Effects of operating variables on the induction period of CaCl2–Na2CO3 system

Effects of operating variables on the induction period of

CaCl

₂

–Na

₂

CO

₃

system

Clifford Y. Tai, Wen-Chen Chien*

Department of Chemical Engineering, National Taiwan University, Taipei 10617, Taiwan

Abstract

The induction period is defined as the time elapsed between the creation of supersaturation and the formation of critical nuclei. A novel data acquisition system is developed in this experiment to determine the induction period on the desupersaturation curve. The effect of several operating conditions, including initial reagent concentration,

temperature, pH, and presence of additive and seeds on the induction period of CaCO3were studied experimentally.

The results showed that the induction period decreases with an increase in initial reagent concentration, temperature,

and pH of the solution. The presence of Mg2+_{in solution prolongs the induction period. On the other hand, the}

presence of Na+ in solution has little influence on the induction period. Further, the presence of seed crystals in

solution shortens the induction period. r 2002 Published by Elsevier Science B.V.

Keywords: A1. Impurities; A1. Nucleation; A1. Supersaturated solutions; B1. Calcium compounds; B1. Inorganic compounds; B1. Minerals

1. Introduction

The induction period, tind; is defined as the time

that elapses between the creation of supersatura-tion, Sa; and the formation of critical nuclei. Many

methods have been applied for the determination of induction period, including the conductivity method [1,2], intensity of transmitted or scattered light method [3–5], heat released method [6,7], activity of precipitated ions method [8] and pH method [9]. Among the proposed methods, the conductivity method is frequently used to measure both long and short induction periods due to its good stability and sensitivity. According to this method, the end of the induction period is

determined as the time when the slope of the desupersaturation curve, which is a plot of solution conductivity against time curve, starts to change [10]. However, by using the conductivity method the induction period is not usually clearly determined on the desupersaturation curve. Thus, a novel data acquisition system is applied in a previous study [11] so that the induction period can be easily estimated from the enlarged desu-persaturation curve.

Since the induction period is closely related to the metastable zone width and nucleation rate of a system, many efforts have been devoted to this area. As a result, several important operating variables such as supersaturation, temperature, pH, agitation speed, and impurity [1,2,9,12,13] are found to have an influence on the induction period. So far, the effects of these variables on

*Corresponding author. Fax: +886-2-236-23040. E-mail address:wcchien@kimo.com.tw (W.-C. Chien).

0022-0248/02/$ - see front matter r 2002 Published by Elsevier Science B.V. PII: S 0 0 2 2 - 0 2 4 8 ( 0 1 ) 0 2 2 4 1 - 2

the induction period of CaCO3 in the CaCl2–

Na2CO3 system are not systematically studied.

Therefore, in this article we investigated experi-mentally the effects of several operating variables, including initial reagent concentration, tempera-ture, pH, and presence of additive and seeds, on the induction period of CaCO3 in the aqueous

CaCl2–Na2CO3 solution, using a conductivity

meter linked with the novel data aquisition system. 2. Experimental procedure

The experimental apparatus, which consists mainly of three parts: (I) reagent feeding system, (II) crystallizer with temperature control, and (III) data acquisition system, has been reported in a previous study [11]. Guaranteed grade calcium carbonate (Nacalai Tesque), extra-pure grade anhydrous sodium carbonate (Nacalai Tesque), extra-pure magnesium chloride hexahydrate and sodium chloride (Nacalai Tesque), and high quality deionized water with a specific resisitivity 18 MO cm were used. The water was filtered through a 0.2 mm filter before use. The experi-mental conditions, including vessel size, type, and material of construction, and agitation rate are maintained constant.

The experimental procedures are described briefly below. A desired quantity of water and 0.1 M CaCl2 solution were poured into the glass

beaker and mixed by a magnetic stirrer to form a solution of specified concentration. After the solution temperature became steady at 251C and the conductivity remained constant for several minutes, a required quantity of 0.1 M Na2CO3

solution at 251C was added into the beaker. The solution conductivity increased rapidly to a high level and stayed there for a certain period of time once the mixing is complete. Then, a decrease in conductivity was observed while the solution was still clear. Afterwards, the solution became turbid as detected by naked eyes. The experiment was stopped after the conductivity had no more significant change. After each run, the experimen-tal apparatus was rinsed with 0.1 M aqueous HCl solution to remove residual precipitate.

In the case in which the pH effect is investigated, the procedure differs from the previous case in

having the 0.1 M aqueous HCl solution added into the CaCl2–Na2CO3system to adjust the solution

pH. In all runs, the solution temperature is at 251C with no seed crystal existing in the aqueous solution. The solution pH investigated in the present work ranges from 8.84 to 10.85.

In the case in which the additives effect is investigated, a desired quantity of 0.1 M aqueous MgCl2(or NaCl) solution as the source of Mg

2+

(or Na+) was added into the solution in beaker before the addition of 0.1 M aqueous Na2CO3

solution. Other experimental procedure is the same as the previous case. In all runs, the solution temperature is at 251C with no seed crystal existing in the aqueous solution. The molar concentration ratio between Mg2+ and Ca2+, [Mg2+]/[Ca2+], investigated in the present work ranges from 0.2 to 1.0. On the other hand, the molar concentration ratio between Na+ and Ca2+, [Na+]/[Ca2+], ranges from 0.57 to 8.57.

In the case of seeded experiment, the procedure differs only in having the seed crystals placed in the solution. The seed crystals with a size of 355– 425 mm and total weight of 0.11–0.13 g are added into the solution before the 0.1 M aqueous Na2CO3solution is added.

Typical desupersaturation curves represented by curve (a) for the aqueous CaCl2–Na2CO3–MgCl2

solution and curve (b) for the aqueous CaCl2–

Na2CO3–seed solution at 251C are shown in Fig. 1.

The typical desupersaturation curves of pure aqueous CaCl2–Na2CO3system and the important

features of the desupersaturation curves have been described elsewhere [11]. The tind is identified by

the change of conductivity as shown in Fig. 2, which is an enlarged figure of the part for time interval between 0 and 180 s shown in Fig. 1.

3. Results and discussion

Fig. 3 illustrates the experimental data of induction period obtained at various initial re-agent concentrations of CaCl2and Na2CO3for the

unseeded and seeded case. The results show that the induction period of CaCO3 decreases

expo-nentially with an increase in initial reagent concentration at constant temperature for both

unseeded and seeded case. The results of unseeded case are the same as those obtained by S.ohnel and Mullin [1] at higher initial concentration of CaCl2

and Na2CO3. For a given level of initial reagent

concentration of CaCl2 and Na2CO3, the

induc-tion periods of seeded case are shorter than the

corresponding induction period of unseeded case. The induction period has frequently been used as a measure of the nucleation event and it can be considered to be inversely proportional to the rate of nucleation derived from classical nucleation theory under simplified assumption [14]

log tindp g 3 T3_{ðlog S} aÞ2   : ð1Þ

Eq. (1) shows that the interfacial energy, g; temperature, T ; and supersaturation, Sa; are the

important factors, which can affect the induction period, tind: Therefore, the decrease in induction

period at higher reagent concentration for both unseeded and seeded case can be interpreted as the increase in supersaturation at constant tempera-ture and interfacial energy as shown in Eq. (1). On the other hand, the shorter induction periods of seeded case than that of unseeded case at a given level of initial reagent concentration is caused by the creation of an interfacial supersaturation near the surface of seeds. Due to the adsorption of solute clusters, the interfacial supersaturation is higher than the bulk supersaturation. Therefore, the addition of seeds in solution causes a decrease in induction period. Qian and Botsaris [15] indicate that the decrease in induction period of

Fig. 1. Typical desupersaturated curves for the two different operating conditions at 251C: (a) [CaCl2]i=[Na2CO3]i= 0.0035 M, and [MgCl2]=0.0021 M and (b) [CaCl2]i=[Na2CO3]i= 0.0025 M, seed weight=0.1107 g, and seed size=355–425 mm.

Fig. 2. (a) Enlarged conductivity–time curves showing tind for

the cases [CaCl2]i=[Na2CO3]i=0.0035M, and [MgCl2]i= 0.0021 M. (b) Enlarged conductivity–time curves showing tind

for the case [CaCl2]i=[Na2CO3]i=0.0025 M, seed weight=0.1107 g, and seed size=355–425 mm.

Fig. 3. Induction period as a function of initial reagent concentration of CaCl2and Na2CO3for unseeded and seeded cases at 251C: ’ experimental point for unseeded case and  experimental point for seeded case.

seeded case is mainly caused by higher coagulation concentration of clusters in the region near the seed crystal. The higher coagulation concentration of clusters in the region near the seed crystal results from the van der Waals attractive force between the cluster and seed. Recently, Kuznetsov et al. [16] has also indicated that at the vicinity of a crystal has a relatively high supersaturation than bulk solution in their study by using the atomic force microscopy on protein. Fig. 4 shows that the induction period as a function of initial reagent concentrations of CaCl2 and Na2CO3 at three

different levels of temperature, i.e., 151C, 251C, and 351C. The experimental results show that the induction periods decrease exponentially with an increase in initial reagent concentration of CaCl2

and Na2CO3 for all levels of temperature. For a

given level of initial reagent concentration of CaCl2and Na2CO3, the induction period decreases

with an increase in temperature. The results are the same that reported by Mullin and $Z!a$cek [13] in the precipitation study of potassium aluminum sul-phate over the temperature range 15–351C. Ac-cording to the classical nucleation theory they concluded that the decrease in induction period at higher solution temperature is not only caused by

the effect of temperature as shown in Eq. (1) but also by the smaller interfacial energy of crystal at higher solution temperature. Fig. 5 illustrates the experimental data of induction period obtained at various levels of pH and three different initial reagent concentrations of CaCl2and Na2CO3, i.e.,

0.0015, 0.0025 and 0.0035 M. Fig. 5 indicates that tind decrease exponentially with an increase in

solution pH for all levels of initial concentrations of CaCl2 and Na2CO3. For a given level of

solution pH, the induction period decreases with an increase in initial reagent concentration of CaCl2 and Na2CO3. The decrease in induction

period at higher solution pH is mainly caused by the higher supersaturation, which is resulted from a higher concentration of CO32 in solution. The

supersaturation is related to the concentration of CO32, which is a function of pH at a fixed initial

concentration of total carbonate [17]. Fig. 6 illustrates the experimental data of induction period obtained at different mole concentration ratios of [Mg2+]/[Ca2+] under four different levels of initial reagent concentration of CaCl2 and

Na2CO3, i.e., 0.0040, 0.0035, 0.0025 and

0.0015 M. The experimental results show that the induction periods increase with an increase in

Fig. 4. tind as a function of initial reagent concentration of

CaCl2 and Na2CO3 at three different levels of solution temperature:  experimental point for T ¼ 151C; ’ experi-mental point for T ¼ 251C and E experiexperi-mental point for T ¼ 351C.

Fig. 5. tindas a function of solution pH at three different initial

reagent concentrations of CaCl2 and Na2CO3 at 251C:  experimental point for [CaCl2]i=[Na2CO3]i=0.0015 M; ’ experimental point for [CaCl2]i=[Na2CO3]i=0.0025 M and E experimental point for [CaCl2]i=[Na2CO3]i=0.0035 M.

[Mg2+]/[Ca2+] for all levels of initial reagent concentration of CaCl2and Na2CO3. For a given

level of [Mg2+]/[Ca2+], the induction period decreases with an increase in initial reagent concentration of CaCl2and Na2CO3. On the other

hand, the addition of Na+ in solution has little influence on the induction period of CaCO3. The

mole concentration ratio of [Na+]/[Ca2+] investi-gated in the present work ranges from 0.57 to 8.57 at a fixed initial reagent concentration of CaCl2

and Na2CO3, i.e., 0.0035 M. In general, the

presence of impurities in a system can affect the induction period considerably, but it is virtually impossible to predict. Some impurities increase the induction period, whereas others may decrease it or have no effect. The effect of impurities is generally interpreted by changing the equilibrium solubility or the solution structure, by adsorption or chemisorption on nuclei or heteronuclei, by chemical reaction or complex formation in the solution [18]. So far, explaination to the impurities effect on the induction period for various systems have been proposed in the literature. For CaCO3

formed by mixing solutions of CaCl2and Na2CO3,

S.ohnel and Mullin [19] found that cations K+, Cr3+ and Ni+ exert little influence on tind; only

Mn2+ and Mg2+ lengthen the induction period. They concluded that the increase in tind caused by

the presence of impurity is due to an increase in the crystal–solution interfacial energy. Pokrovsky [20] also found that the induction period of CaCO3 increases with increasing the

Mg2+/Ca2+ activity ratio in solution. They postulated that the increase in tind is due to an

increase in the crystal–solution interfacial energy at higher Mg2+/Ca2+ region and due to an increase in the activity of Mg2+ at lower Mg2+/ Ca2+region. Therefore, we consider that the effect of Mg2+ on the induction period in our present study is caused from an increase in the interfacial energy of CaCO3. The larger interfacial energy

results in a longer induction period when Mg2+is added into the solution. On the other hand, Na+ has little influence on the interfacial energy and thus the induction period.

4. Conclusion

The effects of initial reagent concentration, temperature, pH, the presence of Mg2+ or Na+, and the addition of seed on the induction period of CaCO3 are studied experimentally. The results

show that the induction period increases with a decrease in initial reagent concentration, tempera-ture, and pH. The presence of Mg2+ in solution prolongs the induction period. On the other hand, the presence of Na+in solution has little influence on the induction period. Further, addition of seed in solution shortens the induction period. The results also showed that the present method should be applicable to many systems for determination of induction period.

Acknowledgements

This study is supported by the National Science Council of the Republic of China.

Fig. 6. The effect of concentration ration [Mg2+]/[Ca2+] on the induction period at different initial reagent concentration of CaCl2 and Na2CO3: m experimental point for [CaCl2]i= [Na2CO3]i=0.0015 M; E experimental point for [CaCl2]i=[Na2 CO3]i=0.0025 M; ’ experimental point for [CaCl2]i=[Na 2-CO3]i=0.0035 M and  experimental point for [CaCl2]i= [Na2CO3]i=0.0040 M.

References

[1] O. S.ohnel, J.W. Mullin, J. Crystal Growth 44 (1978) 377. [2] O. S.ohnel, J.W. Mullin, J. Crystal Growth 60 (1982) 239. [3] M.I.H. Wakita, I. Masuda, J. Crystal Growth 61 (1983)

377.

[4] P.A. Carosso, E. Pelizzetti, J. Crystal Growth 68 (1984) 532.

[5] W. Kibalczyc, K. Bondarczuk, J. Crystal Growth 71 (1985) 751.

[6] A. Glasner, M. Tassa, J. Crystal Growth 13/14 (1972) 441. [7] W. Kibalczyc, A. Zielenkiewicz, J. Crystal Growth 82

(1987) 733.

[8] M.C. van der Leeden, D. Verdoes, D. Kashchiev, G.M. van Rosmalen, in: J. Garside, et al., (Eds.), Advance in Industrial Crystallization, Butterworth, Oxford, 1991, p. 1. [9] J. G!omez-Morales, J. Torrent-Burgu!es, R.

Rodr!ıguez-Clemente, J. Crystal Growth 169 (1996) 331.

[10] O. S.ohnel, J. Garside, Precipitation: Basic Principles and Industrial Applications, Butterworth, Oxford, 1992, p. 206.

[11] W.C. Chien, C.Y. Tai, J.P. Hsu, J. Chem. Phys. 111 (1999) 2657.

[12] M.S. Joshi, A.V. Antony, J. Crystal Growth 46 (1979) 7. [13] J.W. Mullin, S. $Z!a$cek, J. Crystal Growth 53 (1981) 515. [14] M.J. Mullin, Crystallization, Butterworth, Oxford, 1993,

p. 198.

[15] R.Y. Qian, G.D. Botsaris, Chem. Eng. Sci. 52 (1997) 3429. [16] Yu.G. Kuznetsov, A.J. Malkin, W. Glantz, A. Mcpherson,

J. Crystal Growth 168 (1996) 63. [17] ASTM Designation D513, 1974, p.416.

[18] M.J. Mullin, Crystallization, Butterworth, Oxford, 1993, p. 193.

[19] O. S.ohnel, J.W. Mullin, J. Colloid Interface Sci. 123 (1988) 43.