Lithium and rubidium extraction from zinnwaldite by alkali digestion process: Sintering mechanism and leaching kinetics

International Journal of Mineral Processing 123 (2013) 9–17

Contents lists available at ScienceDirect

International Journal of Mineral Processing journal homepage: www.elsevier.com/locate/ijminpro

Lithium and rubidium extraction from zinnwaldite by alkali digestion process: Sintering mechanism and leaching kinetics H. Vu a,⁎, J. Bernardi b, J. Jandová a, L. Vaculíková c, d, V. Goliáš e a

Department of Metals and Corrosion Engineering, Prague Institute of Chemical Technology, Prague, Czech Republic Ecole Nationale Supérieure des Industries Chimiques de Nancy, Nancy, France Institute of Geonics of the AS CR, Ostrava-Poruba, Czech Republic d Institute of Clean Technologies for Mining and Utilization of Raw Materials for Energy Use, Ostrava-Poruba, Czech Republic e Institute of Geochemistry, Mineralogy and Mineral Resources, Charles University, Prague, Czech Republic b c

a r t i c l e

i n f o

Article history: Received 17 August 2012 Received in revised form 6 April 2013 Accepted 30 April 2013 Available online 11 May 2013 Keywords: Zinnwaldite Lithium Rubidium Sintering mechanism Leaching kinetics

a b s t r a c t Lithium and rubidium were extracted from zinnwaldite [KLiFe2+Al(AlSi3)O10(F,OH)2] by (1) its sintering with and CaCO3 powders and (2) water leaching the obtained sinters—the alkali digestion process. The experimental results showed that sintering proceeded in three partly overlapping stages: (1) decomposition of zinnwaldite at temperature up to 800 °C, (2) formation of new phases in the temperature range between 750 and 835 °C, and (3) formation of amorphous glassy phase at temperature above 835 °C. Densification of the reaction mixture occurred via a liquid phase sintering at temperatures above 750 °C and diffusion of calcium, potassium, silicon and rubidium resulted in the formation of the new phases. The decomposition of zinnwaldite and the formation of the new phases increased extraction of lithium and rubidium. The formation of glassy phase probably hindered extraction of lithium but did not affect that of rubidium because of its outward diffusion to sinter's surface. The optimal extraction efficiencies of 84% of lithium and 91% of rubidium were achieved at sintering temperature of 825 °C and leaching temperature of 95 °C. The good fit of the hyperbolic and uniform reaction models to the leaching data indicated that dissolution of lithium and rubidium proceeded through two stages. Application of the shrinking core model showed that dissolution of lithium was controlled by diffusion. The formation of a layer of Ca(OH)2 on surface of sinters apparently slowed and then terminated dissolution of lithium and rubidium in the later stage of leaching. © 2013 Elsevier B.V. All rights reserved.

1. Introduction As the lightest metal in nature with the unique electrochemical properties, lithium is becoming the key material in manufacturing of primary and secondary batteries used in various portable devices and hybrid/ electric vehicles (Garret, 2004; Harben, 2002; Thompson, 2011). The annual global demand, measured as lithium carbonate equivalent, is expected to increase from current 100,000 to 160,000 Mt in 2015, with batteries representing about 40,000 Mt of the perceived growth (Hykawy, 2010). Other main applications of lithium include aluminum refining; glass and ceramics production; synthesis of rubber, plastics, pharmaceuticals and organic compounds; and manufacture of specialty greases and desiccants. Rubidium has been available commercially as a by-product of lithium production from minerals (Thompson, 2011). The main use of rubidium is in atomic clocks for global positioning

⁎ Corresponding author. Tel.: +420 2 2044 5025; fax: +420 2 2044 4400. E-mail address: [email protected] (H. Vu). 0301-7516/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.minpro.2013.04.014

satellites, magneto-optic modulators, solid-state lasers, phosphors, and glass manufacturing. Rubidium-based chemicals are also used as catalysts or to activate catalysts of various types. Total world demand for rubidium was estimated at about 2 to 4 tones per year but increases in lithium exploration may create new supplies leading to expanded commercial applications (Thompson, 2011; Wagner, 2006). The main commercially viable sources of lithium are natural brines with a high lithium chloride content and the pegmatitic minerals such as spodumene, amblygonite, petalite and lepidolite (Garret, 2004; Ebensperger et al., 2005; Wietelmann and Bauer, 2008). Given the mining, processing and energy requirements, lithium ores contribute only 15% of current lithium production (Roskill Information Services (2009)). As world lithium demand, driven primarily by increased production of batteries, is expected to grow about two fold in the next 10 years and may exceed supply by 2017, minerals get more attention as an alternative source of lithium to offset the deficit (Chen et al., 2011; Dundeen Securities, 2009; McNulty and Khaykin, 2009; Sitando and Crouse, 2012; Yan et al., 2012). Zinnwaldite [KLiFe2+Al(AlSi3) O10(F,OH)2], a variety of lepidolite with high iron content, is one of the minerals of interest that can be exploited to obtain both lithium

H. Vu et al. / International Journal of Mineral Processing 123 (2013) 9–17

and rubidium (Jandova et al., 2010; Jandova and Vu, 2008; Siame and Pascoe, 2011). Lithium is extracted from minerals by high-temperature digestion with acid or alkali, or by ion-exchange processes using salt, followed by acid or water leaching. Rubidium is obtained from a solution containing mixed alkali metal carbonates, a by-product of the production of lithium from lepidolite, by precipitation followed by pyrolysis or thermal oxidation (Habashi, 1997; Kamienski et al., 2005; Wietelmann and Bauer, 2008). The alkali digestion process, originally developed for spodumene and pentalite, was investigated to process zinnwaldite in recent research studies. Our works have shown that about 90% of lithium and rubidium was extracted by sintering a zinnwaldite concentrate with calcium carbonate (weight ratio 1:5) at 825 °C for 60 min and water-leaching the pulverized sinter (liquid to solid ratio 10:1) at 90 °C for 30 min (Jandova and Vu, 2008; Jandova et al., 2009, 2010). The concentrate was obtained by magnetic classification of lithium-rich sand, a byproduct of Sn–W ores processing at Cinovec, Czech Republic. In the contrary, Siame and Pascoe (2011) reported that under the same conditions less than 1% of lithium was recovered from a zinnwaldite concentrate obtained from micas-rich sand, a by-product from china clay production in South–West England. Siame and Pascoe (2011) suggested that formation of eucryptite [LiAlSiO4] and its increased crystallinity with rising sintering temperature might cause the decrease in lithium extraction. But according to Habashi (1997) the formation of an amorphous glassy clinker at high temperature severely hinders extraction of lithium from minerals by the alkali digestion method. In the above mentioned works the mechanism of sintering zinnwaldite with calcium carbonate still remains unclear and the leaching behavior of lithium and rubidium from sinters was not fully studied. The present study was designed to provide an insight into the sintering process and to investigate the leaching kinetics using mathematical models.

2. Experimental A zinnwaldite concentrate containing 1.29% of Li and 0.94% of Rb was prepared by dry magnetic separation of the lithium-rich sand, the by-product after Sn–W ore mining in the Czech Republic. According to the results of XRD analysis, the concentrate consisted of two main phases: dominant zinnwaldite KLiFe2+Al(AlSi3)O10(F,OH)2 and minor quartz SiO2. The concentrate was pulverized in a ball mill for 2 min and mixed with CaCO3 of technical grade at a weight ratio of 1:5. The results of the particle size analysis showed that the size of the pulverized zinnwaldite concentrate varies between 1 and 100 μm with a mean diameter of 30 μm. The mixture was sintered in a muffle furnace at different temperatures for 60 min. The sinter was milled in the ball mill for 1 min and subjected to water leaching. The leaching tests were performed at temperatures ranging from 26 to 95 °C, liquid to solid ratio (l:s) 10:1, stirrer speed of 400 rpm, and leaching time of up to 240 min. The influence of sintering temperature on extraction of lithium and rubidium was determined by leaching sinters at 95 °C for 60 min. After leaching tests finished, the leach residues were separated by filtration, water-washed and dried at laboratory temperature. The lithium and rubidium extraction efficiencies were calculated from the total amount of Li or Rb released into leaching liquor and washing water. The leaching programme was followed by sampling of the solution at a chosen time interval and measuring Li and Rb concentrations by a GBC atomic absorption spectrophotometer (model GBC 932 plus). Mineralogical analysis was carried out by X-ray diffractometer PANalytical's X'Pert PRO. SEM-EDX analysis was carried out using Hitachi S4700 Scanning Electron Microscope or Vega 3 Tescan Scanning Electron Microscope. Differential thermal analysis (DTA) was performed by TA Setaram, model Setsys Evolution in a flowing in argon atmosphere at a heating rate of 10 °C/min. The thermobalance was connected to a mass spectrometer Pfeiffer Vakuum, model OmniStar. The infrared spectra of zinnwaldite and sinters were recorded on FTIR spectrometer

Nicolet 6700 (Thermo Fisher Scientific, USA). The KBr pressed-disc technique was used for routine scanning of the spectra.

3. Results and discussion 3.1. Sintering process A similar sharp increase in extraction of both lithium and rubidium was found when sintering temperatures increased from 750 °C to 800 °C, indicating a rapid decomposition of zinnwaldite in this temperature interval (Fig. 1). Lithium extraction efficiency remained almost constant in the temperature interval from 810 °C to 835 °C, with the maximum of 84% at about 825 °C, and decreased rapidly at temperature above 835 °C. In contrast, extraction efficiency of rubidium increased slightly with increased temperature and leveled off at 95% from 850 °C. Different courses of extraction of lithium and rubidium indicated their different behavior during sintering. X-ray powder pattern analysis of sinters shows a decrease in crystallinity of zinnwaldite with increasing sintering temperatures and a shift of diffraction peaks of zinnwaldite [KLiFe 2+Al(AlSi3)O10(F,OH)2] toward those of polylithionite [KLi2AlSi4O10(F,OH)2] at temperatures higher than 750 °C (Fig. 2a–b). At 800 °C, diffraction peaks of zinnwaldite and polylithionite disappeared, indicating their complete decomposition; the presence of hematite Fe2O3 was detected; and the formation of mainly calcium carbonate-bearing minerals was observed. Most diffraction peaks, except those of lime—CaO and spurrite—Ca5(SiO4)2CO3, decreased in intensity with further increase in temperature and disappeared at temperature higher than 835 °C, suggesting the formation of amorphous glassy phase (Fig. 2b–c). Lithium and rubidium-bearing phases were not observed in diffraction patterns, probably because their content in the sinters (~0.2%) is below the detection limit of XRD technique. Rubidium may have substituted potassium in KCaCO3F because of their similarities in chemical and physical properties. The TG-DTA curves of zinnwaldite show two peaks: an exotherm in the temperature range 450–825 °C, followed by a sharp and narrow endotherm with the maximum around 910 °C (Fig. 3). The MS pattern shows that the H2O evolution took place in two minor steps at the temperature ranges of 210 – 300 °C and 400 – 700 °C, and followed a larger step around 930 °C. These three steps of H2O evolution were characterized by a weight loss of about 0.20, 0.25 and 0.75 wt%. The exothermic peak can be contributed to the formation of Fe2O3 by oxidation of divalent iron with oxygen released from dehydroxylation according to the following reaction: –OH− + –OH− = H2O + O 2− (Gallagher, 2003).

100 90

Extraction efficiency (%)

10

80 70 60 50 40 30 20

Li

Rb

10 0 680 700 720 740 760 780 800 820 840 860 880 900

Fig. 1. Influence of sintering temperature on extraction efficiencies of Li and Rb: zinnwaldite: CaCO3 = 1:5, sintering time 60 min, l:s = 10:1, leaching temperature 95 °C, leaching time 60 min.

H. Vu et al. / International Journal of Mineral Processing 123 (2013) 9–17

11

a) Z - KLiFe2+Al(AlSi3)O10(F,OH)2 C - CaCO3 P - KLi2AlSi4O10(F,OH)2 Q – SiO2

b)

A - Ca(Fe,Mg)(CO3)2 C1 - (CaO)4Fe2O3Al2O3

S - Ca5(SiO4)2CO3 P1 -KCaCO3F

L - CaO I - Fe2O3

c) P2 - Ca(OH)2 D - CaMgCO3

Fig. 2. X-ray powder patterns of sinters obtained at sintering temperatures a) 400–700 °C, b) 750–825 °C and c) 810–870 °C: zinnwaldite:CaCO3 = 1:5, sintering time 60 min.

12

H. Vu et al. / International Journal of Mineral Processing 123 (2013) 9–17

Fig. 3. TG/DTA/MS analysis of decomposition of zinnwaldite [KLiFe2+Al(AlSi3)O10(F,OH)2] in Ar, heating rate 10 °C/min.

Fig. 4. FTIR spectra of zinnwaldite and sinters obtained at 825 and 870 °C.

The formation of Fe2O3 as a discrete crystalline phase evidently caused a change in the mineralogical structure of zinnwaldite to that of polylithionite, a mica mineral without iron. The endothermic peak was attributed to dehydroxylation, lattice destruction and melting of zinnwaldite (Vandermarel, 1956; Vedder and Wilkins, 1969). The formation of glass phase probably caused the decrease in extraction efficiency of lithium at sintering temperature above 835 °C. The infrared spectra of zinnwaldite and the sinters obtained at 825 and 870 °C show a disappearance of weak O–H stretching bands near 3568, 3590, 3624 and 3645 cm−1, and a decrease in intensity of the O–H stretching (near 3440 and 2919 cm−1) and O–H bending (near 1640 cm−1) bands with increasing sintering temperature (Fig. 4). It was the evidence of the dehydroxylation in zinnwaldite, in an agreement with results from the TG-DTA analysis. A new sharp O–H stretching band near 3640 cm−1 in sinters was attributed to Ca(OH)2, formed by reaction of H2O from the atmosphere with CaO from decomposition of CaCO3. The bands near 1444 and 1470 cm−1 were attributed to CO32−, confirming the observation of carbonate compounds in the results from X-ray analysis. The bands between 400 and 570 cm−1 are mainly assigned to bending vibrations of SiO4 and AlO6, the bands between 700 and 850 cm−1 are attributed to T–O–T (T = Si/Al) bending vibrations, and vibrations of SiO4 group and T–O–T stretching bands are located between 900 and 1200 cm−1 (Zhang et al., 2005). The disappearance of the bands in zinnwaldite and the appearance of the new bands in sinters between 400 and 1200 cm−1 indicate the lattice destruction of zinnwaldite with increasing sintering temperature. SEM pictures of cross-section of sinters show that sintering mixture experienced densification starting at 750 °C when calcium carbonate melted, flowed around and bridged zinnwaldite particles (Fig. 5). This observation indicates that the formation of calcium carbonate-bearing minerals resulted from solid–liquid reactions between calcium carbonate and intermediate products of zinnwaldite decomposition. SEM-EDX mapping of zinnwaldite and selected sinters shows the inward diffusion of calcium and the simultaneous outward diffusion of potassium toward the surface of the zinnwaldite particle (Fig. 6a–c). A sharp boundary between a reacted outer layer and an unreacted core indicates that sintering proceeded according to the shrinking-core model. Besides calcium and potassium, a partial diffusion of silicon toward the surface was also observed. Rubidium and lithium were not detected because lithium

Fig. 5. SEM micrographs showing cross sections of sinters obtained at: a) 750 °C, top right - CaCO3 melts, flows around and bridges zinnwaldite particles; b) 780 °C, CaCO3 covers a zinnwaldite particle; c) 870 °C, a hollow structure as a result of CaCO3 decomposition.

Fig. 6. SEM/EDX micrographs of cross sections of zinnwaldite and sinters obtained at 780 and 825 °C: a) zinnwaldite, homogenous elemental distribution; b) 780 °C, inward diffusion of Ca, outward diffusion of K and Si; c) 825 °C, complete diffusion of Ca and K, partly outward diffusion of Si.

H. Vu et al. / International Journal of Mineral Processing 123 (2013) 9–17

13

14

H. Vu et al. / International Journal of Mineral Processing 123 (2013) 9–17

Fig. 7. SEM/EDX micrograph of cross sections of a SiO2 particle surrounded by CaSiO3.

1 0,9

Extraction yield of Li

0,8 0,7 0,6 0,5 0,4 0,3 0,2

W

H

E

U

0,1 0 0

20

40

60

80

100

120

140

160

180

Time (min) Fig. 8. Comparisons of leaching models to experimental data of Li: W - Weibull's model; H - Hyperbolic model; E - Elovitch's equation; U - Uniform reaction model.

1 0,9

Extraction yield of Rb

0,8 0,7 0,6 W

H

E

U

1

0,5

70

oC

0,9

0,8

0,4

0,7

0,6

0,3

05

50 o C

0,9

0,2

0,8

0,7

0,1

0,6

0,5

0 0

20

40

60

80

100

120

140

160

180

Time (min) Fig. 9. Comparisons of leaching models to experimental data of Rb: W - Weibull's model; H - Hyperbolic model; E - Elovitch's equation; U - Uniform reaction model.

H. Vu et al. / International Journal of Mineral Processing 123 (2013) 9–17

15

Table 1 Parameters of models for leaching Li and Rb from zinnwaldite sinters. Model

Constants

Leaching temperature 95 °C

Weibull's model X B ¼ 1− exp−


t m δ

δ m S r C1 C2 C1/C2 S r A B S r XB,1 k1 XB,2 k2 S r

Hyperbolic model C 1t X B ¼ 1þC 2t

Elovitch's equation (Logarith Fit) XB = A + Bln(t)

Uniform reaction model XB = XB,1(1 – e-k1t) + + XB,2(1 – e-k2t) + + (1 – XB,1 – XB,2)

70 °C

50 °C

26 °C

Li/Rb

Li/Rb

Li/Rb

Li/Rb

0.897/0.661 0.075/0.122 0.010/0.019 0.997/0.933 1.089/4.216 1.378/4.614 0.789/0.913 0.006/0.019 0.9997/0.935 0.677/0.804 0.025/0.028 0.021/0.023 0.814/0.900 0.228/0.180 0.0001/0.281 0.224/0.097 0.189/0.002 0.005/0.009 0.993/0.990

1.395/0.992 0.112/0.111 0.025/0.015 0.993/0.976 0.419/2.220 0.592/2.824 0.707/0.786 0.004/0.025 0.9998/0.931 0.511/0.639 0.060/0.038 0.027/0.016 0.889/0.969 0.348/0.186 0.0009/0.224 0.651/0.236 0.314/0.0016 0.016/0.010 0.998/0.991

2.915/1.210 0.198/0.139 0.031/0.023 0.988/0.965 0.1639/1.351 0.259/1.779 0.633/0.759 0.011/0.030 0.998/0.938 0.394/0.565 0.0425/0.486 0.014/0.025 0.976/0.959 0.450/0.251 0.0006/0.0009 0.550/0.252 0.174/0.170 0.011/0.011 0.999/0.994

4.690/4.019 0.270/0.353 0.046/0.043 0.970/0.978 0.067/0.129 0.121/0.174 0.553/0.741 0.012/0.032 0.997/0.987 0.113/0.174 0.088/0.115 0.040/0.035 0.938/0.985 0.521/0.377 0.0007/0.0021 0.470/0.482 0.096/0.100 0.010/0.026 0.999/0.994

XB: the fraction reacted, t: leaching time, S: the standard error, r: the correlation coefficient.

a)

b)

1-(1 – XB)1/3

95°C

70°C

50°C

1+2(1-XB) -3(1- XB)2/3

0.5 26°C

0.4 0.3 0.2 0.1

0.5 95°C

0.4

70°C

50°C

40

50

26°C

0.3 0.2 0.1 0

0 0

10

20

30

40

50

60

70

Time (min)

0

10

20

30

60

70

Time (min)

Fig. 10. Plots of 1 − (1 − XB)1/3 and 1 − 3(1 − XB)2/3 + 2(1 − XB) versus t for Li at different leaching temperatures: sinter obtained at 825 °C, l:s = 10:1.

Fig. 11. SEM/EDX micrograph of surface of a leached sinter's particle.

is not detectable by the EDX mapping technique, and rubidium and potassium have overlapping peaks. As a substitute for potassium in zinnwaldite structure (Shaw, 1968), rubidium was presumed to diffuse out toward to the sinters' surface. This explanation is consistent with the previous finding that extraction of rubidium, unlike to that of lithium, remained steadily at sintering temperatures above 835 °C, when the melting process was taking place (Fig. 1).

The diffusion of calcium, potassium, and silicon indicates a plausible formation mechanism of some phases in sinters. 4CaO · Al2O3 · Fe2O3 (C4AF), a constituent in Portland cement clinker, was probably formed by the following reaction (Zhu et al., 2011):

4CaO þ Al2 O3 þ Fe2 O3 ¼ 4CaO·Al2 O3 ·Fe2 O3

16

H. Vu et al. / International Journal of Mineral Processing 123 (2013) 9–17

Ca5(SiO4)2CO3 was evidently formed by the reaction between CaCO3 and CaSiO3 as reported in the work of Treiman and Essene (1983). The reaction is 3CaCO3 þ 2CaSiO3 ¼ Ca5 ðSiO4 Þ2 CO3 þ CO2 As shown in Fig. 7, CaSiO3 was formed by the reaction of SiO2, a minor phase in the zinnwaldite concentrate, with CaO from decomposition of CaCO3. The disappearance of SiO2 phase at sintering temperature above 800 °C and the formation of Ca5(SiO4)2CO3 from 780 °C confirmed the hypothesis (Fig. 2b–c). Ca(Fe +2, Mg)(CO3)2 could be formed from calcite by the reaction (Boles, 1978): 2þ

4CaCO3 þ Fe

2þ

þ Mg

2þ

¼ 2CaMg0:5 Fe0:5 ðCO3 Þ2 þ 2Ca

lithium in the later stage is likely controlled by diffusion, apparently through a different layer. The SCMs with reaction and diffusion controls did not fit to the leaching data of rubidium. The plausible reason was that rubidium was supposed to be found at the outer skin of the particle while these models describe a movement of a reaction zone into a solid particle. An SEM-EDX micrograph of a reacted particle shows that its surface was covered by a layer of Ca(OH)2, formed evidently by hydrolysis of CaO (Fig. 11). Mainly Ca(OH)2, larnite Ca2SiO4 and negligible spurrite Ca5(SiO4)2CO3 were found in the X-ray powder analysis of leached residues. The formation of this new layer may have slowed and terminated the release of lithium and rubidium at the later leaching stage when the layer was becoming thicker and more compact with leaching time and temperature. 4. Conclusions

Iron and magnesium, an impurity in zinnwaldite, were apparently released by the breakdown of zinnwaldite. At sintering temperature above 850 °C, all divalent iron probably oxidized and as a result only CaMg(CO3)2 was formed (Fig. 2c). The formation of KCaCO3F can be explained by the outward diffusion of potassium and fluoride into a layer of CaCO3, which covered the surface of sinters' particles. As reported in Chen et al. (2004) and Sun et al. (2006), the following reaction occurred: KF þ CaCO3 ¼ KCaCO3 F 3.2. Leaching process The sinter obtained at 825 °C was used to determine the leaching kinetics of lithium and rubidium. The following models were used to examine leaching behaviors: (1) hyperbolic model, (2) the Weibull's equation, (3) the Elovitch's equation, (4) the uniform reaction model (URM) and (5) the shrinking core model (SCM). The development and verification of these models were previously described in details (Furman et al., 2006; Kim et al., 2002; Kitanovic et al., 2008; Kunii and Levenspiel, 1991; Levenspiel, 1999). The comparisons of empirical models applied to describe experimental results are shown in Figs. 8–9. The models and their calculated quotients are shown in Table 1. Dissolution of lithium and rubidium went through an initial fast stage, followed by a period in which extraction curves increased progressively before achieving a steady-state condition. The rates and the extents of extraction of lithium and rubidium increased with increasing temperature. In general, the extraction efficiency of rubidium was higher than that of lithium at a given temperature. Increasing temperature increased the extraction of lithium more than that of rubidium. The optimal extraction of lithium (84%) and rubidium (91%) was achieved at leaching temperature of 95 °C. The URM with two parameters gave the best fit to the extraction results of both lithium and rubidium. While this model is applicable only for porous materials with the assumption of uniform chemical reaction throughout the solids, the good fit of the model to the leaching data indicates that two different steps controlled the leaching behavior of lithium and rubidium. The hyperbola formula model could be used to describe the leaching behavior of lithium. According to this model, the release of lithium is a nearly first-order initially and rapidly decreases to zero-order with time, confirming that there are 2 different ratecontrolling steps. Plots of 1 − (1 − XB) 1/3 and 1 − 2(1 − XB) − 3(1 − XB) 2/3 versus time for lithium extraction are shown in Fig. 10a–b. The linear relationship at the initial stage of leaching at 26 and 50 °C indicates that the dissolution of lithium is controlled by the diffusion of reactants and products through the shell. In the later stage, linear relationships were obtained for both reaction and diffusion control models. Because reaction rate was controlled by diffusion in the initial stage, the release of

Sintering zinnwaldite with CaCO3 powders proceeds through three partly overlapping stages: (1) decomposition of zinnwaldite at temperature up to 800 °C, (2) formation of new phases in the temperature range from 750 to 835 °C, and (3) formation of glass phase at temperature above 835 °C. In the initial stage, zinnwaldite decomposes gradually without the formation of any intermediate compounds until temperature reaches 750 °C, above which zinnwaldite transforms to polylithionite as the result of dehydroxylation and the formation of Fe2O3. In the second stage, new phases are formed as the result of wetting zinnwaldite particles by liquid calcium carbonate and subsequently the inward diffusion of calcium and the outward diffusion of potassium, silicon and rubidium. The presence of lithium and rubidium is not detected by X-ray and SEM-EDX techniques but rubidium is presumed to substitute potassium in KCaCO3F at the outer skin of sinters and lithium remains in the solid, probably in the form of aluminates as reported in literature (Habashi, 1997). The decomposition of zinnwaldite and the formation of new phases accompany the acceleration of extraction of lithium and rubidium in the temperature range from 750 °C to 800 °C. As temperature increases to 835 °C, the extraction efficiency of lithium stays at around 84% and that of rubidium increases slightly to 92%. In the final stage of sintering, the formation of glassy phase appears to be responsible to the decreased extraction of lithium but does not affect that of rubidium because of its outward diffusion toward the sinter's surface. The good fit of the hyperbolic and URMs to the leaching data demonstrates that dissolution of lithium and rubidium from sinters proceeds through two stages. Application of the SCM indicates that dissolution of lithium is controlled by diffusion. In the later stage, the formation of Ca(OH)2 on the surface of reacted sinters probably slows and terminates dissolution of lithium and rubidium. Acknowledgments Authors wish to thank the Ministry of Education, Youth and Sports of the Czech Republic (projects No. MSM6046137302 and MSMT No. 21/2011). The FTIR analysis was performed in connection with project “Institute of Clean Technologies for Mining and Utilization of Raw Materials for Energy Use”, reg. no. CZ.1.05/2.1.00/03.0082 supported by Research and Development for Innovations Operational Program, financed by Structural Founds of Europe Union and from the means of state budget of the Czech Republic. References Boles, J.R., 1978. Active ankerite cementation in the subsurface eocene of Southwest Texas. Contrib. Mineral. Petrol. 68 (1), 13–22. Chen, X.L., He, M., Xu, Y.P., Li, H.Q., Tu, Q.Y., 2004. KCaF(CO3) from X-ray powder data. Acta Crystallogr. Sect. E 60, I50–I51. Chen, Y., Tian, Q., Chen, B., Shi, X., Liao, T., 2011. Preparation of lithium carbonate from spodumene by a sodium carbonate autoclave process. Hydrometallurgy 109 (1–2), 43–46.

H. Vu et al. / International Journal of Mineral Processing 123 (2013) 9–17 Dundeen Securities, 2009. Lithium - hype or substance? A look at lithium demand and supply. http://research.dundeesecurities.com/Research/Lithium102809.pdf. Ebensperger, A., Maxwell, P., Moscoso, C., 2005. The lithium industry: Its recent evolution and future prospects. Resour. Policy 30 (3), 218–231. Furman, S.A., Scholes, F.H., Hughes, A.E., Lau, D., 2006. Chromate leaching from inhibited primers - Part II: Modelling of leaching. Prog. Org. Coat. 56 (1), 33–38. Gallagher, P.K., 2003. first ed. Handbook of Thermal Analysis and Calorimetry, vol. 2. Elsevier, Amsterdam, pp. 278–279. Garret, D.E., 2004. Part 1 - Lithium, Handbook of Lithium and Natural Calcium Chloride. Academic Press, Oxford, pp. 1–235. Habashi, F., 1997. Handbook of Extractive Metallurgy, vol. 4. Wiley-VCH, Weinheim; New York, pp. 2038–2040. Harben, P.W., 2002. The Industrial Minerals Handbook: A Guide to Markets, Specifications, and Prices. Industrial Minerals Information, UK, pp. 184–192. Hykawy, J., 2010. Looking at lithium. Mater. World 18 (5), 34–35. Jandova, J., Vu, H.N., 2008. Processing of zinnwaldite wastes to obtain lithium and rubidium compounds. Minerals, Metals & Materials Society, pp. 923–929. Jandova, J., Vu, H.N., Belkova, T., Dvorak, P., Kondas, J., 2009. Obtaining Li2CO3 from zinnwaldite wastes. Ceram. Silik. 53 (2), 108–112. Jandova, J., Dvorak, P., Vu, H.N., 2010. Processing of zinnwaldite waste to obtain Li(2) CO(3). hydrometallurgy 103 (1–4), 12–18. Kamienski, C.W., McDonald, D.P., Stark, M.W., Papcun, J.R., 2005. Lithium and Lithium Compounds. John Wiley & Sons, Inc. 120–153. Kim, J.Y., Kim, C.L., Chung, C.H., 2002. Modeling of nuclide release from low-level radioactive paraffin waste: a comparison of simulated and real waste. J. Hazard. Mater. 94 (2), 161–178. Kitanovic, S., Milenovic, D., Vejkovic, V.B., 2008. Empirical kinetic models for the resinoid extraction from aerial parts of St. John's wort (Hypericum perforatum L.). Biochem. Eng. J. 41 (1), 1–11. Kunii, D., Levenspiel, O., 1991. Fluidization engineering. Butterworth-Heinemann Series in Chemical Engineering. Butterworth-Heinemann, Boston, pp. 449–451. Levenspiel, O., 1999. Chemical Reaction Engineering. Wiley, New York 556–559. McNulty, J.P., Khaykin, A., 2009. Lithium - extracting the details of the lithium market. Credit Suisse. http://www.docstoc.com/docs/12415608/Lithium (accessed04.05.12).

17

Roskill Information Services, 2009. The Economics of Lithium, 11th ed. Roskill Information Services, London. Shaw, D.M., 1968. A review of K-Rb fractionation trends by covariance analysis. Geochim. Cosmochim. Acta 32 (6), 573–601. Siame, E., Pascoe, R.D., 2011. Extraction of lithium from micaceous waste from china clay production. Miner. Eng. 24 (14), 1595–1602. Sitando, O., Crouse, P.L., 2012. Processing of a Zimbabwean petalite to obtain lithium carbonate. Int. J. Miner. Process. 102–103, 45–50. Sun, Y.P., Huang, Q.Z., Wu, L., He, M., Chen, X.L., 2006. A neutron powder investigation of the structure of KCaCO3F at various temperatures. J. Alloys Compd. 417 (1–2), 13–17. Thompson, C.A., 2011. Mineral Commodity Summaries 2011. Nova Science Publishers, Hauppauge, N.Y. Treiman, A.H., Essene, E.J., 1983. Phase equilibria in the system CaO-SiO2-CO2. Am. J. Sci. 283(A), 97–120. Vandermarel, H.W., 1956. Quantitative differential thermal analyses of clay and other minerals. Am. Mineral. 41 (3–4), 222–244. Vedder, W., Wilkins, R.W.T., 1969. Dehydroxylation and rehydroxylation, oxidation and reduction of micas. Am. Mineral. 54 (3–4), 482–509. Wagner, F.S., 2006. Rubidium and Rubidium Compounds. John Wiley & Sons, Inc. 816–825. Wietelmann, U., Bauer, R.J., 2008. Lithium and Lithium Compounds, seventh ed. Ullmann's Encyclopedia of Industrial Chemistry, vol. 20. Wiley-VCH, pp. 33–60. Yan, Q., Li, X., Wang, Z., Wu, X., Wang, J., Guo, H., Hu, Q., Peng, W., 2012. Extraction of lithium from lepidolite by sulfation roasting and water leaching. Int. J. Miner. Process. 110–111, 1–5. Zhang, M., Wang, L., Hirai, S., Redfern, S.A.T., Salje, E.K.H., 2005. Dehydroxylation and CO2 incorporation in annealed mica (sericite): an infrared spectroscopic study. Am. Mineral. 90 (1), 173–180. Zhu, Z., Jiang, T., Li, G., Guo, Y., Yongbin Yang, Y., 2011. Thermodynamics of reactions among Al2O3, CaO, SiO2 and Fe2O3 during roasting processes. In: Moreno-Pirajan, J.C. (Ed.), Thermodynamics - Interaction Studies - Solids, Liquids and Gases. Intech Publisher, pp. 835–836.