WO2024038429A1

WO2024038429A1 - Method for preparing lithium bromide

Info

Publication number: WO2024038429A1
Application number: PCT/IL2023/050820
Authority: WO
Inventors: Eyal Barnea; Tamer ANDREA; Basam Masri
Original assignee: Bromine Compounds Ltd.
Priority date: 2022-08-14
Filing date: 2023-08-08
Publication date: 2024-02-22

Abstract

A process for preparing lithium bromide comprising contacting lithium carbonate with hydrogen bromide gas at a temperature of not less than 200 °C and obtaining anhydrous lithium bromide in a solid form.

Description

Method for preparing lithium bromide

There is a growing need for lithium compounds, for example, as raw materials in production of lithium batteries or in other fields. Lithium bromide is a good candidate material for such applications, e.g., to take part in the synthesis of ionic conductors/solid electrolytes.

Lithium bromide is prepared industrially by neutralizing lithium carbonate or lithium hydroxide with hydrobromic acid, namely, the reaction takes place in aqueous medium. Lithium bromide in fact exists in anhydrous and hydrated forms, e.g., as a monohydrate. The recovery of the anhydrous form from an aqueous solution is not straightforward: concentration of the lithium bromide solution by evaporation to precipitate the salt and isolation of it by filtration yields the hydrated form. Dehydration of the monohydrate to produce the anhydrous form is difficult to achieve, as the monohydrate does not release its water molecule easily. For that reason, a water-free lithium bromide formation reaction that could give the anhydrous form in a direct manner at an acceptable quality (e.g., >97% purity, and ideally, >99% purity) , could offer significant benefits.

Preparation of lithium halides based on a reaction of lithium compound in a solid form with gaseous halogen source in the absence of water was mentioned in an old American patent which dates back to the 1960s. US 2,968,526 shows the manufacture of anhydrous lithium halides by direct halogenation of lithium hydroxide. Excellent yields were reported for LiCl obtained by a reaction of LiOH with chlorine gas, as an analysis of the resulting product indicates >98% of LiCl . However, in the case of lithium bromide, the results were not that good; the passing of bromine vapors over heated LiOH resulted at best in a product consisting of 95.1% LiBr; the amount of unreacted LiOH detected in the product was not insignificant (4.5%) . A recent example of a related approach is found in WO 2022/049123, showing dehydration of lithium hydroxide monohydrate by contact with a carrier gas, followed by reaction of the anhydrous lithium hydroxide with hydrogen halide gases, specifically HC1 gas diluted in the carrier gas, to give LiCl .

We have now found that anhydrous lithium bromide can be prepared by the action of gaseous hydrogen bromide on solid lithium carbonate (L12CO3) at high temperatures, as shown by the reaction equation below:

When the starting materials are of high purity, pure lithium bromide product is formed. Results reported below indicate that a surprisingly high product yield is achievable even in a simple experimental setup where the contact between the L12CO3 powder particles and the HBr gaseous stream was not optimal, i.e., just by heating powdered L12CO3 in a tube furnace under an environment created by flowing HBr<_g) over a few hours.

Accordingly, the invention is primarily directed to a process for preparing lithium bromide, comprising contacting lithium carbonate with hydrogen bromide gas at elevated temperature, e.g., above 200°C, e.g., above 300°C, e.g., >350°C, e.g., >400°C, and obtaining anhydrous lithium bromide in a solid form.

As mentioned above, the starting materials (HBr gas and lithium carbonate) both need to be of high purity, when pure lithium bromide is desired.

Lithium carbonate suitable for use in the invention (for the purpose of reaching pure LiBr) is available on the marketplace, e.g., with >99.0%, >99.3%, >99.5%, >99.8% purity from manufacturers such as Albemarle (formerly Rockwood Lithium) ; impurities profile consists of <0.09% Na; <0.02% Ca; < 0.01 Mg; <0.05% sulfate; < 30 ppm Fe2Os; <10 ppm K and <30 ppm B. It is also possible to purify technical grade lithium carbonate by known methods, and then use the purified material in the process of the invention. Technical grade lithium carbonate can be treated to reach an acceptable purity level so as to be used in the invention by the method described in US 6,592,832 (bubbling CO₂ through Li2CC>3 suspension, to convert Li2CC>3 into the water soluble lithium bicarbonate, filtrating the lithium bicarbonate solution, passing the filtrate through an ion-exchange resin to remove metal impurities, decomposing the lithium bicarbonate under heating to precipitate L12CO3 and collecting L12CO3 of high purity) ; or by the method described by Xu et al. in Metals 2021, 11, 1490 (where, instead of passing the lithium bicarbonate - containing filtrate through an ion-exchange resin, the solution is evaporated to crystallize lithium carbonate followed by airstream pulverization) ; or by the method described in US 6, 048,507.

It is not uncommon for commercial L12CO3 powder to show bimodal particle size distribution (PSD) with D90 > 400 pm. Experimental results reported below indicate that although such grades of commercial L12CO3 can be used in the process of the invention, better conversion of the starting material is achieved with Li₂CO₃ powder possessing reduced, particle size distribution with D90 < 100 pm, e.g., 30 pm < D90 < 100 pm, or 30 pm < D90 < 70 pm (and D50 < 15 pm) , and thus, commercial L12CO3 powder can undergo particle size reduction by customary milling/grinding techniques to adjust the PSD within the ranges set out above, and then the powder with reduced PSD is contacted with HBr gas in a suitable reactor where the synthesis takes place. With increasing surface area of lithium carbonate exposed to the action of HBr gas, the conversion to lithium bromide is improved. The symbols d<o.i), d < o .5 > , d<o.9) and Dio, D50 and D90, respectively, are used herein interchangeably.

High-purity HBr (>99.9%, >99.99%, >99.999%) is available from various manufacturers, such as ICL-IP, Showa Denko and Linde. Methods of preparing hydrogen bromide gas with purity exceeding 99.99% are described, for example, in US 5, 685,169 and US 6, 335, 222.

Turning now to the reaction between the HBr gas and the LiiCOs particles, it can take place in any reactor suited for solid/gas reactions, i.e., non-catalytic gas-solid reactors which provide good contact between the gas and a solid reactant, including stationary and moving packed bed reactors, fluidized bed reactors, rotary drum reactors and rotary kiln reactors, to name a few major industrially common reactor configurations for this type of reaction. The reaction is preceded by purging the air from the system with nitrogen, following which the feed of HBr to the reactor is initiated. Usually, neat HBr gas stream is supplied to the reactor, such that LiiCOs powder is heated in an environment consisting of the flowing HBr<_g) stream, but dilution of the incoming HBr stream in an inert gas is possible. The flow rate of the incoming HBr gas stream is adjusted, depending on the reactor configuration, to supply, over one hour or over a few hours, the total amount of HBr gas needed, which may be in excess relative to the amount of LiiCOs. Stoichiometry dictates that two moles of HBr are consumed by each mole of LiiCOs. Usually, a molar excess of HBr to L12CO3 is used, i.e., above the 2:1 stoichiometric ratio, for example, > 2.2:1, e.g., > 3:1, e.g., > 5:1, e.g., in the range from 10:1 to 70:1, depending on the reactor design. Unreacted, outgoing HBr gas is led to a suitable absorbent medium, such as aqueous alkali hydroxide, to undergo neutralization, or recycled to the reactor . Thus, in another aspect of the invention, there is provided a process wherein an HBr stream is supplied to a reaction vessel in which lithium carbonate is heated at a flow rate and over time such that the molar ratio of HBr supplied relative to lithium carbonate is not less than 2.2:1.

Temperature is yet another important factor influencing the efficiency of the lithium bromide formation reaction. The literature melting point of the starting material - lithium carbonate - is 723°C, whereas the literature melting point of the product - lithium bromide - is 552 °C. Experimental work reported below shows that when the reaction was run in a tube furnace at 400°C, a product in a pulverized form was obtained, e.g., a free-flowing powder. But when the reaction between lithium carbonate and HBr was conducted at 500°C, the reaction mass exhibited texture change and turned into a lumpy product. Without wishing to be bound by theory, it is assumed that the texture change observed indicates a melting process. It is possible that lithium carbonate/lithium bromide forms a eutectic mixture, and that the melting point of the eutectic mixture, consisting of the progressively formed lithium bromide and the gradually consumed lithium carbonate - is in the range from 400°C to 500°C. Thus, fusion of the particles occurs. Subsequently, upon cooling the reaction zone, the molten particles solidify and coalesce such that the solidified mass requires comminution to recover a powdery lithium bromide. But running the reaction at a sufficiently high temperature to bring about the texture change of the reaction mass is not without its benefits, as it has been observed that substantially complete conversion to lithium bromide could be achieved via the "melting/solidif ication route" (i.e., >96%; >97%; >98% and even >99% purity) . Accordingly, a specific aspect of the invention is a process comprising contacting lithium carbonate with hydrogen bromide gas at a temperature above 400°C but below a temperature associated with a change of texture of the reaction mixture and/or a change of phase of the reaction mixture and obtaining anhydrous lithium bromide in a particulate solid form.

Another specific aspect of the invention is a process comprising contacting lithium carbonate with hydrogen bromide gas at a temperature exceeding the temperature associated with a change of texture of the reaction mixture and/or a change of phase of the reaction mixture, recovering a reaction product consisting essentially of anhydrous lithium bromide and optionally comminuting the solid product to obtain anhydrous lithium bromide in a pulverized form.

As indicated by the reaction equation (1) , water and carbon dioxide are by-products of the conversion reaction of lithium carbonate to lithium bromide. Based on the design of the reactor, the progress of the reaction can be monitored, e.g., by observing the condensation of water on the cold sides of a furnace tube, or by determination of heat release.

Experimental work reported below shows that high percentage yields are attainable over a broad temperature range, e.g., 300°C < T, provided that the lithium carbonate starting material is reduced to an appropriate particle size. That is, these two process variables (temperature and PSD) counterbalance each other in that the use of fine lithium carbonate particles permits relatively low reaction temperatures (~300-400°C) . Or vice versa; sufficiently high reaction temperature (close to 500°C) guarantees high percentage yield even if lithium carbonate particles population possessing large PSD is reacted with the HBr gas. Accordingly, the invention provides a process for preparing lithium bromide by the reaction of lithium carbonate with hydrogen bromide gas, wherein the temperature of the reaction and the particle size distribution of the lithium carbonate are adjusted so as to obtain anhydrous lithium bromide at high yield .

For example, the temperature of the reaction is adjusted in the range of 300°C < T (e.g., the subranges of 300°C < T < 550°C, 300°C < T < 500°C, 300°C < T < 400°C, 400°C < T < 500°C, 400°C

< T < 450°C, 450°C < T < 500°C, 450°C < T < 480°C and 480°C < T < 500°C) , and the particle size distribution of the lithium carbonate starting material is adjusted within the range of D50

< 300pm (e.g., 1 pm < D50 d 300pm, 3 pm < D50 d 300pm, 5 pm < D50

< 300pm, 100 pm < D50 d 300pm, 100 pm < D50 d 250pm, 5 pm < D50

< 100pm, D50 < 15 pm, e.g., 5 pm < D50 < 15 pm, as determined by laser diffraction particle size analysis, so as to obtain anhydrous lithium bromide at a yield of not less than 85%, e.g.,

>90%, >95%, >97% and even >99%.

The efficiency of the lithium bromide formation reaction in a tube furnace was studied at the lower and upper ends of the 300°C < T < 500°C interval, using two populations of lithium carbonate particles, one with small PSD (Dso~5 pm) and the other with large PSD (Dso~3OO pm) . As a good rule of thumb, when lithium carbonate with PSD characterized by 100 pm < D50 d 300pm (e.g., 100 pm < D50 d 250pm) is available, then its conversion to lithium bromide by the action of HBr gas is effectively achieved at a temperature in the range of 400°C < T < 500°C, whereas lithium carbonate with smaller PSD of 5 pm < D50 d 100pm (e.g., D50 < 15 pm, such as 5 pm < D50 < 15 pm) is transformed into lithium bromide at surprisingly high yield even at a temperature in the 300°C < T < 400°C. The limitations of the milling equipment available and the reactor configuration may be taken into consideration to adjust the T/PSD variables to maximize yield and minimize costs, by conducting trial and error experiments .

Alternatively, a linear/nonlinear regression model can be fitted to the experimental data to create a prediction formula showing the relationship between the explanatory variables (which are reaction temperature and lithium carbonate particle size distribution) , and the dependent variable (which is the yield of lithium bromide) , to adjust the T/PSD within the ranges set out above, so as to obtain anhydrous lithium bromide at a yield of not less than 90% (>95%, >99%) .

For example, the JMP software was applied to fit a model which contained a term for each process variable and an interaction term (a product of the two independent variables) . That is, in its most general form, the model fitted to the process by the JMP software is of the formula:

LiBr yield % = po + piT + P2D50 + P12 (T' x D50')

The best model prediction formula (with R² value of 0.997) has the following coefficients:

Po = 71.4

Pi = 0.14 p₂ = -0.046

Pi2= 0.00056

T' is the temperature T, corrected by a constant; T'= T-425

D50' is D50 corrected by a constant; D50' = Dso-151.28

A model shown in the experimental section below has the formula: Yield % = 35.47 + (-0.046 x D₅₀) + 0.14 x T + ( ( D₅₀ - 151.28) x (T - 425) ) x 0.00056 In the drawings :

Figure 1 is a particle size distribution curve of the raw material L12CO3 powder.

Figure 2 is a particle size distribution curve of L12CO3 after fine grinding in tungsten carbide grinder.

Figure 3 is a particle size distribution curve of L12CO3 after fine grinding in a vortex mill.

Figure 4 is a particle size distribution curve of L12CO3 after fine grinding in a vortex mill.

Figure 5 shows percentage yield versus PSD (D50) plots at constant temperature (left-hand plot corresponds to 300°C and the right-hand plot corresponds to 500°C) .

Figure 6 is an 'actual by predicted plot' generated by the JMP software .

Figure 7 is a contour plot generated by the JMP software.

Examples

Materials

High-purity HBr is available from ICL IP (99.999%) .

Pure Li₂CO₃ powder was purchased from Albemarle (formerly Rockwood Lithium) .

Methods

Li⁺ measurement: the sample was dissolved in water, acidified with HNO3 and injected into ICP model Agilent 5110.

Bn measurement: the sample was dissolved in water and the Bn was determined by potentiometric titration with 0. IN AgNO₃ volumetric solution.

CO₃ ²~ measurement: CO3²- was determined by manometric analysis.

Preparation 1

Finely ground Li2COs powder with unimodal PSD

PSD of high purity commercial L12CO3 powder was determined by laser diffraction particle size analysis with Mastersizer 2000 (dispersant: isopropanol) . The results indicate a bimodal distribution, with d<o.i) = 5.1 pm; d<o.5) = 23.2 pm and d<o.9) = 495.8 pm, as shown in Figure 1.

Particle size reduction of the L12CO3 powder was carried out in a tungsten carbide ring grinder machine. First, the machine was "washed" by operating on 10 g of L12CO3. The as-milled powder was removed from the machine and a fresh sample of commercial Li₂CO₃ (30 g) was placed in the grinder. The grinder operated for ten minutes to obtain reduced particle size L12CO3. PSD measurement of the powder collected shows a unimodal distribution with d<o.i) = 1.7 pm; d<o.5) = 10.6 pm and d<o.9) = 44.2 pm, as shown in Figure 2. Examples 1-7

Preparation of lithium bromide

A series of experiments were performed in a tube furnace to evaluate the conversion of solid lithium carbonate to lithium bromide by the action of gaseous HBr under different conditions.

A porcelain crucible was loaded with 5-10 g of pure L12CO3 powder (either the commercial powder (Examples 1 to 4) or the reduced particle sized powder of preparation 1 (Examples 5 to 7) ) and placed in the center of a quartz tube. The quartz tube was sealed from both sides with stainless steel flanges. One side, through which the incoming gas enters the tube, was connected to a flowmeter whereby N2 or HBr<_g) were fed to the tube, and the opposite side, through which the outgoing gas stream exits the tube, was connected to an empty trap (to guard against accidental back-suction) followed by NaOH trap to neutralize the excess HBr leaving the tube.

The tube was placed in a tube furnace and flushed with N2 for Ih at a rate of 30 L/h to remove residual air. The N2 feed was stopped and HBr<_g) was fed at the same flow rate. The temperature was raised to 400 °C (Examples 1 and 2) or 500 °C (Examples 3 to 7) at a rate of 10 °C/min and the target temperature was maintained for 2-6 h as tabulated below.

Drops of water coming out of the reaction (see reaction equation

1 above) were observed on both unheated sides of the tube.

After the designated time elapsed, the heating was stopped, and the furnace was allowed to cool to room temperature at 10 °C/min under HBr atmosphere. At 200 °C, HBr was replaced with N2. A white free flowing powder was obtained at T = 400 ° C, while a white chunk solid was obtained at 500 ° C . The crucible with the solid was taken out of the tube under nitrogen, trans ferred to a zip-lock plastic bag and weighed . The solid was grinded quickly to obtain a homogenous powder, a sample was taken into a vial and tightly closed and submitted to elemental analysis . All products were stored in a desiccator .

The yield of the reaction was calculated based on determination of Bn ( determined by titration) and veri fied by Li⁺ determination (by TCP ) , and CO3²- measurement (manometric analysis ) in the product sample . The number of moles of Bn was equal to the number of moles LiBr . The number of moles of LiBr was subtracted from the number of moles of Li⁺ measured in the product . The number of remaining moles of Li⁺ was checked to ascertain that it was twice the number of moles of CO3²- .

The PSD and amount of lithium carbonate starting material loaded into the tube furnace , the reaction conditions ( temperature and time of feeding HBr ( g) at 30 L/h flow rate ) , the amount of product collected, and results of the analysis are summari zed in Table 1 .

Table 1 .

Note : LiBr% calculated based on moles of Br~ as above The reactions of Examples 1 and 2 performed at 400 °C for 3h produced LiBr in the form of a powder at 72-76% yield. The molar ratios HBr : Li⁺ were 13 and 50 for Examples 1 and 2, respectively. The moles of HBr supplied to the reaction was calculated based on the feed flow rate (30 L/h, over three hours) , applying the ideal gas law.

Increasing the temperature to 500 °C and the reaction time to 6 h improved the yield to 93-97% (Examples 3-4) . The mole ratios HBr : Li⁺ were 25 and 27 for Examples 3 and 4, respectively. The product was different in appearance, as a hard white chunk was recovered from the reaction, rather than a powder, perhaps due to melting/ solidification which occurred in the tube furnace (as mentioned above, literature melting points of Li₂CO₃ and LiBr are 723 °C and 552 °C respectively; it is possible that the two components produce a mixture with a lower melting point close to 500 °C) . Apparently, the melt is more accessible to the action of the HBr stream flowing through the reaction zone, explaining the noticeably higher yield in Examples 3 and 4.

Examples 5 to 7 illustrate the effect of lithium carbonate with reduced particle size distribution. Namely, L12CO3 of Preparation 1 in a finely ground form with d<o.9) = 44.2 pm and d(o.5> = 10.6 pm was the starting material. The reaction was performed at 500 °C for 6h, and the HBr : Li⁺ mole ratio was 54 in Examples 5 and 6. The yield increased to 98.4% and 99%, respectively. But with this grade of L12CO3 used as a starting material, even shorter reaction times were sufficient to achieve surprisingly high yield (Example 7) , with HBr : Li⁺ mole ratio of about 18. Examples 8-14

Preparation of lithium bromide

The commercial L12CO3 powder showing the bimodal particle size distribution of Figure 1, with d<o.i) = 5.1 pm; d<o.5) = 23.2 pm and d(o.9> = 495.8 pm, was subjected to particle size reduction in a vortex mill. After the milling, the powder was separated by sieving into two populations of particles, one possessing the particle size distribution shown in Figure 3 (d<o.i) = 1.2 pm, d(o.5> = 5.5 pm and d<o.9) = 30.8) and the other possessing the particle size distribution shown in Figure 4 (d<o.i) = 103 pm, d(o.5> = 297 pm and d<o.9) = 615) . A series of experiments were performed in a tube furnace to evaluate the conversion of these two lithium carbonate particles populations (fine and coarse) to lithium bromide by the action of gaseous HBr under different conditions.

The chemical reaction was conducted according to the procedure described for Examples 1 to 7. The conditions of each reaction and the yield of lithium bromide recovered are tabulated in Table 2.

Table 2.

Note: LiBr% calculated based on moles of Br~ as explained above The results indicate that the major factors influencing the efficiency of the reaction are the reaction temperature and the particle size distribution of the lithium carbonate starting material. High percentage yield can be achieved across the entire temperature range of 300°C < T < 500°C, with the aid of suitably milled starting material. At the lowest end of the temperature range, lithium carbonate reduced to small particle size is needed, to reach satisfactory percentage yield (the finer the L12CO3 particles, the higher the reaction yield; see Example 8 versus Examples 11 and 12) . At the highest end of the temperature range, the reaction goes to completion irrespective of the particle size distribution of the lithium carbonate starting material (reduced particle size L12CO3 samples, with d(o.5)=5.5pm and d(o.5)=297 pm, converted completely into lithium bromide (~ 99-100% yield) after two hours (Examples 9A, 9B and 13) . Reaction time has a slight effect on the reaction yield (see Examples 11 and 12) .

The results are also shown graphically in Figure 5 as percentage yield % versus PSD (d<o.5) ) plots at constant temperature. The left-hand graph represents the reactions conducted at T=300°C, and the right-hand graph represents the reactions conducted at T=500°C (the lower and upper straight lines represent yields versus PSD for the reactions that lasted two hours and five hours, respectively) . It is seen that at T=300°C, the PSD strongly influences the reaction yield. But at T=500°C, excellent yield is achieved even with a coarse fraction of lithium carbonate, such that it is not necessary to use lithium carbonate reduced to PSD with d<o.5) < 10 pm. Example 15

Reaction between Li2COs particles and HBr(_g> : adjusting reaction temperature and PSD of Li2COs to maximize LiBr percentage yield

The data shown in Table 2 suggests that increase in reaction temperature could offset the use of relatively coarse lithium carbonate starting material, or vice versa, that fine lithium carbonate could balance the effect of low reaction temperature. That is, these two process variables could be adjusted to achieve industrially acceptable yields.

The data tabulated in Table 2 and the data of Examples 5-7 from Table 1 were fed into JMP software to fit a regression model and generate a prediction formula showing the relationship between percentage yield (the independent variable) and the reaction temperature and PSD of L12CO3 (T and Dso, respectively; these are the independent variables of the model) . The best model that was fitted by the JMP software had a term for each process variable and an interaction term (a product of the two independent variables) . The following prediction formula (with R² value of 0.997) was generated:

LiBr yield % = po + piT + P2D50 + P12 (T' x D50') where the coefficients were:

Po =71.4

Pi = 0.14 p₂ = -0.046

Pi2=0.00056

T'=T-425 d(50) ’= d(50) -151.28

The prediction formula can also be given by the following form:

Yield % 35.47 + (-0.046 x D₅₀) + 0.14 x T + ( ( D₅₀ - 151.28) x (T - 425) ) x 0.00056 Figure 6 is an 'actual by predicted plot' generated by the JMP software (the abscissa and ordinate represent the predicted values and observed values, respectively) . The slanted red line goes through the middle of the data points, indicating that the model is unbiased. Figure 7 is a contour profiler generated by the JMP software, showing how to obtain a yield between 95% and 100%. To this end, the temperature and D50 need to be adjusted within the unshaded region of the plot (for example, if it is decided to run the reaction at T=450°C, then L12CO3 starting material with particle size distribution showing Dso<9Opm is used) . To reach percentage yield of not less than 99%, the temperature and D50 are adjusted to lie on the dotted line (for example, if the lithium carbonate starting material available shows D50 in the range from 100 to 150 pm, then the reaction temperature should be not less than 490°C to reach the targeted yield) .

Claims

1 ) A process for preparing lithium bromide comprising contacting lithium carbonate with hydrogen bromide gas at a temperature of not less than 200 ° C and obtaining anhydrous lithium bromide in a solid form .

2 ) A process for preparing lithium bromide according to claim 1 , wherein the lithium carbonate and hydrogen bromide gas are contacted at a temperature of not less than 300 ° C .

3 ) A process according to claim 2 , comprising contacting lithium carbonate with hydrogen bromide gas at a temperature above 400 ° C but below a temperature associated with a change of texture of the reaction mixture and/or a change of phase of the reaction mixture and obtaining anhydrous lithium bromide in a particulate solid form .

4 ) A process according to claim 2 , comprising contacting lithium carbonate with hydrogen bromide gas at a temperature exceeding the temperature associated with a change of texture of the reaction mixture and/or a change of phase of the reaction mixture , recovering a reaction product consisting essentially of anhydrous lithium bromide and optionally comminuting the solid product to obtain anhydrous lithium bromide in a pulveri zed form .

5 ) A process according to any one of the preceding claims , wherein the lithium carbonate used possesses particle si ze distribution showing D90 < 100pm determined by laser di f fraction particle si ze analysis .

6 ) A process for preparing lithium bromide according to claim 1 , wherein the temperature of the reaction and the particle si ze distribution of the lithium carbonate are adj usted so as to obtain anhydrous lithium bromide at high yield . ) A process according to claim 6, wherein the temperature of the reaction is adjusted in the range of 300°C < T and the particle size distribution of the lithium carbonate is adjusted within the range of D50

300pm as determined by laser diffraction particle size analysis, so as to obtain anhydrous lithium bromide at a yield of not less than 85%.

8) A process according to claim 7, comprising contacting lithium carbonate possessing particle size distribution showing 5pm < D50

300pm as determined by laser diffraction particle size analysis, with hydrogen bromide gas, at a temperature in the range from 300°C to 500°C, so as to obtain anhydrous lithium bromide at a yield of not less than 90%.

9) A process according to any one of claims 6 to 8, wherein the reaction temperature and the particle size distribution of the lithium carbonate are adjusted in the ranges of T > 300 and D50 300pm, respectively, with the aid of a prediction formula showing the relationship between explanatory variables, which are reaction temperature and lithium carbonate particle size distribution, and a dependent variable, which is the yield of lithium bromide, so as to obtain anhydrous lithium bromide at a yield of not less than 90%.

10) A process according to claim 9, wherein the reaction temperature and the particle size distribution of the lithium carbonate are adjusted with the aid of a prediction formula of the form:

LiBr yield % = Po + PiT + p₂D₅₀ + P12 (T’ x D₅₀’ ) wherein the coefficients po, pi, P2, and P12 are obtained by fitting a regression model, T' is the temperature T corrected by a constant, D50' is D50 corrected by a constant; so as to obtain anhydrous lithium bromide at a yield of not less than 90% .

11) A process according to any one of claims 1 to 10, wherein an HBr stream is supplied to a reaction vessel in which lithium carbonate is heated, the HBr being supplied at a flow rate and over time such that the molar ratio of HBr relative to lithium carbonate is not less than 2.2:1.

12) A process according to claim 11, wherein the molar ratio of HBr relative to lithium carbonate is not less than 5:1.