Semi-batch evaporative crystallization and drying of cobalt sulphate hydrates

Semi-batch evaporative crystallization of CoSO 4 was investigated at various underpressures and with different heating powers. The underpressures for boiling aqueous CoSO 4 solutions at various temperatures and concentrations were measured and compared to predicted values obtained by the Aspen Plus Version 11 process simulation software. The presence of CoSO 4 solute decreased the underpressure for boiling solution in comparison to pure water. In the evaporative crystallization experiments, CoSO 4 crystallized as CoSO 4 .7H 2 O at evaporation temperatures of 30 ◦ C and 40 ◦ C, whereas CoSO 4 .6H 2 O crystallized at 60 ◦ C and 80 ◦ C. The crystals obtained at 30 ◦ C and 40 ◦ C were mainly agglomerates of octahedron shaped particles. Heating power and mixing speed had significant effects on the evaporation flux and particle size distribution of the products. Smaller crystals were produced with higher heating power and higher mixing speed due to the enhancement of evap-oration. Dehydration of the formed cobalt sulfate heptahydrate in the drying process was also investigated with two variables: drying temperature and drying time. The crystallization at 40 ◦ C yielded unstable CoSO 4 .7H 2 O, which tended to dehydrate in a temperature range between 23 ◦ C and 60 ◦ C; these crystals were used in the study of crystal drying. At a drying temperature of 23 ◦ C, CoSO 4 .7H 2 O transferred and stabilized as CoSO 4 .6H 2 O. At drying temperatures of 40 ◦ C and 60 ◦ C, the stable hydrate form was CoSO 4 .H 2 O. Gravimetric analysis, crystal habit, XRD, and Raman analyses were successfully utilized for the investigation of the transformation of the hydrate form during drying.


Introduction
Cobalt sulfate, which commonly exists as cobalt sulfate heptahydrate (CoSO 4 .7H 2 O), is one of the most important cobalt salts.It is widely used in electrochemical materials, ceramic applications, and even in the agricultural sector.Nowadays, the use in the rechargeable battery industry has become the most common use for cobalt sulfate, which is applied as a cathode active material.Adding cobalt into the cathode of rechargeable batteries can efficiently improve a battery's life, stability, inhibited corrosion, and performance (Li and Lu, 2020).The demand for cobalt sulfate is continuously rising due to the demand for 5G devices, consumer electronics, and passenger electric vehicles (EVs).Alloy manufacturing is another major application of cobalt sulfate.The addition of cobalt sulfate alloys makes the treated materials much stronger, with higher melting points, superior hot corrosion resistance, and oxidation resistance (Coutsouradis et al., 1987;Sato et al., 2006).
Cobalt sulfate is also employed as a pigment in the porcelain and glass industries and as a coloring agent in ceramics, enamels, and glazes to prevent discoloration (Yang et al., 2002).In the farming sector, cobalt sulfate has been used as an additive in soils and animal feeds to supply trace minerals (Suslick, 1998).According to an EU report (European Commission, 2020), cobalt has already been identified as a critical material in the EU.
Cobalt sulfate is a water-soluble divalent cobalt(II) salt.The generation and purification of cobalt sulfate solution have attracted a lot of interest, especially in the battery recovery industry (Devi et al., 1998;Huang et al., 2019;Petranikova et al., 2019;Reddy et al., 2009;Swain et al., 2007).Vielma (2021) modelled thermodynamics of aqueous CoSO 4 solutions and various hydrates in temperature range between 270 K and 374 K.However, the studies on cobalt sulfate crystallization from aqueous solution are limited.Aktas et al. (2013) investigated cobalt sulfate precipitation by anti-solvent crystallization with ethanol and Abbreviations: PSD, particle size distribution; SEM, Scanning Electron Microscopy; TG-DTA, Thermogravimetry and Differential Thermal Analysis; ΔT, Temperature difference between heating jacket and solution; XRD, X-ray Powder Diffraction.methanol.In their study, a high cobalt sulfate precipitation efficiency was obtained based on the short precipitation times required and good filterability of the obtained CoSO 4 .7H 2 O crystals.This process also had low power consumption as it did not need power for heating or cooling, whereas the consumption of ethanol or methanol was high.The recycling of organic solvents could be a challenge and may make the process less sustainable or energy efficient.In contrast, the crystallization of cobalt sulfate by evaporative crystallization and the evaporation concentrate-cooling crystallization process have been widely applied on industrial scale.In the evaporation process, the solvent is partially separated from solution by consuming a large amount of heat, so that the cobalt sulfate crystallizes out from the solution.However, the process has one major shortcoming: the energy consumption is high due to the high latent heat of evaporation.In the case of water, the latent heat of evaporation is 2260 kJ/kg at ambient pressure and 100 • C. In order to minimize the energy demand in the evaporation process, many technological improvements have been made in recent years, such as multieffect evaporation and mechanical vapor recompression crystallization (Lu et al., 2017;Said and Louhi-Kultanen, 2019).Although evaporative crystallization of cobalt sulfate is widely applied in industrial processes, there are quite few studies reported in literature about the factors affecting the crystal properties and their stability.In the present work, comprehensive studies on evaporative crystallization of cobalt sulfate were carried out with the aim to gain deeper understanding on the influence of operational conditions on crystal properties.The present work focuses on investigating the influences of various factors, including heat energy input, temperature, and mixing, on the evaporation kinetics and crystal properties of cobalt sulfate.In addition, the transformation of the cobalt sulfate hydrate form during drying was studied as well as the changes in drying temperature and drying time.

Materials
Semi-batch evaporative crystallization was studied with aqueous cobalt sulfate solutions.The synthetic cobalt sulfate solutions were prepared with analytical grade CoSO 4 .7H 2 O (purity ≥99%, ACROS ORGANICS) and deionized water.The initial concentration for synthetic cobalt sulfate solution was slightly lower than the solubility of CoSO 4 at different temperatures reported in the literature (Mullin, 2001), as shown in Fig. 1.

Experimental methods
The experiments were carried out in a 400 ml glass reactor equipped with baffles in an EasyMax 402 stirred tank system (Mettler Toledo).The inside diameter of the reactor was 72.6 mm and the upper surface area of the liquid in the reactor was A = 4.15 × 10 -3 m 2 .The stirring speed and heating energy were adjusted with iControl software.The solution was mixed by using a pitched-blade turbine with four blades (Φ = 38 mm).A pressure transmitter (WIKA, A-10, error < 0.14%) and an MPC 301Z vacuum pump (ultimate pressure < 8 mbar, pumping speed 38 l/min) equipped with a valve were used to control the pressure in the crystallizer.A Liebig condenser and distilling receiver (Anschutz-Thiele, straight, 50 ml) was used for vapor condensation and condensate collection.A diagram of the experimental set-up is shown in Fig. 2. In this study, we investigated the effects of crystallization temperature, mixing speed, and heating power on crystal properties.Different evaporation temperatures were controlled by adjusting the pressure to reach the boiling point.Different heating power values were achieved by changing the jacket temperature.The vacuum evaporation step was stopped 30 min after the solution had reached boiling point.After the vacuum evaporation step, the solution temperature was kept constant during an ageing period for another 30 min under atmospheric pressure.

Characterization
Samples obtained from the crystallization and drying processes were characterized by X-ray powder diffraction (XRD; X'Pert PRO) using a copper radiation source (λ Kα1 = 1.5406Å) operated at 40 kV, with 40 mA.The morphology of the products was examined with a scanning electron microscope, (SEM, TESCAN MIRA3).The particle size distribution (PSD) was analyzed with a particle size analyzer (Malvern Master Sizer 2000).Ethanol (99.5%, absolute, VWR) solution was used as the dispersant because of the low solubility of CoSO 4 in ethanol (Stephen and Stephen, 1963).The samples obtained in the various drying processes were also analyzed by using a Raman spectrometer (inVia™ confocal Raman microscope, Renishaw) to verify the hydrate forms.The Thermogravimetry and Differential Thermal Analysis (TG-DTA) measurements were carried with a TA Instruments (TGA Q500) in nitrogen atmosphere in the temperature range between 30 • C and 500 • C at the heating rate of 10 • C/min.
The cobalt sulfate yield percentage in the experiments was calculated using Eq. ( 1): where m p is the mass of the product (on the basis of anhydrous cobalt sulfate), [g]; m i. s the mass of the initial solution, [g]; and w i the mass fraction of cobalt sulfate in the initial solution, [g/g initial solution].
The evaporation flux was calculated with Eq. ( 2): where R is the evaporation flux of water from the cobalt sulfate solution, [g water m −2 s −1 ]; m e, s the mass of solution weighed after evaporative crystallization, [g]; t the evaporation time initiated when boiling started, [s]; and S the surface area of the upper liquid surface, which in the present work was 0.00415 m 2 .

Determination of underpressure for evaporative crystallization at selected boiling points
In order to conduct evaporative crystallization experiments at a certain boiling temperature, the operational underpressure had to be adjusted to a certain level with a vacuum pump and sufficient heating power had to be provided to achieve the boiling point.The adjusted underpressures and related boiling points were estimated by Aspen Plus V11 and measured experimentally.The boiling points and saturated vapor pressure of water were taken from the literature (Lide, 2005).The results are shown in Fig. 3.The predicted underpressure for boiling cobalt sulfate solution with various concentration obtained by Aspen Plus V11 were slightly higher than the measured value under the same temperature, but the trends are in good agreement with experimental results.In comparison to the results for water, the dissolved cobalt sulfate decreased the underpressure for boiling the solution.Moreover, the underpressure difference between water and cobalt sulfate solution increased with the increase of concentration.
Based on this result, the operational pressure for vacuum evaporative crystallization at various temperatures can be determined: at • C, 40 • C, 60 • C, and 80 • C, the saturated cobalt sulfate solution boils at 35 mbar, 64 mbar, 174 mbar, and at 423 mbar, respectively.

Effect of heating power
Beside the crystallization temperature, heating power is one of the principle operational factors affecting the evaporation flux, supersaturation, and product properties.In this study, the variation in heating power was achieved by changing the difference between the temperature of the solution (T r ) and the heating jacket (T j ).Here, we set the temperature difference (ΔT) to 10, 20, and 30 • C. The other conditions were as follows: evaporation temperature of 40 • C, pressure of 64 mbar (abs), initial solution of 46 g anhydrous CoSO 4 /100 g water, evaporation time of 30 min, and mixing speed of 650 rpm (tip speed of 1.29 m/s).
The evaporation fluxes and product yields obtained at different heating powers are illustrated in Fig. 4. The evaporation flux and product yields showed a significant increase with the increase in ΔT.The larger ΔT gave a higher heating power and thus led to higher evaporation fluxes.As expected, a higher yield was obtained at a higher ΔT.When the ΔT was 30 • C, over 55% cobalt sulfate was crystallized after evaporation for 30 min, although only about 35% water was evaporated.The reason is that the cobalt sulfate crystallized in hydrate form and therefore, water was transferred from the solution to hydrate crystals.The particle size distribution (PSD) analysis results for products obtained with different ΔT values are shown in Fig. 5, which depicts that the product obtained at a high ΔT is smaller than that obtained from a low ΔT.The mean particle size for a ΔT of 10, 20, and 30 • C was 509, 480, and 310 μm, respectively.Moreover, the product obtained at low heating power had a narrower size distribution compared to the products obtained with higher heating power.The higher evaporation flux provides a higher supersaturation degree which can increase both the crystal growth rate and nucleation rate.If nucleation dominates crystallization, the ratio of smaller crystals in the crystalline product becomes larger.
The heat energy input at different ΔT was calculated based on the heat transfer between the jacket and solution in the vacuum evaporation process, following Eq.( 3).The heat energy input was 36.3, 72.7, and 109 kJ with a heating power of 20.1, 40.3, and 60.5 W for a ΔT of 10, 20, and 30 • C, respectively.
where Q is the heat energy input, [J]; q the heating power,

Effect of evaporation temperature
In the evaporative crystallization process, it is evident that temperature is one of the key factors affecting the energy consumption and product properties.In the crystallization of cobalt sulfate, temperature affects the hydrate form of the products, which was investigated more comprehensively in the present work.In this study, four temperatures of 30, 40, 60, and 80 • C were selected.The operating pressure was varied by changing the evaporation temperature based on the determination of vapor pressure for different boiling temperatures; namely, 35 mbar for 30 • C, 64 mbar for 40 • C, 174 mbar for 60 • C, and 423 mbar for 80 • C. The other conditions were as follows: ΔT (heating power) of 20 • C, evaporation time of 30 min, and mixing speed of 650 rpm.The initial solution concentrations of 39.8, 46, 58, and 68 g anhydrous CoSO 4 /100 g water were used in the experiments carried out at temperatures of 30, 40, 60, and 80 • C, respectively.
As shown in Fig. 6, the evaporation fluxes decrease significantly with an increase in temperature.More water evaporated at lower temperature, 34.7 g at 30 • C and 18 g at 80 • C, whereas the product yields were 37.4% at 30 • C and 27% at 80 • C, respectively.It seems that the hydrate formation starts to affect here: evaporation of more water at 30 • C, along with higher degree of hydration leads to greater product yield.On the other hand, it is expected that more cobalt sulfate crystallizes from solutions of higher initial concentration at higher temperature, corresponding to the same amount of water evaporated.The XRD patterns for products obtained at different temperatures are shown in Fig. 7.It can be seen in Fig. 7 that cobalt sulfate crystallized as heptahydrate at 30 • C and 40 • C and hexahydrate at 60 • C and 80 • C. Based on the thermodynamics studies of Vielma (2021), cobalt sulfate should crystallize as cobalt sulfate monohydrate at 80 • C. The reason we obtained hexahydrate at 80 • C is probably because of relatively high overall crystallization rate.The SEM images of solid crystals produced at various temperatures are presented in Fig. 8.The crystals formed at 30 • C and 40 • C had a similar morphology of octahedron crystal shape, and agglomerates were formed during the process.By contrast, the product   crystallized at 60 • C and 80 • C did not have a uniform shape and the agglomeration tendency was higher than that at the lower temperature.The mother liquor with a higher concentration at higher temperature could promote the agglomeration due to higher degree of supersaturation.The PSD of products crystallized at different temperatures is illustrated in Fig. 9.It shows that the crystals obtained at 60 • C and 80 • C had a broader size distribution compared to the crystals obtained at 30 • C and 40 • C, which agrees well with the SEM analyses.The average particle size of products obtained at different temperatures were very similar, with 470, 480, 466, and 481 μm for 30, 40, 60, and 80 • C, respectively.

Effect of mixing
Mixing has a significant effect on the crystallization process.For the evaporative crystallization process, mixing will affect the heat transfer in solution, evaporation flux, and further product properties.In this work, three different mixing speeds for cobalt sulfate crystallization were investigated, namely 400, 550, and 650 rpm with corresponding tip speeds of 0.79 m/s, 1.09 m/s, and 1.29 m/s, respectively.The other crystallization conditions were as follows: evaporation temperature of 40 • C, ΔT of 20 • C, initial solution of 46 g anhydrous CoSO 4 /100 g water, evaporation time of 30 min, and pressure of 64 mbar.
The evaporation flux and product yields at different mixing speeds are presented in Fig. 10.The evaporation flux and product yields increased gradually with a rise in mixing speed from 400 to 650 rpm, which means that a higher mixing rate enhances the heat transfer in the solution, thus promoting evaporation and solid crystallization.The overall heat transfer coefficient at different mixing speeds was measured by calorimetry.The results were 205.5, 227.1, and 243.3 W/(m 2 • C), at mixing speeds of 400, 550, and 650 rpm, respectively.The results showed that heat transfer was enhanced by increasing mixing intensity when the ΔT was the same.The particle size was smaller when the mixing speed was higher, as can be seen in Fig. 11.This is because a higher mixing speed promotes evaporation, which increases the degree of supersaturation.In turn, higher supersaturation enhances nucleation, resulting in the formation of small particles.Moreover, the higher mixing speed can enhance the diffusion of the solute in the solution and the supersaturation and therefore, the driving force, supersaturation degree, usually becomes more uniform in the crystallizer.The crystals obtained at higher mixing speed are more uniform based on PSD results, as shown in Fig. 11.

Dehydration studies
The crystalline products obtained proved to be unstable, which could be seen from the dehydration occurring during the drying of the crystals.The heptahydrate products were able to transform into hexahydrate or even monohydrate at a certain temperature and drying time (Rakuzin and Brodski, 1927).To investigate the dehydration of the cobalt sulfate heptahydrate obtained during the drying process, we carried out a comprehensive study with the samples obtained with an evaporation temperature of 40 • C, ΔT of 20 • C, mixing speed of 650 rpm, at different drying temperatures and times.The drying experiments were carried out with a ventilated oven at temperatures of 23 • C, 40 • C, and 60 • C. The samples were collected after the drying times of 8, 24, 48, and 168 h.
The images of samples in different drying conditions are displayed in Fig. 12.The color of the samples varied clearly over time.At a drying temperature of 23 • C, the color of the sample changed from red to pink after 168 h of drying.The pink particles already appeared after 24 h drying at 23 • C. The weight loss curves in Fig. 13(a) and XRD patterns in Fig. 13(b) show that the weight losses showed a significant increase during the drying time from 8 h to 24 h; however, after 48 h the weight remained stable.The XRD patterns for samples obtained at different drying times also showed the transformation of hydrate with the appearance of diffraction peaks for CoSO 4 .6H 2 O occurring after 24 h of drying.However, the diffraction peaks for CoSO 4 .7H 2 O disappeared when the samples were dried at 23 • C over 48 h, showing that the transformation from heptahydrate to hexahydrate started to occur after crystals had dried for 24 h.Dehydration was completed when the crystals had dried for 48 h.It is pointed out here that CoSO 4 .6H 2 O crystals did not dehydrate during 168 h at 23 • C.
As the drying temperature increased, the final hydrate form and dehydration kinetics changed.In Fig. 13(a) and (c), it can be seen that the original sample transformed from heptahydrate into hexahydrate very rapidly and almost ended after 8 h of drying when the samples were dried at 40 • C. The weight loss changed slightly between 8 h and 24 h.Cobalt sulfate hexahydrate was still the main phase in the sample obtained after 24 h drying at 40 • C, which was verified by XRD analyses.After 48 h of drying, the weight loss had increased significantly during this period.This is because the sample dehydrated from hexahydrate to monohydrate.The sample dried after 48 h had some reddish particles, as  .The total weight loss was 41.4%, which is lower than the theoretical weigh loss of 44.8% for CoSO 4 .7H 2 O.This is probably caused by the dehydration of CoSO 4 .7H 2 O at a temperature lower than 30 • C, where the data was not collected.Compared to the results reported by Sinha et al. (1989) and Nandi and Kher (1980), the crystallized CoSO 4 .7H 2 O in this study seems to be less stable than the samples obtained by slow evaporative crystallization in their study.The dehydration temperatures differed as well.According to Sinha et Chio et al., 2007).In the crystallographic arrangement of cobalt sulfate hydrates (Elerman, 1988;Giester and Wildner, 1991;Redhammer et al., 2007) 2− affect the peak position of the υ 1 (SO 4 ) mode.The selected Raman spectra for the υ 1 (SO 4 ) model shown in Fig. 14(a), (b), and (c) reveal that the wavenumbers of the bands (υ 1 (SO 4 )) increased during the drying process.As the XRD analyses indicated, the changes in the υ 1 (SO 4 ) mode could be a consequence of the dehydration process, where 984.3, 990.3, and 1022.4 cm −1 correspond to heptahydrate, hexahydrate, and monohydrate, respectively.The increasing vibrational wavenumber of the υ 1 (SO 4 ) mode in these lower hydrates agrees with the Raman studies on various hydrates of CuSO 4 by Fu et al. (2012), CaSO 4 by Buzgar et al. (2009), and FeSO 4 by Chio et al. (2007), as shown in Table S.1.

Conclusions
The semi-batch vacuum evaporative crystallization of cobalt sulfate was successfully carried out in various crystallization conditions.Cobalt sulfate crystallized as monoclinic CoSO 4 .7H 2 O, with octahedron shaped crystals at temperatures of 30 • C and 40 • C, whereas a lower hydration degree monoclinic crystal CoSO 4 .6H 2 O was obtained at higher temperatures of 60 • C and 80 • C. In addition, the evaporation temperature, heat energy input, and mixing speed had a significant effect on the evaporation flux, therefore affecting the particle size distribution of the products.The higher heat energy input and mixing intensity resulted in smaller crystals.
The Moreover, a higher drying temperature accelerated the dehydration of CoSO 4 .7H 2 O. Based on the TG-DTA results of the crystals obtained by slow evaporative crystallization, the CoSO 4 .7H 2 O crystals showed to be less stable at lower dehydration temperature.Raman analysis was used to identify the hydrate forms of cobalt sulfate by comparing the vibrational wavenumber of the υ 1 (SO 4 ) bands.The wavenumber of υ 1 (SO 4 ) bands showed an increasing trend with a decrease in the degree of hydration.The combination of weight loss measurements, XRD analysis, Raman analysis, and sample images proved to be an efficient way to study the hydrate forms and their dehydration.
The findings of the present research work provide a deeper understanding on the evaporative crystallization process of cobalt sulfate and the factors affecting the crystallized product.Evaporation at a certain underpressure and temperature range with appropriate heat energy input is a feasible way to crystallize cobalt sulfate into the desired hydrate form.Furthermore, the laboratory-scale results obtained offer a basis for achieving suitable process-scale performance.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Fig. 1 .
Fig. 1.Solubility data for cobalt sulfate and initial solution concentration at different temperatures.

Fig. 3 .
Fig. 3. Underpressure at boiling point of saturated cobalt sulfate solutions and water.

Fig. 4 .
Fig. 4. Evaporation fluxes, heating power and product yields at different ΔT at 40 • C, mixing speed of 650 rpm, and evaporation time of 30 min.

Fig. 5 .
Fig. 5. PSD analyses of cobalt sulfate obtained at different ΔT values at 40 • C, mixing speed of 650 rpm, and evaporation time of 30 min.

Fig. 6 .
Fig. 6.Evaporation fluxes and product yields at different evaporation temperatures, at a ΔT of 20 • C, evaporation time of 30 min, and mixing speed of 650 rpm.
[W]; U the overall transfer coefficient (in the present work U = 243.3W/(m 2 • C), which was measured by calorimetry built in the EasyMax system with a mixing speed of 650 rpm and solution temperature of 40 • C); A is the heat transfer surface area, in the present work A = 8.3 × 10 -3 m 2 ; ΔT the temperature difference between jacket and solution, [ • C]; and t the time needed for evaporation, t = 1800 s.

Fig. 9 .
Fig. 9. Particle size distribution of cobalt sulfate obtained at different evaporation temperatures, a ΔT of 20 • C, evaporation time of 30 min, and mixing speed of 650 rpm.

Fig. 10 .
Fig. 10.Evaporation flux and product yields at different mixing speeds at a temperature of 40 • C, ΔT of 20 • C, and evaporation time of 30 min.

Fig. 11 .
Fig. 11.Particle size distribution of cobalt sulfate obtained with various mixing speeds at a temperature of 40 • C, ΔT of 20 • C, and evaporation time of 30 min.

Fig. 12 .
Fig. 12. Sample images with different drying times at various drying temperatures.
al., CoSO 4 .7H 2 O dehydrated firstly to CoSO 4 .6H 2 O at around 120 • C. At higher temperatures of 162 • C and 185 • C, CoSO 4 .6H 2 O transferred firstly to CoSO 4 .3H 2 O and then CoSO 4 .H 2 O.The dehydration of CoSO 4 .H 2 O occurred approximately at 295 • C. The Raman analysis results for samples obtained from different drying process times are shown in Figs.S.2-S.4 in supplementary materials.The vibrational wavenumber of the υ 1 (SO 4 ) mode (symmetric crystallized CoSO 4 .7H 2 O was unstable at higher temperatures.It dehydrated easily and transformed into CoSO 4 .6H 2 O and CoSO 4 .H 2 O during drying.In ambient conditions (23 • C), CoSO 4 .7H 2 O dehydrated and stabilized as CoSO 4 .6H 2 O.When the drying temperature was increased to 40 • C and 60 • C, dehydration of CoSO 4 .7H 2 O was nearly completed to produce the monohydrate CoSO 4 .H 2 O in stable form.