Mineralogical and geomicrobiological investigation of phosphorite from Ervenik , Croatia

Phosphate minerals hydroxylapatite, fl uorapatite and crandallite were identifi ed in nodules within phosphorites from Ervenik, Croatia. The minerals were identifi ed using optical microscopy, XRD, SEM and EDS analyses. The presence of fungi was recognized only in association with phosphate-rich phases. Fungal activity resulted in the dissolution of apatite, producing hollow crystals, particularly in hydroxylapatite–enriched zones. A substantial number of hyphae were observed on the surface of phosphate minerals, in addition to saprophytic bacteria and bacterial spores. Induced activity of phosphate-accumulating bacteria in an aquatic environment caused dissolution of the phosphate minerals. The aqueous phase contained increased concentrations of several elements, including Ca, Sb, U, V and As. These elements are important constituents of minerals of the apatite group. As a consequence of the crystallization of apatite, the concentration of phosphate decreases with a corresponding increase in aluminium concentration, resulting in the prevalence of crandallite as the stable phase, forming the outer sector of the spherulites.


INTRODUCTION
Phosphorite deposits commonly occur in caves, caverns and other karst phenomena (DAHANAYAKE & SUBASINGHE, 1989;and references therein).They were identifi ed in limestones near the village of Ervenik, in Croatia during a 2009 fi eld campaign.It is generally assumed (IVANOVIĆ et al., 1973) that these limestones are of Cretaceous age, while the exact age of the deposits is unknown, but assumed to be Pleistocene.The source of phosphorus could be the phosphate-rich guano material originating from bats and birds inhabiting the surrounding caves (MARKOVIĆ, 2002).Several outcrops of these deposits were found at the foot of the SE part of Mt.Velebit near Ervenik, Croatia (POSILOVIĆ et al., 2010).
Phosphates are considered distinct in that they are able to incorporate more than half of the elements of the periodic table into their atomic structures (PASERO et al., 2010).The apatite group minerals and crandallite are assumed to be secondary alteration phases and the main phosphorus-bearing minerals.
The aim of this study was to defi ne the paragenesis of phosphorite from Ervenik, Croatia and to study bacteriological infl uence to this phosphorite.

Sampling and preparation of minerals
Phosphorite samples were collected at a location close to the village of Ervenik (Fig. 1).The former phosphorite mine is no longer operational, so the samples were collected from old mine dumps.Thin sections were prepared for optical investigation.The collected rock samples were crushed, sieved and the particle size of 0.212-0.500mm was prepared in the laboratory for separation and bacteriological experiments, as well as other analyses.

Experimental methods
A polarising microscope was used for optical investigations.Mineral phase determinations of the bulk samples were done using a Philips X'pert powder diffractometer, with CuKα radiation fi ltered with a graphite monocrystal monochromator running at 40 kV and 40 mA.An X-ray diffraction data set was collected from 4 to 65º2θ.Powder diffraction patterns were identifi ed and indexed using X'Pert HighScore Plus v. 2.1.software (PANALYTICAL, 2004), and ICDD Powder diffraction fi le database (PDF 2, ICCD, 2004).Unit cell parameters were calculated using "Unitcell" software (HOLLAND & REDFERN, 1997).Crushed samples of phosphorite were examined by scanning electron microscopy (Tescan Vega TS 5136), coupled with an INCA 250 system for elemental composition analysis.The water content in the minerals was determined by drying a 1.000 g aliquot of the mineral at 105°C/4 h.
In sieved samples, the number of natively present bacteria was established.Since the mineral samples were originally found under aerobic conditions and were maintained in an appropriate environment until analysis, the number of aerobic/facultatively anaerobic bacteria was checked.The number of natively present heterotrophic bacteria either viable or in the form of spores was determined in each mineral.The presence of faecal streptococci was tested in order to check the possible anthropogenic infl uence in the form of faecal contamination in the minerals (BOUVY et al., 2010).In order to check whether the minerals facilitated bacterial growth in contact with water, bacteriological tests were performed (1) after 3 min, and (2) after 4 days of incubation.
In order to check if the phosphate-accumulating bacteria use the minerals as the source of phosphorus, the performance of phosphate-accumulating bacteria Acinetobacter junii DSM no.1532 was tested (HRENOVIĆ et al., 2010a).
For determination of the presence of bacteria in minerals, a 1.0 g aliquot of each mineral was placed in a tube containing 9 mL of 0.05 molL -1 NaCl.Samples were vigorously shaken on a mechanical shaker (45Hz/3min, Kartell TK3S) in order to detach the adsorbed bacteria from the mineral (DUR-HAM et al., 1994).The number of aerobic/facultatively anaerobic bacteria was determined in the supernatant.The number of faecal streptococci was determined on Slanetz-Bartley agar (Biolife, Italy) after 3-day incubation at 37°C.The number of heterotrophic bacteria and bacterial spores was determined on the nutrient agar (Biolife, Italy) after 3-day incubation at 22 °C.In order to determine the number of bacterial spores, the minerals were pasteurized at 80 °C for 10 minutes.

Bermanec et al.: Mineralogical and geomicrobiological investigation of phosphorite from Ervenik, Croatia
In order to establish the likelihood of multiplication of the bacteria present on mineral surfaces in contact with water, a 1.0 g aliquot of each mineral was placed in a tube which contained 9 mL of 0.05 molL -1 NaCl and incubated for 4 days in the dark at 22 °C, with continual stirring (70 rpm) without additional aeration.The bacteriological analysis was performed after this incubation period.
In order to check if the phosphate-accumulating bacteria A. junii use the minerals as a source of phosphate, a 1.0 g of autoclaved (121°C/20 min) mineral was placed in a tube which contained 9 mL of 0.05 molL -1 NaCl.A suspension of A. junii was added (1.06±0.04x 10 7 CFU/mL) to each tube and these were aerobically incubated in darkness for 24 h at 30°C with stirring (70 rpm).After incubation, samples were vigorously shaken on a mechanical shaker in order to detach the adsorbed bacteria from the mineral.The number of A. junii was determined from the supernatant on the nutrient agar after incubation at 30 °C/24 h.The Neisser stain was utilised to confi rm the presence of polyphosphate granules in the cells of A. junii while the Gram stain was used to confi rm the shape of bacteria and immobilization of the cells on the mineral surfaces.All measurements were done in triplicate.The fi nal pH of the suspension was measured using a WTW 330 pH-meter.
Statistical analyses were carried out using Statistica Software 9.1 (STATSOFT, 2009).The numbers of bacterial CFU were logarithmically transformed beforehand to normalize the distribution and to equalize variances of the measured parameters.The comparisons between samples were done using the one-way analysis of variance (ANOVA), and subsequently the post-hoc Duncan test was performed for calculations concerning pair-wise comparisons.Statistical decisions were made at a signifi cance level of p<0.05.

1. Mineralogical analyses
The phosphate mineral occurrences found in caves are usually compact and reddish or greyish in colour, with a high Al 2 O 3 and FeO content, but low in carbonate.The minerals occur as white to greyish spherulitic aggregates or single crystals <2 mm in size.Optical microscopy and SEM of spherulitic aggregates showed at least two different phosphate phases.
Thin sections of phosphate containing samples show clear zonality.In the inner core of the spherulites, apatite crystallises in the form of crystal sprays up to 1.5 mm in length; the outer parts of the spherulites contain very fi nely grained radial to fi brous crandallite.Larger spherulites are banded and composed of alternating apatite and crandallite layers.
Samples of phosphorite recently collected at the mine dumps, left during the 1950's were crushed to grains, up to 2mm diameter and separated into three fractions named: Er-venik_white, Ervenik_pink and Ervenik_mix.
On the basis of combined SEM, EDS and XRD analyses, the phosphate phases were identifi ed as hydroxylapatite, fl uorapatite, and crandallite.
Broken samples of phosphorite were used for SEM investigations.Idiomorphic apatite crystals were found in the cores of the nodules.The crystal habit is a simple combination of hexagonal prisms and basal faces.They are <50 µm long and <35 µm wide (usually 25 x 10 µm).Some crystals show parallel growth, while others have more irregular growth patterns (Fig. 3).Some crystals are hollow.The smaller ones have empty cores, but larger crystals have a hexagonal core within the hollow outer part of the crystal.Tiny fi bres connecting the core and crust of the crystals can occasionally be seen (Fig. 3a).
A closer look at the fi bres identifi ed them as fungal hyphae that grow on the surface of apatite crystals and infl uence these with their metabolic processes (Fig. 4).
The outer zones of nodules consist of smaller crystals of crandallite.Crandallite crystals are of rhombohedral habit <5 µm in size.They usually have small basal faces (Fig. 5), and are mutually separated by fi ne-grained anhedral bauxitic material.
Compositions of apatite and crandallite were confi rmed by EDS analyses.EDS analyses of apatite, in the cores of nodules, commonly show the presence of Ca, P, and O.This is a strong indication for hydroxylapatite (Fig. 6A).In contrast to hydroxylapatite, the EDS spectrum of crandallite shows the additional presence of aluminium (Fig. 6B).

Bacteriological analyses
The absence of faecal streptococci (bacteria of the genus Enterococcus), used as surrogates for human and animal pathogens in assessing water quality (BOUVY et al., 2010), con-fi rm ed that the minerals were not contaminated with mu ni cipal wastewater (Table 2).
The original minerals contained higher numbers of viable saprophytic bacteria than the bacterial spores (Table 2).

ARTICLE IN PRESS
Geologia Croatica 65/1 The presence of viable bacteria in the materials may be explained by the 0.92-0.98%water content (which was lost at 105 °C/4 h) in the minerals, which was suffi cient to maintain the bacteria.The sample Ervenik_pink naturally contained a signifi cantly lower number of viable saprophytic bacteria than the other two samples.After 4 days of incubation, the number of viable saprophytic bacteria signifi cantly increased in samples Ervenik_white and Ervenik_mix, but not in sample Ervenik_pink.The sample Ervenik_pink still contained the lowest number of viable saprophytic bacteria.The multiplication of saprophytes in samples decreased in order Er-venik_white = Ervenik_mix > Ervenik_pink (Table 2).It seems that the number of viable saprophytic bacteria and its multiplication was dependent of the content of crandallite in the minerals.In particular, the Al contained in crandallite can have a negative infl uence and a toxic effect against bacteria (HRENOVIĆ et al., 2010b).
The number of bacterial spores (Table 2) in the samples did not change on incubation of minerals in 0.05 mol L -1 NaCl for 4 days.This suggested that the minerals in the presence of nutrient depleted water were not suitable for induction of the germination of bacterial spores.
The fi nal pH values (Table 2) of the aqueous media became alkaline after 3 min of contact of minerals with 0.05 mol L -1 NaCl, and slightly decreased to neutral after 4 days of incubation.Although signifi cantly higher in the aqueous medium which contained the sample Ervenik_pink than Ervenik_ white or Ervenik_mix, the maximum difference was only 0.19 pH units.Therefore, the pH value of the medium can be eliminated as a cause of different bacterial numbers in minerals.
In experiments with phosphate-accumulating bacteria, all inoculated bacteria A. junii were immobilized onto minerals during 24 h of incubation (Fig. 7).The number of immobilized bacteria per mass of dry mineral was 1.04±0.09x 10 7 CFU/g.There was no signifi cant change among the initial and fi nal numbers of A. junii in the tubes.It is probable, that the mineral and nutrient depleted water failed to provide enough sources of nutrients for multiplication of bacteria which were isolated from activated sludge treating the wastewater (HRENOVIĆ et al., 2003).All inoculated bacteria survived and accumulated polyphosphate inside the bacterial cells.The phosphate accumulated inside the cells of A. junii in the form of polyphosphate granule was confi rmed by the Neisser stain (Fig. 8).This experiment confi rmed that the phosphate-accumulating bacteria A. junii were able to use the minerals as the source of phosphate for their metabolism.These fi ndings are in agreement with SMITH et al. (1978), where suffi cient phosphate was released by the partial dissolution of apatite crystals at pH 7.8 to support the growth of bacteria.The fact that all inoculated bacteria survived and were spontaneously immobilized onto minerals confi rmed that the investigated phosphorites were excellent substrates for the immobilization of selected bacteria.After the contact of the sample Ervenik_white with A. junii, the solution was fi ltered through a Sartorius nitrocellulose fi lter with a pore diameter of 0.2 μm and the water was analysed by ICP-MS to compare it with a blank solution of 0.05M NaCl, in order to detect the bacteriological activity on the phosphate minerals (Table 3).A set of 22 trace elements were measured: Al, As, Ba, Cd, Co, Cr, Cs, Cu, Fe, Li, Mn, Mo, Ni, Pb, Rb, Sb, Sn, Sr, Tl, U, V, and Zn, (most of these elements are common constituents of apatite group minerals), as well as the macroelements Ca, K, Mg and Na which were expected at higher concentrations.
Na was measured in both blanks and in sample waters in the same order of concentrations.The concentrations of Ca, Mg and K in the aqueous phase of the samples were higher after bacteriological activity.The highest increase in concentration was observed for Ca (124-fold) when compared to concentrations of Mg and K (5 and 14 times, respectively).This observation is in agreement whit the statement that Mg and K ions were essential elements for A. junii while Ca was not (HRENOVIĆ et al., 2010c).Additionally, Ca is the dominant cation in apatite and Mg and K are minor constituents.Among the trace elements, the bacterial activity has the highest infl uence on the concentrations of Sb, U, V and As, which increased 973, 105, 452 and 226 times, respectively.A signifi cant increase in the concentrations was also established for Ni, Cu and Sr (24, 61 and 57 times, respectively).Li, Cd, Cr, Mn and Co increased their concentrations about 10 times, Rb, Mo, Tl, and Al increased about 5 times, Sn, Fe and Ba about 2 times, Pb concentration remained the same, while Cs concentration signifi cantly decreased.It is possible that A. junii has the ability to accumulate Cs within its cells or on their surface, which has been observed for other bacteria (NAKAO et al., 2005), but this needs to be further examined.Increased concentrations of elements may be explained as a result of the dissolution of hydroxylapatite, which can contain these elements in its crystal lattice.These elements are not involved in bacterial metabolic processes.

DISCUSSION AND CONCLUSION
From the results of the X-ray diffraction analysis it is possible to identify that the minerals investigated (according to their structure) belong to the apatite group of minerals (hexagonal apatites).Apatite group minerals consist of 15 different minerals.Most of them ( 11) have the same hexagonal structure and the space group P6 3 /m (PASERO et al., 2010).By calculating unit cell parameters, it is possible to conclude that these are apatite group minerals (Ca 5 (PO 4 ) 3 X, X=F, OH, Cl), having unit cell parameters comparable to fl uorapatite and hydroxylapatite.Unit cell dimensions of apatite group minerals in the separated parts of spherulites signifi cantly differ from each other, suggesting that there are two end members present in these nodules, partly making solid solutions.The most interesting members of the apatite group are hydroxylapatite and fl uorapatite, as they can be affected by biological activity.
The plumbogummite subgroup of the alunite-jarosite group of minerals (BACK & MANDARINO, 2008) contains phosphates that are very important as weathering products, commonly present as constituents of soils.According to unit cell calculations and EDS measurements, this is crandallite CaAl 3 (PO 4 ) 2 (OH, H 2 O) 6 .This mineral phase is related to the presence of aluminium in bauxitic material surrounding the spherulites.
The identifi ed phosphate minerals are assumed to be associated with biomineralization processes, these being a consequence of organic matter interaction with the carbonate phases of the host rock limestones (POSILOVIĆ et al., 2010).
Careful investigation of the collected samples prior to bacteriological/biological experiments, resulted in the interesting observation that fungi only grow on phosphate rich sections of the mineral substrate, and not on carbonate nor iron and aluminium rich oxide phases.
Due to the higher solubility of hydroxylapatite, with respect to fl uorapatite, hydroxylapatite enriched zones were preferentially dissolved resulting in hollow apatite crystals.
Crandallite formation on the contact of phosphorite nodul es with the surrounding bauxite matrix is the result of increased aluminium availability.Apatite and crandallite spheru lit es were formed during early diagenesis of the cave sediment; each single spherulite was precipitated in a partially closed microenvironment of the sediment voids and pores.Apatite occupying the central portions of the spherulite centre, formed prior to crandallite in the outer spherulite region.The calculated stability fi elds for apatite and crandallite may be used to explain the bi-phase precipitation in the spherulites.Higher phosphate and lower aluminium concentrations with higher pH will favour apatite precipitation; such conditions prevailed during precipitation of the spherulite core.As a consequence of apatite precipitation, phosphate concentration is decreased and aluminium is increased, the stability fi eld of crandallite prevails and crandallite becomes the stable phase precipitating in the spherulite outer sectors (POSILOVIĆ et al., 2010).

Figure 4 :
Figure 4: The partial dissolution of an apatite crystal is a result of fungal activity on zonal apatite crystals (solid solution of hydroxylapatite and fl uorapatite).SE mode allows better view of the organic hyphal tissue (H).

Figure 5 :
Figure 5: Crandallite crystals (Cra) from phosphorite are developed on the contact with the bauxitic host rock.The crystals are of rhombohedral habit and up to 5 µm in size.SEM BSE image.

Figure 7 :
Figure 7: Cells of A. junii immobilized onto mineral Ervenik_white.Figure 8: Granules of polyphosphate inside the cells of A. junii immobilized onto mineral Ervenik_white.

Figure 8 :
Figure 7: Cells of A. junii immobilized onto mineral Ervenik_white.Figure 8: Granules of polyphosphate inside the cells of A. junii immobilized onto mineral Ervenik_white.

Table 1 :
Unit cell parameters for the identifi ed phases.

Table 2 :
Number of bacteria per 1.0 g of dry mineral after 3 min and after 4 days of contact with water.* signifi cantly diff erent at 4 days when compared to 3 min; A signifi cantly diff erent to Ervenik_white; B signifi cantly diff erent to Ervenik_mix.

Table 3 :
Concentrations of macro and trace elements in water after dissolution experiments in the presence (sample) and absence (blank) of A. junii bacteria.