Use of Phosphate Solubilizing Bacteria to Leach Rare Earth Elements from Monazite-Bearing Ore

In the present study, the feasibility to use phosphate solubilizing bacteria (PSB) to develop a biological leaching process of rare earth elements (REE) from monazite-bearing ore was determined. To predict the REE leaching capacity of bacteria, the phosphate solubilizing abilities of 10 species of PSB were determined by halo zone formation on Reyes minimal agar media supplemented with bromo cresol green together with a phosphate solubilization test in Reyes minimal liquid media as the screening studies. Calcium phosphate was used as a model mineral phosphate. Among the test PSB strains, Pseudomonas fluorescens, P. putida, P. rhizosphaerae, Mesorhizobium ciceri, Bacillus megaterium, and Acetobacter aceti formed halo zones, with the zone of A. aceti being the widest. In the phosphate solubilization test in liquid media, Azospirillum lipoferum, P. rhizosphaerae, B. megaterium, and A. aceti caused the leaching of 6.4%, 6.9%, 7.5%, and 32.5% of calcium, respectively. When PSB were used to leach REE from monazite-bearing ore, ~5.7 mg/L of cerium (0.13% of leaching efficiency) and ~2.8 mg/L of lanthanum (0.11%) were leached by A. aceti, and Azospirillum brasilense, A. lipoferum, P. rhizosphaerae and M. ciceri leached 0.5–1 mg/L of both cerium and lanthanum (0.005%–0.01%), as measured by concentrations in the leaching liquor. These results indicate that determination of halo zone formation was found as a useful method to select OPEN ACCESS Minerals 2015, 5 190 high-capacity bacteria in REE leaching. However, as the leaching efficiency determined in our experiments was low, even in the presence of A. aceti, further studies are now underway to enhance leaching efficiency by selecting other microorganisms based on halo zone formation.

Caustic soda decomposition and concentrated sulfuric acid digestion have been widely used to decompose monazite for many decades [2,4].Due to its high chemical and thermal stability, monazite is very difficult to decompose; therefore, it is essential to eliminate the phosphate present in the ore by chemically attacking the mineral with sulfuric acid or sodium hydroxide at high temperature, so as to enhance the capacity to dissolve the REE.The sulfuric acid process results in a loss of phosphate as H3PO4, corrosion of the processing facilities, toxic gas and wastewater generation, as well as yielding impure products; the process is therefore no longer in commercial use.Caustic soda decomposition has some advantages in terms of the recovery of unreacted alkali and phosphorous, low energy consumption, and simplicity; however, the process also has limitations, such as the need for high-grade ore sources [5,6].
Biohydrometallurgical technology is an attractive alternative emerging green technology for the recovery of metals due to its environmental friendly, simple, and economic processing.However, very few works have been published on the biological recovery of rare earth metals, in particular, from monazite.Recently thorium, uranium, and REE extraction by microorganisms from monazite concentrate was reported [7,8].The authors used Aspergillus ficuum, organic acid producing fungi, and Pseudomonas aeruginosa, organic acid/siderophore producing bacteria.They found that those microorganisms produced citric, oxalic, or 2-ketogluconic acid and dissolved 55% and 47% of REE from monazite by A. ficuum and P. aerunoginosa, respectively.Another study on REE leaching from phosphate minerals apatite and monazite by organic acids such as citric, oxalic, phthalic, and salicylic was also published, even though chemical organic acids were used in this study, which were not biologically produced [9].Organic acid producing microorganisms secrete organic acids such as malic, gluconic, or oxalic acids [10,11], and the mechanism of metal dissolution by the microorganisms is both of acidolysis (protons dissociated from the organic acids) and complexolysis (metal-complexing anions from the acids) [12].
In a soil improvement process, phosphate solubilizing bacteria (PSB) have been widely used because they affect the release of phosphorous (P) from inorganic and organic P pools through solubilization and mineralization [13][14][15] by organic acid production.In the present study, we investigated the feasibility to use PSB in REE leaching from monazite-bearing ore as an exploratory study.Screening studies were performed by determining halo zone formation on agar media and phosphate solubilization in liquid media to predict REE leaching capacity of bacteria.The REE leaching abilities of the PSB from monazite-bearing ore were determined and compared with the screening data.

Comparison of Halo Zone Formation, Organic Acid Production, and Phosphate Solubilization by Phosphate Solubilizing Bacteria
Halo zone formation by test PSB strains were tested using Reyes minimal medium supplemented with 1.5% agar, 5.39 g/L Ca3(PO4)2 as a mineral phosphate, 30 g/L glucose or sucrose as a carbon source, and bromocresol green (BCG) as a pH indicator.The pH of the medium was adjusted to 6.5 using 1 N KOH.A 0.5% BCG solution was prepared in 70% ethanol as a stock solution and a 0.5 mL aliquot of the solution was added to 100 mL of Reyes minimal agar medium before autoclaving.Twenty microliters of PSB pregrown in nutrient broth was dot inoculated onto the surface of the agar plate prepared as above, incubated at 30 °C for 48 h.The total halo diameter and colony diameter were measured; halo size was calculated by subtracting the colony diameter from the total halo diameter.
A quantitative estimate of organic acid production by the PSB was performed using a 50-mL conical tube containing 30 mL of Reyes minimal medium supplemented with 0.45 g/L KH2PO4 and 30 g/L glucose, and inoculated with the bacterial strain (3 mL inoculum with approximately 10 8 CFU (colony forming unit)/mL).To measure the organic acid producing ability of the test PSB without any interference, KH2PO4 was used as an easily edible form of phosphate source because the availability of a phosphate source may affect organic acid production.An autoclaved uninoculated medium served as a control.The vials were incubated for 11 days in a shaking incubator at 30 °C and 180 rpm.At 1-day intervals, individual cultures were centrifuged at 5000 g for 30 min, and the supernatant was filtered with a 0.22-μm pore syringe filter.Concentrations of citric acid (C6H8O7), malic acid (C4H6O5), tartaric acid (C4H6O6), and acetic acid (C2H4O2) in the culture were determined by high-performance liquid chromatography (HPLC) (2690 Separation Module, Waters Alliance System, Milford, MA, USA) with 20 mM NaH2PO4 (pH 2.7, adjusted with 20% phosphoric acid), using an Atlantis T3 column (5 μm; 4.6 mm × 250 mm), and quantified using a Waters 996 Photo-diode Array Detector at 210 nm.

Ore Characterization
The elemental composition of the raw monazite-bearing ore used in this study is shown in Table 1, which was measured by alkali fusion with Na2CO3 for silica, permanganametric titration for calcium, and ICP after multi-acid digestion for the other elements [21].The mineralogical and morphological data of the ore were obtained by X-Ray Diffraction (XRD; Model RTP 300 RC, Rigaku Co., Tokyo, Japan) and a scanning electron microscope (SEM, JEOL 6400; JEOL Ltd., Tokyo, Japan) equipped with energy dispersive spectroscopy (EDS).The semi-quantification spot analysis SEM-EDS was performed using Thin Film Standardless Standard Quantitative Analysis (Oxide).

Monazite-Bearing Ore Leaching by Phosphate Solubilizing Bacteria
The monazite-bearing ore was ground and then sieved below 150 μm size.A solution of 30 mL Reyes minimal medium (prepared as above) containing 30 g/L of glucose was added to 50-mL conical tubes containing 5 g of the autoclaved ore.Leaching was run for 9 days in a shaking incubator at 30 °C and 180 rpm.On each of the 9 days, samples were taken and prepared independently for further analysis.Total cerium and lanthanum concentrations in the residue and the liquor were determined by ICP and the mineralogical change of the residue was analyzed by SEM-EDS and XRD.

Comparison of Phosphate Solubilization by Phosphate Solubilizing Bacteria
Ten PSB strains (as stated above) were selected based on the literature reviews of their potential for phosphate solubilization.The halo zone formation was determined as a screening study to investigate the phosphate solubilizing ability of the PSB.Due to the production of organic acids in the surrounding medium, the halo zone on the agar plate was formed by converting insoluble phosphate into soluble phosphate [22].The converting efficiencies vary for different forms of phosphate (i.e., Ca3(PO4)2, AlPO4, and FePO4); among them, calcium phosphate was most aggressively attacked by PSB.Higher phosphate solubilization from calcium phosphate than from aluminum and iron phosphate minerals has been reported, based on the solubility equilibrium and acidity constants of these compounds [23], thus Ca3(PO4)2 was used in this study as a mineral phosphate.Bromocresol green (BCG) was also added in Reyes minimal medium supplemented with glucose and sucrose to improve the clarity and visibility of the yellow-colored halo zone [11].The test PSB formed different sizes of distinct halo zones on the media depending on the bacterial strain and carbon source (Table 2).The widest halo zone was observed in the presence of A. aceti (diameter, 41.0 mm) on media supplemented with glucose; P. rhizosphaerae, P. fluorescens, B. megaterium, and M. ciceri also formed halo zones, while P. polymyxa, E. meliloti, A. brasilense, and A. lipoferum did not form halo zones on the media supplemented with either glucose or sucrose.Wider halo zones were observed in most samples when glucose was present rather than when sucrose was present; therefore, glucose was used as a carbon source throughout this study.Also, the use of modified Reyes agar medium supplemented with BCG and glucose was the most effective condition to determine halo zone formation.Concentrations of solubilized phosphate and calcium ion in the Reyes minimal medium supplemented with glucose and Ca3(PO4)2 were analyzed for a period of five days, to compare the phosphate solubilizing ability of PSB.The viable PSB cell numbers were maintained at approximately 10 7 CFU/mL medium during the experimental periods.The leaching efficiency was calculated by subtracting the mass of calcium and phosphate in the uninoculated control from the mass in the leaching liquor and dividing the leached mass by the total mass.The leaching efficiencies were shown to be highest generally within two to three days of incubation; thereafter, concentrations of leached phosphate and calcium decreased, possibly due to uptake by the PSB or absorption onto bacterial surfaces.The highest leaching efficiencies after three days of leaching were presented by A. lipoferum, P. rhizosphaerae, and B. megaterium (6.4%, 6.9%, and 7.5%, respectively, for calcium; and 3.7%, 6.8%, and 5.7%, respectively, for phosphate (Figure 1); leaching efficiencies by A. brasilense and E. meliloti were less than 2% for calcium; as expected, A. aceti showed the highest leaching efficiency of calcium (~32.5%).

Bacterial strain P . r h i z o s p h a e r a P . f l u o r e s c e n s P . p u t i d a B . m e g a t e r i u m P . p o l y m y x a E . m e l i l o t i A . b r a s i l e n s A . l i p o f e r u m M . c i c e r i A . a c e t i
Leaching efficiency (%) Inorganic phosphate solubilization by PSB depends on organic acid secretion from PSB [14].Citric, malic, tartaric, and acetic acids were measured by HPLC in 30-mL of the PSB cultures containing KH2PO4 and glucose (Figure 2).The test PSB strains showed organic acid production below ~0.2 mmol, except for A. aceti.The maximum malic acid produced by A. brasilense, A. lipoferum, M. ciceri, and P. rhizosphaerae occurred at 0.07 mmol at six days, 0.09 mmol at nine days, 0.06 mmol at eight days, and 0.18 mmol at six days, respectively (maximum production).Acetic acid was produced by A. aceti, with concentrations reaching a maximum of 15.8 mM (0.48 mmol) at four days of incubation.Hwangbo et al. [24] reported that Enterobacter intermedium produced gluconic acid from glucose and subsequently converted it to 2-ketogluconic acid, which has a strong ionic strength and therefore easily solubilizes rock phosphate into soluble forms.The same situation was reported when Acetobacter was used to produce gluconic acid [25].Because of the conversion of gluconic acid as stated above, gluconic acid might not be detected in the microbial cultures in this study.After six days, the amount of malic and acetic acid decreased.The pH of the media during incubation showed similar trends with organic acid production (Figure 3) because the pH decrease is due to the excreted metabolites, which included proton from organic acids, amino acids, and other metabolites.The pH in the medium of A. aceti decreased to about pH 3 (corresponding pH to ~15 mM of acetic acid, which was the acid produced) until eight days of incubation, and subsequently increased.The acidities of the media for A. brasilense, A. lipoferum, and P. rhizosphaerae also showed similar trends with acetic acid, with pH decreasing (i.e., 3-4) until six to seven days of incubation and increasing thereafter.

REE Leaching by the Test PSB from the Monazite-Bearing Ore and Comparison with the Screening Study Data
Ten PSB were introduced to leach REE from monazite-bearing ore (Figure 4).We detected ~5.7 mg/L of cerium (0.13% of leaching efficiency) and 2.8 mg/L of lanthanum (0.11%) in the leaching liquor of A. aceti, and approximately 0.5-1 mg/L of cerium and lanthanum (0.005%-0.01%) in the leaching liquors of A. brasilense, A. lipoferum, P. rhizosphaerae, and M. ciceri.Since acetic and malic acid were produced by A. aceti, A. brasilense, A. lipoferum, P. rhizosphaerae, and M. ciceri as stated above, REE such as cerium and lanthanum may form complexes with these acids.When REE cations (Re2O3) are present in a system and acetic or malic acid is fully dissociated in aqueous solution, a complexation reaction may take place (at 25 °C) [26]; C2H4O2 ↔ C2H3O2 − + H + (pKa = 4.756) 3C2H3O2 − + Re2O3 ↔ Re(C2H3O2)3 (REE acetate complex) Also, since malic acid is a diprotic acid, the possible complexes of REE reaction with malate anions are as follows; C4H6O5 ↔ C4H5O5 − + H + (pKa = 3.40) 3C4H5O5 − + Re2O3 ↔ Re(C4H5O5)3 (Rare earths malate complex) C4H5O5 − ↔ C4H4O5 2− + 2H + (pKa = 5.11) 3C4H4O5 2− + 2Re2O3 ↔ Re2(C4H4O5)3 (Rare earths malate complex) The role of low molecular weight organic acids such as acetic, malic, oxalic, and citric acids in metal dissolution has been discussed in many studies [7][8][9]27,28].The ligands of the organic acids play a dominant role in dissolving metals, and the differences in the effect of metal dissolution is related to their chemical structure.For instance, citric acid has great chelating ability with metals because citrate (i.e., three carboxylic groups) can form stable chelates with 6-membered ring structures.While malic acid (dicarboxylic acid) is more stable than acetic acid (monocarboxylic acid) [29], REE leaching was observed in the A. aceti culture (where most of the leaching agent was acetic acid) because acetic acid production was higher than other organic acids production.
Mineralogical changes of the monazite-bearing ore before and after PSB leaching were also considered.As a variety of minerals that entrap monazite may be present, and these minerals may prevent contact of the organic acids produced by PSB as shown in the XRD patterns and the SEM images (Figures 5 and 6; Table 3).The X-ray pattern of the monazite-bearing ore shows the presence of dolomite (CaMg(CO3)2), quartz (SiO2), siderite (FeCO3), pyrite (FeS2), and monazite ((Ce, La, Pr, Nd, Th, Y)PO4).The cerium and lanthanum-rich region (spot 1 and 2) had a spotty distribution surrounded by other minerals including Mg, Al, Si, or Fe (spot 3 and 4).High carbon ratio in the spots was because of epoxy impregnation in preparing flat-polished specimens.XRD patterns of the monazite-bearing ore before and after leaching in this study showed a reduction in all peak intensities (Figure 6).High peaks of carbonate minerals (dolomite (CaMg(CO3)2) and siderite (FeCO3)) were found and these minerals might consume the organic acids produced by PSB and prevent REE dissolution.
Differently from the phosphate solubilization test in liquid media (Figure 1), P. fluorescens, P. putida, B. megaterium, and P. polymyxa did not leach REE from the monazite-bearing ore.In the case of A. brasilense and A. lipoferum, even though they leached REE from the ore in spite of small leaching efficiency (Figure 4) and produced similar amounts of malic acid with M. ciceri and P. rhizosphaerae (Figure 2), they did not form halo zones (Table 2).This discrepancy sometimes happens because of the difference between cultivation in liquid media and on agar media.A liquid medium is preferred for the production of such organic acids because excreted products are readily available from a liquid culture and the cells are uniformly exposed to conditions of the medium [30].However, the strongest leaching of REE was observed in the presence of A. aceti, as expected by both the results of the halo zone formation and the phosphate solubilization test.In the case of low-capacity bacteria, some errors may occur as stated above in determination of halo zone formation, but this method is still useful to roughly estimate phosphate solubilizing and organic acid producing abilities, as well as REE leaching capacities.Thus we believe that determination of halo zone formation can be used as a simple and rapid screening method to select high-capacity bacteria in REE leaching.Biological leaching technology has some advantages over other methods, as it is relatively simple, inexpensive, and environmental friendly.However, even though A. aceti showed the highest leaching efficiency of the PSB tested, the efficiency of extraction was only ~0.13% (at least in batch-type experiments), which is still very low.The PSB used in this study have been used in soil improvement applications because they adapt well to soil environments; however, their organic acid producing and phosphate solubilizing ability is not that high.Thus, further studies are underway to explore the use of other microorganisms with good organic acid producing abilities for REE extraction applications, as well as to design continuous leaching systems so as to enhance leaching efficiency because the ore particles might not be exposed or contacted by the organic acid produced by the test PSB strains in a batch culture system.

Conclusions
The present study was conducted to use PSB to leach REE from monazite-bearing ore and to predict the REE leaching ability of PSB by determining halo zone formation on agar media and phosphate solubilization in liquid media.Among the ten test PSB strains, the phosphate solubilizing ability of A. aceti was the highest, based on results of halo zone formation and the phosphate solubilization test.Based on REE leaching data from the raw monazite-bearing ore by the test PSB, halo zone formation was found to be a useful method to select high-capacity bacteria in REE leaching.

Figure 1 .
Figure 1.Phosphate and calcium solubilization from Ca3(PO4)2 in Reyes basal liquid medium, associated with microbial growth after 3 days.

Figure 5 .
Figure 5. Scanning electron microscopic image of the raw monazite-bearing ore (3000× magnification).The detailed investigations are given in the text and Table3.

Table 1 .
Elemental analysis of monazite-bearing ore used in this study.

Table 2 .
Comparison of tricalcium phosphate solubilization by phosphate solubilizing bacteria (PSB) in Reyes minimal agar medium supplemented with bromocresol green (BCG) and 30 g/L of glucose or sucrose, associated with microbial growth for 24 h.The total halo diameter and colony diameter were measured; halo size was calculated by subtracting the colony diameter from the total halo diameter.** "-" represents no halo zone formation. *