Nickel Recovery Optimization and Kinetic Study of Morowali Laterite Ore

Nickel plays a critical role in the mining industry. However, the presence of nickel in the primary source of sulfide minerals is decreasing. Focus has since turned to laterite ore, which contains up to 80% Ni metal. The purpose of this study was to optimize nickel leaching using sulfuric acid and conduct a kinetic analysis to discover the mechanism that best controls the leaching process. To optimize the operating conditions, the response surface method (RSM) with a Box–Behnken design was used. The shrinking core and Zhuravlev, Leshokin, and Templeman (ZLT) models were used to assess the kinetics of the nickel leaching process. Mineral characterization was also performed to gain a better understanding of the sample characteristics. At 2 M sulfuric acid concentration, 10% solid–liquid ratio, and 90 °C temperature, the highest observed Ni recovery was 87% and the apparent activation energy was 32.75 kJ/mol.


Introduction
Nickel is one of the most important materials used worldwide and has seen increasing annual consumption. It can be found naturally in sulfide and laterite ores. In the production of stainless steel, up to 66% of total nickel is used as an alloy [1,2]. The rapid surge in the production of lithium batteries has increased the number of nickel refining operations for battery cathodes. To date, commercial nickel refining has relied on sulfide ores, supplying at least 60% of nickel demand, although laterite ores compose up to 80% nickel [3,4]. The shortage of nickel sources from sulfide ores has necessitated the development of an optimum method for extracting nickel from other sources, such as laterite ores with lower nickel content. Nickel is one of the most important materials used worldwide and has seen increasing annual consumption and production since 2009. This has prompted further process improvements to obtain nickel more effectively [5].
Nickel extraction and refining techniques from laterite ore have been investigated as an alternative to nickel sulfide ores [1,[6][7][8]. Laterite ores are classified into two types based on their iron and magnesium content: limonite and saprolite. Recently, saprolite ore, which contains more Ni, has been treated using a pyrometallurgical technique that requires more capital investment than that required for hydrometallurgy [9]. Many leaching methods have been developed to discover the most efficient and cost-effective approach for processing nickel laterite, particularly acid-based atmospheric leaching.
Considerable research has been conducted on atmospheric dissolution of laterite ores. Santos et.al. [10] studied the influence of the leaching temperature, reagent concentration, and reducing agents on laterite nickel recovery. Thubakgale et al. [3] investigated the features of atmospheric leaching of Ni Mafic overburden using sulfuric acid with the addition of NaCl as a refractory mineral decomposition agent. Using citric acid, Astuti et al. [11] studied the kinetics of Ni leaching from laterite ores under atmospheric pressure. Tzeferis [12] investigated the dissolution of low-grade laterite using microbially produced citric acid, whereas Panda et al. [13] studied the characteristics and leaching process of Ni from ferruginous laterite using sulfuric acid. Several studies have been conducted on the optimization of Ni leaching from laterite ores. Nasab et al. [14] investigated the optimization of Ni dissolution using pre-roasted laterite ores by varying sulfuric acid concentration, solid-liquid (S/L) ratio, stirring speed, and time. Mohammadreza et al. [15] investigated the optimization of nickel extraction from low-grade laterite using a two-level full-factorial experimental design. Ochromowicz and Leśniewicz [16] investigated the optimization of laterite leaching using a sequential design of experiments. Johnson et al. [17] investigated the optimization of Ni extraction from laterite using high-pressure acid leaching.
In this study, optimization of the atmospheric direct leaching process was conducted using a Box-Behnken experimental design with three independent variables (concentration of sulfuric acid, S/L ratio, and leaching duration) and kinetic studies. However, in order to scale production from a lab-scale to a pilot plant, optimization is crucial. Response surface methodology (RSM) was used in this study.

Material Characterization
The laterite ore was collected from a mining site in Morowali, Indonesia. The ore was crushed and milled before being sieved to 180 µm. The ore was processed in a laboratory-scale ball mill for 30 min at an ore/steel ball mass ratio of 20% and rotating speed of 8 rpm. X-ray diffraction (XRD) spectrometry (SmartLab) with a 2-θ range of 3-90° was used to characterize the samples. X-ray fluorescence (XRF) (Rigaku Primini Benchtop) spectroscopy was used to identify the samples. Equation (1) calculates the metal recovery from the leaching process: where m 0 and m i are the initial feed mass and residual mass (g), respectively, and C 0 and C i are the Ni grades in the feed and residue (%), respectively.

Design of Experiment and Leaching Process
A Box-Behnken experimental design was used in this study. Fifteen experimental points were used to assess the impact of three experimental variables: acid concentration, S/L ratio, and leaching period. Table 1 summarizes the experimental design of the experiment (DOE). To prevent evaporation, leaching was conducted in a three-neck flask with a reflux condenser. The stirring process was performed using a stirrer, and the heating process was performed using an external heater (hotplate). The optimization of the leaching process was conducted with the following fixed parameters: stirring speed of 200 rpm, particle size < 180 µm, solution volume of 100 mL, and temperature of 60 °C. The leaching product was recovered via vacuum filtration, and the leaching residue was washed in distilled water at 80 °C for 30 min.

Kinetic Analysis
Kinetic studies were conducted to identify the rate-controlling process and the activation energy (E A ) of the leaching process using the shrinking core model (SCM). The film diffusion model was not used in this study because the effect of diffusion on the particle film layer was negligible considering the rapid agitation and large particle size [18]. The SCM equation is presented in Eqs. (2)(3)(4) [19] where Eqs. (2, 3, and 4) are the chemical reaction control, diffusion control, and the mixed model presented by Zhuravlev, Lesokhin, and Templeman (ZLT), respectively [20]. where k r , k d , and k z represent the reaction rate constants for the chemical-controlled processes, diffusion, and the ZLT model, respectively. Meanwhile, t represents the leaching duration (min) and x is the extracted nickel fraction. The activation energy was determined using the Arrhenius equation, as shown in Eqs. (5)(6). The value of the reaction rate constant was obtained from a kinetic model that best characterized the nickel leaching process from the laterite ore.
where k is the reaction rate constant, k 0 is the frequency factor, E A is the activation energy (J/mol), R is the gas constant (J/(mol K)), and T is the leaching temperature (K).

Mineralogy
The elemental concentrations of total Ni, Fe, and Mg in the samples were 1.9%, 13.82%, and 5.13%, respectively, as shown in Table 2. These three components can be used to detect ore types in laterites earlier. The sample possesses a high Fe content and low Mg and Ni content, indicating that it is a type of limonite ore, as has been widely discussed regarding various types of laterite ore [13,21,22]. This is further confirmed by the XRD pattern ( Fig. 1 4 ). Particle size distribution analysis was conducted to determine the particle size properties of the ore. As shown in Fig. 2, approximately 50% of the total mass of the sample was less than 90 µm. This fine particle size indicates a limonite ore type, as previously observed [23][24][25]. The limonite ore is finer than the saprolite ore, which typically has 50% of the total mass of ore greater than 100 µm. Particle size also affects the leaching process [23]. Therefore, a particle size of 90 µm was used for the leaching optimization process in this study.

Nickel Leaching and Optimization Using RSM
The response surface method (RSM) was used to investigate leaching parameters such as sulfuric acid concentration, S/L ratio, and leaching process duration. RSM explanations have been widely published in numerous studies [26][27][28][29][30].
The RSM model that relates the independent variables and Ni recovery is given by Eq. (7). Second-order polynomial equations were used in this study. This model correlates the relationship of three variables to Ni recovery, with a correlation factor (R 2 ) value of 95.76%. Table 3 compares the nickel recovery in the experiment with the predictions using Eq. (7).
where A is the H 2 SO 4 concentration (M), B is the S/L ratio (%), and C is the leaching duration (min). Nickel in limonite ore is associated with the mineral goethite in nickel oxide (NiO) [1]. Nickel remained in the aqueous phase as nickel sulfate in the leaching filtrate, as shown in Eq. (8).
Nickel recovery is greatly influenced by the H 2 SO 4 concentration. Figure 3a shows that the maximum Ni recovery was 58% at 2 M H 2 SO 4 and a S/L ratio of 5%. Nickel recovery increased as the H 2 SO 4 concentration increased, and the S/L ratio decreased. According to Panda et al. [13], low leaching recovery occurs when the slurry has a high S/L ratio because it becomes viscous and inhibits ion mobility. This can also be caused by an insufficient concentration of H + ions leaching the available solid. Figure 3a and c shows that the increase in Ni recovery was practically constant in the 5-10% S/L ratio range. Considering that more amount of ore may be dissolved for the same solvent volume, a 10% S/L ratio was optimal for this study. Thubakgale et al. [31] also reported that the 10% S/L ratio was optimal for nickel extraction from limonite ore using sulfuric acid.
With increasing sulfuric acid concentration and leaching duration, the nickel recovery increased significantly (Fig. 3b). The most significant increase in recovery was obtained at a leaching duration of 240 min and between 0.5 and 2 M sulfuric acid. The difference in the increase was greater for each concentration range than for shorter leaching durations. In a preliminary study, leaching was carried out at acid concentrations of 4 and 6 M, but there was no increase in Ni recovery. A higher concentration of H 2 SO 4 leads to higher Fe recovery, resulting in lower selectivity of Ni/Fe, which is uneconomical [13]. The optimal concentration in this investigation was determined to be 2 M H 2 SO 4 concentration, indicating that extending the leaching duration significantly impacts Ni recovery from goethite minerals in limonite-type ores. Nickel recovery varied from 41 to 63% in the range of 0.5 to 2 M sulfuric acid during 240 min of leaching.   With decreasing S/L ratio and increasing leaching duration, the Ni recovery increased (Fig. 3c). By reducing the S/L ratio from 25 to 5% and extending the leaching duration from 60 to 240 min, Ni recovery increased from 33 to 59%. This demonstrates that the influence of the S/L ratio and leaching duration on the nickel leaching process is insignificant compared with that of the acid concentration.

Kinetic Study
The reaction rate in solid-liquid system reactions is controlled by one of many mechanisms, including diffusion through the ash, product, or film layer, or via reaction on the surface of the core of reacting particles [19,32,33]. However, diffusion through the film layer for the solid-liquid system reaction did not significantly affect the leaching process [34]. Therefore, the kinetic model provided by Zhuravlev was adopted in this study. The ZLT model is derived from the diffusion of reactants into the solid particles [20]. Figure 4 shows the experimental results for the effect of temperature on Ni recovery. At 30 °C, Ni recovery increased significantly when leached for 0-60 min, reaching 31% Ni recovery. Nickel recovery can only reach 41% during extended leaching durations. Nickel recovery increased to 83% at 60 °C for 180 min and then to 85% after 240 min of leaching. At a temperature of 90 °C, the Ni recovery peaked in the shortest time compared to lower temperatures. Figure 5a shows a poor fit of the reaction control model, indicating that the reaction on the unreacted core surface did not affect the leaching process. Meanwhile, two diffusion models, diffusion through the ash layer (Fig. 5b) and the ZLT model (Fig. 5c), provided a fit to the data. The R 2 in  Table 4 indicates that the two diffusion models show a better fit than the reaction control model. Based on its high R 2 value, the ZLT model was selected as the best kinetic model to represent the Ni leaching process using sulfuric acid.
The apparent activation energy was derived from the slope of 1/T versus ln k z (Fig. 6). The R 2 fit the data. The obtained activation energy and k 0 were 32.75 kJ/mol, 459.8 × 10 3 , respectively. This value implies that product diffusion through the ash layer controls the leaching process.
Prior investigations have claimed that nickel leaching of limonite ore with sulfuric acid is controlled by product diffusion through the ash layer [11,34]. Agacayak and Zedef [8] reported an E A value of 68.66 kJ/mol for dissolving Ni from Turkey's laterite ore using sulfuric acid. Luo et al. [24] also reported an E A value of 53.9 kJ/mol for saprolite-type laterite ore leaching using sulfuric acid under atmospheric conditions. Javanshir et al. [35] reported an E A value of 46.9 kJ/mol for leaching Ni from low-grade nickelbearing ores using sulfuric acid. This study obtained a smaller E A value because the kinetic analysis was carried out under optimal conditions. Diffusion usually controls the leaching process, which has an activation energy of 1-40 kJ/mol [36]. The activation energy for the Ni leaching process using sulfuric acid varies between 20 and 60 kJ/mol for various types of ores [37,38]. The activation energies of the chemical reaction control and diffusion through ash layer models were   19.04 and 31.70 kJ/mol, respectively, indicating that the diffusion process did control the leaching process [38]. The ZLT model describes the rate kinetics of this process.

Conclusion
The response surface method was applied in this study to assess the impact of operating parameters such as sulfuric acid concentration, S/L ratio, and leaching duration. The ideal operating parameters were 2 M sulfuric acid concentration and a 10% S/L ratio. The effects of the temperature and kinetics were investigated in this study. The maximum nickel recovery was 87% at 90 °C, 2 M sulfuric acid concentration, and a 10% S-L ratio for 180 min. Product diffusion through the ash layer controls the leaching process, and the ZLT model is ideal for describing this process. The kinetic model