a calorimetric study of early hydration of

8/13/2019 A Calorimetric Study of Early Hydration Of

1/14

PergamonCement and Connete Research, Vol. 25. No. 6. pp. 1333-1346.1995

Copyright 8 1995 Elsevier Science LtdPrinted in the USA. All rights reserved

0008~S&W95 9.50+.00

0008.8846(95)00126-3

A CALORIMETRIC STUDY OF EARLY HYDRATION OFALKALI-SLAG CEMENTS

Caijun Shi* and Robert L. Day*** Wastewater Technology Centre (operated by RockCliffe Research Management)867 Lakeshore Road, Burlington, Ontario, Canada L7R 4L7

** Department of Civil Engineering, The University of Calgary, Calgary,Alberta, Canada T2N lN4(Refereed)

(Received December 6,1994: in final form April 4.1995)

ABSTRACTThis paper examines the early hydration of alkali-slag cements activated by differentsodium compounds, such as NaOH, Na&O,, Na$iO,.SH,O, Na,PO.,, N%HPO, andNaF, at 25 and 50C. A conduction calorimeter was used to monitor hydrationkinetics. It was found that the initial pH of activator solution has an important rolein dissolving the slag and in promoting the early formation of some hydrationproducts. However, further hydration of alkali-slag cements is dominated by thereaction of the anion or anion group of activator and the Ca dissolved from slagrather than the initial pH of the activator solution. Finally, three models wereproposed to describe the early hydration of alkali-slag cements based on heatevolution measurements.

INTRODUCTIONAlkali-slag cements consist of two components: a cementitious component such as blastfurnace slag, phosphorus slag, etc., and an alkaline activator. The cementitious component mayshow little or no cementing properties when they are mixed with water alone. However, theycan develop very high strength, even higher than portland cement, in the presence of a properalkaline activator. For example, ground granulated phosphorus slag pastes show a strength of

only 0.8 MPa in the absence of activators, while reach a strength of 160 MPa in the presence of3% (by mass of N%O) N+O.nSiO,, after curing 90 days at room temperature [ 1, 21.

Shi et al. [3] found that all caustic alkalies and the salts whose anions, or anion groups, reactwith Ca* to form insoluble, or less soluble, calcium compounds can be used as alkalineactivators. However, alkaline activators demonstrate selectivity - that is, different activatorsexhibit different effects on slags from different sources. Glukhovsky et al. [4] classified theactivators into six groups:

1333


2/14

C. Shi and R.L. Day Vol. 25. No. 6334

1)2)(3)(4)(5)(6)

caustic alkalies, MOH;non-silicate weak acid salts, M&O,, M,SO,, M,PO,, MF etc.;silicates, M,O.nSiO,;aluminates, M,O.nAl,O,;aluminosilicates, M,O.Al,0,.(2-6)Si02 andnon-silicate strong acid salts, M,SO,.

Many publications have dealt with the mixing proportions, mechanical properties anddurability of alkali-slag cements and concretes [5-g], but less work has been done on hydrationcharacteristics of these cements [g-lo]. Some controversial phenomena have been observed dueto the difference in the source of slag and the selectivity of activators. The early hydrationcharacteristics of alkali-slag cements and the activation mechanism of different activators are stillnot well known. This study deals with the early hydration features of alkali-slag cements whichare activated by the first three groups of alkalies classified above, using a conduction calorimeter.An attempt is made to explain the hydration characteristics of alkali-slag cements under theactivation of different activators.

EXPERIMENTATIONRaw Materials

The characteristics of the granulated blastfurnace slag (BFS) used are given in Table 1.X-ray diffraction testing indicated that this slagconsists mainly of glassy phase except for tracecrystalline merwinite (3CaO.Mg0.2SiOJ. A CSA(Canadian Standard Association) Type 10 portlandcement (PC) was used as a reference cement.The chemical composition and some physicalproperties of the PC are also given in Table 1.

Analytical grade chemicals NaOH, Na&O,,Na$iO,.SH,O, Na,PO,, N+HPO, and NaF werestudied as alkaline activators. It was found thatthe mass of NqO, instead of the total mass ofactivators, correlated best with physical propertiesof alkali-slag cement pastes [2,4]. Accordingly,all alkaline activators were added as 4% by massof NhO. These activators were dissolved inwater first and then were mixed with the slag.

Measurement of Heat Evolution Rates

Table 1 % Chemical Composition andPhysical Properties of BFS and PCItem 1 BFS 1 PC 11

WiO I 0.31 I 0.14 11KZO l 0 44 I 0 59 III.L. 0 2.70Total 99.89 100.50Density

(kg/m3)BlaineFineness(m2kg)

2920

340

3140 II425

Evolution of heat of hydration was monitored using a Seebeck Isothermal ConductionCalorimeter, in accordance with previous work by Adams [ll]. The test sample of slag wasweighed and uniformly distributed in the stainless steel sample container. The solution of water


3/14

Vol. 25. No. 6 CAJBRIMETRY, ALKALI-SLAG CEMENTS, HYDRATION 1335

and activator was weighed in a syringe. The mass of the slag used was 14 g and a water-to-slagratio 0.45 was used. Measurements of heat of evolution were performed at constant temperaturesof 25 and 5O+lC. The sample container with the slag and the syringe with the activator solutionwere heated to testing temperatures prior to mixing. When thermal equilibrium was achieved,the slag and the solution containing the alkaline activator were mixed by injecting the solutioninto the slag. The heat evolution was recorded as a voltage and then converted to KJ/kg.h bya constant obtained from calibration tests. The calibration base line was established by recordingthe heat evolution of a mixture of water and ground quartz of the same water-to-solid ratio.

EXPERIMENTAL RESULTSPortland Cement

As shown on the heat evolution curves of portland cement (Fig. l), an initial peak appearswithin the first two or three minutes. The heat evolution curves then decline to an inductionperiod lasting from one to two hours. This is followed by an accelerated hydration peak whichresults from the hydration of C,S to C-S-H. Usually, the height of the accelerated hydration peakis lower than that of the initial peak. A third peak, which appears after the accelerated hydrationpeak, is attributed to the transformation of ettringite (GA.3CaS0,.32HZO) to monosulphate(C,A.CaS0,.13&0); at 25C, the transformation peak is very diffuse, but the temperature risesto 50C all peaks sharpen.

Slag + WaterAt both 25 and 50C the mixture of ground blast furnace slag and water gave only a verysmall initial heat evolution peak, no other peaks were recorded until the end of the test at 72

hours at 50C. The recorded curves are not given here.

120

8060

0.0 0.2 0.4 0.6 0.8 1 .O 8 16 24 32Hydration Time hour)Fig. 1 Heat Evolution of Hydration of Type 10 Portland Cement

40 48


4/14

120

100

80

60

40

20

0

C.Shind R.L.Day VoL25 No.6

0.0 0.2 0.4 0.6 0.8 1 .O 8 16 24 32 40 48Hydration Time hour)

Fig. 2 Heat Evolution Curve of Slag+NaOHSlap: + NaOH

The heat evolution curves of the slag activated by NaOH at 25 and 50C are presented inFig. 2. It can be seen that the trend of heat evolution of this cement is similar to that of portlandcement. In contrast to portland cement, NaOH activated slag has a lower initial peak than theaccelerated hydration peak, especially at 50C. Also, NaOH activated slag does not show anappreciable induction period.Slag + NaSiO,

The heat evolution curves of the alkali-slag cement activated by Na$iO, are shown in Fig.3. At 25C, a small peak appears just after the addition of the activator solution, which is120100

80804020

0

I I I I , I I I ,I I I I I I I I II_--_~---;-__~---~--_----~_-~__-;-_~-__ _--I I I I I I I I II I I I I I I I I__-- -_- -__ ---~---_--- -- __- -_-j------ _-_I I I I I I I I II I I I I I I I I--- ~_-- ___~-__~_-----_~_--j--_c-__


Fig. 3 Heat Evolution Curve of Slag+Na,SiO,


5/14

Vol. 25, No. 6 CAU ~Y. ALKALMLAG CEMENTS, HYDRATION 1337

followed immediately by a strong peak during the initial period or pre-induction period (seediscussion of hydration period later). The first peak is called initial peak, like portland cementand NaOH activated slag cases. The second peak that occurs during the pre-induction period ofalkali-slag cements is called additional initial peak. As temperature rises to WC, only one peakduring the initial period occurs, which is much higher than the initial or additional initial peakat 25C. This peak could be regarded as the combination of the initial and additional initialpeaks detected at 25C [9, 121. At the same time, elevating temperature to 50C shortens theinduction period from 18 hours at 25C to 6 hours.Slag + Na,CO,

The heat evolution curves of the slag activated by Na$O, are shown in Fig. 4. The heatevolution curves of this cement are somewhat similar to those of Na$iO, activated slag in thatthere are two early peaks at 25C and one strong initial peak at 50C. However, there are somedifferences between the two cements: (1) the initial peak of Na+Z03 activated slag at 25C ismore noticeable than that of Na.$iO, activated slag; (2) the additional initial peak of Na+ZO,activated slag appears later and is more diffuse than that of Na$iO, activated slag at 25C; (3)elevating temperature has a more significant effect on the initial peak of Na$O, activated slagthan on Na$iO, activated slag; and (4) Na$O, activated slag shows a shorter induction periodthan that of NqSiO, activated slag at both 25 and 50C. All these differences may be attributedto the formation of different substances, which corresponds to the additional initial peak at 25C.This will be discussed later.Slag + Na,PO,, or Na,HPO,

Fig. 5 shows the heat evolution curves for the cements activated by either Na,PO, orN%HPO,. Since Na,PO,, has a very low solubility at 25C, only the test at 50C was undertaken.The cement activated by NqPO, exhibits two initial peaks, even at 50C. After a long periodof induction, a lengthy and diffuse second peak ensues.

120

100

80

60

40

20

00.0 0.2 0.4 0.6 0.8 1.0 8 16 24 32 40 48

Hydration Time hour)Fig. 4 Heat Evolution Curve of Slag+NaJO,


6/14

1338

0

C. Shi and R.L . Day Vol. 25, No. 6


Fig 5 Heat Evolution Curves of Slag+N%PO, and Slag+N%HPQThe results of using N%HPO, activation are quite different from those of the Na,PO, case.

Only one strong initial peak materializes at both 25 and 50C and no other peaks were detected.Visual examination of slag samples at the end of the calorimetry test showed no indication of thehydration of slag grains.

Slag + NaFThe heat evolution curves of the cement activated by NaF, as shown in Fig. 6, are unique.

Two peaks occur during the initial period, and the initial peak is stronger than the additionalinitial peak at both 25 and 50C. The additional initial peak during the pre-induction period at25C is very diffuse and not obviously observed, and no obvious accelerated hydration peak isobserved. At 50C the additional initial peak can be clearly noticed but is still very diffuse.

120

0

c_.__+--_+___+___q____c--+__+--i___+----j_--+---II I I I I I I I II I I I I I I I I-- -_- -___ __-~_____-- __~--- _-~--- -_-I I I I I I I I II I I I I I I I Ic---A__-J___J-_____-c-_d--A_--~---l--L---

I I I I II I--i---L___I II II I I I I

0.0 0.2 0.4 0.6 0.8 1.0 8 16 24 32 40 48Hydration Time hour)

Fig. 6 Heat Evolution Curves of Slag+NaF


7/14

Vol.25 No.6 CALORlMETRY ALKALI-SLAGCEMENTS.HYDRATION 1339

Compared with the other activators, the conspicuous feature of this cement is that it exhibits along induction period and a broad second peak for hydration at WC.

GENERAL DISCUSSIONSThe principal heat evolution characteristics of the alkali-slag cements activated by the various

alkaline activators are summarized in Table 2. The following paragraphs will discuss theactivation process of different activators based on the measured heat evolution and the publishedresults:Effect of Temperature

By comparing the heat evolution characteristics which are summarized in Table 2, such asthe onset time and the height of the accelerated hydration peak, and the total heat during the first24 hours at 25C with those at 50C it is found that temperature has a more significant effect onthese characteristics of all alkali activated slags than on those of portland cement except for theheight of the accelerated hydration peak of NaOH activated slag and the total heat during the first24 hours of N%HPO, activated slag. This means that most alkali activated slags have a higherhydration activation energy than portland cement [13].

Effect of Initial pH of Activator SolutionSome researchers have concluded that the activation of slag depends on the initial pH ofactivator solutions [14]. A recent study indicated that the early hydration of slag is accelerated

as the pH of the activator solution increases [15]. In the current study, NaOH solution had thehighest pH among all studied activators, yet NaOH activated slag did not give the highest totalheat released in its first 24 hours. Furthermore, although the pH of Na.$iO, solution. is higherthan that of NqCO, solution, and Na$O, activated slag shows a longer induction period and alower accelerated hydration peak than Na$iO, activated slag. Also, although NaF solution hasa pH of about 10, it can still activate the hydration of slag. A previous work indicated that NaFactivated ground phosphorus slag gave a high strength at room and high temperatures [3]. Thecurrent research confirms early findings that hydration characteristics of alkali activated slag relymore upon the nature of anion or anion group of the activator than upon the initial pH of theactivator solution [3].Heat Evolution Rate and Total Heat Evolution

It has been reported that alkali-slag cements has lower heat evolution rate and total heatevolution than portland cement and can be regarded as a low-heat cement [16]. However, theheat evolution rate of Na&O, activated slag and the total heat evolution of NaF activated slagat 50C surpass those of portland cement. Similar results were obtained in another work, it wasfound that both the heat evolution rate and total heat evolution of the same alkali-slag cementcould be higher than that of portland cement [ 171. Thus, for alkali-slag cements, the contributionof heat of hydration depends on the nature of the slag, and the nature and dosage of activatorsused; they cannot generally be regarded as low-heat cements.


8/14

Te2SmyoHaoCaescoAkSaCmesAvebVoAvo

3

NSO

4

NC

5

NPO

6

NHO

7

NF

NA-nace

5

35

NA

NA

52

01

25

2

73

30

00

72

17

61

18

5

75cmn

0

49

21

29

2

12

24

00

63

52

18

11

5

131cmn

0

571

05

39

16

5

68

14

01

91

15

19

2

41

NA

NA

NA

NA

44

91

5

89

NA

NA

NA

NA

63

2

10

NA

NA

NA

NA

51

11

5

47

25

00

27

59

38


9/14

Vol. 25. No. 6 CALORIME IRY. ALKALI -SLAG CEMENTS, HYDRATION 1341

Effect of Anions or Anion Grouns of Activators on Hydration Chemistrv of Alkali-slag Cements

Calorimetry has played an important role in understanding the early hydration chemistry ofportland cement. It is generally accepted that the early hydration of portland cement can bedivided into five periods based on heat evolution curves, as schematically represented in Fig.7:(I) initial (pre-induction) period; (II) induction period; (III) acceleration (post-induction) period;(IV) deceleration period; and (V) diffusion period.

One distinctive feature of some alkali-slag cements is that two peaks may appear during theinitial or pre-induction period. The peak during the pre-induction period of portland cement isattributed to the rapid dissolution of alkali sulphates and aluminates, initial hydration of C,S andformation of AFt [18]. Bensted [ 191 believes that the peak can also be attributed partly to thecombined effects of heat of wetting of the cement, the hydration of free lime and partly to thehydration of CaSO,.O.SH,O to CaS0,.2H,O:

CaO + H,O ----> Ca(OH), + 15.58 KJ/mol andCaSO,.O.SH,O + 1.5H,O ----> CaS0,.2H,O + 4.10 KJ/mol

The CaSO,.O.SH,O originates from the decomposition of CaS0,.2H,O during grinding.Based on the measured heat evolution characteristics of alkali-slag cements activated bydifferent activators in this study and the broad classification of hydration period of portland

cement, the hydration of alkali-slag cements can be generally represented by three models asshown in Fig.8.

For Type I, one peak occurs during the first few minutes and no more peaks appearthereafter. Hydration of the slag in water or in the N%HPO, solution at both 25 and 50C is auexample. In this case, the slag usually does not set and harden.

Period. . . . ..~............................~.............._.__~_.........__._.......,.I / II i III ; IV f V

+Minutes HCUrs Days

Hydration TimeFig.7 Schematic Representation of Hydration Periods of Portland Cement


10/14

1342 C. Shi and R.L. Day Vol. 25. No. 6

Periodno other hydration features

LMinutes Days

Hydration Time

a> Type 1

V:.

: :: :: : III ; IV ;: :: :: f: :: :: :: :: :: :: I:: :II : :

Minutes HOUS Days

Period

Hydration Time

b) Type IIPeriod

Minutes HOWS DaysHydration Time

Fig.8 Schematic Representation of Hydration Models of Alkali-Slag CementsFor Type II, only initial peak appears before the induction period and one accelerated

hydration peak appears after the induction period. In this study, hydration of the slag activatedby NaOH at 25 and 50C belongs to this type.

For Type III, two peaks, the initial and the additional initial peaks, appear before theinduction period and one accelerated hydration peak appears after the induction period. Theinitial peak could be lower or higher than the additional initial peak depending on the nature ofactivators and hydration temperature. In this study, this type of hydration includes the slagactivated by NqSiO, and NaJO, at 25C, Na,PO, at 50C and NaF at both 25 and 50C.

Because the initial and the additional initial peaks appear very close, they may merge intoone peak depending on the activator dosage, slag activity and hydration temperature. The mergerwas observed from the hydration of Na.$iO, and Naj.ZO, activated slag at 50C.

The early hydration characteristics of alkali-slag cements will be discussed in detail asfollows:


11/14

Vol. 25 No. 6 CALORMEIRY. ALKAU-SLAG CEMENI-S. HYDRATION 1343

(a) in Water

When ground granulated slag contacts water, the Si-0, Al-O, Ca-0 and Mg-0 bonds on mesurface of the glassy slag break under the polarization effect of OK [20]. Because the Ca-0 andMg-0 bonds are much weaker than Si-0 and Al-O bonds, more Ca and Mg enter into water thanSi and Al, and a Si-Al-rich layer forms quickly on the surface of slag [lo]. The Si-Al-rich layermay adsorb some H in the water which results in an increase of OH- or pH of the solution.However, this concentration of OH still could not break enough Si-0 and Al-O bonds for theformation of significant amount of C-S-H, C-A-H or C-A-S-H and M-A-H. Although elevatingtemperature can enhance the polarization effect [20], no accelerated hydration peak can bedetected even at 50C. The only very small initial peak is attributed to the wetting anddissolution of slag grains and adsorption of some ions onto the surface of slag grains. Only asmall amount of C-S-H forms even at 150 days [lo].(b) in NaOH Solution

NaOH solution has a high concentration of OH-, which can break not only Ca-0 and Mg-0bonds, but also a significant number of Si-0 and Al-O bonds. Table 3 lists these Ca-compoundsassociated with the anions or anion groups of alkaline activators investigated. Because Ca(OH),has a much higher solubility than C-S-H, C-A-H and C-A-S-H, Ca(OH), cannot precipitate fromthe solution, a very thin layer consisting of low C/S ratio C-S-H, C&H,, and C,ASH,, which hasa very low solubility, precipitates very quickly through the solution. M&-J, orM@l,CO,(OH),,.4H,O appears instead of C&I,, if MgO content of the slag is high [25]. Thequick hydration of slag and precipitated hydration products were observed under a scanningelectron microscope in NaOH activated slag hydrated for 20 minutes at 25C [26]. It seems thatthe precipitation of these hydration products does not lead to a noticeable induction period, whichmay be determined by the nature of these hydration products. This needs further research.

Table 3 Solubility Product of Some Ca-Compounds

* tht

ComDound I Solubilitv Product Reference

CaCO,WPQ), I 2.ox1o-2g

CZlF I 2.7x10-CaHPO, 1.0x10-

lower the C/S ratio of C-S-H, the higher th

WI 2[231v41u51Pl lPllI T

solubility pr duct.


12/14

1344 C Shi and R L Day Vol 25 No 6

(c) in Na$iO, Solution

NqSiO, solution has a high pH, but the pH is lower than that of NaOH solution. As statedabove, when ground granulated slag contacts the solution, OK break Ca-0, Si-0 and Al-Obonds. Because this solution has a high concentration of [Si0414-, primary C-S-H, resultingfrom the reaction of [SiO,] from Na,SiO, and Ca dissolved from the surface of slag grains,forms very quickly when the concentration of Ca*+reaches a critical value [12]. Thus, two heatevolution peaks (initial and additional initial) can be identified on heat evolution curves ofNqSiO, activated slag during the initial period at 25C. The initial peak corresponds to thewetting and dissolution of slag grains and adsorption of some ions onto the surface of slaggrains. The additional initial peak is mainly due to the formation of the primary C-S-H asdescribed before [ 121.

Because the initial and the additional initial peaks appear very close, some actions maycontribute to both the peaks. Also, some [Si0414- in the solution may polymerize if theconcentration is too high (especially when a high ratio waterglass is used) and no enough Cais available to react. This can be explained by the bigger diffuse heat evolution peak whenhigher modulus waterglass is used [12]. Raising temperature accelerates the formation of theprimary C-S-H and the two peaks merge into one. The precipitation of primary C-S-Hresults in a very long induction period.(d) in NqCO, Solution

Two peaks are well identified during the initial hydration period of Na&O, activated slagat 25C. The reasons for the initial peak can be regarded the same as described in the Na$iO,case. Since CaCO, has a lower solubility than low C/S ratio C-S-H as shown in Table 3, theadditional initial peak here may be due to the formation of CaCO,. Because the pH of Na$O,solution is lower than that of Na.$iO, solution, it takes longer for the concentration of Ca* inthe solution to become high enough to form CaCO,. The additional initial here peak is higherand more diffuse than the NqSiO, case.

The induction period of Na ZO, activated slag is shorter than that of Na$iO, activated slag.There are two possible two reasons: one is that CaCO, crystals are different from C-S-H gel; theother is that CaCO, crystals involve the formation of hydration product, such as C,AH,,. CaCO,cannot be identified in Na+ZO, activated slag at 1 day [26]. This means that CaCO, is only atransitional product at very early age. As temperature rises to 5OC, one very strong initial peakis detected, instead of two at 25C. The reason for that is the same as in the Na$iO, case.(e) in Na,PO, Solution

The pH of NqPO, solution is lower than NqSiO, solution but higher than NqCO, solution.However, a very strong additional initial peak corresponding to the formation of C%(PO4)2 canstill be observed at about 15 minutes at 50C. Differences in the time to initiation, peak heightand sharpness of the accelerated hydration peak during the hydration of Na.$iO,, Na+ZO, andNa,PO, activated slags may be attributed to the formation mechanism, reaction heat and natureof primary C-S-H, CaCO, and Ca,(PO,),.

It is well known that phosphates can effectively retard the hydration of portland cement.From this study, it seems that the formation of Ca,(PO,), also retards the activation of slag as


13/14

Vol. 25. No. 6 CALORIMETRY. ALKALLSLAG CEhENTS, HYDRATION 1345

usually observed in during the hydration of portland cement. However, the addition of a smallamount of Na,PO., did not retard the setting and activation of slag when NaOH and NqSiO, wereused as activators of phosphorus slag [9].(f) in NaF Solution

For NaF activated slag, no appreciable peaks of heat of evolution occurs after the first 10minutes of hydration at 25C (Fig.@. Tables 2 and 3 indicate that NaF solution has a pH of onlyabout 10 and CaF, has a higher solubility compared with other Ca-compounds. At this pH, notenough slag dissolves to enable appreciable formation of CaF,. At the elevated temperature ofWC, the additional initial peak can be clearly observed but is still very diffuse, and a strong andwide accelerated hydration peak also appears.(g) in NqHPO., Solutions

N%HPO, solution has a pH of about 9 and the solubility product of CaI-IPO., is similar tothat of Ca(OH),. The low pH of N%HPO, solution is not high enough to break enough Ca-0,Si-0 and Al-O bonds for the formation of CaHPO.,, C-S-H, C-A-H and C-A-S-H. Thus NhHPO,solution can not activate the hydration of slag even at 50C.

CONCLUSIONSThis work has examined and discussed the hydration characteristics of ground blast furnace

slag activated by various sodium compounds. The following conclusions are drawn based on theexperimental results and results reported in the literature:

(1) The hydration of alkali-slag cements can be described by three models: Type I - onlyinitial peaks can be identified during the first few minutes and no more peaks appearthereafter; Type II, one peak appears during before the induction period and oneaccelerated hydration peak appears after the induction period; Type III - two peaksappear before the induction period and one accelerated hydration peak appearsafter the induction period.

(2) The initial pH of activator solution has an important role in dissolving the slag forthe early formation of Ca-compounds that result from the reaction between the anionor anion group of the activator and the Ca*+ dissolving from the surface of slaggrains. However, the further hydration of alkali-slag cement is dominated by theCa-Compounds rather than by the initial pH of the activator solution.

(3) The heat evolution rate and the total heat evolution of alkali-activated slag at 24hours can surpass those of portland cement. Thus, alkali-slag cements cannot begenerally regarded as low-heat cements.

REFERENCES1. Tang, X. and Shi, C., Silicate Construction Products (in Chinese), No. 1, pp.28-32, 1988.


14/14

1346 C. Shi and R.L. Day Vol. 25, No. 6

2.3.4.5.6.7.8.9.10.11.12.13.14.15.16.17.18.19.20.21.22.23.24.

25.26.

Shi, C., An Investigation on the Activation of Granulated Phosuhorus Slag, M. SC.Thesis (in Chinese), Nanjing Institute of Technology, Nanjing, P. R. China, 1987.Shi, C., Li, Y. and Tang, X., Journal of Southeast University (in Chinese), Vo1.19,No.1, pp.141-145, Nanjing, P. R. China, 1989.Gluhovsky, V. D., Slag-alkali concretes produced from fine-grained aggregates (inRussian). Vishcha Shkola, Kiev, U.S.S.R., 1981, 223~.Hakkinen, T., Nordic Concrete Research, No.6, pp.81-94, 1987.Malolepszy, J. and Petri, M., Proceedings of 8th International Congress on the Chemistryof Cements, Vo1.4, pp.108-111, Brazil 1986.Douglas, E. and Brandstetr, J., Cem. Concr. Res., Vo1.20, No.5, pp.746-756, 1990,Shi, C., Day, R. L., Wu, X. and Tang, M., Proceedings of 9th International Congress onthe Chemistry of Cement, Vo1.3, pp.298-304, New Delhi, 1992.Shi, C. and Li, Y., Cement and Concrete Research, Vo1.19, No.4, ~~-527-533, 1989.Rajaokarivony-Andriambololona, Z., Thomassin, J. H., Baillif, P. and Touray, J. C.,Journal of Materials Science, Vo1.25, pp.2399-3410, 1990.Adams, L. D., Cement and Concrete Research, Vo1.6, No.3, pp.293-308, 1976.Shi, C. and Li, Y., IL Cemento, Vo1.86, No.3, pp.161-168, 1989.Shi, C., Tang, X. and Li, Y., 11Cemento, Vo1.88, No.4, pp.219-225, 1991.Yuan, R., Gao, Q. and Ouyang, S., Journal of Wuhan University of Technology, No.3,pp.297-303, Wuhan, P.R. China, 1987.Roy, A., Schilling, P. J., Eaton, H. C. Malone, P. G., Brabson, W. N. and Wakely, L. D.,Journal of American Ceramic Society, Vo1.75, No.12 pp.3233-3240, 1992.Gluhovsky, V. D., Rostovkaja, G. S., and Rumyna, G. V., Proceedings of 7th InternationalCongress on the Chemistry of Cements, Vol.III, pp.V-164-168, Paris, 1980.Shi, C. and Day, R. L., Some Factors Affecting Early Hydration of Alkali-Slag cements,to be published.Jawed, I., Skalny, J. and Young, J. F., in Structure and Performance of Cements, ed. byP. Barnes, pp.237-317, Applied Science Publishers, London and New York, 1983.Bensted, J. Advances in Cement research, Vol.1, No.1, pp.35-44, 1987.Teoreanu, I,, IL Cemento, Vo1.8, No.2, pp.91-97, 1991.Yin, Y., Universitv Chemistrv Handbook (in Chinese), Shandong Science and TechnologyPress, Shandong, 1985.Glasser, F. P., Lachowski, E. E. and Macphee, D. E., Journal of American CeramicSociety, Vo1.70, No.7, pp.481-485, 1987.Reardon, E. J., Cement and Concrete Research, Vo1.20, No.2, pp.175-192, 1990.Atkins, M., Bennett, D., Dawes, A., Glasser, F., Kindness, A. and Read, D., AThermodynamic Model for Blended Cements, Research Report for Department ofEnvironment, DoE/HMlP/RR/92/005, 1991.Regourd, M., Proceedings of 7th International Congress on the Chemistry of Cement,Vol. 1, pp.III-2/l l-26, Paris, 1980.Shi, C. and Day, R. L., Early Hydration and Microstructure Formation of Alkali-SlagCement Pastes, to be published.

a calorimetric study of early hydration of

Documents