CEMENT and CONCRETE RESEARCH. Vol. 16, pp. 695-699, 1986. Printed in the USA 0008-8846/86 $3.00+00. Copyright (c) 1986 Pergamon Journa]s, Ltd.
KINETIC STUDY OF THE HYDROTHERMAL REACTION IN CaO-QUARTZ SYSTEM
V.Alujevid, A.Bezjak and A.Glasnovid + Faculty of Pharmacy and Biochemistry, Zagreb, Yugoslavia + Institute of Chemical Engineering, Faculty of Technology, Zagreb, Yugoslavia (Communicated by Z. Sauman) (Received June 25, 1986) ABSTRACT The kinetics of the hydrothermal reaction in CaO-SiO 2 and ~-C~S-SiO~ systems at 2O0°C for initial CaO/SiO? molar ratio I were studie~ by z applying a method that takes into account the effect of simultaneous action of different rate-determining processes. According to the obtained results, at the beginning the reaction rate is determined by the nucleation and growth process, the interaction and diffusion in all the cases. In the later phase the reaction rate is controlled by fast diffusion, except when the reaction of the compressed CaO/SiO? mixture is concerned; in the latter case a period throughout which s~ow diffusion controls the reaction rate precedes the said phase. In all reactions the period of value increase of the diffusion rate constant coincides with the period of xonotlite crystallisation and the period of the value decrease with the period of the crysta]lisation of the CSH(A) phase.
Introduction The article analyzes the kinetics of the hydrothermal reaction of quartz in the mixture with CaO~with CaO/SiO? molar ratio = I at 200°C. One of the quartz samples was of identical particle siz~ as the quartz used to react with ~t-C2S ( i ) , and the text compares not only the reaction kinetics of two samples of quartz differing in granulometry in the powdered and compacted mixture with CaC), but also the kinetics of quartz reaction in the mixtures with CaO and ~"-C2S. The analyses has been carried out by the method taking into account the simultaneous action of the nucleation and growth process, the interaction and the diffusion reaction as rate-determining processes (2). Theoretical Part The method for determining the kinetic data for the systems where the three said mechanisms act simultaneously is based on the comparison of experimentally determined degrees of hydration ~/. (t) and ~ . . ('r) values calculated for a given particle size distribution and foreXv°a'riously assuCr~d transition points from
Vol. 16, No. 5 V. AlujeviE, et al.
one rate-determining process to another. The degree of hydration of a particle with radius R i throughout the nucleation and growth controlled period, the interaction and the diffusion controlled reaction can be represented by the following equations: E-In (t - o t i ) ] t / n
[1 - ( l
-~i )1/3] -El
- (1 -0~1i)1/3]= BRLRi -1. ( i ~ ' - q i )
El - (l - e L i ) l / 3 ] 2 - E l
- (l - o(2i)1/3j 2= c R 2 R [ 2 . (~'-"Z'2i)
In the above equations o~.. and o(^. are transition degrees of hydration at which one rate-determining process subst~utes another. The relative time T equals t/t_ where t is the refe,.'ntial time belonging to the maximum rate of the nucleation O ., and growth process. R / and R n are referential radn of particles for which the transition from the nucleation ~nd growth controlled process to the process controlled by the phase-boundary interaction and to the diffusion process respectively t a kes place at the maximum rate of nucleation and growth. A, B and C are constants dependent on exponent n which amounts to 3 in the present calculations. In the mathematical procedure the transition degrees of hydration o~.. and c~^. and o(.-T mterdependences were calculated for each partmle size group and for v~nous combinations of R L and Rr~ values applying conditions of equal rates of processes which must be satisfied at t~eir exchange points. The total calculated degrees of hydration ( ~ c a t c ) for various ~'values were obtained by •
('~') = ~
wiO(, i ( ~ )
where w i is the weight fraction of the particle with original radius R. The parameters t required for calculating the rate constants for the nucleation t"and growth O process ( k i ) , the interaction (k?) and diffusion (k 3) reactions according the equations 5, 6 and -/ are obtained fro'Pn the linear relationship between ~ ( ~ c a l c . ) and
t (%xp. = °%aic.) k 1 = A/t °
k2=B "RL/t o
k3 = C'RD/t o
In the above equations A = 0.873, B = 0.534, C = 0.213. The analysis of the kinetics of the reaction proceeding throughout the two acceleration periods was carried out applying the described method which was m o dified (3). The analysis was performed independently for the first and second acceleration period. When the linear ~ ( ~ . = c~ ) - t (~ ) relation had been •
established for the hrst acceleration period of r'e~a~tlon, the ~a~e procedure was repeated for the second interval of reaction but with recalculated particle size distribution and with new sets of R I' and R' D values. Prior to calculating ~. ic values within the second acceleration period, the total advancement of react~Bn "for particle R i was determined for particu]aroL~ values by the equation. o~ i = C~gi + ~
( I -OLg i)
16, No. 5
697 Ca0, QUARTZ, HYDROTHERMALREACTIONS, KINETICS
Experimental Procedure CaO was prepared by a two-hour calcination of calcium carbonate of reagent-grade purity up to a temperature of 1000°C. Quartz samples Q. and Q, differing in granulometry were prepared I by the ZsediTABLE ] mentation procedure. The belonging particle size distributions as determined by Coulter Counter method Particle Size Distribution are shown in Table I. CaO and Q1 respectively Q,~ of Ql and G)2 Samples quartz samples were mixed at CaO/SiO? molar ratio = I. Prior to hydrotherma] treatni-ent a part Particle Weight Fraction of the homogeneous ~aO-quartz Q9 mixture was comSize d/AJm Ol G)2 pacted at 300 kg/cm pressure. Thee three prepared reaction mixtures were treated hydrotherma]ly at 2.0 0.03 0.18 200°C under saturated steam pressure. The reaction 5.0 0.05 0.19 of CaO and quartz Q9 was analysed within a 2,5-7 7.0 0.08 0.09 hour period, that of CaO and quartz Q/ within a 9.0 0.09 0.07 5-12 hour period, and the reaction of the compacted l 1.0 O. I 1 0.05 CaO-quartz Q9 mixture within a 8 - ] 7 hour period. 13.0 O. I0 0.04 A f t e r the hycrfothermal treatment the samples were 15.0 0.08 0.04 ground, washed in absolute alcohol and dried in the 17.0 0.08 0.04 atmosphere of nitrogen. The qualitative and quantita19.0 0.08 0.03 tive analysis of the samples was carried out by 2 l.O 0.05 0.03 X-ray diffraction method. The conversion degree of 23.0 0.05 0.02 quartz was determined from the intensity ratio be27.0 0.! I 0.05 tween the diffraction maximum of quartz at d= 3.34~ 35.0 0.07 0.05 and of anatas at d = 3.52 ~ - anatas being used as internal standard - and the loss of ignition. Results and Discussion The analysis results of the kinetics of hydrothermal reaction of the quartz samples Qi and Q2 which differ in granu]ometry, with CaO and a'-C~S are shown in Fig.l,2,3,4. Figs. la, 2a, 3a, 4a show the experimentally determined changes in the conversion degree of quartz related to the reaction time and Figs. Ib, 2b, 3b, 4b the linear ~ ' - t interdependences obtained by given combinations of R j and R D values. The same Figures show also the reaction products formed within ]0articular reaction intervals. Figs. Ic, 2c, 3c, 4c present the distribution of rate-determining processes expressed as the percentage of particles reacting according to the law of nucleation and growth, of phase-boundary reaction and of diffusion respectively. The rate constants for the rate-determining processes within particular reaction intervals are given in Table 2. The reaction of quartz Q] with maximal particle size distribution --,12 AJm in compressed mixture with CaO proceeds throughout two acceleration periods (Fig. ]a, b, c). The induction period of reaction ends after approximately eight hours of hydrotherma] treatment and in the interval between 8 and I].5 hours all the three processes are simultaneously rate-determining. The i reaction rate conslants determined in that pe~io~ of reaction were k I = 0.462 h- , k. = 0 494 ~Jmh- and --l L " k:~ = 0.852 /Jm h . A f t e r ]].5 hours ~he diffusion rate constant (diffu ~!on being the o6]y rate-deterrr~ni~g process within the period from I].5 to ]3.2 hot-'r was reduced to 0.201 jura h- . During the said period the reaction of quartz vv~s very slow. The period of lowering of the diffusion rate constant coincides with the period of CSH(A) phase crystallisation, indicating that the formation of this phase causes lowering in the porozity of the reaction product.
Vol, 16, No. 5 V. A l u ] e v i ~ , et a l .
. . . . . . . . . .
~L-.,~-~-~., . . . . . . . I
^~ XONOTLITE tlh) CSH(~¶ HILLEBRANDITE
)( ONOTL~ C,~"I(A)
"~" " ~ ' ~ 8 9 10 11 12 13 l& 15 16 l'Ttih)
S 6 7 e9
....... b ~ - d i f f u s i o n
m ~ lZ 131~.,~
Compacted mixture of CaO and Q]
Powdered mixture of CaO and Q]
The same effect was noticed also in the reaction of the powdered mixture of identical reactants (Fig. 2b), where the diffusion rate cogst~nt diminished~ , from 2.72 /um~h - ' to 0.962/um~h -I (Table 2) ~vhen CSH(A) phase had appeared in the reaction product. In the compressed mixture of reactants the second acceleration period of reaction began after ]3.2 hours. During that period all the three processes controlled simultaneously the rate of guartz reaction (.~]= 0.359 h - ' , ~ =l.3191umh- , k3= = 4.30/urn h ' t ) . Ttie increased value o f diffusion rate constants indicates an increase in the permeability of the layer of the reaction product due to the formation of crystalline phase. At the beginning of the second acceleration period, xonotlite and hiliebrandite (4) were identified in the reaction product. With the advancement of the reaction, the amount of xonoUite was not changing considerabiy, while the amount of hi]lebrandite was continuously increasing.
The reaction of quartz Qj in the powdered mixture 100 .-,, with CaO began between the 50 "~ ' ". . . . . . . . . . nucleation / ' . . . . interachon fourth and the fifth hour of 50[ : / -r.- : : ,,',~i, __. --diffusion hydrothermal treatment (Figs. 2a,b,c). In the period from 4 8 tlh) to 8.5 hours the nucleation and growth process, the interaction and the diffusion reactions conFIG. 4 FIG. 3 trol and rate of reaction simulPowdered mixture Powdered mixture taneously. After 8.5 hours once of CaO and Q2 of ~'-C2S and Q2 the crystallization of xonotlite 2 -I had begun, ~helvalue of the diffusion rate constant increased from 0.64l /urn h to 2.7l /um h- , but after 10.5 hours in t~e I~eciod of crystallisation of CSH(A) phase its vb]ue was reduced to 0.962 /um~h --. . . . . . . . . Ihf C-S-H XONOTLITE
By comparing Figures 2a,b,c and 3a,b,c it can be noticed that the reaction of t h e finer quartz (Q2) is faster than the reaction of the coarser quartz (QI). The induction period of the reaction ends after 2.5 hours. Within the perioc~ from 2.5 to 5.6 hgurs all the thre% prpcesses are rate-determining (k 1 = 0.51 h - ' , k 9 = =l.23 / u m h - ' , k~ = 1.97 .um"h-'~. ~fer 5.6 hours the reaction rate is controlled only b~/ diffusio¢~ (k~ = 8./31 /urn h- ). The period of value increase of the diffusion rate constant'coincide~ also in this reaction with the period of xonoUite crystallisation.
Vol. 16, No. 5
Ca0, QUARTZ, HYDROTHERMAL REACTIONS, KINETICS
TABLE 2 Rate Constants Samples Constant
Q] + CaO (compacted)
QI + CaO
Q2 + CaO
Q2 + ~"C2S
k?(/umh -] ) - " m 2 h -1,) k3(/u
kg(/umh- l )
k3('/um2h - I )
k~ (/umh- l )
k;('/um2h -I )
The influence of the type of reactant on the kinetics of quartz reaction was analysed by comparing the reaction of quartz Q2 with CaO and with E-CoS (Figs. 3a,b,c and 4a,b,c). For the reaction of quartz and ~C9S three t n reaction intervals were determined. During the period from 5 to 7.5 h'ours the r~action ratle is controlled by the nucleation and growth process and by the interaction (k,=0.28 h- , k~ = 0.]7 ,umh- ). In the second t reaction interval, in addition to the 'said two p~ccesses, /diffusion through the layer of product starts acting as limiting process, too. In that reaction period the semicristailine CSH phases formed at the beginning of the reaction changed to CSH phases similar to tobermorite. While the assessed values of the nucleation and growth and of diffusion rate constants are identical to the values obtained for the first t interval of quartz reaction in the mixture with CaO (Table 2), the interaction r°ate constant is two times lower than the value obtained in the said reaction. At the -,,80% conversion degree of quartz (]0.2 hours of reaction) xonotHte crystallisation began. In that period of reaction diffusion is the only rate-determi~.in~ process; the2val~Je of the diffusion rate constant has increased from ].97 /um h- to 5.24 /urn h . References I. V. Aiujevid and A. Bezjak, Cem. Concr. Res. ]_33, No. l, 33 (]982). 2. A. Bezjak and L Jelenid, Cem. Concr. Res. 10, No. 4, 553 (1980). 3. A. Bezjak and V. Alujevid, Cem. Concr. Res. ]._[l, No. l, ]9 (1981). 4. T. Mitsuda and S. Banno, Cem. Concr. Res. 7, No. 4, 457 (]977).