This article was downloaded by: [University of Auckland Library] On: 16 October 2014, At: 17:49 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Environmental Technology Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/tent20
Growth kinetics of calcium fluoride at high supersaturation in a fluidized bed reactor a
a
a
a
K. Jiang , K.G. Zhou , Y.C. Yang & H. Du a
School of Metallurgical Science and Engineering, Central South University, Changsha 410083, People's Republic of China Accepted author version posted online: 12 Jun 2013.Published online: 02 Jul 2013.
To cite this article: K. Jiang, K.G. Zhou, Y.C. Yang & H. Du (2014) Growth kinetics of calcium fluoride at high supersaturation in a fluidized bed reactor, Environmental Technology, 35:1, 82-88, DOI: 10.1080/09593330.2013.811542 To link to this article: http://dx.doi.org/10.1080/09593330.2013.811542
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
Environmental Technology, 2014 Vol. 35, No. 1, 82–88, http://dx.doi.org/10.1080/09593330.2013.811542
Growth kinetics of calcium fluoride at high supersaturation in a fluidized bed reactor K. Jiang, K.G. Zhou∗ , Y.C. Yang and H. Du School of Metallurgical Science and Engineering, Central South University, Changsha 410083, People’s Republic of China
Downloaded by [University of Auckland Library] at 17:49 16 October 2014
(Received 30 August 2012; final version received 29 May 2013 ) Crystallization process in a fluidized bed reactor (FBR) has been regarded as an environmentally friendly technology for the removal and recovery of fluoride from industrial wastewater. The growth kinetics of calcium fluoride at high supersaturation was studied for design, control, and operation of an FBR. The main variables, including supersaturation, superficial velocity, pH value, and particle size of seed that influenced the crystal growth were investigated. Then, a growth model was used to predict the linear growth rate of calcium fluoride at a high influent concentration of fluoride. The pressure difference in the FBR was used as a feature to characterize the growth rate of calcium fluoride. The aggregation and adsorption between seeds and fine particles were proven to be a possible mechanism for growth of calcium fluoride. Keywords: crystallization; calcium fluoride; growth kinetics; high supersaturation; aggregation
Introduction The traditional method for fluoride removal from industrial wastewater generally involves chemical precipitation and coagulation.[1,2] The process generates large amounts of sludge with high water content, low quality, and non-reusability, which has to be disposed with increasing costs.[3,4] Meanwhile, fluorite resources are suffering a shortage crisis worldwide given their extensive use as a raw material in manufacturing processes.[5] Hence, the recovery of fluoride from industrial wastewater has become a research hotspot to contribute to the raw material conservation of the fluorine industry.[6–9] Traditional stirred-tank reactor and fluidized bed reactor (FBR) are widely used for calcium fluoride recovery. Calcium fluoride recovered in traditional-stirred tank reactors (such as continuous stirred-tank reactor and batch reactor) cannot be dehydrated and reused economically due to its high water content and fine particle size. FBRs are developed based on the crystallization process.[5,10,11] The water content of sludge is greatly decreased as the seed retention time is prolonged. The recovered calcium fluoride can be easily separated and reused as a raw material for the production of hydrofluoric acid and other fluoride salts.[5,12,13] The core part of the crystallization process is the FBR, in which reaction and solid–liquid separation take place simultaneously. In addition, the seed plays an important role in inducing the crystallization of calcium fluoride. The control features mainly include the following: (1) controlling the fluid velocity between the critical fluidization velocity and terminal velocity so that the seeds can be kept suspended ∗ Corresponding
author. Email: [email protected]
© 2013 Taylor & Francis
in a fluidized state; the hydraulic conditions can be determined by fluidization experiments; and (2) obtaining the optimal crystal growth conditions by kinetic experiments so that calcium fluoride particles can grow in size. The variables that influence crystal growth are supersaturation, particle size of seed, superficial velocity, and pH. The growth kinetics of calcium fluoride based on the aggregation and molecular growth mechanisms at low supersaturation has been studied by many researchers.[14–16] To avoid the primary nucleation and formation of fine particles, the influent fluoride concentration needs to be controlled to below 150 mg/L. However, the fluoride concentration in industrial wastewater is usually above 500 mg/L, which indicates that controlling appropriate supersaturation conditions at the inlet of the reactor is not economical. Therefore, the kinetics models at low supersaturation ratio do not fit when wastewater has a high fluoride concentration. This study aimed to investigate the feasibility of recovering calcium fluoride from high fluoride-containing wastewater in an FBR. The influences of supersaturation, particle size of seed, superficial velocity, and pH on the growth rate of calcium fluoride were investigated. The kinetics model and mechanism of crystal growth were also studied. Materials and methods FBR and process description Synthetic wastewater prepared with sodium fluoride (analytical grade) was used as influent. The precipitant of calcium fluoride was prepared with calcium chloride
Downloaded by [University of Auckland Library] at 17:49 16 October 2014
Environmental Technology
83
Figure 1. The schematic of the FBR. Note: 1 wastewater container; 2 precipitant container; 3 mixing zone of FBR; 4 reaction zone of FBR; 5 clarification zone of FBR; 6-8 magnetic pumps; 9-11 flowmeters; 12 pH meter; 13 U-shaped manometer; a-b outlet.
(industrial grade). The synthetic calcium fluoride particles were used as the seed material. The reaction pH was adjusted by adding NaOH or HCl to the precipitant. The schematic of the FBR is shown in Figure 1. The poly methyl methacrylate reactor consisted of a clarification zone (inner diameter = 100 mm; height = 750 mm), a mixing zone, and a reaction zone (inner diameter = 50 mm; height = 800 mm). The total volume was about 7.7 L. A liquid distributor was set up between the mixing zone and the reaction zone to maintain good dispersion of the influent and seeds. A typical run consisted of adding about 250 g of seeds and controlling the superficial velocity so that the seed material did not settle down and flow out of the reactor. The synthetic wastewater and precipitant were directed to flow to the bottom of the reactor by two magnetic pumps. The solution in the clarification zone was partly recycled to the mixing zone by a magnetic pump. The water flow was adjusted and measured by regulation valves and flowmeters, respectively.
Sampling and analysis During the operation, the effluent was sampled from the outlet (b), whereas calcium fluoride grains (pellets) were sampled from the outlet (a). The pressure difference was
measured by a U-shaped manometer (11), and the reaction pH was measured by a pH meter. The fluoride concentration of the effluent was analysed by ion-selective electrodes. The calcium concentration of the effluent was analysed by EDTA Titration. The zeta potential of the effluent was measured by a Zetasizer Nano ZS. After drying at 313 K for 10 h, the pellets were analysed using a scanning electron microscopy (SEM, JSM 6360LV). The krummbein diameter (Lk ) of the pellets was measured by the computer programme Smile View based on the SEM image. The area mean diameter was defined as the particle size of the pellet (L), which was calculated as: L=
Lk2
1/2 .
(1)
The overall crystal linear growth rate [17] of calcium fluoride was defined as: G=
dL , dt
(2)
where t is the operational time. G is calculated by the plot of the particle size versus the operational time. The relative supersaturation ratio was defined as: σ =
cCa2+c2 F− Ksp
1/3 − 1,
(3)
84
K. Jiang et al.
where cCa2+ and cF − are the concentrations of calcium and fluoride, respectively, and Ksp is the solubility product of calcium fluoride (Ksp = 10−10.31 ). The superficial velocity was defined as: V =
Fin , S
(4)
where Fin is the sum of influent flow, precipitant flow, and reflux flow. S is the cross-sectional area of the reaction zone of the reactor.
Downloaded by [University of Auckland Library] at 17:49 16 October 2014
Results and discussion Determination of the crystal growth rate of calcium fluoride Influence of the supersaturation Experiments were conducted at an influent flow of 17 L/h, a precipitant flow of 25 L/h, reflux flow of 0 L/h, a Ca/F molar ratio of 1:1, and reaction pH 7 ± 0.2. The particle size of seed was 200–400 mesh. The supersaturation was adjusted by the fluoride concentration of the influent. Figure 2(a) shows the influence of the supersaturation on the linear growth rate of calcium fluoride. It can be observed that the linear growth rates of calcium fluoride are 8.56 × 10−11 , 5.53 × 10−10 , 5.22 ×
10−1 and 2.87 × 10−10 m/s when σ = 6.01, 25.91, 54.03 and 84.09 respectively. The linear growth rate reaches higher at σ = 25.91 and then decreases with increasing supersaturation. This phenomenon can be explained by the relationship between nucleation and supersaturation. At low supersaturation (σ = 6.01–25.91), the calcium fluoride particles favour growth on the surface of the seeds by heterogeneous nucleation, and the growth rate of calcium fluoride is proportional to the generation amount of particles.[18] In addition, the generation amount of particles increases with increasing fluoride concentration per unit time, which leads to increased growth rate of calcium fluoride. At high supersaturation (σ = 25.91–84.09), the homogeneous nucleation of calcium fluoride dominates the crystallization process.[18] Although a large number of fine particles is generated, most of them are brought out of the reactor, which leads to decreased growth rate of calcium fluoride.
Influence of the particle size of seed Experiments were conducted at an influent flow of 17 L/h, a precipitant flow of 25 L/h, a reflux flow of 0 L/h, a Ca/F molar ratio of 1:1, and reaction pH 7 ± 0.2. The fluoride concentration of the influent was 900 mg/L. Figure 2(b)
Figure 2. Influence of the supersaturation (a), particle size (b), superficial velocity (c), and pH (d) on the particle size of the pellets as a function of time.
Downloaded by [University of Auckland Library] at 17:49 16 October 2014
Environmental Technology shows the influence of the particle size on the linear growth rate of calcium fluoride. It can be observed that the linear growth rates of calcium fluoride are 5.22 × 10−10 , 7.17 × 10−10 , 1.04 × 10−9 and 9.61 × 10−10 m/s when L0 = 51.26, 81.82, 118.17 and 169.39 μm, respectively. Particles of a large seed have a higher a growth rate than those with a smaller seed when L0 is 51.26–118.17 μm. This finding can be explained by the fact that the surface area per seed increases with increasing particle size, which favours the heterogeneous nucleation of fine particles. However, the linear growth rate slightly decreases at L0 = 169.39 μm, which can be attributed to the fact that the superficial velocity (V = 0.0059 m/s) is less than the fluidization velocity for some large particles. Therefore, the fluidization quality of seeds is reduced, which leads to decreased growth rate. It can be concluded that the particle size and fluidization quality of seeds significantly affect the growth process of calcium fluoride. Two of the key parameters of operation are maintaining good fluidization quality of seeds as well as prolonging the contact time between the seeds and fine particles. Influence of the superficial velocity Experiments were conducted at an influent flow of 17 L/h, a precipitant flow of 25 L/h, a Ca/F molar ratio of 1:1, and reaction pH 7 ± 0.2. The fluoride concentration of the influent was 900 mg/L. The particle size of the seed was 200–400 mesh. The superficial velocity was adjusted by the reflux flow. Figure 2(c) shows the influence of the superficial velocity on the linear growth rate of calcium fluoride. It can be observed that the linear growth rates of calcium fluoride are 5.22 × 10−10 , 3.58 × 10−10 , 2.61 × 10−10 m/s when V = 0.0059, 0.0095, and 0.0130 m/s, respectively. The linear growth rate of calcium fluoride decreases with increasing superficial velocity. The possible reasons are as follows: (1) fine particles are rapidly brought out of the reactor, thereby reducing the contact time between seeds and fine particles; (2) the fluidized bed expansion ratio becomes so high that many seeds are retained in the junction area of the clarification zone and the reaction zone of the reactor; this leads to the decreased number of seeds in the reaction zone and surface area of seeds per volume [14,15]; and (3) the layer abrasion of seeds is more serious at a higher superficial velocity; therefore, the superficial velocity must be decreased while ensuring the fluidization quality of seeds. Influence of the reaction pH Experiments were conducted at an influent flow of 17 L/h, a precipitant flow of 25 L/h, a reflux flow of 0 L/h, and a Ca/F molar ratio of 1:1. The fluoride concentration of influent was 900 mg/L. The particle size of the seed was 200–400 mesh. Figure 2(d) shows the influence of the reaction pH on the linear growth rate of calcium fluoride.
85
Figure 3. Parity graph for the predicted values of the overall linear growth rate by the proposed semiempirical model of calcium fluoride particle growth in the FBR.
It can be observed that the linear growth rates of calcium fluoride are 2.95 × 10−10 , 5.22 × 10−10 , 3.90 × 10−10 , 5.06 × 10−10 m/s when pH = 6, 7, 8, and 9, respectively. The linear growth rate slightly increases with increasing pH. To analyse the mechanism of the pH effect, the zeta potential of the effluent was measured. The zeta potentials are 32.95, 19.25, 14.05, 14.00 mv when pH = 6, 7, 8, and 9, respectively. The aggregation between the seeds and fine particles is favoured to take place when zeta potential is below 20 mv at pH 7–9. The repulsive force of particles increases at pH 6, which leads to decreased growth rate. Therefore, the growth of calcium fluoride is more favourable at pH 7–9. Growth model of calcium fluoride The general kinetics equation for growth [19] is expressed as follows: G = Kg L0m V n σ j .
(5)
As shown in Figure 2, the growth rate of fluoride calcium is fitted to Equation (5) to obtain the overall crystal growth equation, as shown in Equation (6). G = 3.82 × 10−4 × L00.82 × V −0.86 × [(6.06 × 10−10 − 3.84 × 10−14 )σ −2 ].
(6)
The correlation between the experimental and predicted values of G is shown in Figure 3. The correlation coefficient of the fitting is 0.976, which agrees well with the experimental data. Therefore, Equation (6) can be used to predict the growth rate of calcium fluoride under the following conditions: L0 = 51.26–118.17 μm, V = 0.0059–0.0130 m/s and σ = 25.91–84.09 (fluoride concentration of influent 400–1400 mg/L).
86
K. Jiang et al. It can be observed that the pressure difference in the FBR increases with increasing particle size of pellets. The pressure difference in the FBR can be determined [20] as follows:
Downloaded by [University of Auckland Library] at 17:49 16 October 2014
P =
Figure 4. Relationship between the particle size of calcium fluoride and the pressure difference in the FBR.
Characterization of the growth process of calcium fluoride Experiments were conducted at an influent flow of 17 L/h, a precipitant flow of 25 L/h, a reflux flow of 0 L/h, a Ca/F molar ratio of 1:1, L0 = 51.26 μm, and reaction pH 7 ± 0.2. The fluoride concentration of the influent was 900 mg/L. The relationship between particle size of pellets and the pressure difference in the FBR was determined during the operation, as shown in Figure 4.
Figure 5.
H (1 − ε)(ρp − ρl )g = (1 − ε)(mp − ml )g , S
(7)
where H is the height of bed; ε is the porosity of bed; ρp and ρl are the density of solid and liquid, respectively; mp is the mass of solid; ml is the mass of the liquid with the same volume as the solid; and S is the cross-sectional area of the reaction zone of the FBR. Assuming that the heterogeneous nucleation of calcium fluoride is dominating, and that fine particles are adsorbed on the seed surface, the relationship between the particle size (L) and mp is as follows: mp ∝ L3 .
(8)
Assuming that the homogeneous nucleation of calcium fluoride is dominant, the change in solid mass is attributed to the increase in the number of particles. By contrast, the mean particle size decreases due to the formation of fine particles. Therefore, the relationship between L and mp is as follows: (9) mp ∝ Lα (0 < α < 3). According to Equations (7)–(9), the relationship between L and P is as follows: P = γ + βLα .
SEM images of calcium fluoride precipitated in the FBR (V = 0.0130 m/s, operation time = 16 h).
(10)
Environmental Technology The data in Figure 4 are fitted to Equation (10), as shown in Equation (11). 1.624 P L = 0.602 + 0.402 × . (11) P0 L0
Downloaded by [University of Auckland Library] at 17:49 16 October 2014
It can be seen that the correlation coefficients is 0.998, which indicates that the experimental data match well with the fitting equation. Therefore, the pressure difference in the FBR can be used to estimate the particle size of pellets and to characterize the growth rate of calcium fluoride. The operation state of the growth process can be indirectly monitored, which is significant in maintaining process operation. Growth mechanism of calcium fluoride To analyse the growth mechanism of calcium fluoride, the pellets obtained at V = 0.0130 m/s and σ = 54.03 in Section 3.1.3 were observed by SEM, as shown in Figure 5. The area surrounded by the white line shows the locally amplified part of the image. It can be seen that both fine particles and the seeds are retained in the reactor during operation. The fine particles range in size from 0.5 to 2.5 μm, which is attributed to homogeneous precipitation. The seed surface is rough with some fine particles. When the seeds are kept suspended in a fluidized state, the ‘seed zone’ with a high holdup is formed in the reaction zone of the reactor. Meanwhile, the fine particles axially rise and pass through the ‘seed zone’. Under low zeta potential in solution (less than 20 mv), some fine particles are adsorbed onto the surface of seeds [18] by fluid shear stress and collision/contact interaction, which is attributed to the low repulsive force of particles (Figure 5(b)–(c)). By contrast, the other unadsorbed fine particles are retained in solution (Figure 5(c)) and then flow out of the reactor. Therefore, the formation of fine particles is unavoidable at high supersaturation due to discrete precipitation and primary nucleation. The adsorption and aggregation between seeds and fine particles is the main mechanisms for growth of calcium fluoride, which differs from the mechanisms of aggregation and molecular growth at low supersaturation.[14–16] Conclusions The growth kinetics of calcium fluoride was studied at high supersaturation in an FBR, and the main conclusions are as follows: (1) Supersaturation (σ ), particle size of seed (L0 ), superficial velocity (V ), and reraction pH are the main factors influencing the growth rate of calcium fluoride. When the seeds are maintained with a high fluidization quality, the growth rate of calcium fluoride is higher with large seed size, lower superficial velocity, and lower supersaturation.
87
(2) When L0 = 51.26–118.17 μm, V = 0.0059– 0.0130 m/s, σ = 25.91–84.09, and reaction pH = 7, the kinetics model given by the equation G = 3.82 × 10−4 × L00.82 × V −0.86 × [(6.06 × 10−10 − 3.84 × 10−14 )σ −2 ] can be used to predict the growth rate of calcium fluoride. (3) The pressure difference is well fitted with the particle size of pellets by the equation P/P0 = 0.602 + 0.402 × (L/L0 )1.624 . The pressure difference of the FBR can be used to predict the particle size of pellets and to characterize the growth rate of calcium fluoride. (4) The formation of fine particles is unavoidable at high supersaturation. The calcium fluoride grows in size by adsorption and aggregation. This finding demonstrates the feasibility of calcium fluoride recovery from high fluoride containing wastewater in the FBR. Acknowledgements This research was supported by the Major Science and Technology Program of Hunan (NO. 2009FJ-1009).
References [1] Turner BD, Binning P, Stipp SLS. Fluoride removal by calcite: evidence for fluorite precipitation and surface adsorption. Environ Sci Technol. 2005;39:9561–9568. [2] Wu J, Lu ZY. Combined treatment process for high fluoride content wastewater. China Water Wastewater. 2003;19:3–4. [3] Aldaco R, Irabien A, Luis P. Fluidized bed reactor for fluoride removal. Chem Eng J. 2005;107:113–117. [4] Van den Broeck K, Van Hoornick N, Van Hoeymissen J, de Boer R, Giesen A, Wilms D. Sustainable treatment of HF wastewaters from semiconductor industry with a fluidized bed reactor. IEEE T. Semiconduct. M. 2003;16:423–428. [5] Aldaco R, Garea A, Irabien A. Calcium fluoride recovery from fluoride wastewater in a fluidized bed reactor. Water Res. 2007;41:810–818. [6] Gouidera M, Fekia M, Sayadib M. Separative recovery with lime of phosphate and fluoride from an acidic effluent containing H3PO4, HF and/or H2SiF6. J Hazard Mater. 2009;170:962–968. [7] Lisbona D, Steel KM. Recovery of fluoride values from spent pot-lining: precipitation of an aluminium hydroxyfluoride hydrate product. Sep Purif Technol. 2008;61:182–192. [8] Lee TC, Liu FJ. Recovery of hazardous semiconductorindustry sludge as a useful resource. J Hazard Mater. 2009;165:359–365. [9] Wang LS, Wang CM, Yu Y, Huang XW, Long ZQ, Hou YK, Cui DL. Recovery of fluorine from bastnasite as synthetic cryolite by-product. J. Hazard. Mater. 2012;209–210:77–83. [10] Giesen A. Crystallisation process enables environmental friendly phosphate removal at low costs. Environ Technol. 1999;20:769–775. [11] Battistoni P, Boccadoro R, Fatone F, Pavan P. Autonucleation and crystal growth of truvite in a demonstrative fluidized bed reactor (FBR). Environ Technol. 2005;26: 975–982. [12] Lee MS, Chang WK, Shao H, Liao CC. Volume reduction and resource recovery of fluoride-containing wastewater.
88
K. Jiang et al.
Downloaded by [University of Auckland Library] at 17:49 16 October 2014
IWA Sludge Management Entering the 3rd Millennium, Taipei, Taiwan, 2001. [13] Rubén A, Aurora G, Angel I. Fluoride recovery in a fluidized bed: crystallization of calcium fluoride on silica sand. Ind Eng Chem Res. 2006;45:796–802. [14] Tai CY, Chen PC, Tsao TM. Growth kinetics of CaF2 in a pH-stat fluidized-bed crystallizer. J Cryst Growth. 2006;290:576–584. [15] Aldaco R, Garea A, Irabien A. Modeling of particle growth: application to water treatment in a fluidized bed reactor. Chem Eng J. 2007;134:66–71.
[16] Aldaco R, Garea A, Irabien A. Particle growth kinetics of calcium fluoride in a fluidized bed reactor. Chem Eng Sci. 2007;62:2958–2966. [17] Mersmann A. Crystallization technology handbook. 2nd ed. New York: CRC Press; 2001. [18] Ding XH, Tan D. Industrial crystallization. Beijing: Chemical Industry Press; 1985. [19] Bravi M, Mazzarotta B. Size dependency of citric acid monohydrate growth kinetics. Chem Eng J. 1998;70:203–207. [20] Guo MS, Li HZ. Handbook of fluidization. Beijing: Chemical Industry Press; 2008.
Environmental Technology Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/tent20
Growth kinetics of calcium fluoride at high supersaturation in a fluidized bed reactor a
a
a
a
K. Jiang , K.G. Zhou , Y.C. Yang & H. Du a
School of Metallurgical Science and Engineering, Central South University, Changsha 410083, People's Republic of China Accepted author version posted online: 12 Jun 2013.Published online: 02 Jul 2013.
To cite this article: K. Jiang, K.G. Zhou, Y.C. Yang & H. Du (2014) Growth kinetics of calcium fluoride at high supersaturation in a fluidized bed reactor, Environmental Technology, 35:1, 82-88, DOI: 10.1080/09593330.2013.811542 To link to this article: http://dx.doi.org/10.1080/09593330.2013.811542
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
Environmental Technology, 2014 Vol. 35, No. 1, 82–88, http://dx.doi.org/10.1080/09593330.2013.811542
Growth kinetics of calcium fluoride at high supersaturation in a fluidized bed reactor K. Jiang, K.G. Zhou∗ , Y.C. Yang and H. Du School of Metallurgical Science and Engineering, Central South University, Changsha 410083, People’s Republic of China
Downloaded by [University of Auckland Library] at 17:49 16 October 2014
(Received 30 August 2012; final version received 29 May 2013 ) Crystallization process in a fluidized bed reactor (FBR) has been regarded as an environmentally friendly technology for the removal and recovery of fluoride from industrial wastewater. The growth kinetics of calcium fluoride at high supersaturation was studied for design, control, and operation of an FBR. The main variables, including supersaturation, superficial velocity, pH value, and particle size of seed that influenced the crystal growth were investigated. Then, a growth model was used to predict the linear growth rate of calcium fluoride at a high influent concentration of fluoride. The pressure difference in the FBR was used as a feature to characterize the growth rate of calcium fluoride. The aggregation and adsorption between seeds and fine particles were proven to be a possible mechanism for growth of calcium fluoride. Keywords: crystallization; calcium fluoride; growth kinetics; high supersaturation; aggregation
Introduction The traditional method for fluoride removal from industrial wastewater generally involves chemical precipitation and coagulation.[1,2] The process generates large amounts of sludge with high water content, low quality, and non-reusability, which has to be disposed with increasing costs.[3,4] Meanwhile, fluorite resources are suffering a shortage crisis worldwide given their extensive use as a raw material in manufacturing processes.[5] Hence, the recovery of fluoride from industrial wastewater has become a research hotspot to contribute to the raw material conservation of the fluorine industry.[6–9] Traditional stirred-tank reactor and fluidized bed reactor (FBR) are widely used for calcium fluoride recovery. Calcium fluoride recovered in traditional-stirred tank reactors (such as continuous stirred-tank reactor and batch reactor) cannot be dehydrated and reused economically due to its high water content and fine particle size. FBRs are developed based on the crystallization process.[5,10,11] The water content of sludge is greatly decreased as the seed retention time is prolonged. The recovered calcium fluoride can be easily separated and reused as a raw material for the production of hydrofluoric acid and other fluoride salts.[5,12,13] The core part of the crystallization process is the FBR, in which reaction and solid–liquid separation take place simultaneously. In addition, the seed plays an important role in inducing the crystallization of calcium fluoride. The control features mainly include the following: (1) controlling the fluid velocity between the critical fluidization velocity and terminal velocity so that the seeds can be kept suspended ∗ Corresponding
author. Email: [email protected]
© 2013 Taylor & Francis
in a fluidized state; the hydraulic conditions can be determined by fluidization experiments; and (2) obtaining the optimal crystal growth conditions by kinetic experiments so that calcium fluoride particles can grow in size. The variables that influence crystal growth are supersaturation, particle size of seed, superficial velocity, and pH. The growth kinetics of calcium fluoride based on the aggregation and molecular growth mechanisms at low supersaturation has been studied by many researchers.[14–16] To avoid the primary nucleation and formation of fine particles, the influent fluoride concentration needs to be controlled to below 150 mg/L. However, the fluoride concentration in industrial wastewater is usually above 500 mg/L, which indicates that controlling appropriate supersaturation conditions at the inlet of the reactor is not economical. Therefore, the kinetics models at low supersaturation ratio do not fit when wastewater has a high fluoride concentration. This study aimed to investigate the feasibility of recovering calcium fluoride from high fluoride-containing wastewater in an FBR. The influences of supersaturation, particle size of seed, superficial velocity, and pH on the growth rate of calcium fluoride were investigated. The kinetics model and mechanism of crystal growth were also studied. Materials and methods FBR and process description Synthetic wastewater prepared with sodium fluoride (analytical grade) was used as influent. The precipitant of calcium fluoride was prepared with calcium chloride
Downloaded by [University of Auckland Library] at 17:49 16 October 2014
Environmental Technology
83
Figure 1. The schematic of the FBR. Note: 1 wastewater container; 2 precipitant container; 3 mixing zone of FBR; 4 reaction zone of FBR; 5 clarification zone of FBR; 6-8 magnetic pumps; 9-11 flowmeters; 12 pH meter; 13 U-shaped manometer; a-b outlet.
(industrial grade). The synthetic calcium fluoride particles were used as the seed material. The reaction pH was adjusted by adding NaOH or HCl to the precipitant. The schematic of the FBR is shown in Figure 1. The poly methyl methacrylate reactor consisted of a clarification zone (inner diameter = 100 mm; height = 750 mm), a mixing zone, and a reaction zone (inner diameter = 50 mm; height = 800 mm). The total volume was about 7.7 L. A liquid distributor was set up between the mixing zone and the reaction zone to maintain good dispersion of the influent and seeds. A typical run consisted of adding about 250 g of seeds and controlling the superficial velocity so that the seed material did not settle down and flow out of the reactor. The synthetic wastewater and precipitant were directed to flow to the bottom of the reactor by two magnetic pumps. The solution in the clarification zone was partly recycled to the mixing zone by a magnetic pump. The water flow was adjusted and measured by regulation valves and flowmeters, respectively.
Sampling and analysis During the operation, the effluent was sampled from the outlet (b), whereas calcium fluoride grains (pellets) were sampled from the outlet (a). The pressure difference was
measured by a U-shaped manometer (11), and the reaction pH was measured by a pH meter. The fluoride concentration of the effluent was analysed by ion-selective electrodes. The calcium concentration of the effluent was analysed by EDTA Titration. The zeta potential of the effluent was measured by a Zetasizer Nano ZS. After drying at 313 K for 10 h, the pellets were analysed using a scanning electron microscopy (SEM, JSM 6360LV). The krummbein diameter (Lk ) of the pellets was measured by the computer programme Smile View based on the SEM image. The area mean diameter was defined as the particle size of the pellet (L), which was calculated as: L=
Lk2
1/2 .
(1)
The overall crystal linear growth rate [17] of calcium fluoride was defined as: G=
dL , dt
(2)
where t is the operational time. G is calculated by the plot of the particle size versus the operational time. The relative supersaturation ratio was defined as: σ =
cCa2+c2 F− Ksp
1/3 − 1,
(3)
84
K. Jiang et al.
where cCa2+ and cF − are the concentrations of calcium and fluoride, respectively, and Ksp is the solubility product of calcium fluoride (Ksp = 10−10.31 ). The superficial velocity was defined as: V =
Fin , S
(4)
where Fin is the sum of influent flow, precipitant flow, and reflux flow. S is the cross-sectional area of the reaction zone of the reactor.
Downloaded by [University of Auckland Library] at 17:49 16 October 2014
Results and discussion Determination of the crystal growth rate of calcium fluoride Influence of the supersaturation Experiments were conducted at an influent flow of 17 L/h, a precipitant flow of 25 L/h, reflux flow of 0 L/h, a Ca/F molar ratio of 1:1, and reaction pH 7 ± 0.2. The particle size of seed was 200–400 mesh. The supersaturation was adjusted by the fluoride concentration of the influent. Figure 2(a) shows the influence of the supersaturation on the linear growth rate of calcium fluoride. It can be observed that the linear growth rates of calcium fluoride are 8.56 × 10−11 , 5.53 × 10−10 , 5.22 ×
10−1 and 2.87 × 10−10 m/s when σ = 6.01, 25.91, 54.03 and 84.09 respectively. The linear growth rate reaches higher at σ = 25.91 and then decreases with increasing supersaturation. This phenomenon can be explained by the relationship between nucleation and supersaturation. At low supersaturation (σ = 6.01–25.91), the calcium fluoride particles favour growth on the surface of the seeds by heterogeneous nucleation, and the growth rate of calcium fluoride is proportional to the generation amount of particles.[18] In addition, the generation amount of particles increases with increasing fluoride concentration per unit time, which leads to increased growth rate of calcium fluoride. At high supersaturation (σ = 25.91–84.09), the homogeneous nucleation of calcium fluoride dominates the crystallization process.[18] Although a large number of fine particles is generated, most of them are brought out of the reactor, which leads to decreased growth rate of calcium fluoride.
Influence of the particle size of seed Experiments were conducted at an influent flow of 17 L/h, a precipitant flow of 25 L/h, a reflux flow of 0 L/h, a Ca/F molar ratio of 1:1, and reaction pH 7 ± 0.2. The fluoride concentration of the influent was 900 mg/L. Figure 2(b)
Figure 2. Influence of the supersaturation (a), particle size (b), superficial velocity (c), and pH (d) on the particle size of the pellets as a function of time.
Downloaded by [University of Auckland Library] at 17:49 16 October 2014
Environmental Technology shows the influence of the particle size on the linear growth rate of calcium fluoride. It can be observed that the linear growth rates of calcium fluoride are 5.22 × 10−10 , 7.17 × 10−10 , 1.04 × 10−9 and 9.61 × 10−10 m/s when L0 = 51.26, 81.82, 118.17 and 169.39 μm, respectively. Particles of a large seed have a higher a growth rate than those with a smaller seed when L0 is 51.26–118.17 μm. This finding can be explained by the fact that the surface area per seed increases with increasing particle size, which favours the heterogeneous nucleation of fine particles. However, the linear growth rate slightly decreases at L0 = 169.39 μm, which can be attributed to the fact that the superficial velocity (V = 0.0059 m/s) is less than the fluidization velocity for some large particles. Therefore, the fluidization quality of seeds is reduced, which leads to decreased growth rate. It can be concluded that the particle size and fluidization quality of seeds significantly affect the growth process of calcium fluoride. Two of the key parameters of operation are maintaining good fluidization quality of seeds as well as prolonging the contact time between the seeds and fine particles. Influence of the superficial velocity Experiments were conducted at an influent flow of 17 L/h, a precipitant flow of 25 L/h, a Ca/F molar ratio of 1:1, and reaction pH 7 ± 0.2. The fluoride concentration of the influent was 900 mg/L. The particle size of the seed was 200–400 mesh. The superficial velocity was adjusted by the reflux flow. Figure 2(c) shows the influence of the superficial velocity on the linear growth rate of calcium fluoride. It can be observed that the linear growth rates of calcium fluoride are 5.22 × 10−10 , 3.58 × 10−10 , 2.61 × 10−10 m/s when V = 0.0059, 0.0095, and 0.0130 m/s, respectively. The linear growth rate of calcium fluoride decreases with increasing superficial velocity. The possible reasons are as follows: (1) fine particles are rapidly brought out of the reactor, thereby reducing the contact time between seeds and fine particles; (2) the fluidized bed expansion ratio becomes so high that many seeds are retained in the junction area of the clarification zone and the reaction zone of the reactor; this leads to the decreased number of seeds in the reaction zone and surface area of seeds per volume [14,15]; and (3) the layer abrasion of seeds is more serious at a higher superficial velocity; therefore, the superficial velocity must be decreased while ensuring the fluidization quality of seeds. Influence of the reaction pH Experiments were conducted at an influent flow of 17 L/h, a precipitant flow of 25 L/h, a reflux flow of 0 L/h, and a Ca/F molar ratio of 1:1. The fluoride concentration of influent was 900 mg/L. The particle size of the seed was 200–400 mesh. Figure 2(d) shows the influence of the reaction pH on the linear growth rate of calcium fluoride.
85
Figure 3. Parity graph for the predicted values of the overall linear growth rate by the proposed semiempirical model of calcium fluoride particle growth in the FBR.
It can be observed that the linear growth rates of calcium fluoride are 2.95 × 10−10 , 5.22 × 10−10 , 3.90 × 10−10 , 5.06 × 10−10 m/s when pH = 6, 7, 8, and 9, respectively. The linear growth rate slightly increases with increasing pH. To analyse the mechanism of the pH effect, the zeta potential of the effluent was measured. The zeta potentials are 32.95, 19.25, 14.05, 14.00 mv when pH = 6, 7, 8, and 9, respectively. The aggregation between the seeds and fine particles is favoured to take place when zeta potential is below 20 mv at pH 7–9. The repulsive force of particles increases at pH 6, which leads to decreased growth rate. Therefore, the growth of calcium fluoride is more favourable at pH 7–9. Growth model of calcium fluoride The general kinetics equation for growth [19] is expressed as follows: G = Kg L0m V n σ j .
(5)
As shown in Figure 2, the growth rate of fluoride calcium is fitted to Equation (5) to obtain the overall crystal growth equation, as shown in Equation (6). G = 3.82 × 10−4 × L00.82 × V −0.86 × [(6.06 × 10−10 − 3.84 × 10−14 )σ −2 ].
(6)
The correlation between the experimental and predicted values of G is shown in Figure 3. The correlation coefficient of the fitting is 0.976, which agrees well with the experimental data. Therefore, Equation (6) can be used to predict the growth rate of calcium fluoride under the following conditions: L0 = 51.26–118.17 μm, V = 0.0059–0.0130 m/s and σ = 25.91–84.09 (fluoride concentration of influent 400–1400 mg/L).
86
K. Jiang et al. It can be observed that the pressure difference in the FBR increases with increasing particle size of pellets. The pressure difference in the FBR can be determined [20] as follows:
Downloaded by [University of Auckland Library] at 17:49 16 October 2014
P =
Figure 4. Relationship between the particle size of calcium fluoride and the pressure difference in the FBR.
Characterization of the growth process of calcium fluoride Experiments were conducted at an influent flow of 17 L/h, a precipitant flow of 25 L/h, a reflux flow of 0 L/h, a Ca/F molar ratio of 1:1, L0 = 51.26 μm, and reaction pH 7 ± 0.2. The fluoride concentration of the influent was 900 mg/L. The relationship between particle size of pellets and the pressure difference in the FBR was determined during the operation, as shown in Figure 4.
Figure 5.
H (1 − ε)(ρp − ρl )g = (1 − ε)(mp − ml )g , S
(7)
where H is the height of bed; ε is the porosity of bed; ρp and ρl are the density of solid and liquid, respectively; mp is the mass of solid; ml is the mass of the liquid with the same volume as the solid; and S is the cross-sectional area of the reaction zone of the FBR. Assuming that the heterogeneous nucleation of calcium fluoride is dominating, and that fine particles are adsorbed on the seed surface, the relationship between the particle size (L) and mp is as follows: mp ∝ L3 .
(8)
Assuming that the homogeneous nucleation of calcium fluoride is dominant, the change in solid mass is attributed to the increase in the number of particles. By contrast, the mean particle size decreases due to the formation of fine particles. Therefore, the relationship between L and mp is as follows: (9) mp ∝ Lα (0 < α < 3). According to Equations (7)–(9), the relationship between L and P is as follows: P = γ + βLα .
SEM images of calcium fluoride precipitated in the FBR (V = 0.0130 m/s, operation time = 16 h).
(10)
Environmental Technology The data in Figure 4 are fitted to Equation (10), as shown in Equation (11). 1.624 P L = 0.602 + 0.402 × . (11) P0 L0
Downloaded by [University of Auckland Library] at 17:49 16 October 2014
It can be seen that the correlation coefficients is 0.998, which indicates that the experimental data match well with the fitting equation. Therefore, the pressure difference in the FBR can be used to estimate the particle size of pellets and to characterize the growth rate of calcium fluoride. The operation state of the growth process can be indirectly monitored, which is significant in maintaining process operation. Growth mechanism of calcium fluoride To analyse the growth mechanism of calcium fluoride, the pellets obtained at V = 0.0130 m/s and σ = 54.03 in Section 3.1.3 were observed by SEM, as shown in Figure 5. The area surrounded by the white line shows the locally amplified part of the image. It can be seen that both fine particles and the seeds are retained in the reactor during operation. The fine particles range in size from 0.5 to 2.5 μm, which is attributed to homogeneous precipitation. The seed surface is rough with some fine particles. When the seeds are kept suspended in a fluidized state, the ‘seed zone’ with a high holdup is formed in the reaction zone of the reactor. Meanwhile, the fine particles axially rise and pass through the ‘seed zone’. Under low zeta potential in solution (less than 20 mv), some fine particles are adsorbed onto the surface of seeds [18] by fluid shear stress and collision/contact interaction, which is attributed to the low repulsive force of particles (Figure 5(b)–(c)). By contrast, the other unadsorbed fine particles are retained in solution (Figure 5(c)) and then flow out of the reactor. Therefore, the formation of fine particles is unavoidable at high supersaturation due to discrete precipitation and primary nucleation. The adsorption and aggregation between seeds and fine particles is the main mechanisms for growth of calcium fluoride, which differs from the mechanisms of aggregation and molecular growth at low supersaturation.[14–16] Conclusions The growth kinetics of calcium fluoride was studied at high supersaturation in an FBR, and the main conclusions are as follows: (1) Supersaturation (σ ), particle size of seed (L0 ), superficial velocity (V ), and reraction pH are the main factors influencing the growth rate of calcium fluoride. When the seeds are maintained with a high fluidization quality, the growth rate of calcium fluoride is higher with large seed size, lower superficial velocity, and lower supersaturation.
87
(2) When L0 = 51.26–118.17 μm, V = 0.0059– 0.0130 m/s, σ = 25.91–84.09, and reaction pH = 7, the kinetics model given by the equation G = 3.82 × 10−4 × L00.82 × V −0.86 × [(6.06 × 10−10 − 3.84 × 10−14 )σ −2 ] can be used to predict the growth rate of calcium fluoride. (3) The pressure difference is well fitted with the particle size of pellets by the equation P/P0 = 0.602 + 0.402 × (L/L0 )1.624 . The pressure difference of the FBR can be used to predict the particle size of pellets and to characterize the growth rate of calcium fluoride. (4) The formation of fine particles is unavoidable at high supersaturation. The calcium fluoride grows in size by adsorption and aggregation. This finding demonstrates the feasibility of calcium fluoride recovery from high fluoride containing wastewater in the FBR. Acknowledgements This research was supported by the Major Science and Technology Program of Hunan (NO. 2009FJ-1009).
References [1] Turner BD, Binning P, Stipp SLS. Fluoride removal by calcite: evidence for fluorite precipitation and surface adsorption. Environ Sci Technol. 2005;39:9561–9568. [2] Wu J, Lu ZY. Combined treatment process for high fluoride content wastewater. China Water Wastewater. 2003;19:3–4. [3] Aldaco R, Irabien A, Luis P. Fluidized bed reactor for fluoride removal. Chem Eng J. 2005;107:113–117. [4] Van den Broeck K, Van Hoornick N, Van Hoeymissen J, de Boer R, Giesen A, Wilms D. Sustainable treatment of HF wastewaters from semiconductor industry with a fluidized bed reactor. IEEE T. Semiconduct. M. 2003;16:423–428. [5] Aldaco R, Garea A, Irabien A. Calcium fluoride recovery from fluoride wastewater in a fluidized bed reactor. Water Res. 2007;41:810–818. [6] Gouidera M, Fekia M, Sayadib M. Separative recovery with lime of phosphate and fluoride from an acidic effluent containing H3PO4, HF and/or H2SiF6. J Hazard Mater. 2009;170:962–968. [7] Lisbona D, Steel KM. Recovery of fluoride values from spent pot-lining: precipitation of an aluminium hydroxyfluoride hydrate product. Sep Purif Technol. 2008;61:182–192. [8] Lee TC, Liu FJ. Recovery of hazardous semiconductorindustry sludge as a useful resource. J Hazard Mater. 2009;165:359–365. [9] Wang LS, Wang CM, Yu Y, Huang XW, Long ZQ, Hou YK, Cui DL. Recovery of fluorine from bastnasite as synthetic cryolite by-product. J. Hazard. Mater. 2012;209–210:77–83. [10] Giesen A. Crystallisation process enables environmental friendly phosphate removal at low costs. Environ Technol. 1999;20:769–775. [11] Battistoni P, Boccadoro R, Fatone F, Pavan P. Autonucleation and crystal growth of truvite in a demonstrative fluidized bed reactor (FBR). Environ Technol. 2005;26: 975–982. [12] Lee MS, Chang WK, Shao H, Liao CC. Volume reduction and resource recovery of fluoride-containing wastewater.
88
K. Jiang et al.
Downloaded by [University of Auckland Library] at 17:49 16 October 2014
IWA Sludge Management Entering the 3rd Millennium, Taipei, Taiwan, 2001. [13] Rubén A, Aurora G, Angel I. Fluoride recovery in a fluidized bed: crystallization of calcium fluoride on silica sand. Ind Eng Chem Res. 2006;45:796–802. [14] Tai CY, Chen PC, Tsao TM. Growth kinetics of CaF2 in a pH-stat fluidized-bed crystallizer. J Cryst Growth. 2006;290:576–584. [15] Aldaco R, Garea A, Irabien A. Modeling of particle growth: application to water treatment in a fluidized bed reactor. Chem Eng J. 2007;134:66–71.
[16] Aldaco R, Garea A, Irabien A. Particle growth kinetics of calcium fluoride in a fluidized bed reactor. Chem Eng Sci. 2007;62:2958–2966. [17] Mersmann A. Crystallization technology handbook. 2nd ed. New York: CRC Press; 2001. [18] Ding XH, Tan D. Industrial crystallization. Beijing: Chemical Industry Press; 1985. [19] Bravi M, Mazzarotta B. Size dependency of citric acid monohydrate growth kinetics. Chem Eng J. 1998;70:203–207. [20] Guo MS, Li HZ. Handbook of fluidization. Beijing: Chemical Industry Press; 2008.
