Recovery of lithium carbonate by acid digestion and hydrometallurgical processing from mechanically activated lepidolite

Accepted Manuscript Recovery of lithium carbonate by acid digestion and hydrometallurgical processing from mechanically activated lepidolite

Nathália Vieceli, Carlos A. Nogueira, Manuel F.C. Pereira, Fernando O. Durão, Carlos Guimarães, Fernanda Margarido PII: DOI: Reference:

S0304-386X(17)30441-3 doi:10.1016/j.hydromet.2017.10.022 HYDROM 4681

To appear in:

Hydrometallurgy

Received date: Revised date: Accepted date:

29 May 2017 11 October 2017 20 October 2017

Please cite this article as: Nathália Vieceli, Carlos A. Nogueira, Manuel F.C. Pereira, Fernando O. Durão, Carlos Guimarães, Fernanda Margarido , Recovery of lithium carbonate by acid digestion and hydrometallurgical processing from mechanically activated lepidolite. The address for the corresponding author was captured as affiliation for all authors. Please check if appropriate. Hydrom(2017), doi:10.1016/ j.hydromet.2017.10.022

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Recovery of lithium carbonate by acid digestion and hydrometallurgical processing from mechanically activated lepidolite

Nathália Viecelia,* , Carlos A. Nogueira b , Manuel F. C. Pereira c, Fernando O. Durão c, Carlos

Center for Innovation, Technology and Policy Research – IN+, Instituto Superior Técnico, University of

RI

a

PT

Guimarãesc, Fernanda Margarido a

SC

Lisbon, 1049-001 Lisboa, Portugal. b LNEG - Laboratório Nacional de Energia e Geologia, I.P., Campus do Lumiar, 1649-038 Lisboa, Portugal. cCERENA – Centro de Recursos Naturais e Ambiente, Instituto Superior

NU

Técnico, University of Lisbon, 1049-001 Lisboa, Portugal.

MA

*[email protected]

ED

Abstract

Lithium extraction from hard-rock ores has regained importance due to the increased demand for this metal to supply the growing battery market. Therefore, several studies have been focused on t he lithium extraction from

EP T

ores, however, leaching and purification steps are sparsely studied. Thus, the main objective of this study was to evaluate the main factors affecting the water leaching step and the subsequent purification operations for lithium recovery from a lepidolite concentrate, which was processed by mechanical activation and sulphuric acid digestion. In the leaching step, among the variables studied, only one, the leaching temperature, showed a

AC C

significant effect on the lithium extraction, taking into account the range of values tested. Thus, the recommended operating value for the leaching time and the L/S ratio is the minimum, while for the leaching temperature is 50ºCAfter optimizing the leaching operation, the purification of the leachate obtained, by neutralization, was thoroughly performed by efficient removal of impurities (Fe, Al, Mn and Ca), allowing to obtain lithium carbonate as final product, as well as other relevant by-products, such as rubidium and potassium alums.

ACCEPTED MANUSCRIPT Keywords: lepidolite, lithium extraction, leaching, precipitation, purification.



PT

Highlights

New process for recovery lithium from lepidolite by mechanical activation, acid digestion and

RI

hydrometallurgical processing. Optimization of the water leaching step;



Detailed study of the purification step by neutralization and carbonate precipitation;



Lithium carbonate obtainment, as well as other by-products.

AC C

EP T

ED

MA

NU

SC



ACCEPTED MANUSCRIPT 1. Introduction Lithium batteries are increasingly used in a large number of application s, from small devices appliances, such as portable electronics, to electric transportation in large scale and stationary grid storage. The secondary battery market is dominated by such batteries , due to their characteristics, such as high energy density, lightweight, long storage life and other superior performance compared to other batteries (Meshram et al., 2016; Julien et al., 2016). Rechargeable batteries were the largest potential growth area for lithium compounds and their consumption for

PT

batteries has increased significantly: in 2016, the global end-use in this market was estimated at 35%, while in 2010 represented 23% of the global end-use market (Jaskula, 2016). Moreover, between 2014 and 2015, new

RI

registrations of electric cars (including both electric and plug-in hybrids) increased by 70%, with over 550000 vehicles being sold worldwide in 2015 (OECD/IEA, 2016), which has increasingly boosted the demand for this

SC

metal.

In the late 1990s, subsurface brines became the main raw material for lithium carbonate production worldwide,

NU

due to the lower production costs compared with the mining and processing of hard-rock ores. However, given the growing lithium demand in the past years, hard-rock ores have regained market share (Jaskula et al. 2016). After mining, processing of lithium ores involves their comminution, followed by beneficiation using

MA

techniques such as flotation, optical sorting, magnetic separation or heavy media separation, in order to upgrade the lithium content and to produce concentrates. Then, these concentrates can be subjected to roasting and leaching to extract lithium into solution (Chagnes and Swiatowska, 2015; Brandt & Haus, 2010; Siame &

ED

Pascoe, 2011; Banks et al.¸ 1953; Vieceli et al., 2017a; Menéndez et al., 2004; Munson and Clarke; 1955; Orocobre, 2012).

EP T

Lithium ores have been traditionally submitted to a pre-treatment step of calcination or decrepitation followed by acid or alkaline digestion (Meshram et al., 2014; Averill & Olson, 1978; Wietelmann & Bauer, 2000). According to Li et al. (2016), traditional methods for extracting lithium frequently have high costs due to the

AC C

inefficient exploitation of other metals present in the ores. Lithium extraction from ores involving a single roasting step using additives has become increasingly important because the time and temperature of the roasting process are reduced. In this context, several studies focused on roasting of lithium ores, with a wide variety of additives have been published (Hien-Dinh et al., 2015; Luong et al., 2013; Luong et al., 2014; Yan et al., 2012a-d; Kondás & Jandová 2006; Jandová et al., 2009; Vu et al., 2013; Jandová et al., 2010; Sitando & Crouse, 2012; Barbosa et al., 2014; Siame & Pascoe, 2011; Vieceli et al., 2017a). Recently, an alternative process for extracting lithium from a lepidolite ore, constituted by mechanical activation and sulphuric acid digestion, avoiding the calcination step, was proposed and studied in detail (Vieceli et al., 2017b). The effects of several factors on the structural transformations of the mineral ph ases were evaluated in this study. In the present paper, the subsequent operations of the mechanical activation and acid digestion are addressed, namely the water leaching, the purification of leachates and the production of lithium carbonate. Regarding water leaching, most of the published work is not dedicated in detail to this operation, which is

ACCEPTED MANUSCRIPT sparsely studied and the variables affecting this operation are normally fixed or studied one at a time, which does not allow a proper assessment of the possible interaction between the effects. Therefore, in this paper, the leaching of a lepidolite concentrate, which was pre-treated by an alternative process of mechanical activation followed by acid digestion is studied in detail, by evaluating the main factors , and their possible interactions, that affect the Li recovery, in order to optimize it. Furthermore, the purification of the leachates and the obtainment of lithium carbonate by precipitation from the resulting solution is also presented and discussed.

PT

2. Methodology

RI

2.1 Mechanical activation and acid digestion

A lepidolite concentrate was obtained by grinding the original ore to -500 µm and by froth flotation, which

SC

allowed obtaining a product containing 1.95% Li (or 4.2% of Li2 O). Afterwards, this lepidolite concentrate was mechanically activated in a disk mill (N.V. Tema) during 15 min and it was digested with sulphuric acid (98%) in a pre-heated oven, for 15 min at a temperature of 130ºC, using an adequate acid/concentrate mass ratio,

NU

normally of 0.96 g H2 SO4 /g concentrate, but in some tests this ratio was varied. The digested product obtained

MA

was used in the subsequent leaching tests.

2.2 Leaching procedure

Digested samples of 2.5 g were set in closed glass flasks and the leaching tests were performed with

ED

demineralized water in a thermostatic orbital shaker (at 100 rpm). The leaching temperature, the liquid/solid ratio (L/S) and the leaching time were varied according to an experimental design, which will be discussed in

EP T

the section 2.3.

After the leaching procedure, samples were filtered (Whatman 52 filter paper) and the solid fraction retained in the filter was dried in an oven at 50ºC for 24ºC, with air circulation and stored for analysis. The leach ate was

AC C

analysed by atomic absorption spectrometry (AAS, SOLAR 969 AA Spectrometer, Thermo Elemental) to quantify the lithium content and to estimate the lithium extraction , which was made by relating the solution concentration with the initial Li content in the solids, taking into account the mass changes occurred during all operations. The experimental errors of the Li recoveries were estimated between 3-5%.

2.3 Leaching experimental design An experimental approach using a first-order design was assumed in order to evaluate the factors that affect the leaching process. It was applied a 2k factorial design, with three process variables or factors (k=3), corresponding to each one 2 levels, respectively low and high. Additionally, four central runs (n C) were performed in the standard/central level of the factors, allowing to estimate the variance of the experimental error. The studied factors were the liquid/solid ratio, L/S (x1 ), the leaching time, t (x2 ) and the leaching temperature, T (x3 ), being the values of the levels presented in Table 1.

ACCEPTED MANUSCRIPT

Table 1 – Studied factors and respective levels in the leaching factorial design experiments. Levels

Factors

Units

Low (-1)

Central (0)

High (+1)

L/S (𝑥1 )

2

3.5

5

Leaching time t (𝑥2 )

0.25

1.125

2

h

Leaching temperature T (𝑥3 )

20

50

80

ºC

PT

L/kg

The significance of the effects of the factors was assessed by analysis of variance (ANOVA) using the Fisher’s

SC

RI

F-test.

2.4 Purification and precipitation procedures

NU

According to (Chagnes & Swiatowska, 2015), the process of lithium extraction from ores involves purification mainly by precipitation, which is carried out to remove impurities, such as Ca, Al, Mn and Fe, followed by lithium concentration using mostly evaporation and finally, crystallization, carbonation or electrodialysis to

MA

produce lithium compounds. In this study, the leachate purification was accomplished by crystallization of alums followed by neutralization/precipitation of impurities and carbonate precipitation (Figure 1). The metals behaviour during the several precipitation steps was followed by analysis of the solution. All data in this section

ED

were based on the initial volume of the leachate, including the concentration of the metals in the solution, which was corrected and referred to the initial volume when a dilution was performed by addition of a solution of any precipitant agent. This procedure was helpful for better understanding the behaviour of the metals during the

precipitation.

EP T

precipitation, avoiding a misleading evaluation when the concentration decreased due to dilution and not to

The purification tests were carried out using a leachate prepared from an adequate amount of lepidolite

AC C

concentrate, in the optimized conditions of mechanical activation, acid digestion and water leaching. Purification operations by precipitation were performed in a 1 L glass reactor with a controlled temperature system and mechanical s tirring (two-blade impeller). Two additives were used, a slaked lime suspension (200 g/L) for the first step (precipitation of hydroxides) and a sodium carbonate solution (300 g/L) for the second step (precipitation of carbonates), both prepared from pro -analysis reagents. After each operation, the formed precipitates were removed from the solution by vacuum filtration, the solids were washed with water, and the solution returned to the reactor for a new step of purification. After performing all steps, the lithium carbonate was finally produced by heating at about 80ºC. The lithium carbonate formed was washed with a hot saturated solution of Li2 CO3 , to avoid losses by dissolution. Intermediate solids and the final product were characterised by X-ray powder diffraction (XRPD, PANalytical X’PERT-PRO diffractometer), under the following conditions: Cu Kα radiation, scan step size of 0.050°2θ, step time 150 s, generator settings of 35 mA and 40 kV. The analytical interpretation was done using the X’PERT HIGHSCORE PLUS software and the PDF4 database. For evaluating the efficiency of the precipitation process, namely regarding Li, Rb, K, Ca, Fe, Mn and Al, the

ACCEPTED MANUSCRIPT aqueous solutions were analysed by atomic absorption spectrometry (AAS, Thermo Elemental SOLAR 969 AA) and the calculations were carried out taking into account all changes of volume and mass occurred in each operation, namely those caused by the addition of precipitant agents.

Concentrate

H2SO4 conc.

Water

MECHANICAL ACTIVATION

DIGESTION

PT

Insoluble silicates LEACHING

Leach solution

Fe, Al, Mn hydroxides

Rb,K Alum

SC

NEUTRALIZATION PRECIPITATION

2nd STEP PRECIPITATION

HOT CRYSTALLIZATION

Ca carbonate

NU

COOLING CRYSTALLIZATION

RI

Soda ash

Lime

Water to treatment

Li2CO3

3. Results and discussion

EP T

3.1 Activation and acid digestion

ED

MA

Figure 1 – Proposed and tested simplified process diagram for recovery of lithium from lepidolite.

Mechanical activation and acid digestion were previously studied in detail and reported in a published work, as referred in section 1 (Introduction). The median value, d 50, and the percentiles d 10 and d 90 of the particle size

AC C

distribution of the activated concentrate sample were 6.34 µm, 1.02 µm and 32.95 µm, respectively. The mechanical activation provides a structural change in the lepidolite mineral, which becomes amorphous, and consequently much more chemically reactive. The digestion performed with an acid/concentrate mass ratio of 0.66, at 130ºC for 15 min, led to the formation of lithium sulphate that is easily soluble in water. The final products obtained were rostite (Al(SO4 )(OH)·5H2 O), potassium alum and lithium sulphate hydrate, whose the diffraction patterns can be seen in Figure 2.

ACCEPTED MANUSCRIPT 14000 R

12000

P R

R p

8000 P

R

6000

R

4000

P

2000

R P P L L

P

L L P L

R LP

R R P P L

R P L P L

R P

R P L

PT

Counts (a.u.)

10000

0 20

25

30

35

Position 2ϴ (º)

40

45

50

RI

15

NU

SC

Figure 2 – XRPD patterns of the products obtained by mechanical activation followed by acid digestion of a lepidolite concentrate. R: rostite, P: potassium alum and L: lithium sulphate hydrate.

The effect of the acid amount was evaluated varying the acid/concentrated mass ratio between 0.66 and 1.10 (Figure 3). For ratios from 0.66 to 0.88, any increase in the lithium extraction was found. Nevertheless, when

MA

this ratio was increased to 0.96 or higher, a rise in the lithium extraction of about 12% was verified. Thus, taking into account that sulphuric acid represents one of the main costs of the employed process, the acid amount choice should consider the resulting costs and revenues. However, mass ratios above 0.96 are not necessary

ED

since they do not reflect a considerable increase in the lithium extraction .

Li extraction (%)

AC C

EP T

100

80

75.2

76.0

0.66

0.88

88.4

86.8

0.96

1.03

91.8

60 40 20 0

1.10

acid/concentrate (g/g) Figure 3 - Effect of the acid/concentrate mass ratio in the lithium extraction. Digestion conditions: activation for 15 min, digestion during 15 min at 130 ºC. Leaching conditions: L/S of 10 L/kg, 80 ºC for 4 hours.

3.2 Preliminary leaching tests Some preliminary water leaching tests varying the liquid/solid ratio (L/S) were performed. In these tests, the lepidolite concentrate was activated in a high energy mill for 15 min and then it was digested during 15 min at

ACCEPTED MANUSCRIPT 130 ºC, using an acid/concentrate mass ratio of 0.66. Afterwards, the digested product was leached for 4 h at 80 ºC. As it is possible to observe in Figure 4, when L/S ratios of 2 or 5 L/kg were used, a slight increase in the lithium extraction was obtained when compared with the test in which L/S was 10 L/kg. However, it is important to highlight that, taking into account the experimental error, this variation cannot be considered very representative. Thus, the variable L/S ratio does not seem to influence greatly the lithium extraction, in the tested conditions.

81.0

81.2

76.8

RI

60

PT

80

40 20 0 5

10

NU

2

SC

Li extraction (%)

100

L/S (L/kg)

MA

Figure 4 – Effect of the L/S in the lithium extraction. Digestion conditions: activation for 15 min, digestion during 15 min at 130 ºC, using an acid/concentrate mass ratio of 0.66. Leaching conditions: 80 ºC for 4 hours

ED

The effect of the leaching time on the lithium extraction was also tested using an L/S of 10 L/kg and a leaching temperature of 80 ºC. The leached sample was previously digested using an acid/concentrate mass ratio of 0.66, for 15 min at 130 ºC. Although better results have been obtained for lower L/S ratios, it was decided to perform

EP T

these experiments at favourable conditions in order to evaluate the time influence without any solubility constraints. Moreover, the optimization of all these processing factors was evaluated in the factorial design of experiments (section 3.3). The leaching time does not seem to affect the lithium extraction (Figure 5) and after

100 80

Li extraction (%)

not increase.

AC C

15 min of leaching, about 80% of extraction was reached. Even after four hours of leaching, the extraction did

79.2

75.2

76.8

2.0

4.0

60 40 20 0

0.5

Leaching time (h) Figure 5 - Effect of the leaching time on the lithium extraction. Digestion conditions: activation for 15 min, digestion during 15 min at 130 ºC. Leaching conditions: L/S of 10 L/kg and 80 ºC.

ACCEPTED MANUSCRIPT

Additional tests for evaluating the effect of the leaching time were carried out, using samples digested with a higher acid/concentrate mass ratio (0.96) and at the same time and temperature already tested (15 min at 130ºC, respectively). The L/S ratio was set at 10 L/kg in the leaching step. One of the experiments was carried out at 80 ºC and samples were collected at four time intervals; the other experiment was performed at 25ºC, to evaluate the potential of performing the leaching at ambient temperature, being samples taken at two time intervals. As shown in Figure 6, the Li extraction at 25ºC was lower than at 80ºC. In both cases the lithium extraction

PT

seems to increase slightly with the leaching time. It is worth noting that the lithium sulphate solubility increases along with the decrease of the temperature. Therefore, the results can be associated with the formation of lithium

RI

sulphate connected with other lithium compounds, which possibly exhibit higher solubility with temperature. Kinetic constraints (slower dissolution rates) can also explain the lower values obtained at room temperature for

SC

initial times. This may explain that only about 75% of lithium is extracted after 30 min of leaching at 25 ºC. 90

80 75

70

NU

80 ºC

MA

Li extraction (%)

25 ºC 85

ED

0.08

0.25

0.50

2.00

Leaching time (h)

AC C

EP T

Figure 6 – Effect of leaching time in the lithium extraction. Digestion conditions: activation for 15 min, digestion during 15 min at 130 ºC using an acid/concentrate mass ratio of 0.96. Leaching conditions: L/S of 10 L/kg.

3.3 Leaching factorial design experiments Water leaching tests were carried out according to the factorial design methodology. Samples used were firstly digested with an acid/concentrate ratio of 0.96 g/g at 130ºC for 15 min. A matrix with the experimental conditions and respective Li extractions is presented in Table 2. The experimental error was calculated using the conditions of the central points (Tests 9 to 12). The effects of the factors and the analysis of variance are shown in Table 3. As observed, only the effect of the main factor 𝑥 3 (leaching temperature) is significant, at a significance level of 5% (p-value < 0.05). In the experimental range studied, the other factors and their interactions do not present a significant effect on the lithium extraction.

ACCEPTED MANUSCRIPT Table 2 – Evaluated factors and responses of the leaching experimental design Coded units Standard order

Original units

Response

𝑥1

𝑥2

𝑥3

L/S (L/kg)

Leaching time (h)

1

-1

-1

-1

2

0.25

Leaching temperature (ºC) 20

2

1

-1

-1

5

0.25

20

73.3

3

-1

1

-1

2

2

20

80.0

4

1

1

-1

5

2

20

82.8

5

-1

-1

1

2

0.25

80

86.6

6

1

-1

1

5

0.25

80

89.0

7

-1

1

1

2

2

80

8

1

1

1

5

2

80 50

3.5

1.125

0

3.5

1.125

11

0

0

0

3.5

1.125

12

0

0

0

3.5

1.125

PT

0

0

89.0 89.3

RI

0

0

77.7

84.1

50

90.8

50

90.2

SC

0

90.6

50

NU

9 10

Lithium extraction (%)

Table 3 - Results of the effects of the factors and analysis of variance (ANOVA). Effect

Sum of squares

DF

MA

Source

F statistic

p-value

𝑥1

0.28

0.16

1

0.16

0.01

0.91

𝑥2

3.65

26.70

1

26.70

2.54

0.21

𝑥3 Interactions

10.03

201.30

1

201.30

19.19

0.02

𝑥1𝑥2

1.27

ED

M ain factors

M ean square

3.21

1

3.21

0.31

0.62

𝑥1𝑥3

2.22

1

2.22

0.21

0.68

10.55 11.13

1 1

10.55 11.13

1.01 1.06

0.39 0.38

Residuals

31.47

3

10.49

Total

286.73

10

265.75

EP T

1.05

𝑥2 𝑥3 𝑥1𝑥2 𝑥3

-2.30 -2.36

AC C

DF – degrees of freedom

The plot of lithium extraction versus the mean responses for the three levels of the main effect (factor 𝑥 3 ) is shown in Figure 7 and it was constructed using the add-in Action Stat for Excel. The bars represent the confidence interval for the averages of the levels of the factor. An average increase in the lithium extraction of about 10% is reached when the temperature is varied from the low level to the standard/central level, while no significant change is found from the standard level to the upper level. Thus, the leaching temperature of 50 ºC seems to be the best condition for the lithium extraction. Moreover, L/S and leaching time do not exhibit a significant effect on the response, in the range of values tested in this study. It is important to highlight that this conclusion does not imply that the L/S ratio and the leaching time have no effect, but it means that the recommended operating value for these variables is the minimum. Therefore, the L/S of 2 L/kg was considered the best value, since allows reducing the amount of water used and to obtain a high concentrated leachate,

ACCEPTED MANUSCRIPT requiring lower energy inputs for the subsequent operations , such as for the concentration of solutions . The concentration of the lithium solutions obtained were 1.9 and 4.8 g/L Li, respectively for L/S = 5 and 2 L/kg.

PT

85

80

RI

Lithium extraction (%)

90

-1

0

SC

75

1

NU

X 3 (levels)

MA

Figure 7– Main effect plot for the factor 𝑥 3 (Constructed using the add-in Action Stat for Excel).

Water leaching experiments of lithium ores using different roasting additives were evaluated by many authors . Table 4 summarizes the type of mineral, the process used, the best leaching conditions proposed and the corresponding yields attained. It is worth noting that in some tests the leaching conditions were varied and the

this is indicated in Table 4.

ED

best conditions are presented in Table 4. In other tests, some parameters were kept fixed during the leaching and

EP T

Lithium was extracted from a zinnwaldite waste obtained from the clay production using different additives by Siame & Pascoe (2011), under the same leaching conditions . Moreover, it was demonstrated that 10 min of leaching lead to the almost full lithium extraction, providing that the Li species form were soluble.

AC C

The lithium extraction from zinnwaldite wastes was also investigated by Jandová et al. (2009), who verified that during the leaching, the lithium dissolution was very fast and it was almost completed after 10 min of leaching. A reduction in the leaching temperature from 90 to 20ºC resulted in an extraction decrease of only 5% and the effect of the L/S was considered negligible in this study. The extraction of lithium from zinnwaldite waste by roasting using CaCO3 was studied by Jandová et al. (2010) and in such case, the leaching conditions were kept fixed, as presented in Table 4. Luong et al. (2013) used Stabcal modelling to investigate the stability of Li species formed during the roasting of lepidolite with Na 2 SO4 and evaluated the water leaching performance. Through the simulation, the authors identified that one of the main lithium products formed was LiKSO 4 , which has a low solubility and regulates the lithium dissolution during the leaching step. The low solubility of LiKSO 4 indicated by Stabcal was confirmed using different leaching temperatures and an increase in the Li extraction was verified by increasing the temperature from 25ºC to 85ºC.

ACCEPTED MANUSCRIPT On the other hand, Luong et al. (2014) also studied the iron sulphate roasting of lepidolite followed by water leaching and in such case, lithium sulphate (Li2 SO4 ) was identified as the main lithium soluble compound. Thus, lithium extraction did not depend on leaching temperature being similar the results obtained at room and higher temperatures. A steady state for the lithium extraction was also verified after 15 min of leaching and similar recoveries were obtained at different L/S ratios (from 1:1 to 10:1). Sitando & Crouse (2012) studied the processing of petalite using sulphuric acid leaching. The decrease in the L/S seemed to have negatively influenced the lithium extraction. According to the authors , the use of higher L/S

PT

ratios decreases the viscosity of the system, which results in a diminution in the mass transfer resistance at the liquid-solid interface. On the other hand, the temperature (50 and 90ºC) showed little effect on the lithium

RI

extraction, which was stable after 60 min for all temperatures.

Yan et al. (2012d) observed that the lithium extraction by leaching, from a lepidolite treated by chlorination

SC

roasting, increases with time, reaching a steady state at 30 min, for temperatures of 60 and 90ºC, while extraction seemed to decreased for lower values of the L/S ratio. Yan et al. (2012c) also studied the extraction of lithium from lepidolite by sulphate roasting and in this case, the leaching was performed at room temperature

NU

for 30 min and the extraction efficiency was independent of the L/S ratio. In this context, the authors considered that a low L/S is helpful and more realistic since allows obtaining higher concentrations. The leaching of defluorinated lepidolite was performed by Yan et al. (2012a), using lime at an autoclave, which allowed

MA

achieving higher temperatures. The extraction of metals from lepidolite using roasting was also carried out by Yan et al. (2012b), but the leaching conditions were not studied, using only the values presented in Table 4. The leaching of a lepidolite previously treated by iron sulph ate roasting was also tested by Hien-Dihn et al.

ED

(2015). In this study, the L/S ratio greatly affected the lithium extraction and the highest recovery was obtained when the L/S was 10:1. However, this ratio leads to a low Li concentration, thus lower L/S ratios (e.g. 5:1) can

EP T

eventually be an alternative. According to the authors, this behaviour could be related to the formation of high amounts of LiKSO4 . In this study, only a slight increase in the extraction was found when the temperature was increased from 50 to 85ºC. The leaching time, tested from 1 to 3 h, did not seem to affect the extraction.

AC C

Amer (2008) used a ball mill type autoclave (allows higher temperatures) for the direct leaching of a clay mineral with H2 SO4 solutions. In this case, the variables temperature, time, acid concentration and particle size seemed to influence the lithium extraction. It is important to highlight that in the mentioned studies, the leaching conditions were tested one variable at time, which does not allow evaluating the simultaneous effect of the several variables that can affect this process. Moreover, since the lithium ore pre-treatment used in the mentioned studies was different than the used in this work, the products obtained may be different and, therefore, the leaching results could differ.

ACCEPTED MANUSCRIPT Table 4 – Results of published studies where a leaching step was performed (leaching conditions and Li extractions presented are corresponding to best combinations proposed).

Reference

M ineral

% Li

Processing

Additives

Leaching temperature (ºC)

Leaching time (h)

L/S ratio

M aximum Li extraction (%)

Liquor Li concentration (g/L)

Hien-Dihn et al. (2015)

Lepidolite

1.55

Roasting

FeS, CaO

50

1.5

10:1

81

0.64

Luong et al. (2014)

Lepidolite

1.79

Roasting

FeSO4.7H 2O, CaO

Room

1

1:1

93

8.7

Luong et al. (2013)

Lepidolite

2.55

Roasting

Na2SO4

85

15:1

~90

1-3

Yan et al. (2012b)

Lepidolite

2

Roasting

Na2SO 4, K2SO 4, CaO

Room *

2.5:1

91.61

4.39

Yan et al. (2012a)

Lepidolite

1.4

Deflurination and leaching in autoclave

lime-Defluorinated lepidolite

C S

0.5*

150

1

4:1

98.9

1.74

Yan et al. (2012c)

Lepidolite

2

Roasting

Na2SO 4, CaCl2

Room*

0.5*

0.8:1*

94.8

8.53

Yan et al. (2012d)

Lepidolite

2

Chlorination roasting

NaCl, CaCl2

60

0.5

2.5:1*

92.86

3.71

Sitando & Crouse (2012)

Petalite

1.9

Pre-heat and roasting

H 2SO4

50

1

7.5:1

97.30

5.72

Jandová et al. (2009)

Zinnwaldite wastes

1.4

Roasting

CaSO 4:Ca(OH)2

90

0.5

10:1

96

0.69

Siame &Pascoe (2011)

Zinnwaldite wastes

0.96

Roasting

Limestone, gypsum and Na2SO 4

85*

0.5*

10:1*

~90

~1

Jandová et al. (2010)

Zinnwaldite wastes

1.21

Roasting

CaCO 3

90-95*

0.5*

5:1*

>90

0.79

Amer (2008)

M ontmorillonitetype clay (hectorite)

leaching in a ball mill autoclave

H 2SO4

250

1.5

1:5

>90

~1.3

D E

* Condition was kept fixed in the study.

E C

T P

C A

0.56

U N

A M

T P

I R 3

ACCEPTED MANUSCRIPT 3.4 Purification of leachates The initial leachate used in the purification step contained 3.8 g/L Li, 1.3 g/L Rb, 14 g/L K, 20 g/L Al, 0.65 g/L Mn and 1.27 g/L Fe. The progression of metals during the several precipitation steps was monitored analysing the solution. The metal concentrations presented in this section were referred to the initial volume of the leachate (therefore, adequately corrected for volume changes), unless when identified that the real concentrations were used. This procedure was helpful to better understanding the behaviour of the metals during the precipitation, avoiding misleading evaluation when the concentrations decreased d ue to dilution (when

PT

adding precipitant agent solutions) and not to precipitation.

RI

3.4.1 Crystallization of alum

From literature data, it was possible to foresee that rubidium alum is much less soluble than other alkali alums ,

SC

such as sodium and potassium (Seidell, 1919). Regarding lithium, there is no reference to the formation of lithium alum. Thus, a possibility of separating Rb and Li from the leachate was considere d, through the

(aq) +

Al3+ (aq) + 2 SO4 2- (aq)  RbAl(SO4 )2

(1)

(s)

MA

Rb +

NU

following chemical reaction (1),

From data of Siedell (1919), it was found that the solubility of this salt is strongly temperature dependent, being much less soluble at low temperature conditions (e.g. the solubility is 0.71 g/100 g H2 O at 0ºC and 4.98 g/100 g H2 O at 50ºC). On the other hand, lithium sulphate presents an opposite behaviour, t rending to remain in solution

ED

at low temperatures. Moreover, the presence of high concentrations of sulphate and aluminium ions, when compared to rubidium, anticipates a substantial decrease of Rb in solution due to the common ion effect. Figure

EP T

8 shows the predicted Rb ion concentrations versus sulphate concentration, calculated from equilibrium data, where it can be observed that below 10ºC and for SO4 2- concentrations above 100 g/L, the expected levels of Rb

1000

Rb equil. conc. (mg/L)

AC C

in solution can be as low as 0.1 mg/L.

50ºC

10ºC

100

0ºC

10 1 0.1 0.01 0.001

0

100 200 2SO4 conc. (g/L)

300

Figure 8 – Theoretical rubidium ion equilibrium concentration in saturated solutions of Rb alum, as a function of total sulphate ion concentration, at three levels of temperature (total Al in solution = 15 g/L).

ACCEPTED MANUSCRIPT The experimental approach followed was to cool the leachate until 5ºC to promote the crystallization of Rb alum, following the concentrations with time by sampling and analysing the solution. This temperature was also proposed by Guo et al. (2016) for the crystallization of potassium alum, who states that the hydrolysis of Al 3+ is endothermic and, therefore, lower temperatures constrain the hydrolysis, releasing more Al3+ to form alum. As shown in Figure 9, after 5 h the Rb concentration decreased from 1.3 g/L to 27 mg/L (98% precipitation yield) and after 72 h the concentration decreased to 5 mg/L (99.6% precipitation yield). Simultaneously, the Al concentration in solution also decreases 25%, from 20 g/L to 15 g/L. However, the amount of Al precipitated is much higher than the stoichiometry of Rb alum, because simultaneous precipitation of potassium alum occurs,

(aq) +

Al3+ (aq) + 2 SO4 2- (aq)  KAl(SO4 )2

(s)

(2)

RI

K+

PT

according to the reaction (2),

concentration from 15 to 13 g/L (overall decrease of 35%).

20

MA

15

Al

10

Li

5

Rb

0

3.5 3

2.5 2 1.5 1 0.5 0

10 20 30 40 50 60 70

EP T

0

4

Li and Rb in Solution (g/L)

NU

4.5

ED

Al in Solution (g/L)

25

SC

confirmed by the reduction of the potassium concentration in the solution from 14 to 3.6 g/L and the aluminium

Crystallization time (h)

AC C

Figure 9 – Evolution of concentrations of Al, Li and Rb with time, during the process of crystallization of alum at 5ºC.

Rubidium is as abundant as metals such as copper, zinc and lead. It is more abundant than cesium and lithium, however, it is only obtained as a by-product of the extraction of these two metals. This is because copper, lead, zinc, lithium and cesium occur in mineral as a main component. On the other hand, rubidium forms no minerals in which it is the predominant element; therefore, there are no known rubidium ores. It occurs widely dispersed in potassium minerals and salt brines. Lepidolite contains up to 3.5% Rb 2 O and is the main source of rubidium. Pollucite, a cesium silicate, can contain up to 1.5% Rb 2 O and some rubidium is produced as a by-product of cesium extraction from this source (Jandová et al. 2012; Butterman and Reese, 2003; Wagner, 2006). Rubidium is also much more difficult to extract than other alkali metals and its production is often linked to the production of cesium and most rubidium is obtained as a by -product of the lepidolite and pollucite processing, being separated from potassium and cesium by several chemical treatments (Jandová , 2012). In this context, the rubidium extraction as a by-product of the process proposed in this study can be considered very important.

ACCEPTED MANUSCRIPT Since aluminium is an important contaminant of the process and should be removed before the recovery of the lithium product, it was tried to precipitate more potassium alum by adding potassium sulphate (200 ml of 70 g/L K2 SO4 , per litre of initial solution). After standing for 24 h at 5ºC, a new precipitate appeared . The two alums produced, the Rb/K alum and the K alum, were analysed by XRPD (Figure 10). The phase KAl(SO4 ) was identified in Figure 10a, which is very similar to the pattern shown in Figure 10b, because the structures of both alums are very similar and the Rb content is relatively low. Additional quantities of potassium sulphate would further decrease the Al content in the solution, but this would be inadequate due to the excessive use of reagents and also to the subsequent dilution of the leachate. Moreover, according to Guo et al. (2016), an excessive use

PT

of K2 SO4 could lead to the formation of KLiSO4 , which would affect the lithium extraction in the process. Nevertheless, during the process of alum formation studied, the Li contents in the solution did not change

RI

significantly (~4 g/L), demonstrating that there was no formation of Li alum. Therefore, this process seems to be efficient and selective for separating Rb from Li. The alum formed can be subsequently treated to recover Rb in

SC

some valuable form, although this was not considered in the present study.

An alum crystallization procedure was also studied by Guo et al. (2016) using a lepidolite treated by a sulphuric acid method. The crystallization with K2 SO4 was based on the characteristic of K+, Rb + and Cs + can form alums

NU

with Al3 + and, the need to remove the considerable amount of Al3 + present in solution, before the Li2 CO3 precipitation. An alternative process for aluminium removal from a leaching solution of lepidolite using

MA

ammonium as (NH4 )2 SO4 was also recently proposed by Li et al. (2016) and seems to be an efficient alternative method to remove aluminium from lepidolite leaching solutions.

ED

K

K

K

EP T

K

K (a) Potassium alum

AC C

Counts (a.u.)

K

~

15

20

25

(b) Rubidium and potassium alum

30

35

40

45

50

55

60

65

70

Position 2θ (º)

Figure 10 – XRPD patterns of the two alums produced in the purification steps: (a) Potassium alum (K) produced in the second step; (b) Rubidium and potassium alum produced in the first step.

ACCEPTED MANUSCRIPT 3.4.2 Neutralization and precipitation of hydroxides or oxides The following purification step was the neutralization and the precipitation of metal oxides/hydroxides (Al, Fe and Mn), using slaked lime as neutralizing agent. A similar procedure using lime to precipitate aluminium and heavy metals, followed by addition of sodium carbonate to remove the remaining lime , was also reported by Yan et al.(2012a); Yan et al. (2012b); Botton et al. (1965). The use of lime allows the formation of low-soluble gypsum, that enables easier settling and solids separation as well as provides removal of some sulphate ions from the solution. For the excess acid and the aluminium, the

(aq) +

Ca(OH)2

(s)

 CaSO4 .2H2 O

2/3 Al3+ (aq) + SO4 2- (aq) + Ca(OH)2

+ H2 O (l)  2/3 Al(OH)3

(s) +

(3)

CaSO4 .2H2 O

(4)

(s)

SC

(s)

(s)

RI

H2 SO4

PT

following reactions (3) and (4) are expected to occur,

NU

Aluminium precipitation would occur at pH above 3.5. For the other two metals (Fe and Mn) the precipitation of hydroxides would also depend on the solution pH, but also on the oxidation state of both metals. Iron can precipitate at relatively low pH values (e.g. from pH 2) provid ing that it is in Fe(III) state, while Fe(II) would

MA

precipitate at higher pH values (e.g. pH 7). The measured redox potential of the solution (E = 0.33 V vs. Ag/AgCl electrode) indicated that the iron should be predominantly at Fe(II) state, since only abo ve E=0.55 V the Fe(III) would become predominant. During the neutralization, H 2 O2 was added to improve the potential.

ED

Values in the range 0.38-0.48 V were achieved, but higher values would require excessive amounts of the oxidant agent and thus were not attempted. Therefore, it is foreseen that the precipitation of iron can occur by a

EP T

series of several possible equations,

2/3 Fe3+ (aq) + SO4 2- (aq) + Ca(OH)2

2 Fe2+ (aq) + 2 SO4 2- (aq) + 2 Ca(OH)2

+ 2 H2 O (l)  2/3 Fe(OH)3

2 H2 O (l)  Fe(OH)2

(s) +

AC C

Fe2+ (aq) + SO4 2- (aq) + Ca(OH)2

(s)

(s) +

(s)

(s)

+ CaSO4 .2H2 O

+ CaSO4 .2H2 O

H2 O2 (aq) + 2 H2 O (l)  2 Fe(OH)3

(5)

(s)

(6)

(s)

(s) +

2 CaSO4 .2H2 O

(s)

(7)

For manganese, present in Mn(II) oxidation state, it can be predicted that several forms of oxides/hydroxides would form, Mn(OH)2 , Mn 3 O4 and Mn(OH)3 . The redox potential achieved in solution was not higher enough to form MnO2 . The possible reactions are, therefore,

Mn 2+ (aq) + SO4 2- (aq) + Ca(OH)2 3 Mn 2+ (aq) + 3 SO4 2- (aq) + 3 Ca(OH)2 2 Mn 2+ (aq) + 2 SO4 2 -(aq) + 2 Ca(OH)2

(s) +

(s) +

(s) +

2 H2 O (l)  Mn(OH)2

H2 O2

(aq) +

(s) +

CaSO4 .2H2 O

(8)

(s)

2 H2 O (l)  Mn 3 O4 (s) + 3 CaSO4 .2H2 O

H2 O2 (aq) + 2 H2 O (l)  2 Mn(OH)3

(s) +

2 CaSO4 .2H2 O

(s)

(s)

(9) (10)

ACCEPTED MANUSCRIPT As referred for iron, the pH at which the precipitation of Mn occurs depends on the oxidation state of the species and thus from the redox potential. Mn(OH)2 usually precipitates at higher pH (e.g above 9), while the other Mn oxides or hydroxides can be precipitated at lower pH. In order to predict more accurately the conditions for precipitation of the species above referred, equilibrium data (E/pH) for the above equations was acquired from literature (Pourbaix, 1974). The resulting predictions are plotted in Figure 11, where it can be seen that the order of precipitation will probably be Al and Fe firstly and Mn only for higher neutralization levels, but this evaluation depends mostly on the oxidation states of Fe and

PT

Mn.

Mn2+ → Mn(OH)3

Fe2+ → Fe(OH)3

Mn2+ → Mn3O4

Fe2+ → Fe(OH)2

Mn2+ → Mn(OH)2

RI

Fe3+ → Fe(OH)3

SC

Al3+ → Al(OH)3

NU

10 1 0.1

MA

Conc. in solution (g/L)

100

0.01

0.001

3

ED

1

5

7

9

11

pH

AC C

EP T

Figure 11 – Theoretical predictions of equilibrium metal ions in solution considering the several precipitation reactions for Al, Fe and M n (see equations 4 to 10 for correspondence). For equations with change in oxidation state, calculations were done assuming E=0.40 V (vs. Ag/AgCl electrode).

The evolution of the metal ion concentrations with the pH and the neutralizer additions are represented in Figure 12. The addition of the two neutralizers is differentiated by the dashed line near pH 9 (lime addition is before the line and sodium carbonate addition after). The initial solution, with pH 0.8, was neutralized until pH reaches 3.5 without any significant change in the metals concentration. This initial step corresponds essentially to the acid neutralization (equation 3), with an addition of 20 g of OH per litre of initial solution. Nonetheless, a slight decrease of iron was also noted (from 1.3 to 1.1 g/L) in this period, probably due to precipitation of some Fe(III) hydroxide above pH 2 (equation 5). At pH near 4 the precipitation of Fe occurs suddenly, reaching a final concentration of 3 mg/L above pH 4.5. The pH where this precipitation occurs, when compared with the predictions of Figure 12, indicates that the reaction can be explained by equation 7. The aluminium also precipitates at near the same pH (from 4 to 5), but the total Al removal from the solution is only achieved at pH above 5.

ACCEPTED MANUSCRIPT Manganese concentrations in solution remain practically unaltered during the previously referred phenomena. Only after pH 5 the Mn concentration started to decrease from 700 g/L until 41 mg/L, observed at pH about 7. According to the prediction reactions for Mn (see Figure 12), both equations 9 and 10 are compatible with the experimental behaviour found. Continuing the addition of lime until pH 9, the final concentrations found in the solution where <3 mg/L Al, 2.9 mg/L Fe and 1.5 mg/L Mn, which corresponds to a very successful precipitation of these contaminants. The alkali hydroxides such as LiOH are quite soluble, and thus the Li content in solution was not substantially changed during the precipitation operations. However, a slight decrease was found

be overcome by an efficient stirring and further washing of the precipitates. Fe

Mn

Ca

RI

8

4 2 0 0

1

2

3

4

5 6 Solution pH

Lime add.

1.2 1.0 0.8 0.6

40 35

Al hydrox. pp.

25

Al+Fe hydrox. pp.

15 10 5 0

(b) 0

0.0 8

1

2

9

10

11

20 Mn oxide pp.

15

Fe hydrox. pp. acid neutr.

AC C

20

0.2

EP T

30

7

0.4

Sodium carbonate add.

ED

45

OH cumulative addition (g/L)

1.4

NU

6

(a)

Na2CO3 addition

SC

10

MA

Corrected concentration Al, Li (g/L)

Ca(OH)2 addition

12

1.6

Corrected concentration Fe, Mn, Ca (g/L)

Li

14

10

Ca carbonate pp.

5

CO3 addition (g/L)

Al

PT

(estimated as 16%), attributed to co-precipitation and adsorption on the solids, which in industrial practice can

0 3

4

5 6 Solution pH

7

8

9

10

11

Figure 12 – Evolution of (a) metal ion concentrations, and (b) neutralizers consumption, as function of pH during the neutralization/precipitation operation. M etal concentrations and addition of neutralizers are corrected/referred to the initial volume of solution.

The total amount of lime consumed was 42 g/L (expressed as OH). According to the precipitation reactions and to the metal contents precipitated, a consumption of about 26 g/L of OH was predicted. This value is nearly that observed in the region of metals precipitation, from pH 3.5 to 9. The remaining OH addition could be attributed to the acid neutralization but the value seems quite excessive (about 16 g/L OH from pH 0.8 to 3.5, when compared to about 11 g/L OH theoretically predicted). However, the efficiency of using lime is generally not

ACCEPTED MANUSCRIPT high due to gypsum precipitation that can passivate the su rface of the lime particles and thus hinder its access to the acid. This phenomenon can explain the excess of the lime addition in the first part of the neutralization. Another possible explanation for the excessive lime consumption is the formation of calcium fluoride, if some fluorine ion present in lepidolite mineral was dissolved during the digestion and leach ing steps. However, the fluorine concentration was not followed in this study. Further work should consider the fluorine behaviour in the whole process, since at least a part of it can be lost during the mechanical activation process due to a p robably defluorination phenomenon, as it was suggested in a previous work (Vieceli et al., 2017b). The calcium ion concentration was also followed during the neutralization step. Since lime has some solubility

PT

in water, the concentration of calcium increased during the process, reaching about 1.5 g/L at pH 7 and decreasing to 0.9 g/L at pH 9. Calcium should be removed from the solution since it would be very deleterious

RI

for the purity of lithium carbonate due to the very low solubility of calcium carbonate. Therefore, calcium precipitation was accomplished by adding sodium carbonate to the solution at an excess (9 g CO 3 2- per litre of

SC

initial solution, as Figure 12b shows). Since the solubility of Li2 CO3 is higher, it was possible to remove the calcium from the solution without precipitating lithium. The final solution was heated to the boiling point and concentrated by evaporation, leading to lithium precipitation as lithium carbonate (Li2 CO3 ). The XRPD patterns

NU

of this product can be seen in Figure 13, where the only compound identified was lithium carbonate, as

MA

zabuyelite.

8000

Z

ED

7000 6000

Z

EP T

Counts (cps)

5000 4000 3000

AC C

2000

Z

Z

Z

Z

1000

Z

Z

Z

Z Z

Z

ZZ

Z

Z

Z

Z Z

0

5

10

15

20

25

30

35

40

45

50

55

60

65

70

Position 2θ (º)

Figure 13 – XRPD patterns of the lithium carbonate obtained through the purification process. Lithium carbonate was identified as zabuyelite (Z).

4. Conclusions

ACCEPTED MANUSCRIPT The leaching step of a lepidolite concentrate, which was previously treated using an alternative process involving mechanical activation followed by acid digestion, was studied using a design of experiments approach. The main variables affecting the leaching step were evaluated and only the temperature presented a significant effect on the lithium extraction. The leaching temperature of 50ºC seemed to be the best one, on the other hand, the L/S ratio and the leaching time did not demonstrate a significant effect on the response, thus, the recommended operating value for these variables is the minimum. Using the minimum L/S ratio (2 L/kg) has also the advantage of reducing the amount of water used in the leaching step and requires lower energy inputs

PT

for the concentration of lithium in solution. Lithium extractions above 90% were obtained. The purification of the leaching solution obtained at optimized conditions was accomplished by crystallization of alums followed by neutralization and carbonate precipitation. After a very efficient removal of impurities (Fe,

RI

Al, Mn, Ca), lithium was precipitated as lithium carbonate and it was identified by XRPD as zabuyelite. Other by-products obtained during the purification were rubidium and aluminium alum. It is important to highlight the

NU

SC

importance of obtaining rubidium compounds by this process, since lepidolite is the main source of it.

MA

Acknowledgements

The author N. Vieceli acknowledges the doctorate grant ref. 9244/13-1 supplied by CAPES Foundation, Ministry of Education of Brazil. The authors are also very grateful to Felmica Minerais Industriais, S.A. for

AC C

EP T

ED

having kindly provided the lepidolite ore.

ACCEPTED MANUSCRIPT

References Averill, W.A., Olson, D.L., A review of extractive processes for lithium from ores and brines, Energy, 3(1978), pp. 305–313. Amer, A.M., The hydrometallurgical extraction of lithium from Egyptian montmorillonite -type clay, JOM, 60(2010), No. 10, pp. 55–57.

PT

Banks, M.K., MacDaniel, W.T., Sales, P.N., A method for concentration of North Caroline spodumene ores, Mining Engineering, (1953), Transactions AIME, pp. 181-186.

RI

Barbosa, L.I., Valente, G., Orosco, R.P., and González, J.A., Lithium extraction from β-spodumene through

SC

chlorination with chlorine gas , Minerals Engineering, 56(2014), pp. 29 – 34.

Botton, R., Delgrange, J.P., Steinmetz, A., Method of reco vering lithium from lepidolite, assignors to Compagnie de Saint-Gobain, US Patent 3,189,407; June 15, 1965.

NU

Brandt, F., Haus, R., New concepts for lithium minerals processing, Minerals Engineering, 23(2010), 8, pp. 659-661.x

MA

Butterman, W.C., Reese Jr., R.G., 2003, Rubidium, Open-file report 03-045, Mineral Commodity Profiles, USGS, 2003.

Recycling, Elsevier, 1st ed.

ED

Chagnes, A., Swiatowska, J., eds., 2015, Lithium Process Chemistry: Resources, Extraction, Batteries, and

Guo, H., Kuang, G., Yang, J., Hu, S., Fundamental Research on a New Process to Remove Al 3 + as Potassium

3557–3564.

EP T

Alum during Lithium Extraction from Lepidolite, Metallurgical and Materials Transactions B, 47(2016), 6, pp.

Hien-Dinh, T.T.; Luong, V.T., Gieré, R., and Tran, T., Extraction of lithium from lepidolite via iron sulphide

AC C

roasting and water leaching, Hydrometallurgy, 153(2015), pp. 154 – 159. Jandová, J., Vu, H.N., Belková, T., Dvorák, P., and Kondás, J., 2009, Obtaining Li2 CO3 from zinnwaldite wastes, Ceramics – Silikáty, 53(2009), 2, pp. 108–112. Jandová, J., Dvorak, P., and Vu, H.N., Processing of zinnwaldite waste to obtain Li2 CO3 , Hydrometallurgy, 103(2010), 1-4, pp. 12 – 18. Jandová, J., Dvořák, P., Formánek, J., Vu, H.N., Recovery of rubidium and potassium alums from lithiumbearing minerals, Hidrometallurgy, 119-120(2012), pp. 73-76. Jaskula, B.W., 2016, Lithium, Mineral Commodity Summaries, USGS. Julien, C., Mauger, A., Vijh, A., Zaghib, K., 2016, Lithium batteries Science and Technology, 1st ed., Springer International Publishing, Switzerland.

ACCEPTED MANUSCRIPT Kondás, J., and Jandová, J., Lithium extraction from zinnwaldite wastes after gravity dressing of Sn-W ores, Acta Metallurgica Slovaca, 12(2006), pp. 197–202. Li, H., Kuang, G., Hu, S., Guo, H., Jin, R., Vekariya, R.L., Removal of Aluminum from Leaching Solution of Lepidolite by Adding Ammonium, JOM, 68(2016), 10, pp. 2653-2658. Luong, V.T., Kang, D.G., An, J.W., Kim J.M., and Tran, T., Factors affecting the extraction of lithium from lepidolite, Hydrometallurgy, 134-135(2013), pp. 54 – 61.

lithium from lepidolite, Hydrometallurgy, 141(2014), pp. 8 – 16.

PT

Luong, V.T., Kang, D.J., An, J.W., Dao, D.A., Kim, M.J., and Tran, T., Iron sulphate roasting for extraction of

Menéndez, M., Vidal, A., Toraño, J., Gent, M., Optimisation of spodumene flotation, The European Journal of

RI

Mineral Processing and Environmental Protection, 4(2004), 2, pp. 130-135.

SC

Meshram, P., Pandey, B.D., Mankhand, T.R., Extraction of lithium from primary and s econdary sources by pretreatment, leaching and separation: a comprehensive review, Hydrometallurgy, 150(2014), pp. 192–208.

NU

Meshram, P., Abhilash, Pandey, B.D., Mankhand, T.R., Deveci, H., Acid baking of spent lithium ion batteries for selective recovery of major metals: A two-step process, Journal of Industrial and Engineering Chemistry,

MA

43(2016), pp. 117-126.

Munson, G.A., Clarke, F.F., Mining and concentrating spodumene in the Black Hills, South Dakota, Mining Engineering, (1955), Transactions AIME, pp. 1041-1043.

ED

OECD/IEA – International Energy Agency, 2016, Global EV outlook 2016 Beyond one million electric cars, Available on: . Orocobre Limited, 2012, Lithium Market, Available on: http://www.orocobre.com.au/Lithium_Market.htm,

EP T

Access: 04th December 2016.

Pourbaix, M., 1974, Atlas of electrochemical equilibria in aqueous solutions, 2 nd Ed., NACE/CEBELCOR.

AC C

Seidell, A., 1919, Solubilities of Inorganic and Organic Compounds: A Compilation of Quantitative Solubility Data from the Periodical, D.Van Nostrand Company, 843 p. Siame, E., Pascoe, R.D., Extraction of lithium from micaceous waste from china clay production, Minerals Engineering, 24(2011), pp. 1595–1602. Sitando, O., and Crouse, P.L., Processing of a Zimbabwean petalite to obtain lithium carbonate, International Journal of Mineral Processing, 102-103(2012), pp. 45–50. Vieceli, N., Nogueira, C.A., Pereira, M.F.C., Durão, F.O., Guimarães, C., Margarido, F., Optimization of lithium extraction from lepidolite by roasting using sodium and calcium sulphates , Mineral Processing and Extractive Metallurgy Review, 38(2017a), pp. 62-72.

ACCEPTED MANUSCRIPT Vieceli, N., Nogueira, C.A., Pereira, M.F.C., Dias, A.P.S., Durão, F.O., Guimarães, C., Margarido, F., Effects of mechanical activation on lithium extraction from a lepidolite ore concentrate, Minerals Engineering, 102(2017b), pp. 1-14. Vu, H., Bernardi, J., Jandová, J., Vaculíková, L., and Goliáš, V., Lithium and rubidium extraction from zinnwaldite by alkali digestion process: Sintering mechanism and leaching kinetics , International Journal of Mineral Processing, 123(2013), pp. 9–17. Yan, Q., Li, X., Yin, Z., Wang, Z., Guo, U., Peng, W., and Hu, Q., A novel process for extracting lithium from

PT

lepidolite, Hydrometallurgy, 121-124(2012a), pp. 54 – 59. Yan, Q., Li, X., Wang, Z., Wu, X., Guo, H., Hu, Q., Peng, W., and Wang, J., Extraction of valuable metals from

RI

lepidolite, Hydrometallurgy, 117-118(2012b), pp. 116 – 118.

SC

Yan, Q., Li, X., Wang, Z., Wu, X., Wang, J., Guo, H., Hu, Q., and Peng, W., Extraction of lithium from lepidolite by sulfation roasting and water leaching, International Journal of Mineral Processing, 110111(2012c), pp. 1 – 5.

NU

Yan, Q., Li, X., Wang, Z., Wang, J., Guo, H., Hu, Q., Peng, W., and Wu, X., Extraction of lithium from lepidolite using chlorination roasting-water leaching process, Transactions of Nonferrous Metals Society of

MA

China, 22(2012d), 7, pp. 1753 – 1759.

Wagner, F.S., 2006, Rubidium and Rubidium Compounds , Kirk-Othmer Encyclopedia of Chemical Technology, DOI: 10.1002/0471238961.1821020923010714.a01.pub2.

ED

Wietelmann, U., Bauer, R.J., 2000, Lithium and lithium compounds, Ullmann's Encyclopedia of Industrial

AC C

EP T

Chemistry, DOI: 10.1002/14356007.a15_393.