Revisiting the Traditional Process of Spodumene Conversion and Impact on Lithium Extraction

Abstract

Since 1950, the traditional process has been dominating the production of lithium compounds from spodumene ores to sustain the lithium market because of its economic viability and the markets need for technical grade (99.5% purity) product. This traditional process includes thermal conversion, acid roasting and lithium leaching . However, this process has not been challenged and very few studies have tried to explain its limitation (95% lithium yield) or optimize the thermal treatment. Here, α to β spodumene conversion and lithium extraction were performed in a rotary kiln on a 2 mm to 2 cm spodumene concentrate instead of a micrometric one using a 1050 °C, 30 min conversion treatment and 250 °C, 30 min, 30% H₂SO₄ excess leaching treatment. X-Ray Diffraction analyses were performed on the converted material to determine the conversion rate by Rietveld analysis. It was observed after thermal treatment that the α-spodumene particles fractured and divided while the impurities remained compact. A sifting was performed after thermal treatment and it was determined that 65% of the initial mass became finer than 180 µm. X-Ray diffraction analyses and lithium content measurement were performed on both fractions and it was determined that the finer fraction had a very high lithium content of 3.24 wt%. Lithium extractions were performed on both fractions separately. While the coarser fraction’s lithium yield was only 61% the finer fraction’s lithium yield went up to 99% without any additional treatment. These observations may open more economical ways for the traditional process by potentially bypassing two third of costly steps such as grinding.

Highlights

Thermal conversion of a 2 mm–2 cm spodumene concentrate gives 65% of the initial mass under 180 µm
Impurities stay unaffected by the thermal treatment
The lithium content can reach 3.24 wt% for the fine fraction
An extraction rate of 99% on the fine fraction can be obtained

Introduction

Lithium and its compounds are essential to modern industry and society. They are used in a wide range of applications such as glasses, ceramics, semiconductors, etc. But the most important application nowadays is lithium -ion batteries (LiBs) used in smartphones or newly developed electric cars. In 2012, the lithium production established itself at 34,000 tons with estimated reserves of 13 million tons [1]. The predicted growth rates for lithium carbonate and lithium hydroxide, two of the LiBs raw materials, are respectively of 10.0 and 14.5% until 2025 [2].

Up to 2012, the salt lake brines have been the main source of lithium due to low production costs. However, between 2003 and 2012, the price per ton of lithium went from 2000 US$ to 6000 US$ [3], which led the lithium -rich minerals to regain market share, now accounting for 50% of the world production [4].

Spodumene is one of the many lithium rich minerals , but is the most common and the most studied among all. A process of lithium extraction from spodumene ore has been patented as early as the middle of the twentieth century [5]. This has led it to be the leader in the commercial mining of lithium rich minerals industry. The former process is still nowadays the most used by companies and is the base of many future spodumene projects. It is a sulphuric acid (H₂SO₄) process that permits 90–95% lithium recovery with 30–140% stoichiometric excess of H₂SO₄ [5, 6]. Several processes have been studied since, using chlorine (Cl₂) [7] or fluoric acid (HF) [8] for example, but these processes concern the lithium extraction . Very few studies concern the thermal treatment step. There has been some study such as microwave heating [9] but the consensus is that the rotary kiln heating of the traditional process is the most efficient method.

Even though the lithium extraction from spodumene has been heavily studied, there is almost no research on improving the traditional process, particularly on improving the thermal conversion of spodumene .

Here, the influence of initial particle size on thermal conversion and lithium extraction yield has been studied using the traditional process. This study is based on the property of spodumene . Spodumene undergoes a phase transition during the thermal treatment which allows its density to go down to 2.37 from 3.15 g/cm³ [10] leading to a volume expansion of 27% of its crystal structure . It is known that micrometric particles expand during conversion, however, the effect of the expansion on coarser particles has never been studied even though they are known to fracture (decrepitation phenomena) [11].

Materials and Methods

Initial Material

The spodumene concentrate used for the experiments is a 2 mm–2 cm DMS concentrate (Fig. 1). It was obtained from fine ore that went through a DMS Hydraulic Separator, a rougher dense media cyclone, a cleaner dense media cyclone before being dried in DMS concentrate rotary electric dryer. The magnetic fraction of the concentrate was then removed by dry magnetic separators [12]. Particle colours ranged from white to green (see Fig. 1) with many particles presenting veins of one colour into the other, suggesting that the concentrate is highly inhomogeneous. Black particles can be seen, their size being in the same range as the spodumene particles. Orange to red areas can also be seen on some particles.

../images/468727_1_En_191_Chapter/468727_1_En_191_Fig1_HTML.gif — Fig. 1
Spodumene concentrate used as base material for the experiments

Thermal Conversion of Spodumene

Approximately 60 g of concentrate was weighed with a Mettler Toledo New Classic MF balance, with attention given to ensure the presence of each type of particles. The concentrate was transferred into a Thermo Scientific Lindberg Blue M Tube Furnace rotary kiln equipped with a 2.5 cm diameter quartz tube (Fig. 2) and then heated at 1050 °C for 30 min. It should be noted that the temperature was not controlled with the furnace controller but with a thermocouple and that the timer was started when the temperature reached 1045 °C on the thermocouple.

../images/468727_1_En_191_Chapter/468727_1_En_191_Fig2_HTML.jpg — Fig. 2
Rotary kiln used to roast the spodumene concentrate with a thermocouple inserted to read temperature

Sifting

After thermal treatment, the converted concentrate was sifted in a Fisher Scientific 180 µm laboratory test sieve using a Tyler RO-TAP RX-29 (Fig. 3) for 5 min. The material was weighed before sifting and each fraction, one being above 180 µm and the other under 180 µm, was weighed after to determine their weight proportion in the heated concentrate.

../images/468727_1_En_191_Chapter/468727_1_En_191_Fig3_HTML.jpg — Fig. 3
Sifting system used to sift the converted concentrate into two fractions

Acid Roasting and Lithium Leaching

Depending on the fraction, between 20 and 30 g of concentrate was weighted with a Mettler Toledo New Classic MF balance. Then, an accurate mass of H₂SO₄ (C = 96.2 wt%) determined by Eq. (1) [12] including stoichiometric and excess amount of acid desired, was added to each fraction. For these experiments %_excess was fixed at 30%. Manual stirring ensured the homogeneity of the mixtures until visual uniformity. The action of sulphuric acid on spodumene follows Eq. (2) [13].

$m_{acid} = m_{\beta - spod} \left( {100\% + \%_{excess} } \right)*\frac{{M_{acid} }}{{2M_{Li} }}*\frac{{C_{Li} }}{{C_{acid} }}$

(1)

$2LiAlSi_{2} O_{6} \left( s \right) + H_{2} SO_{4} \left( l \right) \to 2HAlSi_{2} O_{6} \left( s \right) + Li_{2} SO_{4} \left( s \right)$

(2)

Each mixture was then heated at 250 °C for 30 min in the same furnace used for thermal conversion. After acid roasting, 40 mL of osmosed water was added to the mixture and magnetically stirred for 30 min to leach the lithium sulphate (Li₂SO₄) from the concentrate and solubilized it. The leached Li₂SO₄ solution was separated from the roasted concentrate by filtration on a Buchner using a 1 µm pore filter paper and the roasted concentrate was washed three times with 25 mL of milliQ water.

Chemical Analyses

X-Ray Diffraction (XRD)

XRD patterns were obtained using a PANalytical X’Pert PRO MPD diffractometer equipped with a PIXcel^1D detector. The data were collected between 5° and 70° (2θ angle) with the Cu Kα radiation at a rotation speed set at 1 s⁻¹. Rietveld analyses using MDI Jade2010 were performed when deemed necessary on patterns to determine the chemical composition of the samples.

Atomic Absorption Spectroscopy (AAS)

It has to be noted that the initial DMS concentrate was not analyzed by AAS due to its highly inhomogeneous nature. Approximately 250 mg of each of the heated fractions was weighed with a Mettler Toledo XA204 DeltaRange balance. Then, 3 mL hydrochloric acid (HCl), 4 mL hydrofluoric acid (HF) and 1 mL nitric acid (HNO₃) were added and the mixtures were placed into a CEM Mars 6 microwave digestion system at 200 °C for 20 min with a 20 min ramp to solubilize the solids. A second microwaving was performed after adding 50 mL of hydrogen borate (H₃BO₃) solution (C = 4.5% m/v) at 170 °C for 15 min with a 15 min ramp to neutralize the HF. The obtained solutions were diluted in HNO₃ (C = 2% m/v) and their lithium content was measured using a PerkinElmer AAnalyst 200 AA spectrometer. The filtered lixiviants were diluted in HNO₃ (C = 2% m/v) and analysed using the same spectrometer. The initial converted material underwent the same procedure.

Results and Discussion

Thermal Conversion

The initial DMS concentrate was thermally converted and analyzed, as were each fraction and the initial converted concentrate, by XRD . It has to be noted that most of the particles became much smaller and all lost their green color for a red-brown one (Fig. 4). The coarser red-brown particles were not hard and could be crushed easily. Moreover, the black particles remained hard and compact after thermal treatment and were also analyzed by XRD to determine their nature. Finally, white particles, that were easily crushed, even manually, appeared but were not crystalline enough to be analyzed by XRD, though are supposed to be albite.

../images/468727_1_En_191_Chapter/468727_1_En_191_Fig4_HTML.gif — Fig. 4
Initial concentrate after thermal treatment and before sifting

The converted concentrate was analyzed by XRD and then sifted, before analyzing both fraction by XRD (Fig. 5), the finer fraction weighing 65% of the initial mass while the coarser fraction weighting the other 35%. The XRD analyses showed β-spodumene (LiAlSi₂O₆), virgilite (Li_xAl_xSi_3−xO₆), and quartz (SiO₂) in all patterns. The α-spodumene (LiAlSi₂O₆) is detected in the finer and coarser fractions but not in the initial concentrate. If the pattern of the initial concentrate shows that the conversion is completed, then it appears there is still a bit of unconverted material. Albite (NaAlSi₃O₈) appears exclusively in the coarser fraction. The reason behind this is that the impurities are concentrated in the coarser fraction. Albite particles can be seen in XRD because of their higher concentration in the fraction.

../images/468727_1_En_191_Chapter/468727_1_En_191_Fig5_HTML.gif — Fig. 5
XRD patterns of (1) initial DMS concentrate, (2) initial concentrate converted at 1050 °C during 30 min, (3) coarser fraction of (2) after sifting, (4) finer fraction of (2) after sifting

The appearance of virgilite is very interesting because it is not a mineral reported before in spodumene conversion. It is a naturally occurring mineral that has a stuffed β-quartz structure [14] with one aluminium atom randomly replacing one silicon atom with lithium atoms inserted. The behavior towards acid leaching of this mineral is unknown and so are its effects on the lithium yield. The Rietveld analyses show non-negligible rates of virgilite in the different materials (Table 1). The initial DMS concentrate did not undergo a Rietveld analysis because of the too high preferential orientations in the concentrate.

Table 1

Rietveld analysis showing the composition of initial concentrate converted at 1050 °C for 30 min, coarser fraction after sifting, finer fraction after sifting

	α-spodumene LiAlSi₂O₆ (%)	β-spodumene LiAlSi₂O₆ (%)	Virgilite Li_xAl_xSi_3−xO₆ (%)	Quartz SiO₂ (%)	Albite NaAlSi₃O₈ (%)
Initial converted concentrate	0	88.1	7.9	4.0	0
Coarser fraction (35 wt%)	20.2	35.1	17	25.4	2.2
Finer fraction (65 wt%)	1.3	81.4	14.1	3.1	0

α-spodumene LiAlSi₂O₆ (%)

β-spodumene LiAlSi₂O₆ (%)

Virgilite Li_xAl_xSi_3−xO₆ (%)

Quartz SiO₂ (%)

Albite NaAlSi₃O₈ (%)

Initial converted concentrate

88.1

7.9

4.0

Coarser fraction

(35 wt%)

20.2

35.1

25.4

2.2

Finer fraction (65 wt%)

1.3

81.4

14.1

3.1

The black particles have been identified by XRD after treatment (Fig. 6) as a mixture of diopside (CaMgSi₂O₆), anorthite (CaAl₂Si₂O₈) and augite ((Ca, Mg, Fe, Ti, Al)₂Si₂Al₂O₆), which are alumina silicates containing many metals such as Mg, Ca, Fe. Nevertheless, none of them contain lithium , which makes the fact that they stay hard and compact very interesting for an easy purification , their size being much larger than the spodumene particles. The black particles were not analysed by XRD before the thermal treatment due to many of them being trapped inside the α-spodumene particles, making it impossible to get a representative sample of those particles before thermal treatment.

../images/468727_1_En_191_Chapter/468727_1_En_191_Fig6_HTML.gif — Fig. 6
XRD pattern of the black particles in the initial concentrate after thermal treatment at 1050 °C during 30 min

Lithium Extraction

The initial converted concentrate and both fractions were digested and analyzed by AAS to determine their lithium content. The lithium concentration of the coarser fraction was measured at C_Li = 2.45 ± 0.02 wt% (C_Li2O = 5.27 ± 0.04 wt%) while the finer fraction was measured at C_Li = 3.24 ± 0.01 wt% (C_Li2O = 6.97 ± 0.02 wt%). After calculation, these results give a lithium concentration of C_Li = 2.96 ± 0.03 wt% (C_Li2O = 6.38 ± 0.06 wt%). The initial converted concentrate was measured at C_Li = 2.92 ± 0.08 wt% (C_Li2O = 6.28 ± 0.17 wt%) which fits the previous calculation. The weight concentration of the finer fraction is high and very interesting since it did not undergo additional purification treatment.

The lithium yield for the coarser fraction is 61%. This is a low rate of extraction but was expected due to the high impurity concentration in the fraction and the absence of purification . The interesting result is a lithium yield of 99% for the finer fraction. It shows that purification is not needed for the fine particles obtained after thermal treatment. It also points out that converting the concentrate before grinding and flotation may allow for a reduction in the amount of material used for those two steps by almost two thirds. Given the lithium concentration in each fraction and their respective weights after thermal treatment, this would be equivalent to a lithium yield of 88% on the initial converted concentrate.

The coarser fraction was observed by optical microscopy (Fig. 7) to see if some particles could explain its low lithium yield. It was observed that many transparent particles, supposed to be SiO₂ , contain finer red-brown particles which are characteristic of β-spodumene . Either the quartz particles become viscous enough to allow fine spodumene particles to insert themselves during thermal treatment or the spodumene particles are already contained in the quartz particles and are not big enough to free themselves from their quartz during their decrepitation. Given that quartz is not attacked by sulphuric acid, this explains why the lithium is extracted at a very low rate and confirms that other processing, such as grinding and flotation , could be necessary to improve the lithium recovery from larger particles in order to liberate the trapped β-spodumene particles from their quartz and remove some, if not all, the impurities.

../images/468727_1_En_191_Chapter/468727_1_En_191_Fig7_HTML.jpg — Fig. 7
Quartz coating of β-spodumene particles (circles)

Conclusion

A thermal treatment of a 2 mm to 2 cm DMS spodumene concentrate showed that the particles decrepitated during treatment. Sifting the treated spodumene showed that 65% of the initial mass became smaller than 180 µm. It also showed that most of the impurities were not affected by the thermal conversion and were contained in the coarser fraction of the heated concentrate. An acid leaching was performed on both fraction and while the coarser fraction had a lithium yield of only 61% the finer fraction had an extremely high lithium yield of 99% without any additional treatment. This result opens new possible ways of thermally treating spodumene concentrate that would allow the initial grinding and flotation of the concentrate to be bypassed. It has to be noted that the acid leaching on the coarser fraction was performed also without additional treatment. Its low yield may be related to the absence of grinding and flotation , both of which are to be performed in future works, which will include the study of the influence of the rate of virgilite and sifting on lithium extraction , as well as an economical study of this process.

Acknowledgements

The authors are grateful to the employees of the “Centre de Caractérisation des Matériaux” for their help with the various characterization devices. This work is financed and supported by “Fonds de Recherche du Québec-Nature et Technologies” (FRQNT) and Nemaska Lithium Inc.

References

1.
U.S. Geological Survey (2012) Minerals yearbook—lithium, USGS, Virginia
2.
Christmann P, Gloaguen E, Labbé J-F, Melleton J, Piantone P (2015) Chapter 1—Global lithium resources and sustainability issues A2—Chagnes, Alexandre. In: Światowska J (ed) Lithium process chemistry. Elsevier, Amsterdam, pp 1–40
3.
Lithium: market outlook to 2017, 12th edition, Roskill, 2013
4.
U.S. Geological Survey, Mineral commodity summaries 2016: U.S. Geological Survey Lithium, USGS, Virginia, 2016, p 202
5.
Ellestad RB, Leute KM (1950) Method of extracting lithium values from spodumene ores. In: M. Corporation (Ed.) United States, p 10
6.
Hader RN, Nielsen RL, Herre MG (1951) Lithium and its compounds. Ind Eng Chem 43:2636–2646Crossref
7.
Barbosa LI, Valente G, Orosco RP, González JA (2014) Lithium extraction from β-spodumene through chlorination with chlorine gas. Miner Eng 56:29–34Crossref
8.
Rosales GD, Ruiz MDC, Rodriguez MH (2014) Novel process for the extraction of lithium from -spodumene by leaching with HF. Hydrometallurgy 147–148:1–6Crossref
9.
Khoshdel Salakjani N, Nikoloski A, Singh P (2017) Mineralogical transformations of spodumene concentrate from Greenbushes, Western Australia. Part 2: Microwave heating. Miner Eng 100:191–199Crossref
10.
Hevia RP, Gil C, Carda JB (2006) Mineral treatment of spodumene in the ceramic industry. Characterisation and technological use, Qualicer
11.
C.K. Gupta, Chemical metallurgy—principles and practice, Chapter 5—Hydrometallurgy: pretreatments, p 478, (2003)
12.
Dupéré M, Perez P, et al. (2018) Nemaska Lithium—NI 43–101 Technical report feasibility study on the Whabouchi lithium mine and Shawinigan electrochemical plant, p 235
13.
Xiao M, Wang S, Zhang Q, Zhang J (1997) Leaching mechanism of the spodumene sulphuric acid process. Rare Met. (Beijing) 16:36–44
14.
French BM, Jezek PA, Appleman DE (1978) Virgilite: a new aluminum silicate mineral from the macusani glass. Peru, American Mineralogist 63:461–465