8.21 strength, sorptivity and carbonation of geopolymer concrete

8/11/2019 8.21 Strength, Sorptivity and Carbonation of Geopolymer Concrete

1/6

Challenges, Opportunities and Solutions in Structural Engineeringand Construction Ghafoori (ed.)

2010 Taylor & Francis Group, London, ISBN 978-0-415-56809-8

Strength, sorptivity and carbonation of geopolymer concrete

A.A. AdamTadulako University, Palu, Indonesia

T.C.K. Molyneaux & I. PatnaikuniRMIT University, Melbourne, Victoria, Australia

D.W. LawHerriot Watt University, Edinburgh, Scotland, UK

ABSTRACT: Several studies have shown that alkali activated binders can achieve similar strengths to both

ordinary Portland cement (OPC) and blended cements. However, to date little research has been undertaken ontheir durability properties. This study investigated the influence of activators concentration and alkali modulus onstrength, sorptivity and carbonation of alkali activated slag (AAS) and fly ash (FA) based geopolymer concrete.The same tests were also conducted on blended concrete with 30%, 50%, and 70% OPC replacement with groundgranulated blast-furnace slag (GGBS) along with control concrete. Results indicate that the alkali modulus hasa major effect on sorptivity of both AAS and geopolymer, however no significant effects of the alkali moduluson carbonation was observed on AAS concrete. The phenolphthalein indicator gave no clear indication betweencarbonated and non-carbonated area in geopolymer specimens. The sorptivity of blended concrete reduced butthe carbonation increased as the replacement level increased.

1 INTRODUCTION

It is widely known that the production of Portlandcement consumes high energy and contributes largequantities of CO2 to the atmosphere. However, atpresent Portland cement is still the main binder inconcrete construction and the search for more envi-ronmentally friendly materials is essential.

One possiblealternative is the use of alkali-activatedbinder using industrial by-products containing silicatematerials (Philleo 1989). The most common industrialby-products used as binder materials are fly ash (FA)and ground granulated blast furnace slag (GGBS).GGBS has been widely used as a cement replace-

ment material due to its latent hydraulic properties,while fly ash has been used as a pozzolanic material toenhance physical, chemical and mechanical propertiesof cements and concretes.

GGBS is a latent hydraulic material which can reactdirectly with water, but requires an alkali activator. Inconcrete, this is the Ca(OH)2released from the hydra-tion of Portland cement. While FA is a pozzolanicmaterial which reacts with Ca(OH)2 from Portlandcement hydration forming calcium silicate hydrate(C-S-H) as the hydration product of Portland cement.Thus, when used with Portland cement, GGBS or

FA will not start to react until some Portland cementhydration has taken place. This delay, causes blended

Portland cements to develop strength more slowly atearly ages than Portland cement alone.

Recent research has shown that it is possible touse fly ash or slag as a sole binder in mortar byactivating them with an alkali component, such as;caustic alkalis, silicate salts, and non silicate salts ofweak acids (Talling and Branstetr 1989). There aretwo models of alkali activation. Activation by low tomild alkali of a material containing primarily silicateand calcium will produce calcium silicate hydrate gel(C-S-H), similar to that formed in Portland cements,but with a lower Ca/Si ratio (Bakharev and Patnaikuni1997). The second mechanism involves the activationof material containing primarily silicate and alumi-

nates using a highly alkaline solution. This reactionwill form an inorganic binder through a polymeriza-tion process (Xu 2002). The term Geopolymeric isused to characterize this reaction from the previous,and accordingly, the name geopolymer has beenadopted for this type of binder (Davidovits 1991).The geopolmeric reaction differentiates geopolymerfrom other types of alkali activated materials (suchas; alkali activated slag/fly ash since the product ispolymer rather than a C-S-H gel.

In order to compare the strength, sorptivity, andcarbonation of blended OPC-GGBS, alkali activated

slag (AAS) and geopolymer concrete; some concretespecimens were prepared with a range of OPC and

563


2/6

GGBS ratios of (0, 30, 50 and 70% GGBS), whileothers prepared with GGBS and fly ash activated byalkaline solution with different alkali modulus.

2 MATERIALS

2.1 Cementitious materials

The GGBS supplied conformed to AS 3582.2-2001.A Scanning Electron Microscope (SEM) image of theGGBS is shown in Fig. 1. The fly ash was a class Ffly ash from Gladstone power station conforming toAS 3582.1-1998. A SEM image of the FA is shownin Figure 2. The OPC used in this investigation wasgeneral purpose (GP) cement. The chemical analysisof these materials is given in Table 1.

2.2 Alkaline activators

A Grade D sodium silicate solution (Na2SiO3) of1.53 g/cc density with an alkali modulus (AM) = 2(Na2O = 14.7% and SiO2 = 29.4%) was supplied

Figure 1. SEM image of GGBS used in this study.

Figure 2. SEM image of Fly ash used in this study.

Table 1. Composition of cementitious materials (%).

Component Cement Slag Fly ash

SiO2 19.9 33.45 49.45Al2O3 4.62 13.46 29.61

Fe2O3 3.97 0.31 10.72CaO 64.27 41.74 3.47MgO 1.73 5.99 1.3K2O 0.57 0.29 0.54

Na2O 0.15 0.16 0.31TiO2 0.23 0.84 1.76P2O5 0.12 0.53Mn2O3 0.06 0.40 0.17SO3 2.56 2.74 0.27

S2 0.58 0.21Cl 0.01 0.001

by PQ Australia. Sodium hydroxide solution (NaOH)was prepared by dissolving sodium hydroxide pelletsin deionised water.

3 MIX PROPORTIONS AND TEST SPECIMENS

A w/b ratio of 0.5 was used to prepare all blendedGGBS-OPC and control concrete. Table 2 shows themix design of control and blended GGBS-OPC con-crete. The proportions of GGBS were 30%, 50%, and70% of the total binder

A water/solid ratio of 0.45 was used for AAS and0.29 for geopolymer concrete. In the case of AAS andFA based geopolymer concrete, the amount of waterin the mix was the sum of water contained in thesodium silicate, sodium hydroxide and added water.The amount of solid is the sum of GGBS or FA, thesolid in the Na2SiO3 solution, and the NaOH pel-lets. The detailed mixes of the AAS and FA basedgeopolymer concrete are shown in Tables 3 and 4.

Liquid sodium silicate and sodium hydroxide wereblended in different proportions (Table 5), providingan alkali modulus (AM) in solution (mass ratio of SiO2to Na2O) ranging from 0.75 to 1.25. The alkali con-centrations (Percentage of Na2O by mass of binder) inthe solution, were 5% for AAS and 7.5% for FA basedgeopolymer concrete.

The mixing was performed using a 120-liter mixer,the mix was then poured into 100 mm diameter 200mmhighcylindermouldsandvibratedfor1minute.The blended GGBS-OPC, control, and AAS concretespecimens were demoulded after 24 hours followed bywater curing at 20C for 6 days and then exposed tothe laboratory environment (26C and 40%RH) priorto testing. The geopolymer specimens did not achievestructural integrity at room temperature. As such the

curing regime was 24 hours at room temperature, fol-lowed 24 hours at 80C (covered with clingfilm). Thespecimens were then allowed to cool in the mould

564


3/6

Table 2. Details of the blended concrete mixes (kg/m3).

Binder Aggregate

Mix OPC GGBS Sand 7 mm 10 mm Water

CTL 428 784 346 693 222S30 296 127 784 346 693 220S50 210 210 784 346 693 219S70 125 293 784 346 693 217

Table 3. Details of the AAS mixes (kg/m3).

Agg regate Alkaline sol.

Mix GGBS Sand 7mm 10 mm Na2SiO3 NaOH Water

AAS5-0.75 419 784 346 693 53 56 137

AAS5-1.00 415 784 346 693 71 46 136

AAS5-1.25 412 784 346 693 87 33 135

Table 4. Details of theFAbased geopolymer mixes (kg/m3).

Aggregate Alkaline sol.

Mix FA Sand 7 mm 10 mm Na2SiO3 NaOH Water

G7.5-0.75 476 784 346 693 90 95 40

G7.5-1.00 467 784 346 693 119 75 38G7.5-1.25 461 784 346 693 147 56 36

Table 5. Proportion of alkaline activators.

Binder (%) Alkaline solution

Na2O/binder AMMix GGBS FA (%) (SiO2/Na2O)

AAS5-0.75 100 5 0.75AAS5-1.00 100 5 1.00

AAS5-1.25 100 5 1.25G7.5-0.75 100 7.5 0.75G7.5-1.00 100 7.5 1.00G7.5-1.25 100 7.5 1.25

before being demoulded. The specimens were thenleft in the laboratory environment until testing.

4 TEST PROGRAM

4.1 Compressive strength testCompressive strength measurements of concreteswereperformed on an MTS machine under a load control

regime with a loading rate of 20 MPa/min. Three tofive cylinders were tested for each data point. Thespecimens were tested at 7, 28 and 90 days aftercasting.

4.2 Sorptivity testThe Sorptivity tests were undertaken in accordancewith DIN 52617. The sides of the specimens werecoated with epoxy to allow free water movement onlythrough the bottom face (unidirectional flow). Theresults were plotted against the square root of the timeto obtain a slope of the best fit straight line. Accordingto Hall (1989), the penetration of water under capillaryaction can be modeled by:

I= A + St1/2 (1)

whereIis the cumulative absorbed volume after timetper unit area of inflow surface,I= Dw/ar,Dwbeingthe increase in weight, a the cross-sectional area andrthe density of water.

4.3 Depth of carbonation test

For depth of carbonation test, the 100 mm diameter200 mm high concrete cylinder were cut into threeparts. In order to keep carbonation direction in theradial direction, top and bottom of each specimen werecoated with epoxy. The specimens were transferred to

a specially designed chamber to accelerate the carbon-ation process. The chamber was supplied with carbondioxide (CO2) to maintain a CO2level of 20%. A satu-rated NaCl solution was used to maintain the humiditylevel between 75%80%.

At weekly intervals specimens were split and thensprayed with a phenolphthalein indicator, as pre-scribed by RILEM (1994). An average carbonationdepth was then taken from the cross-sectioned slices.

5 RESULTS AND DISCUSSIONS

5.1 Comparison of strength

The strength of blended, AAS, and FA based geopoly-mer concrete are shown in Tables 6, 7, and 8 respec-tively.

In general, the 28-days compressive strengths ofthe AAS and FA geopolymer concretes are compa-rable with that of 100% OPC concrete and blendedOPC-GGBS concretes as shown in Figures 3, 4 and 5.It should be noted that heat curing was applied to theFA geopolymer concrete to achieve structural integrity.Heat curing in general will result in increased early

strengths. As such a comparison of the 28 and 90 daysstrengths will give a better assessment of the compa-rable strengths, than the 7 days data.

565


4/6

Table 6. Compressive strength of blended concretes.

Cementitious (%) Strength (MPa)

Mix OPC GGBS 7d 28d 90d

CTL 100 0 36 50 57S30 70 30 32 47 50S50 50 50 29 47 53S70 30 70 26 36 43

Table 7. Compressive strength of AAS concretes.

Strength (MPa)Na2O AM

Mix (%) (SiO2/Na2O) 7d 28d 90d

AAS5-0.75 5 0.75 25 33 37AAS5-1.00 5 1.00 36 44 43AAS5-1.25 5 1.25 42 46 47

Table 8. Compressive strength of Geoplymer concretes.

Strength (MPa)Na2O AM

Mix (% ) (SiO2/Na2O) 7d 28d 90d

G7.5-0.75 7.5 0.75 39 44 44G7.5-1.00 7.5 1.00 50 43 54G7.5-1.25 7.5 1.25 52 57 57

0

10

20

30

40

50

60

70

CTL S30 S50 S70

Compressivestrength(MPa)

7 days 28 days 90 days

Figure 3. Strength development of blended GGBS-OPCconcrete.

The AM of the activator has a significant influ-ence on the strength of AAS andFA-based geopolymerconcrete up to AM = 1, beyond this level the influencereduced. The strength of AM = 1.25 geopolymer wasslightly higher than that of AM = 1, by contrast thestrength of AM =1.25 AAS was slightly lower than

that of AM = 1 AAS concrete specimens.In comparison, the blended OPC-GGBS devel-

oped strength slowly at an early age, and decreased

0

10

20

30

40

50

60

AAS5-0.75 AAS5-1 AAS5-1.25

Compressives

trength(MPa)


Figure 4. Strength development of AAS concrete.

0

10

20

30

40

50

60

G7.5-0.75 G7.5-1 G7.5-1.25compressivestrength(MP

a)


Figure 5. Strength development of geopolymer concrete.

in strength as the level of replacement increased. At28 days age, the strength of 30% and 50% blendedOPC-GGBS concretes were constant, but the strengthreduced at 70% replacement. At 90 days the strengthof the 50% blended OPC-GGBS concrete was highest,with the 70% again the lowest. The 100% OPC controlconcrete displayed a higher strength than the blendedconcretes at 7, 28 and 90 days. It is expected that theblended concretes will exhibit higher strengths as thetime increases. It should be noted that the water cur-ing only applied for 7 days, this delayed the strengthdevelopment of blended as the hydration of slag ismore sensitive to water curing than those for OPCconcrete. The hydration of slag requires Ca(OH)2from Portland cement hydration, and it will not startuntil the hydration of OPC has taken place.

For FA based geopolymer concrete most of thestrength were gained by 7 days and no further increasein strength was observed up to 28 days, this wasattributed to the heat curing.

As for the AAS concrete, the alkali modulus of theactivator has a significant influence on the strength of

the FA-based geopolymer concrete up to an AM = 1.Beyond this limit the influence was marginal. Increas-ing the alkali modulus in these examples resulted in

566


5/6

an increase in soluble silicates and consequently anincrease in the reaction rate (a higher concentration ofreactants induces a higher reaction rate). Overall theFA based geopolymer displayed higher strengths thanthe AAS concrete specimens

5.2 Sorptivity

The results of the sorptivity tests are presented inFigures 68. The results for blended GGBS-OPCshows that the reduction of sorptivity was proportionalto the level of GGBS replacement.

0.00

0.02

0.04

0.06

0.080.10

0.12

0.14

0.16

0.18

CTL S30 S50 S70

Sorptivity(m

m/min1/2) 56 days 90 days

Figure 6. Sorptivity of blended GGBS-OPC concrete.

0.00

0.05

0.10

0.15

0.20

0.25

0.30

AAS5-0.75 AAS5-1 AAS5-1.25

Sorptivity(mm/min1

/2) 56 days 90 days

Figure 7. Sorptivity of AAS concrete.

0.00

0.02

0.04

0.06

0.08

0.10

0.12

G7.5-0.75 G7.5-1 G7.5-1.25

Sorptivity(mm/min1/2) 56 days 90 days

Figure 8. Sorptivity of FA-based geopolymer concrete.

There was a large reduction in sorptivity of bothAAS and geopolymer concrete as the alkali modu-lus increased from 0.75 to 1.00. However only a smallreduction in sorptivity was observed as the alkali mod-ulus increased from 1.00 to 1.25 for FA based geopoly-mer concrete, and no further reduction observed for

AAS concrete.The results show that there was an optimum alkali

modulus for both AAS and geopolymer concrete, inthis case the value was AM = 1. Increasing the alkalimodulus, which also means increasing the SiO2 con-tent in the system will reduced the porosity of bothAAS and FA-based geopolymer concrete.

The FA geopolymer concrete specimens displaya significantly lower value than both the AAS con-crete and the control concrete. As the level of GGBSreplacement increases the sorptivity becomes compa-rable with the FA based geopolymer.

5.3 Depth of carbonation

A summary of the results for the accelerated carbona-tion at 20C and 75%80% RH is shown in Figure 9.It can be seen that the rate of carbonation was higherwhen the level of GGBS replacement increased. Theresult shows that in blended OPC-GGBS, the carbon-ation rate is not primarily influenced by the porosityof the concrete, as can be observed in OPC concrete.Carbonation in normal concrete is regarded as a reac-tion that takes place between carbon dioxide (CO2)

and Ca(OH)2. This reaction, that takes place in aque-ous solution, which can be written as (Bertolini et al.2004):

CO2 + Ca(OH)2H2 O, NaOH CaCO3 + H2O (2)

Regardless of the alkali modulus, the depth of car-bonation of alkali activated slag was even higher than

0

5

10

15

20

25

30

35

40

0 1 2 3 4 5 6 7 8 9

week

Depthofcarbonation(mm)

CTL S30 S50 S70AAS5-0.75 AAS5-1 AAS5-1.25

Figure 9. Depth of carbonationof blended OPC-GGBS andAAS concrete at 20% CO2 (20

C and 7580% RH).

567


6/6

that of blended OPC-GGBS, and about three times thatof control concrete. Comparing the rate of carbonationin OPC, blended GGBS-OPC, and AAS concrete, asseen in Figure 9, the carbonation of OPC was almostconstant after week 6, whereas AAS continue to car-bonate at a relatively high rate. The carbonation of

blended OPC-GGBS also continues to occur althoughat a slower rate compared to the AAS concrete. Inblended OPC-GGBS concrete, after initial carbona-tion, the product of this reaction f ills up the poresresulting in a higher density of the matrix and furtherslowing the diffusion of CO2. Since the matrix of theAAS concrete contains marginal volumes of Ca(OH)2,the main carbonation reactant must be coming fromanother source. According to Peter et al. (2008), otherconstituents in the concrete can also carbonate, partic-ularly calcium-silicate hydrates (C-S-H). The latter isbelieved to be the main carbonation reactant in AAS

concrete, whereas in blended OPC-GGBS concrete itis the combination of the two.

The phenolphthalein indicator gave no clear borderbetween colour and colourless area for geopolymer asseen in control, blended GGBS-OPC and AAS con-crete. Therefore it was not possible to measure thecarbonation depth. This was attributed to the poly-meric type reaction of the geopolymer, which did notproduce either the C-S-H gel or the Ca(OH)2. Furthertesting is being undertaken on the pH and carbonationof the geopolymer concretes.

6 CONCLUSIONS

The blended OPC-GGBS concrete gains strengthmore slowly than the OPC concrete for the samewater cement ratio.

The early strength of 1.0 AM and above for FA-based geopolymer concrete was considerably higherthan that of OPC concrete and similar for 28-daysstrength. This is attributed to the heat curing for theFA geopolymer concrete

Increasing the alkali modulus(AM) up to 1 enhances

strength but further increases of AM have mini-mal impact on the strength of FA based geopoly-mer concrete and reduction on strength for AASconcrete.

There was a large reduction in sorptivity of bothAASand geopolymerconcrete as thealkali modulusincreased from 0.75 to 1.00.

The carbonation mechanism in AAS concrete is dif-ferent from that of OPC concrete since the reactantis C-S-H whereas it is Ca(OH)2in OPC. In blended

OPC-GGBS concrete the mechanisms appears to bea combination of both.

The phenolphthalein gave no clear indication bet-ween carbonated and non-carbonated area in geo-polymer specimens.

REFERENCES

Bakharev, T. and Patnaikuni, I., 1997. Microstructure anddurability of alkali activated cementitious pastes. In:K.C.G. Ong (Editor), The Fifth International Confer-

ence on Structural Failure, Durability and Retrofitting.Singapore Concrete Institute, Singapore, p. 200.

Bertolini, L., Elsener, B., Pedeferri, P. and Polder, R., 2004.Corrosion of Steel in Concrete. WILEY-VCH VerlagGmbH & Co. KGaA, Weinheim, p. 392.

Davidovits, J., 1991. GEOPOLYMERS: INORGANICPOLYMERIC NEW MATERIALS. Journal of ThermalAnalysis, 37: 16331656.

Hall, C., 1989. Water sorptivity of mortars and concretes: areview. Magazine of Concrete Research 41(147): 5161.

Peter, M.A., Muntean, A., Meier, S.A. and Bhm, M., 2008.Competition of several carbonation reactions in concrete:A parametric study. Cement and Concrete Research,

38(12): 13851393.Philleo, R.E., 1989. Slag or other supplementary materials?In: V.M. Malhotra (Editor), Third international confer-ence on the use of fly ash, silica fume, slag and nat-ural pozzolan in concrete. American Concrete Institute,Trondheim, Norway, pp. 11971208.

RILEM, 1994. CPC 18. Measurement of hardened concretecarbonation depth, 1988. In: RILEM (Editor), RILEMRecommendations for the Testing and Use of Construc-tions Materials. E & FN SPON, pp. 5658.

Talling, B. and Brandstetr, J., 1989. Present State and Futureof Alkali-Activated Slag Concretes. In: V.M. Malhotra(Editor), third international conference on fly ash, silicafume, slag, and natural pozzolans in concrete. Publica-

tion SP; 114. American Concrete Institute, Trondheim,Norway, pp. 15191545.

Xu, H., 2002. Geopolymerisation of Aluminosilicate Miner-als, The University of Melbourne, Melbourne.

568

8.21 strength, sorptivity and carbonation of geopolymer concrete

Documents