SP-267—4 Influence of TiO2 Nanoparticles on Early C3S Hydration by B.Y. Lee, J.J. Thomas, M. Treager, and K.E. Kurtis Synopsis: The effect of nano-anatase titanium dioxide (TiO2) powder on early age hydration kinetics of tricalcium silicate (C3S) was investigated. Isothermal calorimetry was performed on C3S pastes with 0, 10, and 15% of TiO2 addition by weight, and two mathematical models–the Avrami model and the boundary nucleation model (BN model)–were fitted to the data. The addition of TiO2 accelerated the rate of hydration, increased the peak reaction rate, and increased the degree of hydration at 12 and 24 hours. The model fits demonstrate that the BN model better captures the kinetics of the reaction, particularly in the deceleration period, than the Avrami model. The increase in the ratio of rate parameters (kB/kG) of the BN model with TiO2 addition suggests that hydration product is formed on or near the surfaces of TiO2 particles, as well as on the C3S surface. These results demonstrate that the addition of TiO2 nanoparticles accelerates the early hydration by providing additional nucleation sites, forming the foundation for future optimization of photocatalytic and other nanoparticle-containing cements. Keywords: Avrami model; kinetics; modeling; nucleation; titanium dioxide. 35 36 Lee et al. ACI member Bo Yeon Lee is a PhD Candidate in the School of Civil and Environmental Engineering at Georgia Institute of Technology, Atlanta, GA, where she also received her MS. She received her BS in architectural engineering from Yonsei University in Korea. Jeffrey J. Thomas is a Research Associate Professor in the Department of Civil and Environmental Engineering at Northwestern University. His research interests include modeling of the nanostructure and hydration kinetics of cement-based materials. Matthew Treager is an undergraduate in the School of Civil and Environmental Engineering at Georgia Institute of Technology in Atlanta, GA. ACI member Kimberly E. Kurtis is Associate Professor in the School of Civil and Environmental Engineering at Georgia Institute of Technology. She is Chair of ACI Committee 236, Materials Science of Concrete, and is a member of ACI Committee 201, Durability of Concrete; E802, Teaching Methods and Educational Materials, and the Educational Activities Committee. INTRODUCTION Tricalcium silicate (3CaO·SiO2 or C3S) comprises about 50 to 70% of ordinary portland cement, and it hydrates to form calcium silicate hydrate (C-S-H) gel, the primary binding constituent in ordinary concrete. It is of primary importance to understand the hydration kinetics of C3S to predict the hydration kinetics of Portland cement (Fujii and Kondo 1974; Tarrida et al. 1995; Berliner et al. 1998; Damasceni et al. 2002). There have been attempts to approximate the kinetics of C3S hydration mathematically, and several mathematical models have been developed. The Avrami nucleation and growth model (Avrami 1939, 1940, 1941) is the most widely used model to predict the early age hydration rate (Brown et al. 1985; FitzGerald et al. 1998; Thomas and Jennings 1999). Recently, Thomas (2007) suggested that the boundary nucleation model (BN model), developed by Cahn (1956) to describe a solid-solid phase transformation in a polycrystalline material, could better approximate the process of C3S hydration than the Avrami model. While the Avrami model assumes that nucleation occurs randomly throughout the transforming volume, the BN model assumes that nucleation is favored at grain boundaries, similar to the observation that C3S hydration occurs by the outward growth of hydration products from the surface of the hydrating grains into the surrounding water-filled porosity. The BN model takes into account the effect of the surface area of the starting material, which, in the case of hydrating cement or C3S, is known to have a strong effect on the hydration kinetics. In this research, early hydration of C3S was examined with two different addition levels of insoluble, nanoanatase titanium dioxide (TiO2) particles along with a control C3S without TiO2. The anatase form of TiO2 is known for its photocatalytic smog-abating, self-cleaning, and biocidal capacities, which are especially efficient in nanocrystalline form (Carp et al. 2004; Fujishima and Zhang 2006). The addition of TiO2 to cement-based materials is increasingly being researched and has been used in applications in various parts of the world (Beeldens 2007; Kawakami et al. 2007; Poon and Cheung 2007; Jayapalan et al. 2009). The objective of this research is to verify whether the surface area provided by nano-sized TiO2 affects the cement hydration rate and whether such an effect can be captured by mathematical modeling. The model fit results from both the Avrami model and the BN model will be compared, and the role of TiO2 in early age cement hydration will be discussed. RESEARCH SIGNIFICANCE Photocatalytic cementitious materials hold promise in the construction industry due to their potential for smog-abatement, self-cleaning abilities, and biocidal applications (Carp et al. 2004; Fujishima and Zhang 2006). However, while practical use of these materials continues to grow, little has been published describing the influence of photocatalytic nanoparticles on early age behavior of such cementitious materials (Jayapalan et al. 2009). This research examines the influence of photocatalytic titanium dioxide dosage rates on early age hydration of C3S by isothermal calorimetry, and compares it with two existing models on hydration kinetics. The aim is to better understand the influence of the nanoparticles on hydration of Portland cement, forming the basis for anticipating the influence of these nanoparticles on setting time, rate of strength gain, and durability. THEORETICAL BACKGROUND For the purpose of this research, it is useful to first review the hydration models considered. 1 THEORETICAL BACKGROUND 2 of Concrete: Thereview NexttheBig Thing Is Small 3 ForNanotechnology the purpose of this research, it is useful to first hydration models considered.37 4 Avrami nucleation growth model and growth model 5 and Avrami nucleation The theory was first treated by Kolmogorov (1937), Johnson and Mehl (1939), Avrami (1939, 1941) 6 The theory was first treated by Kolmogorov (1937),and Johnson and Mehl1940, (1939), andtoAvrami (1939, 19 explain the kinetics of phase change of metals. The main assumptions made were that the new phase is nucle7 1941) to explain the kinetics of phase change of metals. The main assumptions made were that the new phase 8 and nucleated by grain germ centers nuclei and thatnew the grain of the new phase arethroughout randomly distributed ated by germ nuclei that the of the phasecenters are randomly distributed the matrix.throughout the mat 9 mathematical Due to its simple mathematical theory has been adapted for C S Due to its simple form, the theory hasform, been the widely adapted for Cwidely S hydration (Brown et al. 1985; (Brown et al. 19 3 hydration 3 101998;FitzGerald et al.Jennings 1998; Thomas final forms theintroduced equations will FitzGerald et al. Thomas and 1999). and OnlyJennings the final1999). formsOnly of thethe equations willofbe here.be introduced here. 11 volume The transformed volume of fraction X asbea written functionas, of time can be written as, The transformed fraction X as a function time can 12 n (1) 13 X = 1 − exp[− (k avrX t=) 1] – exp –(kavrt)n 14 where the effective constant kavr, as rate proposed by kAvrami, is a function of constant linear growth rate, G, and 15 rate where the effective constant avr, as proposed by Avrami, is a function of constant linear growth rate, G, a either of the rate of nucleation per unit of untransformed volume, Iv, or the number of nuclei pernumber unit volume, Nv0,per unit volume, N 16 either of the rate of nucleation per unit of untransformed volume, Iv, or the of nuclei and n is an exponent that depends on the dimensionality of the product that forms. 17 and n is an exponent that depends on the dimensionality of the product that forms. The hydration canhydration be obtained differentiating Eq. (1) respect to time, berespect writtentoas, 18 rate, whichThe rate,by which can be obtained bywith differentiating Eq. (1)can with time, can be written as, 19 n n n = Ankn n(t t0)n–1 (2) 20 R = AnkRavr (t − t 0 avr ) −1 –exp( −[exp(–[k k avr (t −avr t 0(t)]–)t0)] ) 21 where R is the22 hydration is ahydration normalization to match constant isothermal calorimetry data,calorimetry and t0 is the whererate, R isAthe rate, Aconstant is a normalization to match isothermal data, and t0 is the ti time delay between the time of mixing and the start of nucleation and growth kinetics. Note that there 23 delay between the time of mixing and the start of nucleation and growth kinetics. Note are thatfour there are four paramet parameters varying, which are A, kavr , t0A,and Avrami been to describe C3S C S hydration, 24 varying, which are kavrn. , t0While and n.the While the model Avramihas model hasused beenwidely used widely to describe 3 hydration, the25 heterogeneous nucleation process that occurs during cement hydration may not be best described heterogeneous nucleation process that occurs during cement hydration may not be best described by a homogene by a homogeneous (Garraultetetal. al.2006; 2006; Thomas 2007). 26 nucleation nucleation assumption assumption (Garrault Thomas 2007). 27 Boundary nucleation model nucleation model 28 Boundary The boundary nucleation andboundary growth model (BN model) wasmodel first developed bywas Cahn (1956), to describe the 29 The nucleation and growth (BN model) first developed by Cahn (1956), to describe kinetics of nucleation and growth of a polycrystalline material. The key assumption is that nucleation is permitted 30 kinetics of nucleation and growth of a polycrystalline material. The key assumption is that nucleation is permitted to occur only on boundaries, unlike the Avrami model assumes that nucleation occurs at nucleation randomly occurs at random 31internal occur only on internal boundaries, unlikewhich the Avrami model which assumes that 32 distributed locations everywhere within volume. the untransformed Fortransformed the BN model, the transformed volu distributed locations everywhere within the untransformed For the BNvolume. model, the volume 33 fraction function of time, X, is given by: fraction as a function of time,asX,a is given by 34 35 36 ( ( )) Gt X = 1 − exp ⎡− 2OvB ∫ 1 − exp − Y e dy ⎤ ⎢⎣ ⎥⎦ 0 πI B 2 3 ⎡ 3 y 2 2 y 3 ⎤ (if t > y/G) e where Y = 3 G t ⎢1 − 2 2 + 3 3 ⎥ G t ⎦ ⎣ G t (3) 37 (if t < y/G) Ye = 0 38 Notetemporary that y and variables Ye are temporary variablesafter that integration. disappear after integration. In Eq. 3, depends Note that y 39 and Ye are that disappear In Eq. (3), X depends onXjust threeon just three well40 defined physical parameters: G, the linear growth rate of transformed phase, I , the nucleation rate per unit area o B well-defined physical parameters: G, the linear growth B rate of transformed phase, IB, the nucleation rate per unit 41 untransformed boundary, and O , the boundary area per unit volume. v B area of untransformed boundary, and Ov , the boundary area per unit volume. 42 In order to apply the BN model to C3S hydration rate data, some slight modifications must be made (Thom In order to apply the BN model to C3S hydration rate data, some slight modifications must be made (Thomas 43 2007). Eq. (3) must be numerically integrated to obtain the transformation rate, dX/dt. Also, the scaling parame 2007). Eq. (3) must be numerically integrated to obtain the transformation rate, dX/dt. Also, the scaling param44 A and the time constant t0, described above in conjunction with the Avrami model (Eq. 2), are introduced. eter A and the time constant t0, described above in conjunction with the Avrami model (Eq. (2)), are introduced. 45 The threeB physical parameters OvB, IB, and G are correlated such that the kinetic profile is described by t The three physical parameters O , I , and G are such by thatThomas the kinetic profile is described by two v B 46 independent rate constants kB andcorrelated kG as proposed (2007): independent rate constants kB and kG as proposed by Thomas (2007): 1/ 4 47 k B = (I B OvB ) G 3 / 4 kB = (IBOvB)1/4G3/4 48 kG = OvB G (4) k = O BG 49 Thus fits made with the BN model Ghave vfour varying parameters: A, t0, kB, and kG. If any one of the physi 50 parameters in Eq. 4 are known independently, then the other two can be calculated from the fitted values of the r Thus fits made with the BN model have four varying parameters: A, t0, kB, and kG. If any one of the physical parameters in Eq. (4) are known independently, then the other two can be calculated from the fitted values of the rate constants. For the case of a hydrating C3S paste, the value of OvB can be calculated by dividing the measured surface area of the C3S powder by the calculated volume occupied by the3hydration products after complete hydration, allowing G and IB to be determined from the fits. 38 Lee et al. Each of the rate constants represents different physical behavior: kB describes the rate of transformation on the internal boundaries (the surface of particles), whereas kG describes the rate of transformation in the bulk matrix (the pore space between the particles). The ratio of these two rate constants (kB/kG) can be used to identify the type of kinetic behavior. If kB/kG is very large, then the hydrated products will be densely populated on or very near the nucleation sites (i.e., the C3S surface in the case of C3S hydration). If kB/kG is very small, then the hydrated products will form evenly throughout the paste, approaching the condition of the Avrami model. EXPERIMENTAL PROCEDURE Materials Pure tricalcium silicate (C3S) powder was obtained from the Lafarge Research Center in Lyon, France. The chemical composition of C3S was analyzed by quantitative x-ray diffraction (QXRD) using Cu-Ka radiation, which gave C3S = 98.43%, C2S = 1.46%, and CaO = 0.10% by mass fraction (Fig. 1). Particle size distribution was measured in ethanol slurry using Microtrac X-100 laser particle analyzer (Fig. 2). The median diameter of the C3S sample was measured to be 7.754 mm. Using an empirical relationship between surface area of cement and particle size distribution developed by Zhang and Napiermunn (1995), the surface area of C3S was found to be 0.291 m2/g (1420 ft2/lb). The anatase titanium dioxide (AMT-100, Tayca Corp.) used was 93% pure with an average crystal size of 6 nm and a pH of 7. The manufacturer-provided surface area was 280 m2/g (1.37 × 106 ft2/lb). Experiment Three different C3S pastes were prepared at 0%, 10%, and 15% addition rate of TiO2 by mass; note that the TiO2 was dosed in addition to, not by weight replacement, for cement. This approach keeps the cement content constant among the different pastes, which simplifies interpretation of calorimetry data. The water to cement ratio (w/c) was kept constant at 0.50, which resulted in a stiffer mix as more of the high surface area TiO2 was added. The rate of hydration was measured by isothermal calorimetry (TAM AIR, TA instruments) at 20°C (68°F), which has precision of ±20 mW and accuracy greater than 95%. All the materials were equilibrated at 20°C (68°F) for 24 hours prior to the testing. The TiO2 was hand mixed with deionized water for 1 minute, and then the C3S was added and hand mixed for another 2 min. Samples were placed into the calorimeter less than 5 minutes after mixing with C3S. The calorimetry data from the initial 15 minutes after mixing were excluded for both of the model fits as some time is required for the samples to be equilibrated within the instrument. RESULTS AND DISCUSSION Figure 3 shows the rate of hydration per gram of C3S for the first 40 h after mixing for each of the pastes examined; corresponding cumulative heat data for the first 80 hours is presented in Fig. 4. Recall that the C3S content and w/c for each paste remains constant, while the dosage of TiO2 nanoparticles varies among 0, 10, and 15%. As the addition level of TiO2 nanoparticles increased, the rate of hydration was accelerated. For example, the rate peak (or greatest power evolved with time) occurs 123 min. earlier for the 10% TiO2 paste and 200 min. earlier for 15% TiO2 compared to the peak for the ordinary paste (Fig. 3). Furthermore, the peak heights in the rate curve occurring at 6.5, 7.7, and 9.8 hours after mixing, were increased by 11.0% and 11.9%, respectively, for 10 and 15% TiO2 paste, when compared to pure C3S paste. The total heat of hydration (Fig. 4) data demonstrate that the pastes containing TiO2 experience greater heat evolution than the ordinary paste in the first 20 hours before the energy evolved begins to slow. In comparison, the 0% TiO2 (pure C3S) paste releases energy more slowly initially, but continues to steadily increase through the 80 hours of data collected. These data suggest that the addition of TiO2 particles to cement paste may affect early age hydration by decreasing the time to set and increasing the rate of early strength development. The degree of hydration, a, of a C3S paste can be calculated by dividing the heat of hydration at a given time by the enthalpy of reaction of C3S, which has been reported to be DH = –121 kJ/mol (Thomas et al. 2009). The values of a at 12 and 24 hours for each paste are listed in Table 1. The increasing a with higher levels of TiO2 shows that these pastes have achieved greater degrees of hydration by 12 and 24 hours, as suggested by the faster hydration rate described above. In particular, increases in a at 12 hours of 42.86 and 51.43%, respectively, for 10 and 15% TiO2 when compared to the reference paste are notable. By 24 hours, the acceleratory effect of the TiO2 is less than at 12 hours; increases in a at 24 hours are 19.30 and 21.05%, respectively, at 10 and 15% TiO2, when compared to the reference C3S paste. It should be recalled that the TiO2 particles added to these pastes are not known to react with water or calcium silicates. Therefore, an alternative explanation for the observed increase in early age C3S reaction in the presence of these nanoparticles must be sought. Nanotechnology of Concrete: The Next Big Thing Is Small 39 Both the Avrami model and the BN model were fit to the experimental data for each of the samples (see Fig. 5 to 7). Fitting was performed manually using least mean square error to obtain the closest fit around the rate peak. The two rate parameters kB and kG for the BN model were first derived with a fixed OvB calculated from the surface area of the C3S powder (OvB =0.44 m2/cm3 [77.6 ft2/in.3]) and then G and IB for different pastes were calculated using the relationships in Eq. (4). The fit parameters are listed in Table 2 for the Avrami model and in Table 3 for the BN model. Figures 5 to 7 show the rate of hydration data with both fits for each of the three pastes examined. For the C3S paste, the BN model clearly provides a better fit than the Avrami model to the hydration rate behavior, as was previously shown for C3S hydration under different conditions by Thomas (2007). While both models closely approximate the early hydration reaction of C3S in the acceleration period (for all of the pastes), the BN model more closely captures the hydration behavior of the deceleration period after the rate peak. The divergence with the Avrami model occurs in the deceleration period and suggests that the conditions of random nucleation assumed by that model do not apply to pure C3S hydration. The BN model also better represents the early age hydration of C3S mixed with TiO2 than the Avrami model. Again, the improvements in the fit are especially notable in the down slope region. This suggests that nucleation is spatially nonrandom and is likely related to the surface area of the solid phases available for product nucleation and growth. Further, an increase in the kG/kG ratio of the BN model is found as the TiO2 dosage (and solid surface area) increases (Table 3). This suggests that the presence of additional surface area, provided by the increasing amounts of TiO2, promotes hydration product formation, which occurs on or near the surfaces of the particles according to the model assumptions. This implies that the addition of TiO2 powder supplied additional nucleation sites to accelerate the hydration. It is also proposed that by providing additional nucleation sites away from the reacting C3S particles, the start of diffusion-controlled hydration kinetics is delayed, increasing the amount of early nucleation and growth hydration, as has been proposed for other nucleating materials (Thomas et al. 2009). Such behavior would account for both the acceleration in reaction rate and the increase in peak height observed in the presence of TiO2 nanoparticles. The t0 parameter in both the Avrami model and the BN model graphically shifts the curves either to the right or to the left, depending on the sign. A positive t0 value can be interpreted as an induction period, shifting the curves to the right. On the other hand, a negative t0 found in the BN model fits suggests that the TiO2 particles have an effect of promoting the nucleation rate at a very early age. Note that the differences seen in t0 between the 0% paste and the 10 and 15% pastes in the case of the BN model are 2.2 and 3.0 hours, respectively, which are on the order of magnitude of the acceleration noted previously with respect to the peak heights seen in Fig. 3. CONCLUSIONS The early hydration behavior of pure C3S was measured and compared to that of C3S pastes containing 10% and 15% additions of high surface area TiO2. Two nucleation and growth models–the Avrami model and boundary nucleation model–were applied to the experimental data. Based on the results of this study, the following conclusions are drawn: • The addition of high surface area TiO2 powder accelerates the early age hydration of C3S by ~2.0-3.3 hours and increases the rate peak height by ~11-12%, an effect that is attributed to its ability to stimulate nucleation of hydration product. • The BN model is capable of representing the kinetic behavior of C3S paste mixed with TiO2 nano-particles and provides a better fit to the rate data than the Avrami model for all pastes tested. • The increase in the ratio of the rate constants from the BN model (kB/kG) with addition of TiO2 suggests that the formation of hydration products is linked to the increase in available nucleation sites (i.e., increase in solid surface area) provided by the TiO2 nanoparticles. Observations such as these which show that the early rate of hydration can be altered through the addition of unreactive TiO2 particles suggest that the setting behavior, strength development, and permeability of photocatalytic and other Portland cements can be optimized by controlling compositional variables and particle size. ACKNOWLEDGMENTS This material is based upon work supported by the National Science Foundation under Grant No. CMMI-0825373. The authors are grateful to Lafarge for providing the C3S used. The authors thank Matthew Treager and Sarah Fredrich for their help. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation. 40 Lee et al. REFERENCES Avrami, M., “Kinetics of Phase Change 1—General Theory,” Journal of Chemical Physics, V. 7, No. 12, 1939, pp. 1103-1112. Avrami, M., “Kinetics of Phase Change 2—Transformation-Time Relations for Random Distribution of Nuclei,” Journal of Chemical Physics, V. 8, 1940, pp. 212-224. Avrami, M., “Kinetics of Phas Change 3—Granulation, Phase Change and Microstructure,” Journal of Chemical Physics, V. 9, No. 2, 1941, pp. 177-184. Beeldens, A., “Air Purification by Road Materials: Results of the Test Project in Antwerp,” Proceedings of the International RILEM Symposium on Photocatalysis, Environment, and Construction Materials, 2007. Berliner, R.; Popovici, M.; Herwig, K.W.; Berliner, M.; Jennings, H.M.; and Thomas, J.J., “Quasielastic Neutron Scattering Study of the Effect of Water-to-Cement Ratio on the Hydration Kinetics of Tricalcium Silicate,” Cement and Concrete Research, V. 28, No. 2, 1998, pp. 231-243. Brown, P.W.; Pommersheim, J.; and Frohnsdorff, G., “A Kinetic Model for the Hydration of Tricalcium Silicate,” Cement and Concrete Research, V. 15, No. 1, 1985, pp. 35-41. Cahn, J. W., “The Kinetics of Grain Boundary Nucleated Reactions,” Acta Metallurgica, V. 4, No. 5, 1956, pp. 449-459. Carp, O.; Huisman, C.L.; and Reller, A., “Photoinduced Reactivity of Titanium Dioxide,” Progress in Solid State Chemistry, V. 32, No. 1-2, 2004, pp. 33-177. Damasceni, A.; Dei, L.; Fratini, E.; Ridi, F.; Chen, S.H.; and Baglioni, P., “A Novel Approach Based on Differential Scanning Calorimetry Applied to the Study of Tricalcium Silicate Hydration Kinetics,” Journal of Physical Chemistry B, V. 106, No. 44, 2002, pp. 11572-11578. FitzGerald, S.A.; Neumann, D.A.; Rush, J.J.; Bentz, D.P.; and Livingston, R.A., “In Situ Quasi-Elastic Neutron Scattering Study of the Hydration of Tricalcium Silicate,” Chemistry of Materials, V. 10, No. 1, 1998, pp. 397-402. Fujii, K., and Kondo, W., “Kinetics of Hydration of Tricalcium Silicate,” Journal of the American Ceramic Society, V. 57, No. 11, 1974, pp. 492-497. Fujishima, A., and Zhang, X.T., “Titanium Dioxide Photocatalysis: Present Situation and Future Approaches,” Comptes Rendus Chimie, V. 9, No. 5-6, 2006, pp. 750-760. Garrault, S.; Behr, T.; and Nonat, A., “Formation of the C-S-H Layer During Early Hydration of Tricalcium Silicate Grains with Different Sizes,” Journal of Physical Chemistry B, V. 110, No. 1, 2006, pp. 270-275. Jayapalan, A.R.; Lee, B.Y.; and Kurtis, K.E., “Effect of Nano-Sized Titanium Dioxide on Early Age Hydration of Portland Cement,” Nanotechnology in Construction 3, 2009. Johnson, W.A., and Mehl, R.F., “Reaction Kinetics in Processes of Nucleation and Growth,” Transactions of the American Institute of Mining and Metallurgical Engineers, V. 135, 1939, 416 pp. Kawakami, M.; Furumura, T.; and Tokushige, H., “NOx Removal Effects and Physical Properties of Cement Mortar Incorporating Titanium Dioxide Powder,” Proceedings of the International RILEM Symposium on Photocatalysis, Environment, and Construction Materials, 2007, pp. 163-170. Kolmogorov, A.N., “A Statistical Theory for the Recrystallization of Metals,” Bulletin of the Academy of Sciences of the USSR: Physics Series, V. 1, 1937, 255 pp. Poon, C. S., and Cheung, E., “NO Removal Efficiency of Photocatalytic Paving Blocks Prepared with Recycled Materials,” Construction and Building Materials, V. 21, No. 8, 2007, pp. 1746-1753. Tarrida, M.; Madon, M.; Lerolland, B.; and Colombet, P., “An In-Situ Raman-Spectroscopy Study of the Hydration of Tricalcium Silicate,” Advanced Cement Based Materials, V. 2, No. 1, 1995, pp. 15-20. Thomas, J.J., “A New Approach to Modeling the Nucleation and Growth Kinetics of Tricalcium Silicate Hydration,” Journal of the American Ceramic Society, V. 90, No. 10, 2007, pp. 3282-3288. Thomas, J.J., and Jennings, H.M., “Effects of D2O and Mixing on the Early Hydration Kinetics of Tricalcium Silicate,” Chemistry of Materials, V. 11, No. 7, 1999, pp. 1907-1914. Thomas, J.J.; Jennings, H.M.; and Chen, J.J., “Influence of Nucleation Seeding on the Hydration Mechanisms of Tricalcium Silicate and Cement,” Journal of Physical Chemistry C, V. 113, No. 11, 2009, pp. 4327-4334. Zhang, Y.M., and Napiermunn, T.J., “Effects of Particle-Size Distribution, Surface-Area and Chemical-Composition on Portland-Cement Strength,” Powder Technology, V. 83, No. 3, 1995, pp. 245-252. Table 1—Cumulative heat of hydration and degree of hydration, a, of TiO2-blended C3S pastes at 12 and 24 hours TiO2 added (%) Cumulative heat (12 h), kJ/mol a (12 h) 0 10 15 42.65 60.15 64.70 0.35 0.50 0.53 Cumulative heat (24 h), kJ/mol 68.76 82.13 83.85 a (24 h) 0.57 0.68 0.69 Nanotechnology of Concrete: The Next Big Thing Is Small Table 2—Fit parameters A, t0, kavr and n for Avrami model TiO2, % A, kJ/mol t0, h kavr, h–1 n 0 10 15 56.718 64.116 59.184 2.700 0.180 0.042 0.1190 0.1144 0.1320 2.90 2.97 2.80 Table 3—Fit parameters A, t0, kG, kB and kB/kG for boundary nucleation model TiO2, % A, kJ/mol t0, h kB, h–1 kG, h–1 kB/kG IB (mm–2h)–1 G, mm/h 0 10 15 71.514 85.488 87.954 1.2 -1.0 -1.8 0.1030 0.0990 0.1043 0.0824 0.0734 0.0656 1.2500 1.3488 1.5899 0.0390 0.0471 0.0811 0.1871 0.1666 0.1490 Fig. 1—Diffraction pattern for cement sample compared to reference pattern for C3S. Fig. 2—Particle size distribution and cumulative particle size of C3S powder examined. 41 42 Lee et al. Fig. 3—Hydration rate of TiO2-blended C3S pastes. Fig. 4—Cumulative heat of hydration of TiO2-blended C3S pastes. Nanotechnology of Concrete: The Next Big Thing Is Small Fig. 5—Rate of hydration with 0% TiO2 – Experiment versus Avrami and BN models. Fig. 6—Rate of hydration with 10% TiO2 – Experiment versus Avrami and BN models. 43 44 Lee et al. Fig. 7—Rate of hydration with 15% TiO2 – Experiment versus Avrami and BN models.

© Copyright 2019