(min)

Finalsettingtime (min)

3 days 28 days 3 days 28 days

27.3 47.5 6.3 8.7 132 1872.2.6. Micrographs testThe morphological changes that occurred during thermal expo-

sure were obtained by means of JSM-5610LV scanning electronicmicroscopy (SEM). The specimens for morphology observationwere cut into prisms of about 8 8 2 mm3.SEM) of metakaolin.

esign 74 (2015) 1251373. Results and discussion

3.1. Temperature prole

It is believed that geopolymer possesses good thermal stabilityproperty due to re resistance stems from ceramic-like properties.To evaluate thermal resistance of the investigated samples and theinuence of high-temperatures on microstructure, dimension sta-bility and mechanical properties, all samples were calcined fromroom temperature to 1000 C in an electric furnace with a heatingrate of 2 C/min and then cooled naturally to room temperature.Sample characterization was conducted on untreated and calcinedsamples to evaluate the effects of the elevated temperature expo-sure on OPC and geopolymer with size of 40 40 40 mm3 and20 20 20 mm3, respectively.

The temperature of the air inside the furnace was measured byan in-built thermocouple. The temperature at the centre of the spe-cimen was measured by a K-type thermocouple inserted in thespecimens during casting. The thermocouples were connected toelectronic data loggers that recorded the measured temperatures.

The result provided in Fig. 3 reveals the temperature evolutionat different distances from the heated surface of a cubic OPC and yashmetakaolin geopolymer to the center of paste specimensduring heating process. Maximum thermal gradient between theexposure surface and center can be observed to occur between180 and 270 min. This clearly shows that the distribution of tem-perature is not uniform and a thermal gradient exists. This thermalgradient is likely to have adverse effects on the integrity of the OPCand geopolymer paste at elevated temperatures.

Heat resistance is the ability of materials withstanding a hightemperature treatment while not destroying during the treatment.

On the other hand, the compressive strength of OPC specimensdrops considerably between temperatures of 400 C and 600 C.This strength deterioration of OPC is attributed to the Ca(OH)2decomposition that occurs at about 400 C. In geopolymer, alumi-nosilicate gel is the major binding phase that provides interparticlebonding, which in turn enhances the macroscopic strength [7880]and takes over the strength gain behavior of geopolymer paste.

The compressive strength of geopolymer changes dramaticallyat early ages. The compressive strength of OPC and geopolymerpaste specimens at 3 days of curing were also tested and comparedbefore and after elevated temperatures exposure and the resultshas been summarized in Fig. 5.

The results provided in Fig. 5 reveal that the OPC suffers strengthloss after high temperature exposure, however, the y ashmetakaolin geopolymer gains strength around 400 C. Compressivestrength development shows a dramatic drop of nearly 100% inOPC paste specimen after 600 C, and a nal strength of 25 MPafor geopolymer can be observed after the 1000 C temperature

3.3. Thermal stability

P. Duan et al. /Materials and Design 74 (2015) 125137 129Water accumulated inside the geopolymer structure is the maincontributor to re resistance and it would release during theheating process to reduce the heat temperature and also to forma porous microstructure. Therefore, geopolymer shows lower tem-perature at the same distance to the surface compared to OPC.

3.2. Compressive strength at elevated temperatures

The compressive strength of OPC and geopolymer paste speci-mens at 28 days of curing were tested and compared before andafter elevated temperatures exposure and the results has beensummarized in Fig. 4.

The results provided in Fig. 4 show that the OPC suffers strengthloss after high temperature exposure. However, the y ashmetakaolin geopolymer gains strength after the same high temper-ature exposure around 400 C. Compressive strength developmentindicates a dramatic drop of nearly 100% in OPC paste specimenwhile y ashmetakaolin geopolymer shows a strength increaseof about 3% after 400 C, attaining a peak strength of 72 MPa at400 C. Subsequently, this strength can be observed to deteriorategradually. However, a nal strength of 46 MPa is recorded after

Fig. 3. Temperature evolution of geopolymer and OPC at various distances to thesurface.the 1000 C temperature exposure, which is about a 25% decreaseover the strength of the reference specimen at ambient temperature.

0

10

20

30

40

50

60

70

80

20 100 200 300 400 500 600 700 800 900 1000

Temperature (C)

Com

pres

sive

stre

ngth

(MPa

)

Portland cement

Geopolymer

Fig. 4. Compressive strength of geopolymer and OPC pastes at 28 days of curing atvarious temperatures.Thermal stability of the selected geopolymer specimens wasalso investigated in terms of dimension evolution after exposedto elevated temperature (1000 C) for 2 h. Fig. 6 presents the pho-tographs of geopolymer before and after the thermal exposures. Itcan be observed that geopolymer gets a certain degree of volumeshrinkage after the thermal exposures.

Actually, OPC exhibits a higher rate of shrinkage compared togeopolymer [8284], which can cause severe defects when it ispractically applied.

The increase in strength with increasing temperature up to400 C may also be attributed to the hardening of geopolymerexposure.As mentioned above, water accumulated inside the geopolymer

structure is the main contributor to heat resistance and it wouldrelease during the heating to reduce the heat temperature and alsoto form a porous microstructure. The water releases duringgeopolymer formation is expelled from the geopolymer matrixduring heating process, causing discontinuous nano pores through-out the matrix, which improves the strength of the geopolymer[81] before 400 C. On the other hand, the generation of bubbleand foam-like structure subjects to high temperature makes thegeopolymeric system porous and it reduces the density, whichshows a negative effect on the compressive strength after 400 C.Fig. 5. Compressive strength of geopolymer and OPC pastes at 3 days of curing atvarious temperatures.

According to previous published paper [87], the drying shrink-age of cement paste may be due to a higher volume of mesopores,which causes a higher capillary stress by the water meniscusdeveloped in the capillary pores of the paste, resulting in a higher

OPC OPC Geo Geo OPC OPC Geo Geo

Fig. 6. Dimension evolution of OPC and geopolymer after the thermal exposure.

130 P. Duan et al. /Materials and Design 74 (2015) 125137paste caused by drying and further hydration, as indicated by Xuet al. [85], who observed an increase in strength when exposuretemperature increased to 250 C. In our study, the increase in thestrength was recorded up to a temperature of 400 C.

In the beginning of hydration from the rst watercement (ory ash and metakaolin) contact to the initial setting, the capillarystress can be negligible, and the autogenous shrinkage is mainlyrelated to the Le Chateliers contraction and is thus identical tothe chemical shrinkage [86]. Considering the aforementionedresults, it can be stated that the autogenous shrinkage of geopoly-mer may be due to self-desiccation in the hardened state ratherthan volume contraction by chemical shrinkage in fresh state.

3.4. Thermal shrinkage

Thermal shrinkage results for the OPC and geopolymer sampleshave been illustrated in Fig. 7. All samples were tested afterpre-drying at 105 C for 24 h to remove free water and to avoidexcessive shrinkage in the initial stage of the measurement.

As is shown in Fig. 7, all samples exhibit shrinkage up to 1000 Cdue to dehydroxylation of chemically bound water. The shrinkageis particularly evident for cement sample where dehydration of theCASH phase and CSH phase is believed to dominate thedilatometry prole, resulting in loss of contact with the pushrodat about 1000 C. The shrinkage for geopolymer may be due tosintering and further geopolymerization above 600 C.

Polynomial trend lines can be used to show the trend ofshrinkage evolution in geopolymer specimens with increasingtemperature. A polynomial relationship also exists betweenshrinkage evolution and the elevated temperature for cementspecimens. The tting coefcients (R2) are all excess 0.99 forgeopolymer and OPC paste. Obviously, cement exhibits evidentand higher shrinkage values compared to geopolymer pastes.Fig. 7. Thermal shrinkage of Portland cement and geopolymer paste with elevatedtemperatures.level of drying shrinkage [88].

3.5. Mass loss at elevated temperatures

The mass losses of the OPC and geopolymer samples due toelevated temperature were determined from the mass changebefore and after heat exposure. The average values of mass lossafter exposure to different temperatures have been plotted inFig. 8. An increasing trend of the mass loss in OPC up to 600 Ccan be observed. However, the exact values of the mass loss inthe OPC could not be determined after 600 C because of spallingattributed to the Ca(OH)2 decomposition that occurs at this tem-perature. The geopolymer specimen shows similar mass loss curvesas OPC before 500 C but with relatively lower mass loss. It can beseen from Fig. 8 that most of the mass loss takes place at around500 C for geopolymer. The rate of mass loss reduces and becomestable in geopolymer after this temperature while OPC continuesat a similar rate until 600 C. Themass loss of the geopolymer speci-mens at 1000 C is 3.39% in total compared to that at ambienttemperature.

3.6. Electron microscopy after thermal exposure

The morphological changes of OPC and geopolymer during thethermal exposure were illustrated in the micrographs of Fig. 9.

4.5

5.00.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

0 200 400 600 800 1000

Temperature (C)

Mas

s los

s (%

)

Portland cement

Geopolymer

Fig. 8. Changes in mass loss of the OPC and geopolymer at elevated temperatures.

Ond DOPC 20C

P. Duan et al. /Materials aThe geopolymer microstructures become denser with theincrease of temperature up to 400 C. This change has occurred inthemicrostructure because of sintering and further geopolymeriza-tion of y ash and metakaolin with the increasing temperature.As mentioned above, an increasing trend of the mass loss in theOPC up to 600 C can be observed because of spalling, and the

OPC 200 C O

GEO 20 C G

GEO 200 C G

OPC 600 C

Fig. 9. The morphological changes of geopolPC 100C

esign 74 (2015) 125137 131corresponding microstructure at 600 C becomes looser withmicrocracking. However, the microstructure of geopolymerremains to be stable after exposure to high temperatures and therate of mass loss reduces and becomes stable nally.

Geopolymer possesses notable denser microstructure than OPC.The Si/Na atomic ratio in an alkaline silicate solution affects the

PC 400 C

EO 100 C

EO 400 C

GEO 600 C

ymer and OPC at various temperatures.

polymerization degree of the dissolved species [8990]. Thealkaline aluminosilicate system is symbolized by NASH.The formation process of these systems is a polycondensation. Theamorphous NASH gel in geopolymer becomes more compact.

3.7. Resistance to acid attack

Geopolymer and OPC specimens were immerged in the sulfuricacid (2%) plus hydrochloric acid (2%) for 28 days to assess the resis-tance to acid attack. Fig. 10 shows the photographs of geopolymerbefore and after the acid exposure.

Geopolymer specimens show very small change in appearanceafter 28 days of immersion in the acidic solutions. Some softeningof the surface cover and insignicant change of the color can benoticed in the geopolymer specimens after exposure to the acid solu-tion, and there is only 0.7% mass change in geopolymer specimens.

Surfaces of OPC specimens exposed to the acidic solutions indi-cate severe deterioration compared to geopolymer. After 10 days,the surface layer of the samples is converted to some reactionproducts and is corroded to a depth of 4 mm, however, the surfacelayer of geopolymer samples deteriorates to a depth of 2 mm after

severely deteriorated in the rst 28-day exposure. An exponent

Signicant uctuations of strength measured in OPC specimensare possibly connected to the decomposition of hydration productsand migration of alkalis from the specimens into solution imposedby the acid attack. Changes of strength measured in geopolymer

Fig. 11. Compressiver strength evolution of the geopolymer and OPC specimensexposed to sulfuric acid (2%) plus hydrochloric acid (2%) solutions.

132 P. Duan et al. /Materials and Design 74 (2015) 125137relationship is also found between compressive strength and theexposure time of OPC.28 days. After 28 days, OPC samples are severely deteriorated andexhibit severely deteriorated surface layer with 3.3% mass change.

3.8. Compressive strength exposed to acid

The results provided in Fig. 11 uncover the evolution of thecompressive strength of the samples exposed to the acidic solu-tions. The geopolymer samples activated by sodium hydroxideand sodium silicate solution perform well with about 10.4%strength decline in the 28-day exposure, which continues overthe next 28 days, reaching 22.2% after the test. Exponent trendlines can be used to show the trend of strength evolution ingeopolymer specimens because signicant uctuations of strengthare not observed for geopolymer in the acidic solutions.

OPC paste sample shows 34.4% strength loss after 28-day expo-sure, and about 57.8% after the next 28 days. The OPC samples areFig. 10. Photographs of geopolymer and OPC before and after the acid exposures: (a) geoacid exposures; (d) OPC after acid exposures.specimens are possibly connected to the breakdown of somegeopolymer alkali components and migration of alkalis from thespecimens into acidic solutions.

3.9. Water absorption

Sorptivity tests were conducted to determinemass change in thewater absorption process of cement paste and geopolymer. Theresults of saturated water absorption on OPC, metakaolin and yash geopolymer specimens after several days curing have been pre-sented in Fig. 12. It indicates less water absorption in geopolymercompared to cement paste. Minor difference between metakaolinand y ash geoplymer can be observed. Based on the pore structureresults mentioned below, the comparison among metakaolingeopolymer, y ash geopolymer and OPC paste shows that the totalporosity (cumulative intrusion volume) is decreasing for geopoly-mer pastes, which indicates that geopolymer possesses densermicrostructure. It also indicates that geopolymer could withstandoutside medium ingress including water and acid, which relatespolymer before acid exposures; (b) geopolymer after acid exposures; (c) OPC before

Fig. 14. Change of mass in water absorption test of geopolymer and cementsamples after acid exposure.

0.03

0.04

0.05

0.06

0.07

0.08

0.09

mul

ativ

e in

trusi

on (m

L/g)

Cement paste

MetakaolinGeopolymerFly ashGeopolymer

(a)

nd DFig. 12. Mass change of geopolymer and cement samples in water absorption test.

P. Duan et al. /Materials ato better permeability resistance, and the results are consistentwith the water absorption test result. It can be concluded that alkaliactivated condition is optimum for better pore structure modica-tion. The strength is also higher at this optimal condition.

As the compressive strength of OPC specimens drops consider-ably after 500 C (see Figs. 4 and 5) because of spalling, waterabsorption test after thermal exposure was carried out for geopoly-mer and OPC samples exposed to 500 C.

The results of saturated water absorption on OPC, metakaolinand y ash geopolymer specimens exposed to 500 C have beenpresented in Fig. 13. It indicates markedly less water absorptionin geopolymer compared to cement paste. Minor differencebetween metakaolin and y ash geoplymer can be observed, andwater absorption of samples is relatively higher when exposed toelevated temperatures.

Water absorption test after acid exposure was also made forgeopolymer and OPC samples exposed to acidic solutions for28 days. The results of saturated water absorption are presentedin Fig. 14. Similarly, it shows markedly less water absorption ingeopolymer compared to cement paste. Minor difference betweenmetakaolin and y ash geoplymer can be observed, and waterabsorption is relatively higher than samples not exposed toacidic solutions but lower than samples exposed to elevated tem-peratures. These results are consistent with the compressivestrength development results (see Figs. 4, 5 and 11), which uncoverthat better permeability resistance is closely related to the compres-sive strength.

3.10. Pore structure after acid exposure

Mercury intrusion porosimetry (MIP) has been widely used todetermine pore structure and pore size distribution of porous

Fig. 13. Change of mass in water absorption test of geopolymer and cementsamples after thermal exposure.esign 74 (2015) 125137 133materials. As pointed out by Diamond [91], MIP is the onlyavailable procedure that can cover nearly the whole range of sizesthat must be tallied.

The porosity and pore size distribution measurements per-formed by MIP of metakaolin and y ash geopolymer samples andOPC paste are shown in Fig. 15. The results at 28 days are plottedto identify how pore size distribution varies with ages.

0.00

0.01

0.02

Pore diameter (nm)

Cu

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0.16

0.18

1 10 100 1000 10000 100000 1000000

Pore diameter (nm)1 10 100 1000 10000 100000 1000000

dV/d

logD

(mL/

g)

Cement paste

MetakaolinGeopolymerFly ashGeopolymer

(b)

Fig. 15. The porosity and pore size distribution of geopolymer and OPC paste.

nd D134 P. Duan et al. /Materials aThe comparison amongmetakaolin geopolymer, y ash geopoly-mer and OPC paste shows that the total porosity (cumulative intru-sion volume) (Fig. 15(a)) is decreasing for geopolymer pastes, andy ash geopolymer shows a signicant change. The differentialcurves of pore size distribution are almost identical betweendifferent samples. The critical pore diameters, dened as the peaksin the differential curves (Fig. 15(b)), giving the rate of mercuryintrusion per change in pressure (differential curves) [92], do notshow a signicant change. However, the total pore volume in OPCis higher than that in geopolymer. It is evident that exposure ofcement to acidic solutions shifts the pore size peak to a higher valueand also leads to the development of a macropore systemwith poresize exceeding 100 nm.

Unlike cement pastes which hydrate with time, the strength ofthe geopolymer samples remain high after exposure to sulfuric acid(2%) plus hydrochloric acid (2%) solutions. Geopolymer possesses arelatively lower strength loss when compared to conventionalcement paste. This is also conrmed by MIP tests with almost iden-tical results, which show the total porosities of the OPC samples aresignicantly higher than the porosities of the y ash and metakaolinbased geopolymer. This is also extremely true for metakaolin and yash geopolymer samples andOPC paste exposed to high temperature(Fig. 16) and acid (Fig. 17). Geopolymer has a relatively lowerstrength loss when compared to conventional cement paste afterthermal exposure and acid exposure. This is also conrmed by MIPtest, which shows the total porosities of the OPC samples are sig-nicantly higher than the porosities of the y ash and metakaolinbased geopolymer. Additionally, major difference between OPC andgeopolymer can be observed due tomarked strength loss of OPC afterthermal exposure and acid exposure (see Figs. 4, 5 and 11).

Based on the pore structure results by MIP, the total porosity(cumulative intrusion volume) is lower for geopolymer pastescompared to OPC, which means that geopolymer possesses denser

Fig. 16. The porosity and pore size distribution of geopolymer and OPC paste afterthermal exposure.Fig. 17. The porosity and pore size distribution of geopolymer and OPC paste after

esign 74 (2015) 125137microstructure. It indicates that geopolymer could withstand out-side medium ingress including water and acid, which relates tobetter permeability resistance, and the result of MIP is consistentwith the water absorption and acid exposure test. It can beconcluded that alkali activated condition is optimum for betterpore structure modication.

3.11. Formation of microstructure and simulation

MIP tests have been veried from the formation of microstruc-ture aspect, which was simulated with a numerical model. For thesimulation of the reaction processes and of the formation of themicrostructure in cement and geopolymer systems, the numericalmodel HYMOSTRUC was developed [93].

In this model, the degree of hydration was simulated as a func-tion of the particle size distribution, the chemical composition ofgeopolymer materials, the water/solid ratio and the reaction tem-perature. In a computer digitized format of HYMOSTRUC [93], theraw solid particles were modelled as digitized spheres randomlydistributed in a three-dimensional body and the hydratinggeopolymer grains were simulated as growing spheres.

In this model, the geopolymer particles are modelled as spheresof different diameters. Metakaolin and y ash are not regularspheres, and this may be the limitation of this simulation model.However, the hydration proceeded at the particle/water interfaceand the hydration products formed around the original particlesin this model are true when geopolymerization occurs.

The microstructure formation and development in geopolymeris schematically presented in Fig. 18, which can be described viasimplied mechanisms, geopolymer samples show a homogenousmicrostructure with reaction products evenly distributed in thematrix. Some large cavities form and are seemed as isolated byaluminosilicate gels.

acid exposure.

nd DP. Duan et al. /Materials aThe original initial geopolymer particles (Fig. 18(a)) are random-ly distributed in a three-dimensional body. Some hydration prod-ucts form at 1 h (Fig. 18(b)), 2 h (Fig. 18(c)) and 3 h (Fig. 18(d)),

Fig. 18. Formation of the microstructure in geopolymer at various curing times( un-reacted particles; inner products; outerproducts; pores).respectively. Only a little amount of products could be observedat early ages. However, with the reaction time increases, plenty ofhydration products ll the voids in the matrix and the microstruc-ture becomes denser at 6 h (Fig. 18(e)) and 24 h (Fig. 18(f)), respec-tively. Obviously, large amount of outer hydration products form atlonger ages and the microstructure at 3 days (Fig. 18(g)), 7 days(Fig. 18(h)) and 28 days (Fig. 18(i)) can be observed to be muchdenser and only a few voids exist in the matrix.

As mentioned in Fig. 9, geopolymer presents denser micro-structure compared to OPC paste. Additionally, the comparisonamong metakaolin geopolymer, y ash geopolymer and OPC pasteshows that the total porosity (cumulative intrusion volume) inFigs. 1517 is decreasing for geopolymer pastes. The total pore vol-ume in OPC is higher than that in geopolymer. The differentialcurves of pore size distribution also indicate that geopolymer hasoptimized pore structure. All these simulation results are demon-strated by the obtained experimental results.

4. Conclusions

This study intends to broaden application of uidized bed yash and metakaolin in cement or concrete. Durability andmicrostructure of uidized bed y ash and metakaolin geopolymerexposed to elevated temperatures and acid attack were investigat-ed. Compressive strength and several key durability parameters forgeopolymer and ordinary Portland cement (OPC) were assessedand compared. Microstructure formation and development wascharacterized in terms of morphology by SEM and pore structuresby simulation. The results are summarized below:

(1) The distribution of temperature is not uniform and a thermalgradient exists after exposure to high temperature. Wateraccumulated inside the geopolymer structure is the maincontributor to heat resistance, geopolymer presents lowertemperature at the same distance to the surface comparedwith OPC. OPC suffers strength loss, however, the y ashand metakaolin geopolymer gains strength after the samehigh temperature exposure. OPC shows a higher rate ofshrinkage compared to geopolymer after the thermal expo-sures. Polynomial trend lines can be used to show the trendof shrinkage evolution in geopolymer specimens and OPCwith increasing temperature.

(2) The geopolymer microstructures become denser with theincrease of temperature up to 400 C. This change in themicrostructure may be due to sintering and further geopoly-merization of y ash and metakaolin with the increasingtemperature.

(3) Geopolymer specimens have very small change in appear-ance after 28 days of immersion in the acidic solutions, andthere is only 0.7% mass change in geopolymer specimens.Surfaces of OPC specimens exposed to the acidic solutionsindicate severe deterioration compared to geoplymer andthere is 3.3%mass change in the OPC specimens. An exponentrelationship can be used to show the trend of strength evolu-tion in geopolymer specimens and OPC after exposure toacidic solutions.

(4) Sorptivity tests on OPC, metakaolin and y ash geopolymerspecimens after several days curing show the lesser waterabsorption in geopolymer compared to cement paste. Thecomparison among metakaolin geopolymer, y ash geopoly-mer and OPC paste shows that the total porosity is decreas-ing for geopolymer pastes, which is closely related to betterpermeability resistance.

esign 74 (2015) 125137 135(5) The total pore volume in traditional OPC is higher than thatin geopolymer. It is evident that exposure of cement toacidic solutions shifts the pore size peak to a higher value

containing high level y ash and silica fume, Mater. Des. 64 (2014) 261269.[16] W. Chalee, P. Ausapanit, C. Jaturapitakkul, Utilization of y ash concrete in

nd Dmarine environment for long term design life analysis, Mater. Des. 31 (3)(2010) 12421249.

[17] P. Lorca, R. Calabuig, J. Benlloch, L. Soriano, J. Paya, Microconcrete with partialreplacement of Portland cement by y ash and hydrated lime addition, Mater.Des. 64 (2014) 535541.

[18] R.F. Li, G.H. Wu, L.T. Jiang, D.L. Sun, Interface microstructure and compressivebehavior of y ash/phosphate geopolymer hollow sphere structures, Mater.Des. 65 (2015) 585590.

[19] A. Islam, U.J. Alengaram, M.Z. Jumaat, I.I. Bashar, The development ofcompressive strength of ground granulated blast furnace slag-palm oil fuelash-y ash based geopolymer, Mater. Des. 56 (2014) 833841.

[20] T. Phoo-ngernkham, P. Chindaprasirt, V. Sata, S. Hanjitsuwan, The effect ofadding nano-SiO2 and nano Al2O3 on the properties of high calcium y ashand also leads to the development of a macropore system.The compressive strength shows agreement with the resultsof pore structure.

(6) Geopolymer samples show homogenous microstructurewith reaction products evenly distributed in the matrix.Only a little amount of products can be observed at earlyages. However, with the reaction time increases, plenty ofhydration products ll the voids in the matrix and themicrostructure becomes denser after 6 h curing ages.

Acknowledgements

This work was supported by the Fundamental Research Fundsfor the Central Universities (CUGL150806), China University ofGeosciences, Wuhan, Public Service Project of the ChineseMinistry of Land and Resources (201311024) and the NationalNatural Science Foundation of China (NNSF-51202172).

References

[1] K.H. Yang, J.K. Song, K.I. Song, Assessment of CO2 reduction of alkali-activatedconcrete, J. Clean Prod. 39 (2013) 265272.

[2] R.I. Malek, Z.H. Khalil, S.S. Imbaby, D.M. Roy, The contribution of class-F y ashto the strength of cementitious mixtures, Cem. Concr. Res. 35 (2005) 11521154.

[3] X.Y. Wang, Effect of y ash on properties evolution of cement based materials,Constr. Build. Mater. 69 (10) (2014) 3240.

[4] P.K. Sarker, S. Kelly, Z.T. Yao, Effect of re exposure on cracking, spalling andresidual strength of y ash geopolymer concrete, Mater. Des. 63 (11) (2014)584592.

[5] P.S. Deb, P. Nath, P.K. Sarker, The effects of ground granulated blast-furnaceslag blending with y ash and activator content on the workability andstrength properties of geopolymer concrete cured at ambient temperature,Mater. Des. 62 (10) (2014) 3239.

[6] P. Nath, P.K. Sarker, Effect of GGBFS on setting, workability and early strengthproperties of y ash geopolymer concrete cured in ambient condition, Constr.Build. Mater. 66 (9) (2014) 163171.

[7] Z.H. Zhang, H. Wang, Y.C. Zhu, A. Reid, J.L. Provis, F. Bullen, Using y ash topartially substitute metakaolin in geopolymer synthesis, Appl. Clay Sci. 8889(2014) 194201.

[8] T. Sun, J.Y. Chen, X.R. Lei, C.Y. Zhou, Detoxication and immobilization ofchromite ore processing residue with metakaolin-based geopolymer, J.Environ. Chem. Eng. 2 (1) (2014) 304309.

[9] T.W. Cheng, M.L. Lee, M.S. Ko, T.H. Ueng, S.F. Yang, The heavy metal adsorptioncharacteristics on metakaolin-based geopolymer, Appl. Clay Sci. 56 (2) (2012)9096.

[10] Z.T. Yao, X.S. Ji, P.K. Sarker, J.H. Tang, L.Q. Ge, M.S. Xia, et al., A compre-hensive review on the applications of coal y ash, Earth-Sci. Rev. 141 (2015)105121.

[11] R.S. Blissett, N.A. Rowson, A review of the multi-component utilization of coaly ash, Fuel 97 (2012) 123.

[12] R. Kumar, S. Kumar, S.P. Mehrotra, Towards sustainable solutions for y ashthrough mechanical activation, Resour. Conserv. Recy. 52 (2) (2007) 157179.

[13] M. Saridemir, Effect of specimen size and shape on compressive strength ofconcrete containing y ash: application of genetic programming for design,Mater. Des. 56 (2014) 297304.

[14] B. Nematollahi, J. Sanjayan, Effect of different superplasticizers and activatorcombinations on workability and strength of y ash based geopolymer, Mater.Des. 57 (2014) 667672.

[15] W. Wongkeo, P. Thongsanitgam, A. Ngamjarurojana, A. Chaipanich,Compressive strength and chloride resistance of self-compacting concrete

136 P. Duan et al. /Materials ageopolymer cured at ambient temperature, Mater. Des. 55 (2014) 5865.[21] J. Davidovits, Geopolymers-inorganic polymeric new materials, J. Ther. Anal.

37 (8) (1991) 16331656.[22] P. Duxson, J. Provis, G. Lukey, J.S.J. van Deventer, The role of inorganic polymertechnology in the development of green concrete, Cem. Concr. Res. 37 (2007)15901597.

[23] F. Pacheco-Torgal, J.P. Gomes, S. Jalali, Alkali-activated binders: a review. Part2: About materials and binders manufacture, Constr. Build. Mater. 22 (2008)13151322.

[24] F. Pacheco-Torgal, J.P. Gomes, S. Jalali, Alkali-activated binders: a review. Part1: Historical background, terminology, reaction mechanisms and hydrationproducts, Constr. Build. Mater. 22 (2008) 13051314.

[25] M. Shojaei, K. Behfarnia, R. Mohebi, Application of alkali-activated slagconcrete in railway sleepers, Mater. Des. 69 (2015) 8995.

[26] C. Bilim, O. Karahan, C.D. Atis, S. Ilkentapar, Inuence of admixtures on theproperties of alkali-activated slag mortars subjected to different curingconditions, Mater. Des. 44 (2013) 540547.

[27] A. Nazari, Compressive strength of geopolymers produced by ordinaryPortland cement: application of genetic programming for design, Mater. Des.43 (2013) 356366.

[28] P. Posi, C. Teerachanwit, C. Tanutong, S. Limkamoltip, S. Lertnimoolchai, V.Sata, et al., Lightweight geopolymer concrete containing aggregate fromrecycle lightweight block, Mater. Des. 52 (2013) 580586.

[29] P. Sukmak, S. Horpibulsuk, S.L. Shen, Strength development in clay-y ashgeopolymer, Constr. Build. Mater. 40 (3) (2013) 566574.

[30] D. Hardjito, S.E. Wallah, D.M.J. Sumajouw, On the development of y ash-basedgeopolymer concrete, ACI Mater. J. 101 (6) (2004) 467472.

[31] D. Hardjito, S.E. Wallah, D.M.J. Sumajouw, Fly ash-based geopolymer concrete,Aust. J. Struct. Eng. 6 (1) (2005) 7786.

[32] B.V. Rangan, Studies on low-calcium y ash-based geopolymer concrete,Indian Concr. Inst. J. 4 (2006) 917.

[33] D.M.J. Sumajouw, D. Hardjito, S.E. Wallah, Fly ash-based geopolymer concrete:study of slender columns, J. Mater. Sci. 42 (9) (2007) 31243130.

[34] P.K. Sarker, Analysis of geopolymer concrete columns, Mater. Struct. 42 (2009)715724.

[35] A. Bagheri, A. Nazari, Compressive strength of high strength class C y ash-based geopolymers with reactive granulated blast furnace slag aggregatesdesigned by Taguchi method, Mater. Des. 54 (2014) 483490.

[36] E. Papa, V. Medri, E. Landi, B. Ballarin, F. Miccio, Production andcharacterization of geopolymers based on mixed compositions of metakaolinand coal ashed, Mater. Des. 56 (2014) 409415.

[37] S. Pangdaeng, T. Phoo-ngernkham, V. Sata, P. Chindaprasirt, Inuence of curingconditions on properties of high calcium y ash geopolymer containingPortland cement as additive, Mater. Des. 53 (2014) 269274.

[38] J.J. Feng, R.F. Zhang, L.L. Gong, Y. Li, W. Cao, X.D. Cheng, Development of porousy ash-based geopolymer with low thermal conductivity, Mater. Des. 65(2015) 529533.

[39] A.M. Fernandez-Jimenez,A. Palomo,C. Lopez-Hombrados, Engineeringpropertiesof alkali-activated y ash concrete, ACI Mater. J. 103 (2) (2006) 106112.

[40] M. So, J.S.J. van Deventer, P.A. Mendis, Engineering properties of inorganicpolymer concretes (IPCs), Cem. Concr. Res. 37 (2007) 251257.

[41] A.M. Rashad, A comprehensive overview about the inuence of differentadmixtures and additives on the properties of alkali-activated y ash, Mater.Des. 53 (2014) 10051025.

[42] M.O. Yusuf, M.A.M. Johari, Z.A. Ahmad, M. Maslehuddin, Evolution of alkaliactivated ground blast furnace slag-ultrane palm oil fuel ash based concrete,Mater. Des. 55 (2014) 387393.

[43] M.O. Yusuf, M.A.M. Johari, Z.A. Ahmad, M. Maslehuddin, Effects of H2O/Na2Omolar ratio on the strength of alkali activated ground blast furnace slag-ultrane palm oil fuel ash based concrete, Mater. Des. 56 (2014) 158164.

[44] P. Duxson, A. Fernndez-Jimnez, J.L. Provis, G.C. Lukey, A. Palomo, J.S.J. vanDeventer, Geopolymer technology: the current state of the art, J. Mater. Sci. 42(9) (2007) 29172933.

[45] K.C. Goretta, F. Gutierrez-Mora, D. Singh, J.L. Routbort, G.C. Lukey, J.S.J. vanDeventer, Erosion of geopolymers made from industrial waste, J. Mater. Sci. 42(9) (2007) 30663072.

[46] D.L.Y. Kong, J.G. Sanjayan, K. Sagoe-Crentsil, Comparative performance ofgeopolymers made with metakaolin and y ash after exposure to elevatedtemperatures, Cem. Concr. Res. 37 (12) (2007) 15831589.

[47] Sindhunata, J.S.J. van Deventer, G.C. Lukey, H. Xu, Effect of curing temperatureand silicate concentration on y-ash-based geopolymerization, Ind. Eng.Chem. Res. 45 (10) (2006) 35593568.

[48] C.K. Yip, G.C. Lukey, J.S.J. van Deventer, Effect of blast furnace slag addition onmicrostructure and properties of metakaolinite geopolymeric materials,Ceram. Trans. 153 (2004) 187209.

[49] T.W. Cheng, J.P. Chiu, Fire-resistant geopolymer produce by granulated blastfurnace slag, Miner. Eng. 16 (3) (2003) 205210.

[50] J.W. Ginacaspro, P.N. Balaguru, R.E. Lyon, Use of inorganic polymer to improvethe re response of balsa sandwich structures, J. Mater. Civ. Eng. 18 (3) (2006)390397.

[51] R.E. Lyon, P.N. Balaguru, A. Foden, U. Sorathia, J. Davidovits, M. Davidovics,Fire-resistant aluminosilicate composites, Fire Mater. 21 (2) (1997) 6773.

[52] P.G. He, D.C. Jia, T.S. Lin, M.R. Wang, Y. Zhou, Effects of high-temperature heattreatment on the mechanical properties of unidirectional carbon berreinforced geopolymer composites, Ceram. Int. 36 (4) (2010) 14471453.

[53] N. Ranjbar, M. Mehrali, U. Johnson Alengaram, H.S.C. Metselaar, M.Z. Jumaat,

esign 74 (2015) 125137Compressive strength and microstructural analysis of y ash/palm oil fuel ashbased geopolymer mortar under elevated temperatures, Constr. Build. Mater.65 (8) (2014) 114121.

[54] O.A. Abdulkareem, A.M. Mustafa Al Bakri, H. Kamarudin, I. Khairul Nizar, A.A.Saif, Effects of elevated temperatures on the thermal behavior and mechanicalperformance of y ash geopolymer paste, mortar and lightweight concrete,Constr. Build. Mater. 50 (11) (2014) 377387.

[55] L. Zeng, D.Y. Cao, Y. Xu, C.W. Fan, X.Q. Peng, Novel method for preparation ofcalcined kaolin intercalation compound-based geopolymer, Appl. Clay Sci. 101(11) (2014) 637642.

[56] Z. Pan, J.G. Sanjayan, Stressstrain behaviour and abrupt loss of stiffness ofgeopolymer at elevated temperatures, Cem. Concr. Compos. 32 (9) (2010)657664.

[57] Z. Pan, J.G. Sanjayan, F. Collins, Effect of transient creep on compressivestrength of geopolymer concrete for elevated temperature exposure, Cem.Concr. Res. 56 (2) (2014) 182189.

[58] C.C. Mao, H. Ran, H.W. Te, C.F. Toun, Utilization of circulating uidized bedcombustion (CFBC) y ash and coal-red y ash in Portland cement, Key Eng.Mater. 629630 (2014) 306313.

[59] M. Shaswata, M. Saroj, U.B. Adhikari, Study on the physical and mechanicalproperty of ordinary Portland cement and y ash paste, Int. J. Civ. Struct. Eng. 2(3) (2012) 731736.

[60] . Eren, Strength development of concretes with ordinary Portland cement,slag or y ash cured at different temperatures, Mater. Struct. 35 (9) (2002)536540.

[61] W. Sha, G.B. Pereira, Differential scanning calorimetry study of ordinaryPortland cement paste containing metakaolin and theoretical approach ofmetakaolin activity, Cem. Concr. Compos. 23 (6) (2001) 455461.

[73] F.U.A. Shaikh, V. Vimonsatit, Compressive strength of y-ash-basedgeopolymer concrete at elevated temperatures, Fire Mater. (2014), http://dx.doi.org/10.1002/fam.2240.

[74] Y.C. Choi, J.H. Kim, S. Choi, Mechanical performance of y-ash-based greengeopolymer composites subjected to elevated temperatures, Asian J. Chem. 26(17) (2014) 55175521.

[75] J.Z. Kuang, X.C. Zhao, F. Shi, H.Y. Cao, Study on inuence factors of y ash andmetakaolin based geopolymer, Adv. Mater. Res. 538541 (2012) 23222328.

[76] C. Prinya, R. Ubolluk, Utilization of blended uidized bed combustion (FBC) ashand pulverized coal combustion (PCC) y ash in geopolymer, Waste Manage.30 (4) (2010) 667672.

[77] X.H. Yuan, W. Chen, Z.A. Lu, H.G. Chen, Shrinkage compensation of alkali-activated slag concrete and microstructural analysis, Constr. Build. Mater. 66(15) (2014) 422428.

[78] A. Palomo, M.W. Grutzeck, M.T. Blanco, Alkali-activated y ashes: a cement forthe future, Cem. Concr. Res. 29 (8) (1999) 13231329.

[79] H. Rahier, B. van Mele, J. Wastiels, Low-temperature synthesized alumino-silicate glasses. Part II: Rheological transformations during low-temperaturecure and high temperature properties of a model compound, J. Mater. Sci. 31(1) (1996) 8085.

[80] P. Duxson, G.C. Lukey, J.S.J. van Deventer, Physical evolution of Na-geopolymerderived from metakaolin up to 1000 C, J. Mater. Sci. 42 (9) (2007) 30443054.

[81] R.D. Recommendation, P.D.R. De La Rilem, 129-MHT: test methods formechanical properties of concrete at high temperatures, Mater. Struct. 28(1995) 410414.

P. Duan et al. /Materials and Design 74 (2015) 125137 137[62] A.M. Mustafa Al Bakri, H. Kamarudin, M. Binhussain, N.I. Khairul, A.R. Raza, Y.Zarina, Comparison of geopolymer y ash and ordinary Portland cement to thestrength of concrete, Adv. Sci. Lett. 19 (12) (2013) 35923595.

[63] V. Kirk, K. Aditya, S. Gaurav, N. Narayanan, The rheological properties ofternary binders containing Portland cement, limestone, and metakaolin or yash, Cem. Concr. Res. 52 (2013) 196207.

[64] G.S. David, W. Stan, O.F. Martin, Heat of hydration of Portland cementmetakaoliny ash (PCMKPFA) blends, Cem. Concr. Res. 38 (6) (2008) 832840.

[65] V. Kirk, A. Matthew, O. Tandre, S. Gaurav, N. Narayanan, Hydration andstrength development in ternary Portland cement blends containing limestoneand y ash or metakaolin, Cem. Concr. Compos. 39 (2013) 93103.

[66] R.D. Hooton, W. Ma, D. Sample, R. Martin, P.W. Brown, Calorimetric study ofcement blends containing y ash, silica fume, and slag at elevatedtemperatures, Cem. Concr. Aggr. 16 (2) (1994) 9399.

[67] N.I. Fattuhi, B.P. Hughes, Ordinary Portland cement mixes with selectedadmixtures subjected to sulfuric acid attack, ACI Mater. J. 85 (6) (1988) 512518.

[68] S. Wild, B.B. Sabir, J. Bai, J.M. Kinuthia, Self-compensating autogenousshrinkage in Portland cementmetakaoliny ash pastes, Adv. Cem. Res. 12(1) (2000) 3543.

[69] R. Mateusz, O. Jan, Effects of curing conditions on the properties of ternary(ordinary Portland cement/y ash/silica fume) concrete, ACI Mater. J. (2014),http://dx.doi.org/10.14359/51687307.

[70] P. Sandor, Portland cementy ashsilica fume systems in concrete, Adv. Cem.Based Mater. 1 (2) (1993) 8391.

[71] M.S. Sudarshan, R.V. Ranganath, Properties of y ash based geopolymerconcrete exposed to sustained elevated temperatures, Adv. Mater. Res. 250253 (2011) 962968.

[72] D.L.Y. Kong, J.G. Sanjayan, S.C. Kwesi, Comparative performance ofgeopolymers made with metakaolin and y ash after exposure to elevatedtemperatures, Cem. Concr. Res. 37 (12) (2007) 15831589.[82] F. Collins, J.G. Sanjayan, Microcracking and strength development of alkaliactivated slag concrete, Cem. Concr. Res. 23 (4) (2001) 345352.

[83] E. Douglas, A. Bilodeau, V.M. Malhotra, Properties and durability of alkali-activated slag concrete, ACI Mater. J. 89 (5) (1992) 509516.

[84] S.D. Wang, X.C. Pu, K.L. Scrivener, P.L. Pratt, Alkali-activated slag cement andconcrete: a review of properties and problems, Adv. Cem. Res. 7 (27) (1995)93102.

[85] Y. Xu, Y.L. Wong, C.S. Poon, M. Anson, Impact of high temperature on PFAconcrete, Cem. Concr. Res. 31 (2001) 10651073.

[86] M. Bouasker, P. Mounanga, P. Turcry, A. Loukili, A. Khelidj, Chemical shrinkageof cement pastes and mortars at very early age: effect of limestone ller andgranular inclusions, Cem. Concr. Compos. 30 (1) (2008) 1322.

[87] F.G. Collins, J.G. Sanjayan, Effect of pore size distribution on drying shrinkageof alkali-activated slag concrete, Cem. Concr. Res. 30 (9) (2000) 14011406.

[88] Z. Jiang, Z. Sun, P. Wang, Autogenous relative humidity change an autogenousshrinkage of high-performance cement pastes, Cem. Concr. Res. 35 (8) (2005)15391545.

[89] J.L. Provis, J.S.J. van Deventer, Direct measurement of the kinetics of geopoly-merisation by in-situ energy dispersive X-ray diffractometry, J. Mater. Sci. 42(2007) 29742981.

[90] J. Davidovits, Geopolymers and geopolymeric materials, J. Ther. Anal. 35(1989) 429441.

[91] S. Diamond, Mercury porosimetry: an inappropriate method for themeasurement of pore size distributions in cement-based materials, Cem.Concr. Res. 30 (10) (2000) 15171525.

[92] Y. Akkaya, K. van Bui, S.P. Shah, Rheological model for self-consolidatingconcrete, ACI Mater. J. 99 (6) (2002) 549559.

[93] G. Ye, P. Lura, K. van Breugel, A.L.A. Fraaij, Study on the development of themicrostructure in cement-based materials by means of numerical simulationand ultrasonic pulse velocity measurement, Cem. Concr. Compos. 26 (5) (2004)491497.

An investigation of the microstructure and durability of a fluidized bed fly ashmetakaolin geopolymer after heat and acid exposure1 Introduction2 Experimental2.1 Materials2.2 Test procedure2.2.1 Compressive strength2.2.2 Drying shrinkage2.2.3 Acid attack2.2.4 Water sorptivity2.2.5 Mercury intrusion porosimetry2.2.6 Micrographs test

3 Results and discussion3.1 Temperature profile3.2 Compressive strength at elevated temperatures3.3 Thermal stability3.4 Thermal shrinkage3.5 Mass loss at elevated temperatures3.6 Electron microscopy after thermal exposure3.7 Resistance to acid attack3.8 Compressive strength exposed to acid3.9 Water absorption3.10 Pore structure after acid exposure3.11 Formation of microstructure and simulation

4 ConclusionsAcknowledgementsReferences