chemical engineering research and design 8 8 ( 2 0 1 0 ) 1450–1454 Contents lists available at ScienceDirect Chemical Engineering Research and Design journal homepage: www.elsevier.com/locate/cherd The effects of organic additives on induction time and characteristics of precipitated calcium carbonate Raluca Isopescu a,∗ , Carmencita Mateescu b , Mihaela Mihai a , Gabriel Dabija a a b University Politehnica of Bucharest, 1 Gh Polizu Str., 011061 Bucharest, Romania National Institute for Materials Physics, Magurele, Romania a b s t r a c t The effects of an organic additive, polyoxyethylene sorbitan trioleate (Tween 85), on the induction time for the precipitation of calcium carbonate are experimentally and theoretically investigated. Calcium carbonate was precipitated from aqueous solutions of K2 CO3 and Ca(NO)3 at moderate supersaturations ranging between 5 and 16 with and without the organic additive. Experimentally it has been noticed that the induction period for CaCO3 precipitation increases at low supersaturation and is also influenced by temperature. An increase of the induction time was noticed when Tween 85 was added in the system. The “cluster coagulation model” proposed by Qian and Botsaris (1997), which combines nucleation models and coagulation theory, was used to explain the effects of operating parameters on the induction time in terms of interfacial energy and cluster sizes. © 2009 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved. Keywords: Calcium carbonate; Induction time; Homogeneous nucleation; Coagulation theory 1. Introduction Precipitated calcium carbonate is a material of great interest due to its large range of applications: as polymer composite, rubber-filler, additive for plastics and paper. Calcium carbonate morphology depends on the precipitation conditions such as: supersaturation, pH, presence/absence of additives, temperature, mixing and the operating system (batch or continuous). The use of additives, especially complexing agents, is often crucial for tailoring the crystal habit and purity. Low concentrations of these agents influence the kinetics of nucleation and crystal growth. It is assumed that the additives have two functions: they inhibit crystal growth by binding themselves to the growth sites of the crystals; they enhance the heterogeneous nucleation, controlling and stabilizing the precipitated polymorph. It is generally accepted that the effect of minor amount of additives on the precipitation kinetics of mineral salts can be attributed to their preferential adsorption on the solid surface (Reddy and Hoch, 2001). One of the basic quantities which can be defined for the precipitation of sparingly soluble salts is the induction period defined as the time interval between the onset of supersaturation and the formation of critical nuclei. The induction time is affected by many external factors, but it cannot be regarded as fundamental property of the system. The presence of organic additives in the system can considerably affect the induction period, but it is impossible to predict it. The presence of impurities or additives increase the induction period of calcium carbonate precipitation, others cause its decrease or have no effect (Westin and Rasmuson, 2005; Tai and Chien, 2003). The effects are interpreted by a change of equilibrium solubility or the solution structure, by adsorption or chemisorption on nuclei and by chemical reaction or complex formation in the solution (Mullin, 2001). Some experimental methods can be applied to determine the induction period: conductivity method, pH method (Gomez-Morales et al., 1996), activity of precipitated ions method (Verdoes et al., 1992). The aim of this paper is to study the influence of operating parameters, such as supersaturation and temperature on the induction time for calcium carbonate precipitation in absence/presence of an organic additive. The “cluster coagulation model” (Qian and Botsaris, 1997) was used to get a better insight of the dependence of induction time on experimental conditions based on the nucleation theory. According to classical theory of nucleation in a supersaturated solution the monomer start to coagulate and form Corresponding author. E-mail address: r isopescu@chim.upb.ro (R. Isopescu). Received 31 January 2009; Received in revised form 30 September 2009; Accepted 1 October 2009 0263-8762/$ – see front matter © 2009 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.cherd.2009.10.002 ∗ 1451 chemical engineering research and design 8 8 ( 2 0 1 0 ) 1450–1454 calculated as: Nomenclature n1 = Cs D Gg gc g k n1 ng r2 rg S T Vm number of solute molecules per unit volume (m−3 ) diffusivity of clusters (m2 s−1 ) total excess free energy for the formation of gmer cluster (J) number of monomers in a critical nuclei dominant cluster size in solution Boltzmann constant (J/K) number concentration of monomers in the liquid phase (m−3 ) number concentration of g-mer in solution (m−3 ) correlation coefficient radius of g-mer (m) supersaturation temperature (K) volume of monomer (m3 ) 1 8Drg  1 ng  gc −1 g  Gg = (4)  G  g kT  = 2 ı3 32Vm 3(kT ln S) 3 ı(3Vm g) 2/3 − g(kT ln S)  3   4kT × n−1 exp −g ln S + 1 2 ı3 32Vm 3g(kT ln S) (5) 3 ı 1/3 2/3 (4) (3gVm ) kT −1  (6) Eq. (6) can be used for investigating the influence of supersaturation, temperature and additives on the induction period. 2. (1) (2) Through free energy arguments a critical cluster size, gc can be defined to represent the size above which stable nuclei are formed and grow to form a crystal. gc = 1/3 Combining relations (2)–(5) relation (1) can be rewritten as: In this expression  is induction time, D = kT/6rg is the diffusivity of clusters, g is the clusters dominating size, rg is the radius of g-mer, ng is the number concentration of g-mer,  is the viscosity of the liquid phase, k is the Boltzmann constant, T is the absolute temperature. From thermodynamic considerations based on Gibbs theory further equations can be derived. The number concentration of solute g-mers, ng can be evaluated considering the free energy of formation: ng = n1 exp − (4) g(ng /n1 ) where S is the supersaturation and Cs represents the number of molecules per unit volume. The total excess free energy, Gg for the g-mer formation is expressed by: clusters (Qian and Botsaris, 1997). When the size of a cluster exceeds a critical size a nucleus is formed and the growth of the nucleus leads to a crystal. The variation of cluster size distribution can be described by the Smoluchowski theory of coagulation. Initially, only g-mers are present in solution and coagulate to form 2g-mers and so on until gc -mers are formed. In this context three terms: diffusivity, coagulation concentration and critical nuclei size contribute at the evaluation of induction time (Qian and Botsaris, 1997):  SCs g=1 Greek symbols ı interfacial tension of crystal (mJ m−2 )  viscosity of the liquid phase (Pa s)  induction time (s)  exp experimental induction time (s)  comp computed induction time (s) = gc (3) In the above relation, Vm is the volume of a monomer and ı is the interfacial tension of the crystal. The number concentration of monomer, n1 in relation (2) depends on the initial concentration of the solute (Qian and Botsaris, 1997), and is Experimental study Calcium carbonate was precipitated from aqueous solutions of calcium nitrate and potassium carbonate. All chemical reagents were of analytical grade. The initial solutions were filtered through a 0.22 ␮m membrane filter. The temperature was maintained constant by a thermostat bath. The vessel was tightly closed to minimize the exchange of carbon dioxide between the reaction system and the environment. Supersaturation varied in a wide range from 5 to 16 and values for working temperature: 303 K, 318 K, 333 K and 348 K. A nonionogen organic additive-polyoxyethylene sorbitan trioleate (Tween 85), was used. The organic additive was mixed in calcium nitrate solution in a quantity corresponding to critical micellar concentration CCM = 5.2 × 10−5 mol L−1 . To estimate the induction time the conductometric method was used. The procedure for conducting induction period experiments was as follows. A certain quantity of water and 0.1 M K2 CO3 solution were poured into the glass beaker and mixed by a magnetic stirrer (700 min−1 ). After the solution temperature became steady and the conductivity remained constant, a certain quantity of 0.1 M Ca(NO3 )2 solution was added. The solution conductivity jumps to a high level until the mixing is complete. The conductivity remains constant for a certain period, after which it starts to decrease due to the formation of critical nuclei. The solution became turbid as detected by naked eye and in this moment the experiment was stopped. This is considered the “induction period”. For each series of experiments 20 replicates were performed, and the highest and the smallest values were eliminated. After the induction period the suspensions were left 20 min in contact with mother liquor for ageing and ripening. The solid phase was analyzed from the morphological point of view. The final solid phase was analyzed by optical microscopy and by infrared spectra using a PerkinElmer FT-IR spectrophotometer. 3. Results and discussions The experimental results show that, as expected, the induction period, at constant temperature, decreases with increased supersaturation (Fig. 1). In the presence of additive Tween 85 (Fig. 2), at high supersaturation levels the induction times have 1452 chemical engineering research and design 8 8 ( 2 0 1 0 ) 1450–1454 Fig. 1 – Induction period variation with supersaturation for calcium carbonate precipitation without additive. Fig. 3 – log() vs. 1/[log(S)]2 for 303 K. comparable values, while at low supersaturations the value of the induction period is much higher than in the absence of additive. Consequently the presence of the Tween 85 retards the nucleation process. According to Mullin (2001), the classical nucleation relationship between the induction time, supersaturation and temperature is written as: log() ∝ 1 T 3 log (S) 2  (7) At constant temperature the induction time for precipitation is expected to obey a linear relation ship: log() = B 1 2 [log(S)] +C (8) The broken line (Fig. 3) that approximates the variation of induction time proves that there is a change in nucleation mechanism reflected by a steep decrease of the induction time at high supersaturation. Consequently a single linear correlation cannot be derived in terms of classical nucleation theory. For calcium carbonate induction time some other equations were derived (Stamatakis et al., 2005) mainly valid for low supersaturation values. The attempt to correlate our data for a Fig. 4 – Correlation between experimental data and coagulation model for induction period variation with supersaturation without Tween 85. large experimental region leads to rather complex expressions in order to ensure a great accuracy. For the precipitation of calcium carbonate without additives a good correlation is given by relation (9). ln  = −27.66 + 0.299S + 0.180T − 0.005S2 − 0.0003T 2 − 0.0008ST (9) The correlation coefficient, r2 , for relation (9) is 0.988. For the induction times corresponding to the presence of Tween additive relation (8) was derived: ln  = −21.667 − 0.039S1.5 + 0.00075T 2 + 1.5 × 10−6 T 3 Fig. 2 – Induction period variation with supersaturation for calcium carbonate precipitation in presence of Tween 85. (10) The correlation coefficient, r2 , for relation (10) is 0.986. These correlating relations which may be used in practical applications, described very well the evolution of experimental data (Figs. 4 and 5) but they provide no insight of the nucleation process. The influence of supersaturation and temperature on the induction time was also analyzed in terms of the coagulation model presented above (Eqs. (2)–(6)). The interfacial tension and dominant cluster size were estimated by fitting the experimental induction times,  exp with computed values,  comp by 1453 chemical engineering research and design 8 8 ( 2 0 1 0 ) 1450–1454 Fig. 5 – Correlation between experimental data and coagulation model for induction period variation with supersaturation in presence of Tween 85. nonlinear regression. The objective function was: F = min n  2 (exp − comp ) (11) i=1 The objective function minimized using the Luus–Jaakola random search method recommended for multimodal objective functions, which is the general case when solving parameter estimation problems. The dominant cluster size, g was explored for integer values between 1 and 6 and the interfacial tension searching range was 30–90 mJ m−2 , according to similar values reported in literature (Pokrovsky, 1998; Mullin, 2001). In the estimation of the theoretical induction time by means of relation (6), the viscosity of water was considered and the calcite molecular volume (Vm = 6.13 × 10−29 m3 /molecule) as it is the most insoluble calcium carbonate polymorph phase (Westin and Rasmuson, 2005). The correlation results are presented in Tables 1 and 2. A comparison of the experimental values of induction period with calculated values in the absence of additive (Table 1) shows that the model predictions are fairly close to the experimental measurements. The cluster model is best verified for the dominant g-mer size g = 1 and interfacial tension ı = 66–80 mJ m−2 . These results are in agreement with other similar data reported in literature (Mullin, 2001; Tai and Chien, 2003). For each temperature value the critical cluster size decreases with increased supersaturation and an enhanced nucleation rate will occur. The higher interfacial tension identified at low temperatures determine a larger critical size, fact that increased the induction period. When the nonionogen additive was added the di-mer is the dominant cluster size and different values of the superficial tension were identified for each value of the temperature. In the range of 303–333 K there is a lower interfacial tension when the additive is present. At 333 K a significant decrease of the interfacial tension in the presences of Tween 85 can be explained by the nucleation of aragonit, which has a lower interfacial tension (Westin and Rasmuson, 2005). At 348 K an unexpected high value was obtained that can stand for change Table 1 – Experimental and calculated induction time values for calcium carbonate precipitation in pure water. T (K) S  exp (s)  comp (s) 303 16.35 13.70 11.07 8.66 14.25 18.22 8.37 12.17 19.63 318 15.70 12.00 10.00 7.33 12.50 18.33 333 14.70 12.32 9.93 348 11.75 8.27 5.97 ı (mJ m−2 ) g gc Er (%) 78.40 1 37 45 58 3.3 14.5 7.7 7.92 11.75 18.55 76.93 1 29 35 45 7.9 6.0 1.1 6.16 9.16 20.66 5.44 9.89 20.49 71.62 1 19 27 31 11.0 7.8 0.7 5.33 7.50 14.66 4.82 6.77 15.48 66.65 1 21 34 57 12.7 10.6 6.3 Table 2 – Experimental and calculated induction time values for the precipitation of calcium carbonate in the presence of Tween 85 as organic additive. T (K) S  exp (s)  comp (s) ı (mJ m−2 ) g gc Er (%) 303 16.35 13.73 11.07 6.83 10.33 16.33 6.97 10.15 16.38 74.8 1 31 38 50 2.1 1.7 0.3 318 15.47 12.99 10.61 6.83 16.83 23.50 7.39 13.00 25.30 67.6 2 20 26 32 8.1 22.7 7.7 333 14.79 12.32 9.93 9.16 12.50 35.50 8.85 14.23 33.78 52.27 2 30 36 48 11.0 7.8 0.7 348 13.96 11.75 5.98 10.50 15.50 50.16 9.21 16.22 48.12 83.10 2 34 42 62 12.7 10.6 4.1 1454 chemical engineering research and design 8 8 ( 2 0 1 0 ) 1450–1454 in the polymorphic phase that nucleates, probably calcite and vaterite. Calcium carbonate can crystallize as three distinct polymorphs: calcite, aragonite and vaterite. The final product obtained in our experiments was a mixture of calcite and aragonite. The use of the nonionogen additive increased the concentration of aragonite as revealed by FT-IR spectra and microscopic analyses. The final polymorphic phase is not necessary the same as the phase that nucleated, as calcium carbonate undergoes polymorphic transformation all along the precipitation process and during the aging period. Often a mixture of polymorphs is formed. These polymorphs have different solubility and hence different supersaturations are generated. This fact may have a significant influence on the estimation of interfacial energy. In the present work the polymorphic phase in the nucleation stage was not investigated. Therefore, the consideration of calcite nucleation may not stand for all operating conditions and it can be a cause for a less accurate model prediction of the induction time. On the other hand, we must keep in mind that parameters estimated by a correlation technique are affected by experimental errors and the lack of fit. Therefore the estimated interfacial tension and cluster size cannot entirely explain the variation of induction times in different operating conditions. 4. Conclusions The effects of supersaturation, temperature and the presence of an organic additive on the induction period in calcium carbonate precipitation were studied experimentally and theoretically. The organic nonionogen additive (Tween 85) influences the induction period of calcium carbonate precipitation and the final polymorphic phase. Some quantitative evaluation of the induction period is made in terms of the interfacial energy and cluster sizes using the coagulation model. Acknowledgements This study was supported by the Ministry of Education and Research, National Authority for Scientific Research, through CEEX Program, project no. 18 and National Program II, project no. 22116. References Gomez-Morales, J., Torrent-Burgues, J. and Rodrigues-Clemente, R., 1996, Nucleation of calcium carbonate at different initial pH conditions. J Cryst Growth, 169: 331–338. Mullin, J.W., (2001). Crystallization (4th edition). (Butterworths–Heinemann, London, UK). Pokrovsky, O.S., 1998, Precipitation of calcium and magnesium carbonates from homogeneous supersaturated solutions. J Cryst Growth, 186: 233–239. Qian, R.Y. and Botsaris, G.D., 1997, A new mechanism for nuclei formation in suspension crystalizers: the role of interparticle forces. Chem Eng Sci, 52: 3429–3440. Reddy, M.M. and Hoch, A.R., 2001, Calcite crystal growth rate inhibition by polycarboxylic acids. J Colloid Interface Sci, 235(2): 365–370. Stamatakis, E., Stubos, A., Palyvos, J., Chatzichristos, C. and Muller, J., 2005, An improved predictive correlation for the induction time of CaCO3 scale formation during flow in porous media. J Colloid Interface Sci, 286: 7–13. Tai, C.Y. and Chien, W.C., 2003, Interpreting the effects of operating variables on the induction period of CaCl2 –Na2 CO3 system by a cluster coagulation model. Chem Eng Sci, 58: 3233–3241. Verdoes, D., van Kashchiev, D. and Rosmalen, G.M., 1992, Determination of nucleation and growth rates from induction times in seeded and unseeded precipitation of calcium carbonate. J Cryst Growth, 118: 401–413. Westin, K.-J. and Rasmuson, A.C., 2005, Nucleation of calcium carbonate in presence of citric acid, DTPA, EDTA and pyromellitic acid. J Colloid Interface Sci, 282: 370– 379.