1. Introduction
Phosphate rocks are essential in various industries, such as in the production of phosphoric acid and fertilizers. The total phosphate reserves in the world are up to 70 billion tons. The biggest phosphate resources are in Morocco with a reserve of 50 billion tons, accounting for 71.43% of the total amount [
1]. Approximately 75% of phosphate resources have a sedimentary origin [
2]. The marketable sedimentary phosphate ore usually has 28%–30% of P
2O
5 to be economical [
3]. The requirements for phosphate concentrate used in the production of phosphoric acid are P
2O
5 content around 30%, CaO/P
2O
5 ratio <1.6, MgO content <1%, and Fe
2O
3 and Al
2O
3 content ≤2.5% [
4]. However, with an increasing demand for phosphate and a reduction in phosphate reserves, it is necessary to enrich phosphate tailings (e.g., phosphate washing sludges) to concentrate phosphate to be used in the production of phosphoric acid.
Phosphate washing sludges (Ps) constitute a considerable amount of tailing, which are not exploitable by simple washing because they have a low content of P
2O
5 and an abundant amount of carbonates and silica. Moreover, these tailings are becoming a critical issue for the mining industry because of environmental concerns. They need to be processed by separation techniques to remove the problematic portions (e.g., carbonates) and recover the valuable parts (e.g., phosphates). Good beneficiation of Ps can be achieved by various processes, depending on the liberation size of phosphate and gangue minerals. Different processes like screening, scrubbing, milling, washing, calcination, leaching, and flotation may be used. Milling is an important stage in the beneficiation of low-grade phosphates. Indeed, it helps in the liberation of the phosphate grains from the gangue [
5]. In our case, phosphate is recovered from the carbonated gangue.
The calcareous Ps can be also enriched by organic acid leaching. In this process, the carbonates are dissolves by the dilute acid solution, and the phosphates remain in the leached residues. Thus, the beneficiation of the low-grade phosphate is achieved. Strong acids can attack the phosphate minerals when used to selectively remove calcareous material in low-grade Ps. However, according to the literature, most of the organic weak acids show an interesting degree of selective leaching [
6]. The liquid phase can be easily separable after leaching for regeneration and recycling of the acid. The commonly used organic acids are acetic acid [
7], lactic acid [
8], and citric acid [
9]. According to Fei et al. [
10], after leaching optimization of low-grade phosphate ore with a P
2O
5 content of 17.45% using Lactic and Acetic acid, the following optimum conditions were maintained: 11% of Lactic Acid, at a temperature of 55 °C for a 120 min leaching period, and 5% of Acetic acid, at a temperature of 40 °C for a 30 min leaching period. A concentrate of P
2O
5 content above 30% was obtained. In another study, by using 15% of Acid Acetic for leaching a phosphate ore of 10% P
2O
5, Gharabaghi et al. [
7], obtained a concentrate of 32.14% P
2O
5 at a temperature of 40 °C, a reaction time of 60 min, and a solid concentration of 15%. Zafar et al. [
8] have also reported in their study that with 8% of lactic acid, a temperature of 45 °C, a liquid/solid ratio of 7, and a reaction time of 45 min, it was possible to increase the content of P
2O
5 from 22.3% to 35%. Moreover, organic acids are effective and economical for the enrichment of low-grade calcareous phosphates. Organic acid leaching is a very selective process, ecological, and can produce high purity products [
10].
The main purpose of this research is to upgrade a Moroccan low-grade Ps through organic acid leaching by using three organic acids acetic acid, lactic acid, and citric acid as a leaching agent for carbonates. A single-factor experiment was carried out to investigate the influence of various operational factors such as reaction temperature, reaction time, organic acid concentration, solid concentration, and acid type on the concentration and recovery of P2O5 and to optimize the process parameters for maximum beneficiation. Then Taguchi L9 orthogonal array, Signal-to-noise ratio, Desirability function, and Analysis of Covariance ANCOVA were used to determine optimal conditions and the main influence factors on the enrichment of low-grade Ps by leaching. Dissolution kinetics was studied using two reaction models. The dissolution kinetics model that best fit the experimental data was determined, and the activation energy was calculates using Arrhenius’ law.
2. Materials and Methods
In our experiments, we used Ps coming from Khouribga’s washing plant. Ps samples are collected from ponds by chip sampling technique according to the international standard ISO 18400. Then the samples were dried in an electric oven at about 105 °C, cooled to room temperature, homogenized, and quartered through a riffle sampler to obtain a representative sample.
2.1. Characterization Techniques
All the analyses were conducted following quality assurance and analytical quality control program NM ISO/IEC 17025. For the Mineralogical Structure determination, we used X-ray diffraction of Rigaku (XRD) monochromatized CuKα radiation (λ = 1.54 Å) using the international standard (EN 13925-2:2003. Microscopic observation was made on different size fractions of a milled sample to determine the percentage of liberated phosphate mineral in each size range. Sieving was carried out before mineral characterization. Wet sieving is carried out on a set of 5 sieves (NM EN 12948): 160, 100, 71, 53, and 20 μm. The elemental composition of Ps was determined by the X-ray fluorescence spectrometry (XRF) technique using a Philips spectrometer (Philips Industrial Electronics, Almelo, Netherlands, ISO 18227:2014). International standards were used to determine the P2O5 content by a colorimetric method (Spectrophotometer, SECOMAM, Alès, France) using ammonium molybdate and ammonium metavanadate and applying a filter of 430 nm. Atomic Absorption Spectrophotometer (AAS) A-1800 Hitachi (Hitachi, Tokyo, Japan) was used to determining the Ca2+ content in the leach solution.
2.2. Ball Milling
A milling process was carried out on the laboratory scale in a planetary Mono Mill PULVERISETTE 6 classic line (FRITSCH, Idar-Oberstein, Germany) consisting of a metal cylindrical milling bowl of 225 mL capacity with media and rotated on its axis, and the milling load is made up of balls of same size 0.1 mm. The milling bowl and milling balls are made of agate.
2.3. Acid Leaching
The organic leaching experiments were carried out in a 500 mL glass reactor on a hot plate with a magnetic stirrer. A known amount of Ps (30–120 g) was leached with 300 mL of an acid solution with different concentrations (1–15% in mass) at different temperatures. A digital thermometer was used to control the temperature. Moreover, during leaching experiments, all temperatures were ± 3 °C taking into account thermometer error. The leaching mechanism experimental setup is shown in
Figure 1. During the experiments, the stirring speed was fixed at 300 rpm to ensure no sediment at the bottom of the reactor. The stirring intensity effects on leaching were not studied because they are not so significant [
7]. Finally, after the sample reacted completely, the solid phase was separated from the reaction solution by filtration and dried at 105 °C before analysis. In the filtrate, the P
2O
5 content was determined by the colorimetric method and the Ca
2+ content in the leach solution was determined by AAS.
In the analysis, after grinding and leaching, for each experiment, the recovery of P
2O
5 (%) was calculated as follows:
where
C is the weight of the concentrate,
c is P
2O
5 concentration (%) in the concentrate,
P is the weight of original Ps, and
p is P
2O
5 concentration (%) in original Ps.
The dissolution fraction of calcareous material was calculated by Equation (2):
2.4. Taguchi Experimental Design
The Taguchi Parameter method is an experimental design technique that allows reducing the number of experiments by using orthogonal arrays. It reduces time and cost. This had an effect that substitutes the full factorial design of an experiment. Considering four experimental parameters and three levels, the total numbers of combinations of possible experiments are 3
4. It is difficult to determine the best combination among the 3
4 combinations due to the lengthy experimentation required. Taguchi’s method affirms that we need to try out only 9 out of 3
4 experiments, which will help us to achieve the best solution where every factor is varied, one at a time, while all of the other factors remain constant [
11].
In this study Taguchi L
9 (3
4) experimental design was generated using JMP software (Version 11, SAS Institute, Inc., Cary, NC, USA). Each experiment was repeated only three times due to a lack of raw material. The signal-to-noise (S/N) ratio corresponding to each combination is calculated. The (S/N) ratio is a measure of robustness that identifies the control factor settings for the minimum impact of noise on response [
12]. Various types of (S/N) ratios have been developed, e.g., smaller-is-better, larger-is-better, and nominal-is-best. The greater S/N ratio corresponds to better quality characteristics [
11]. As the aim of the present study is to enrich sludges with P
2O
5 and to obtain the highest possible % P
2O
5, the larger-is-better ratio was selected and calculated using JMP software according to the following Equation (3):
where
yi is the experimental response and
n is the experiment replication number.
2.5. Statistical Analysis
In this section, the statistical approach to determine factors influencing the two responses, namely % P
2O
5 mean values and S/N ratio, is discussed. By evaluating statistical analysis tools such as the Pareto plot and Desirability function for each response, the most influential factors and optimum conditions are determined. The Pareto plot is a plot of scaled estimates. The most important factor has the longest horizontal bar [
13]. The desirability function analysis is used for the determination of optimum conditions of different variables that determine optimum performance for different responses [
14].
Analysis of Covariance (ANCOVA) is performed to determine the most effective parameters using the commercial statistics software package IBM SPSS version 25.0 (IBM, New York, U.S.A) for Windows.
4. Conclusions
Phosphate washing sludges from the Khouribga washing plant have, so far, been rejected as non-profitable material. These low-grade sludges (15.84% P2O5), non-recoverable by the washing process, are enriched with organic acid leaching for upgrading. Ball-milling under optimum conditions (rotation speed of 100 rpm, charge ratio of 4/1, a solid concentration of 40%, and a grinding time of 4 min) increases the content of P2O5 up to 18.51%. After studying the influence of the operating parameters (acid type, acid concentration, solid concentration, Temperature, and reaction time) on organic acid leaching of Ps through single-factor experiments and Taguchi orthogonal experiments, we conclude that optimum parameters obtained from single-factor experiments are leaching for 60 min using acetic acid with a concentration of 7%, solid concentration of 25%, and a temperature of 40 °C. Single-factor analysis results helped to improve the grade of P2O5 to 30.1% with a recovery of 80%. It is important to highlight that increasing the Acid concentration above 7% may have a negative influence on phosphate minerals. It tends to attack phosphate minerals besides dissolving carbonates. Moreover, we can affirm that increasing leaching temperature has a good influence on leaching efficiency, and acid selectivity, but the most refractory carbonates remain unreacted above 40 °C; there are no benefits at leaching above this temperature. Then, the Taguchi L9 (34) orthogonal array experimental design is generated and % P2O5 mean values and the signal-to-noise ratio (larger the better) are calculated. Pareto plot of the two responses shows that acid type is the most significant factor in leaching and according to the desirability function, the optimum conditions were measured to be 7% of acetic acid, a solid concentration of 30%, a reaction time of 100 min, and a temperature of 40 °C. By optimization of the leaching parameters, the final test results showed that it was possible to produce a marketable phosphate concentrate with 81.01% recovery and 30.7% of P2O5 content. According to the ANCOVA analysis, the most important factor affecting the % P2O5 of phosphate washing sludges is the organic acid type at a reliability level of 95%, while it did not show effects for all the other leaching factors. These results are considered satisfactory as long as they have allowed the beneficiated low-grade phosphate washing sludges to reach the levels of marketable grades (≥30%). The dissolution kinetics by two different kinetic models of the calcareous material with acetic were proven to fit the shrinking core model for a chemically controlled process. The activation energy was determined to 48.9 kJ/mol, which is consistent with a chemically controlled reaction.