Apatite-(CaOH) in the fossil bat guano deposit from the "dry" Cioclovina Cave, Sureanu Mountains, Romania

Abstract

Apatite-(CaOH) is the most abundant phosphate in the deposit of fossil but guano in the "dry" Cioclovina Cave, Sureanu Mountains, South Carpathians, Romania. Initial deposits, of both biogenic and authigenic origin, were chemically equilibrated during diagenesis. Individual crystals are tabular, roughly hexagonal, platy on (0001) and usually between 3 and 15 mu m across and LIP to 1 mu m thick. The mean indices of refraction measured for 10 representative samples are epsilon 1.645(2) and omega 1.653(1). The mean measured density [D(m) = 3.17(2) g/cm(3)] is in good agreement with the individual calculated values. The unit-cell parameters calculated as an average of 20 sets of values are a 9.436(13), c 6.868(8) angstrom. The mineral is Ca-deficient, carbonate-and sulfate-bearing. Less than 2.26% of the phosphate groups are protonated, and less than 4.66% are replaced by sulfate. The cumulative incorporation of other cations in the Ca sites accounts for only 0.71 to 4.07% (mean 2.07%). Both Unit-Cell parameters and thermal behavior are characteristic for a hydrous A-type carbonated apatite-(CaOH), with molecular H(2)O and carbonate substituting for hydroxyl in the structural channels. The multiplicity Of the hands in the infrared absorption spectrum (3 nu(3) + 1 nu(1) + 3 nu(4) + 2 nu(2)) is consistent with a C(6) point symmetry of the phosphate anion. Whitlockite obtained by thermal breakdown at 1000 degrees C is sulfate-bearing. The authigenesis of the Cioclovina apatite-(CaOH) involved a reaction between Calcium carbonate from the moonmilk flows or the cave floor and phosphoric solutions derived from guano, with brushite or an X-ray amorphous phase as a precursor.

Full text

463
The Canadian Mineralogist
Vol. 46, pp. 000 (2008)
DOI : 10.3749/canmin.46.2.000
APATITE-(CaOH) IN THE FOSSIL BAT GUANO DEPOSIT FROM
THE “DRY” CIOCLOVINA CAVE, ˛SUREANU MOUNTAINS, ROMANIA
De l i a -Ge o r G e t a DUMitraŞ a n D Şt e fa n MarinCea§
Department NII, Geological Institute of Romania, 1 Caransebeş Street, Bucharest, RO–012271 Romania
es s a ï D Bilal
Département GENERIC, Centre SPIN, Ecole Nationale Supérieure des Mines de Saint Etienne,
158, Cours Fauriel, Saint Etienne, Cedex 2, F–42023, France
fr é D é r i C Hatert
Laboratoire de Minéralogie, Université de Liège, Sart-Tilman, Bâtiment B-18, B–4000 Liège, Belgium
aB s t r a C t
Apatite-(CaOH) is the most abundant phosphate in the deposit of fossil bat guano in the “dry” Cioclovina Cave, Şureanu
Mountains, South Carpathians, Romania. Initial deposits, of both biogenic and authigenic origin, were chemically equilibrated
during diagenesis. Individual crystals are tabular, roughly hexagonal, platy on (0001) and usually between 3 and 15 mm across and
up to 1 mm thick. The mean indices of refraction measured for 10 representative samples are  1.645(2) and v 1.653(1). The mean
measured density [Dm = 3.17(2) g/cm3] is in good agreement with the individual calculated values. The unit-cell parameters calcu-
lated as average of 20 sets of values are a 9.436(13), c 6.868(8) Å. The mineral is Ca-decient, carbonate- and sulfate-bearing.
Less than 2.26% of the phosphate groups are protonated, and less than 4.66% are replaced by sulfate. The cumulative incorpora-
tion of other cations in the Ca sites accounts for only 0.71 to 4.07% (mean 2.07%). Both unit-cell parameters and thermal behavior
are characteristic for a hydrous A-type carbonated apatite-(CaOH), with molecular H2O and carbonate substituting for hydroxyl
in the structural channels. The multiplicity of the bands in the infrared absorption spectrum (3n3 + 1n1 + 3n4 + 2n2) is consistent
with a C6 point symmetry of the phosphate anion. Whitlockite obtained by thermal breakdown at 1000°C is sulfate-bearing. The
authigenesis of the Cioclovina apatite-(CaOH) involved a reaction between calcium carbonate from the moonmilk ows or the
cave oor and phosphoric solutions derived from guano, with brushite or an X-ray amorphous phase as a precursor.
Keywords: apatite-(CaOH), physical properties, crystal chemistry, unit-cell parameters, thermal behavior, infrared absorption
data, authigenesis, “dry” Cioclovina Cave, Romania.
so M M a i r e
L’apatite-(CaOH) est le phosphate le plus abondant du dépôt fossile de guano de chauve-souris de la grotte «sèche» de Cioclo-
vina, monts Şureanu, Carpates Méridionales, en Roumanie. Les dépôts initiaux, d’origine aussi bien biogène qu’authigène, ont été
portés à l’équilibre chimique durant la diagenèse. Les individus cristallins sont tabulaires, vaguement hexagonaux, aplatis selon
(0001). En général, ils ont entre 3 et 15 mm de diamètre et une épaisseur jusqu’à 1 mm. Les indices de réfraction moyens, mésures
sur 10 échantillons représentatifs, sont  1,645(2) et v 1,653(1). La densité mesurée moyenne [Dm = 3.17(2) g/cm3] concorde
bien avec les valeurs de densité calculées. Les paramètres réticulaires, obtenus comme moyenne de 20 déterminations de valeurs
individuelles, sont a 9.436(13) et c 6.868(8) Å. Le minéral est décitaire en Ca et contient du carbonate et du sulfate. Jusqu’à
2.26% des groupes phosphate sont protonés et jusqu’à 4.66% sont remplacés par des groupes sulfate. Le remplacement du Ca
par d’autres cations est mineur, ne comptant que pour 0.71 à 4.07% des sites occupés normalement par Ca (2.07% en moyenne).
Les paramètres de la maille, ainsi que le comportement thermique, sont caractéristiques d’une apatite-(CaOH) carbonatée de
type A, contenant de l’eau moléculaire et du carbonate en tant que remplaçants de l’hydroxyle dans les tunnels structuraux. La
multiplicité des bandes dans le spectre d’absorption en infrarouge (3n3 + 1n1 + 3n4 + 2n2) concorde avec une symétrie ponctuelle
C6 de l’anion phosphate. La whitlockite obtenue après la décomposition thermique à 1000°C de l’apatite-(CaOH) contient du
§ E-mail address: marincea@igr.ro
463_vol_46-2_art_15.indd 463 08/04/08 12:05:12

464 tHe CanaDian MineraloGist
sulfate. Si authigène, l’apatite-(CaOH) de Cioclovina est issue des réactions entre le carbonate de calcium des coulées de “lait
de lune” ou du socle de la grotte et les solutions phosphatées dérivées du guano, ayant la brushite ou des phases amorphes aux
rayons X comme précurseurs.
Mots-clés: apatite-(CaOH), propriétés physiques, chimie cristalline, paramètres de maille, comportement thermique, données
d’absorption infrarouge, authigenèse, grotte «sèche» de Cioclovina, Roumanie.
phosphate association displays all the characteristics
of a “dry” system, favored by a mean temperature of
about 10°C, with small variations over the seasons, and
a relatively low humidity (80–88%). The mined part of
the cave (Fig. 1) accounts for about 800 m out of a total
length of 2002.5 m known so far (Tomuş 1999).
Apatite-(CaOH) is the most abundant calcium phos-
phate in the fossil deposit of bat guano from Cioclovina;
it makes up to 70% by volume of the phosphate mass.
The mineral is most abundant near the cave floor,
which is visible where the carbonate substratum was
not disturbed by mining.
Apatite-(CaOH) from Cioclovina is commonly
found in intimate association with quartz, illite 2M1,
X-ray-amorphous iron sesquioxides and, in some
cases, with brushite or ardealite. Apatite-(CaOH) is
the earliest formed of the phosphate phases. Other
associated minerals in the fossil guano deposit include
taranakite, monetite, crandallite, tinsleyite, leucophos-
phite, variscite, francoanellite, gypsum, bassanite,
calcite, vaterite, aragonite, hematite, goethite, birnes-
site, romanèchite, todorokite and kaolinite. Onac et
al. (2002, 2005) and Onac & White (2003) reported
the presence of some exotic mineral species, such as
berlinite, burbankite, churchite, chlorellestadite, foggite,
paratacamite, collinsite and sampleite, but some of
these species were probably misidentied (Marincea
& Dumitraş 2005).
an a l y t i C a l Pr o C e D U r e s
The mineral was investigated using scanning
electron microscopy (SEM), combined with energy-
dispersion spectroscopy (SEM–EDS), optical study,
wet-chemical analysis, wavelength-dispersive electron-
microprobe analysis (EMPA), inductively coupled
plasma – atomic emission spectrometry (ICP–AES),
ion chromatography, X-ray powder diffraction (XRD),
thermally assisted X-ray powder diffraction, thermal
analysis and infrared absorption spectrometry. Most
of the analytical facilities, procedures and experi-
mental details are similar to those already described by
Marincea et al. (2002), Marincea & Dumitraş (2003)
and Dumitraş et al. (2004). We will focus on the tech-
niques not previously used.
Wavelength-dispersive electron-microprobe anal-
yses were carried out on a fully automated CAMECA
SX–50 apparatus, using the same technique, standards,
operating conditions and corrections as described by
Marincea et al. (2002).
in t r o D U C t i o n
Apatite-(CaOH), ideally Ca5(PO4)3(OH), a mineral
investigated by mineralogists, crystallographers and
solid-state physicists over many years, is of great
interest in material science, mainly because of its use in
biomedical applications (e.g., bone and teeth implants).
Despite the wealth of documented research on apatite-
(CaOH), and its notorious abundance in the fossil bat
guano deposits in the cave environment, it is perhaps
surprising that there are very few reports of integrated
mineralogical data on apatite-(CaOH) from caves (e.g.,
Balenzano et al. 1974, Fiore & Laviano 1991).
Normally, apatite-(CaOH) from caves is carbonate-
bearing. A description using a specific term like
“carbonate–hydroxylapatite” is undesirable, as this is an
invalid species (Burke 2008). The use of terms such as
“dahlite” (e.g., Diaconu & Medeşan 1975) has caused
a considerable amount of confusion in the past.
As part of a wider attempt to better characterize the
phosphate minerals from the type locality of ardealite
(Schadler 1932), we document the occurrence and some
fundamental chemical, crystallographic and physical
aspects of apatite-(CaOH) from the “dry” Cioclovina
Cave, in the Şureanu Mountains, Romania. Our aims
are twofold: (1) to present mineralogical data on the
mineral, and (2) to investigate the chemical peculiarities
of apatite-(CaOH) from a “dry” cave hosting a fossil bat
guano deposit, as well as the relations between apatite-
(CaOH) and the other phases.
Ge o l o G i C a l se t t i n G
The “dry” Cioclovina Cave has received consider-
able attention over the years (e.g., Schadler 1929, 1932,
Constantinescu et al. 1999, Marincea et al. 2002, Onac
et al. 2002, 2005, Marincea & Dumitraş 2003, 2005,
Onac & White 2003, Dumitraş et al. 2004). Descrip-
tions of the geological setting were presented by all
these authors. The cave is located in the upper basin of
Luncanilor Valley, at the base of the western slope of
the Şureanu Mountains, in the Southern Carpathians, at
about 16 km east–southeast of the city of Haţeg. It is
developed in Tithonian – Neocomian reef limestones,
which form a ridge that penetrates into the Haţeg
Basin.
The huge fossil guano deposit inside the cave was
mined during the rst half of the last century. This
resulted in magnificent exposures of the phosphate
layers and greatly facilitated the investigation. The
463_vol_46-2_art_15.indd 464 08/04/08 12:05:12

a P at i t e -(Cao H ) i n fossil B a t GUano DePosit, roMania 465
For all the bulk analyses, the mineral was sepa-
rated by hand-picking with a preliminary screen on
the basis of the mineral association in order to choose
only samples devoid of brushite and ardealite. The
nal purity was checked by XRD. X-ray-amorphous,
carbonate-bearing “moonmilk” was identied in many
of the samples by infrared spectrometry.
The wet-chemical analysis was carried out on about
1 g of carefully hand-picked sample, using combined
gravimetric, titrimetric and atomic absorption tech-
niques, as described by Iosof & Neacşu (1980). The
presence of S was checked by ion chromatography. The
amount of CO2 should normally be in excess because of
the presence in many samples of the X-ray amorphous
“moonmilk”. However, a test for the CO2 content was
carried out on about 50 mg of high-purity powder
tested by infrared spectrometry, using a conventional
volumetric method coupled with a passage through
KOH of the evolved gas, modied from that described
by Shapiro & Brannock (1962).
The ICP–AES analyses were carried out following
the method described by Marincea et al. (2002), using a
Jobin–Ivon 138 Ultrace spectrometer. The sulfate anion
was determined from the same solutions as used for
ICP–AES analysis using a Dionex DX–500 ion chro-
matograph with standard AG 11–HC and AS 11–HC
columns. The estimated accuracy of the analytical data
is Å1% of the amount present for each of the major
cations.
X-ray powder-diffraction (XRD) analysis was
performed on a Siemens D–5000 Kristalloex automated
diffractometer equipped with a graphite diffracted-beam
monochromator (CuKa radiation, l = 1.54056 Å). The
analytical conditions were the same as described by
Dumitraş & Marincea (2004). The unit-cell parameters
were calculated by least-squares renement of the XRD
data, using the computer program of Appleman & Evans
(1973) modied by Benoit (1987). Synthetic silicon
(NBS 640b) was used as an external standard in order
to verify the accuracy of the measurements.
The same operating conditions were maintained
during thermally assisted diffractometry, which was
carried out using a Brucker (AXS) D8 Advance diffrac-
tometer, under a constant helium ow (ow rate 20
mL/min). The powders were deposited on a stainless
steel support [Ni–Cr–Fe, a 3.5545(5) Å at 25°C]. The
heating rate was of 10°C/min, and each record was
made 10 minutes after the desired temperature was
attained. Three lines belonging to the support (those at
~2.05, 1.78 and 1.26 Å) were used to correct for shift
of the main lines.
The thermal analysis was accomplished with a
double furnace SETARAM TAG 24 thermobalance
coupled with a DSC 111 thermal analyzer, with simul-
taneous registration of thermogravimetric (TGA),
differential thermogravimetric (DTG) and differential
scanning calorimetric (DSC) curves. The heating rate
was 10°C/min under a constant ow of argon (ow
rate 30mL/min).
The infrared absorption spectra were recorded using
both a SPECORD M–80 spectrometer in the frequency
range between 250 and 4000 cm–1 (Marincea & Dumitraş
2003), and a Fourier-transform THERMO NICOLET
NEXUS spectrometer, in the frequency range between
400 and 4000 cm–1. In both cases, we used standard
pressed disk techniques and KBr pellets.
Luminescence tests were performed using a portable
Vetter ultraviolet lamp, with 254 and 366 nm lters.
fiG . 1. Sketch of the “dry” Cioclovina Cave (the mined area) showing the location of apatite-(CaOH) samples used for this
study (marked with stars). Redrawn from Constantinescu et al. (1999).
463_vol_46-2_art_15.indd 465 08/04/08 12:05:13

466 tHe CanaDian MineraloGist
Mo D e o f oC C U r r e n C e
a n D Mo r P H o l o G y o f Cr y s ta l s
Macroscopically, apatite-(CaOH) from Cioclovina
generally forms porous multilayered crusts or mamillary
encrustations. The layered crusts attain 3 cm thick. They
are usually deposited directly on the carbonate oor of
the cave. The mineral also occurs as nodules in the fossil
guano mass or in fossil bone fragments surrounded by
guano. Veins of apatite-(CaOH) may also transect the
fossil guano mass and locally the detrital sequences of
terra rossa type.
Under the microscope, in transmitted light, the
mineral, very ne-grained, has invariably a lamellar
appearance and appears as rosette-like to radiating
aggregates of tabular crystals. Individual crusts,
bunches of crystals and radiating aggregates do not
exceed 3 mm in thickness, and most are typically much
smaller. Diffuse alteration of some apatite-(CaOH)
crusts spreads out from microscopic fractures, which
are marked by extremely ne veins of brushite or iron
sesquioxides.
The SEM study shows that the masses macroscopi-
cally perceived as crusts are composed of randomly
oriented or compact radiating aggregates of lamellar
crystals, roughly hexagonal and attened on (0001).
The individual crystals are usually between 3 and
15 mm across, and up to 1 mm thick. Five modes of
occurrence may be distinguished: (1) Aggregates of
subspherical shape or isolated, nearly perfect spheres,
generally varying in diameter between 10 and 100 mm.
These aggregates (Fig. 2A) consist of very ne, densely
packed radiating crystals and seem to be formed by the
crystallization of an amorphous material. (2) Networks
of reticulate aggregates (Fig. 2B), probably representing
fragments of fossil spongy bones or other tissues. (3)
Rosette-like radial aggregates with diameters up to 500
mm, built of blade-like, roughly hexagonal crystals (Fig.
2C). These could also result from the recrystallization
of the initial gels and are quite common. Illite (Fig. 2D)
and quartz were commonly found as inclusions in this
kind of aggregates and apparently served as seeds for
crystallization during the transition of an amorphous
Ca phosphate toward apatite-(CaOH). (4) Aggregates
of randomly oriented interlocking or subparallel platy
crystals, having locally rosette-like nuclei (Fig. 2E).
This kind of aggregate constitutes the major amount
of the mass of layered crusts deposited directly on the
carbonate substrate. (5) Aggregates of tabular hexagonal
crystals intergrown on (0001), that are apparently
relics of cortical (“compact”) fossil bones affected
by advanced recrystallization after “digestion” by the
guano mass (Fig. 2F). The crystals composing these
aggregates have larger dimensions (up to 100 mm across
and 5 mm thick) and are the only ones to have distin-
guishable crystal forms: in addition to the basal pinacoid
{0001}, only {100} prisms could be identied.
PH y s i C a l Pr o P e r t i e s
Some of the samples uoresce slightly in short wave
UV light (l = 254 nm), the response color being a light
blue-violet. None of them uoresces in long wave (366
nm) ultraviolet radiation.
Because of the small size of the crystals, no direct
measurement of the indices of refraction could be made.
For this reason, they were measured on thin, platy
bundles of crystals that show hexagonal peripheral faces
in monochromatic light (l = 589 nm) by immersion in
Cargille oils. The values were determined as minimum
and maximum values for separates from 10 selected
samples (the rst ten in Table 1), assuming that nmin
=  and nmax = v. The indices of refraction obtained
were  = 1.645(2) and v = 1.653(1), resulting in a mean
index of refraction n = (2v + )/3 of 1.6503 for use in
Gladstone–Dale calculations.
The mean density of the same representative samples
of carefully hand-picked apatite-(CaOH) from Cioclo-
vina was measured by the sink-oat method at 25°C,
using a mixture of methylene iodide and toluene as
immersion liquid. The average of ve determinations
yielded Dm = 3.17(2) g/cm3, which compares well
with the calculated densities in Table 1, based on the
chemical formulae (Table 2) and unit-cell parameters
(Table 1), accepting Z = 1 (PDF 86–1201).
Gladstone–Dale calculations with the constants of
Mandarino (1981) gave compatibility indices indicating
excellent compatibilities in all cases (Table 1).
CH e M i C a l Da t a
Preliminary SEM–EDS analyses showed Ca and P
as principal elements, with minor S, Mg, Fe, Mn, K
and Na.
Results of ICP–AES analyses of 20 selected samples
are presented in Table 2. The H2O contents were calcu-
lated so as to fulll charge balance, and the totals were
recalculated to 100 wt.%. The amount of CO2 was
ignored. The formulae were normalized on the basis
of 6 (S + P) and 26 (O,OH) per formula unit (pfu), as
recommended by Fiore & Laviano (1991). This basis
is appropriate for emphasizing the Ca-deciency of our
samples, that probably belong to a series of general
formula Ca10–x(HPO4)x(PO4)6–x(OH)2–x, where 0 ≤ x ≤ 1
(Berry 1967, Elliott 1994) and presumably also contain
(CO3)2– and (SO4)2– groups.
A few remarks must be drawn on the basis of the
results in Table 2: (1) No signicant compositional
differences occur between the various textural types of
apatite-(CaOH). (2) As expected, all but six samples in
Table 2 are Ca-oversaturated, showing seven-fold plus
nine-fold cation totals higher than the theoretical value
of 10 apfu. If we assume that the CO2 content in each
sample is 0.81 wt.%, as determined by wet-chemical
analysis of the purest sample, we must consider
463_vol_46-2_art_15.indd 466 08/04/08 12:05:13

a P at i t e -(Cao H ) i n fossil B a t GUano DePosit, roMania 467
fiG . 2. SEM photomicrographs of typical aggregates of apatite-(CaOH). A. Densely packed aggregate of subspherical shape.
B. Reticulate network of apatite-(CaOH) aggregates, probably a fossil fragment of spongy bone or of another tissue. C.
Rosette-like radial aggregates of apatite-(CaOH). Individual crystals, roughly hexagonal, are clearly visible. D. Rosette-like
aggregates of apatite-(CaOH) overgrown on platy crystals of illite. E. Aggregates of randomly oriented interlocking and
subparallel platy crystals of apatite-(CaOH). The arrow indicates a perfectly shaped crystal. F. Fractured aggregate of tabular
crystals of apatite-(CaOH) intergrown on (0001), representing probably a fragment of a fossil cortical bone.
463_vol_46-2_art_15.indd 467 08/04/08 12:05:14

468 tHe CanaDian MineraloGist
that part of the Ca (about 0.02 apfu in equivalent)
is consumed for charge-balancing carbonate. In this
case, all our samples are in reality Ca-decient, which
suggests that (HPO4)2– groups partially replace (PO4)3–.
However, apatite-(CaOH) samples in Table 2 contain
less than 2.26% of the protonated phosphate groups,
the maximum value being recorded in Sample D 18 D.
(3) Sulfate groups, already reported in apatite-(CaOH)
from caves (e.g., Fiore & Laviano 1991), also substitute
for phosphate in apatite-(CaOH) from Cioclovina: S
accounts for 0.32% to 4.66% of the tetrahedral sites
normally occupied by P. Consequently, we could
imagine that coupled substitutions such as (PO4)3– +
(OH)– 7 (SO4)2– + O2– or (PO4)3– + (OH)– 7 (SO4)2– +
(CO3)2– play an important role. (4) The substitutions
in the cation sites normally occupied by Ca are less
important. The main substitution is Na-for-Ca, but the
cumulative substitution in these sites accounts for only
0.71 to 4.07% (mean 2.07%). Other potential cations
substituting for Ca were evaluated with the ICP–AES
results, but were found as minor elements, as shown by
the results in Table 3.
As in the case of brushite (Dumitraş et al. 2004),
apatite-(CaOH) from Cioclovina contains less Ba
than Sr, which is quite normal in apatite-(CaOH) in
fossil bone [see Trueman & Tuross (2002) and refer-
ences therein], but also encountered in the authigenic
samples. The REE contents are relatively low. On the
basis of results of partial analyses in Table 3, we can
state that apatite-(CaOH) from Cioclovina is enriched
in the light REE relative to the heavy REE + Y; all but
three samples respect this rule, which also parallels the
behavior of brushite (Dumitraş et al. 2004). In addi-
tion to these elements, trace concentrations of some
transitional metals (i.e., Cu, V, Cr and Ni) were also
identied. The higher concentration of these elements
in two of the analyzed samples (D 36A and D 38C B),
is probably due to admixed X-ray amorphous phases
and correlates with higher contents of manganese and
iron.
Attempts to check the accuracy of the conclusions
drawn from the results of bulk ICP–AES analyses were
done using EMPA. We chose for study a fragment of
fossil bone engulfed in the guano mass, which consists
of pure apatite-(CaOH). Representative results for this
sample (D 40 A, equivalent of D 40 B in Table 2) are
summarized in Table 4. All iron was considered to be
divalent, and the amount of H2O was calculated to fulll
charge-balance conditions. A parallel wet-chemical
analysis was carried out on the same sample (Table
4). A few conclusions may be drawn: (1) The propor-
tions of halogens in the EMP analyses are probably
463_vol_46-2_art_15.indd 468 08/04/08 12:05:15

a P at i t e -(Cao H ) i n fossil B a t GUano DePosit, roMania 469
463_vol_46-2_art_15.indd 469 08/04/08 12:05:18

470 tHe CanaDian MineraloGist
overestimated owing to analytical interferences with
the araldite used for xing the samples on the support;
it does not seem possible that uorine and chlorine
occupy up to 34.77% and 6.44%, respectively, of the
positions normally occupied by hydroxyl, even if the
analyzed apatite-(CaOH) represents a bone fragment.
The wet-chemical analysis in Table 4 seems more accu-
rate on this point. (2) EMP analyses reveal insignicant
variations in composition among different micromounts.
(3) The chemical peculiarities established by ICP–AES
analyses are conrmed: up to 2.32% of the tetrahedral
positions are occupied by sulfur, and the incorpora-
tion of Na, K, Mn, Fe, Mg in the Ca sites is minor,
summing up to 2.11%. (4) Because the EMP analysis
does not offer any indication for the incorporation of
(CO3)2– into the apatite-(CaOH) lattice, it is difcult to
evaluate the deciency in Ca of the analyzed sample;
accepting, however, a CO2 content of 0.81 wt.% as
determined by wet-chemical analysis, all but one EMP
datassets in Table 4 are representative of Ca-decient
apatite-(CaOH).
X-r a y Po w D e r Da t a
X-ray powder diffractometry was carried out on
about 100 powders of apatite-(CaOH) from Cioclovina
in order to refine the cell parameters. The primary
patterns are available upon request from the rst author.
the index of crystallinity calculated by the method
recommended by Simpson (1964), involving the ratio
between the total width of the diffraction peaks (211)
and (112) at their half height and the greatest height
463_vol_46-2_art_15.indd 470 08/04/08 12:05:19

a P at i t e -(Cao H ) i n fossil B a t GUano DePosit, roMania 471
of the two peaks, vary between 0.037 and 0.158. The
indices of crystallinity of the samples whose chemical
analyses are given in Table 2 are reported in Table 1.
Only about 12% of the analyzed samples have high
indices of crystallinity, indicating that as a whole
they are poorly crystalline (Simpson 1964). The unit-
cell parameters were successfully rened based on a
hexagonal P63/m cell, after indexing the patterns in
analogy with PDF 86–1199, 86–1201 and 86–1203.
Sets of 36 to 60 reections in the 2u range between
10 and 70° (CuKa, l = 1.54056 Å) for which unam-
biguous indexing was possible were used to rene the
cell dimensions of 20 representative samples, which
are summarized in Table 1. The unit-cell parameters
calculated as an average of these 20 datasets of values
are a 9.436(13), and c 6.868(8) Å, where the errors in
the brackets correspond to the standard deviations of the
mean (2s) for each set of data. These values, as well as
the individual values in Table 1, and particularly c, differ
signicantly from those determined by various authors
for synthetic apatite-(CaOH) of stoichiometric compo-
sition [e.g., a = 9.4176(5), c 6.8814(5) Å according to
Elliott (1994), a 9.421(2), c 6.882(2) Å according to
Brunet et al. (1999), and a 9.4302(5), c 6.8911(2) Å
according to Dorozkhin & Epple (2002)].
The unit-cell dimensions of carbonate-bearing
apatite-(CaOH) depend, to a rst approximation, on the
incorporation of (CO3)2– in the structure (Elliott 1994).
From this point of view, our samples generally behave
like the A-type carbonated apatite-(CaOH), which are
characterized by increased a unit-cell parameters due to
carbonate substituting for hydroxyl group in the struc-
tural channels (Elliott 1994). The presence of carbonate
in the analyzed samples could explain why the c value,
6.867(5) Å, does not differ signicantly from that deter-
mined by Neweseley (1963) for a synthetic “carbonate–
hydroxylapatite”. As apatite-(CaOH) samples from
Cioclovina have uniformly low chlorine and uorine
contents, the occupancy by halogens of the X sites in
the structure (notation after Hughes et al. 1989, and
Brunet et al. 1999), could not explain the change in
lattice parameters.
tH e r M a l De C o M P o s i t i o n
The differential thermal analysis and differential
thermogravimetric curves of apatite-(CaOH) are gener-
ally smooth and featureless until 1000°C, in spite of
the losses of H2O and CO2 recorded on the thermogra-
vimetric curve (Fiore & Laviano 1991, Ivanova et al.
2001). This was the reason for recording the thermal
curves of apatite-(CaOH) from Cioclovina using a
DSC–TG–TGA coupling, in spite of the lower limits
of working temperature (up to 600°C). The thermal
curves recorded for a representative sample of authi-
genic apatite-(CaOH) from Cioclovina (D 301) are
given in Figure 3. On the basis of previous studies on
the thermal decomposition of the Ca-decient carbon-
ated apatite-(CaOH) (Elliott 1994, Ivanova et al. 2001),
we can presume that the main effects on the DTG and
DSC curves are due to the following causes: (1) the
loss of surface-bound H2O (–1.43 wt.% on the TGA
curve), marked by an endothermic effect at ~97°C, (2)
the loss of channel H2O (–3.61 wt.% cumulative on the
TGA curve), marked by a small endothermic effect at
~198°C, and (3) the start of the carbonate migration (see
below), marked by a small endothermic effect on the
DTG curve at ~520°C, accompanied by an exothermic
effect on the DSC curve. The loss of the lattice H2O is
continuous, whereas the loss of carbonate starts after
fiG . 3. Thermal curves recorded for a apatite-(CaOH) sample
from Cioclovina: differential thermogravimetric (top),
differential scanning calorimetric (middle) and thermo-
gravimetric (bottom).
463_vol_46-2_art_15.indd 471 08/04/08 12:05:20

472 tHe CanaDian MineraloGist
520°C. The TGA curve gradually declines, the total
loss-in-weight at 600°C being of 6.62 wt.%.
An XRD study of the breakdown product at 600°C,
after cooling, reveals that only a phase with the apatite-
(CaOH) pattern persists, without any supplementary
crystalline associated phase. According to Brunet et al.
(1999), the breakdown of Ca-decient apatite-(CaOH)
at high temperature must, however, result in the growth
of b-Ca3(PO4)2 according to the reaction:
Ca10–x(HPO4)x(PO4)6–x(OH)2–x ! (1–x)
Ca10(PO4)6(OH)2 + 3x Ca3(PO4)2 + x H2O.
In fact, no traces of a Ca3(PO4)2 phase were detected,
even after heating of many selected powdered samples
of apatite-(CaOH) from Cioclovina at 600°C in a resis-
tance furnace; these were held at this temperature for
24 hours in order to favor the loss of H2O, then cooled
down to 25°C. The breakdown products, analyzed by
XRD, gave patterns similar with those of the starting
materials. This nding apparently disagrees with our
results indicating a substantial loss of H2O, with the
results of Berry (1967), which indicate the dehydration
of Ca-decient apatite-(CaOH) at 600°C, and also with
the results of the thermal expansion measurements of
Brunet et al. (1999), which indicate a breakdown at
about 550°C. In fact, it is more reasonable to accept that
the loss of H2O due to the breakdown of the protonated
phosphate in our sample can be explained by another
reaction, proposed by Mortier et al. (1989):
Ca10–x(HPO4)x(PO4)6–x(OH)2–x !
Ca10–x(P2O7)x–z(PO4)6–2x+2z(OH)2(1–z) + (x +z) H2O.
A weak and broad band centered at ~735 cm–1 in the
FTIR spectra of the heated samples (not given, but
available upon request from the rst author) can be
assigned to (P2O7)4– groups, indicating the polymeriza-
tion of some of the structural orthophosphate groups,
without structural breakdown.
As proven by the differential weighting, the mass
loss continues until 1000°C, involving a continuous
dehydration and the expelling of carbonate: the total
loss in mass of sample D 301 heated to 1000°C is 7.43
wt.%. X-ray-assisted heating of this sample shows,
however, that a phase with a apatite-(CaOH) pattern still
persist at 900°C, allowing the renement of its unit-cell
parameters (Fig. 4, Table 5). Taking into account the
dehydration, this phase must be an “oxyapatite”, whose
crystal structure is identical with that of apatite-(CaOH)
(Alberius-Henning et al. 1999), generating a similar
XRD pattern (PDF 01–089–6495). This inference
agrees with the hypothesis that by heating the (OH)–
ions in the apatite-(CaOH), anion channels are replaced
by O2– through the scheme 2 (OH)– ! O2– + □, where
□ is a vacancy, without any disruption of other struc-
tural constituents (Alberius-Henning et al. 1999).
As expected, and in agreement with experimental
studies on Ca-deficient carbonated apatite-(CaOH)
(Ivanova et al. 2001) and with studies of the thermal
expansion of stoichiometric apatite-(CaOH) (Brunet et
al. 1999), both a and c unit-cell parameters generally
dene an increasing trend with increasing temperature
(Table 5). Two contractions of the cell can be observed,
however: (1) at up to 250°C, owing to the loss of the
channel H2O, and (2) at 550–600°C, owing to the loss
of carbonate (cf. Ivanova et al. 2001). The increase of
a and the striking decrease of c between 600 and 650°C
agree with the changes in the location of (CO3)2– from
(OH)– to (PO4)3– sites, in a process contrary to that
inferred by Ivanova et al. (2001) in the B-type carbon-
ated apatite-(CaOH).
Whitlockite occurs as breakdown product at 700°C,
the transformation being complete at 1000°C. The
resulting whitlockite [a 10.438(8), c 37.51(7) Å at
1000°C and a 10.430(6), c 37.47(7) Å after cooling
at 25°C] is, as proven by ion chromatography, sulfate-
bearing, indicating that sulfate was not (entirely)
expelled at 1000°C.
in f r a r e D aB s o r P t i o n Da t a
Figure 5 displays the infrared absorption spectra
obtained for two representative samples of apatite-
(CaOH) from Cioclovina with and without a Fourier-
transform spectrometer, whereas Table 6, give the
wavenumbers, characters and intensities of the infrared
absorption bands as well as attempts to assign the bands
on the basis of the works of Baddiel & Berry (1966),
Bhatnagar (1968), LeGeros et al. (1970) and Fowler
(1973). The positions of the bands in Table 6 were
established as the mean of the wavenumbers recorded
for similar bands in ve (IR) and, respectively, nine
(FTIR) different spectra of representative samples, with
standard deviations given into brackets. The primary
records are available on request from the rst author.
As the band assignment is generally agreed upon, only
the following remarks are considered important:
463_vol_46-2_art_15.indd 472 08/04/08 12:05:20

a P at i t e -(Cao H ) i n fossil B a t GUano DePosit, roMania 473
(1) Although both the mineral powder and the KBr
were previously stored in a desiccator, both IR and
FTIR spectra clearly show absorption bands due to
molecular H2O (Table 6). This may be due to the very
fine-grained nature of our samples, whose specific
surfaces are large enough to adsorb molecules of H2O,
or to the presence of molecular H2O in the structure
of the analyzed samples, if they are Ca-decient. In
fact, in Ca-decient apatite-(CaOH), the protonation
of some of the phosphate groups respects a charge-
balance scheme [Ca2+ + (OH)– + (PO4)3–] !
← (HPO4)2–,
resulting in vacancies in both Ca and hydroxyl sites. It
is very likely that molecules of H2O occupy some of
the vacated hydroxyl sites, being present in the c*-axis
anion channels (ivanova et al. 2001, Elliott 2002). The
presence in our samples of bands due to molecular H2O
also agrees with the result of the wet-chemical analysis
(Table 4).
(2) A band at ~3570 cm–1 is characteristic for the
O–H stretching involving the hydroxyl group in apatite.
According to the bond distance – frequency correlation
of Libowitzky (1999) [nO–H = 3632 – 1.79 3 106• exp
(–d/0.2146)], this frequency corresponds to a H...O
distance of 0.957 Å, identical to that determined by Kay
et al. (1964) for the O–H distance in the (OH) group
of apatite-(CaOH). The corresponding librations are
expressed by the band at 635 cm–1.
fiG . 4. X-ray powder patterns of a representative sample of apatite-(CaOH) from Cioclovina, as a function of temperature.
Representative lines of austenitic steel (a.s.), whitlockite (Wh) and apatite-(CaOH) (Hap) are marked on the gure.
463_vol_46-2_art_15.indd 473 08/04/08 12:05:21

474 tHe CanaDian MineraloGist
(3) The splitting into two fundamental vibrations
of the antisymmetric stretching of the carbonate
group (bands at ~1456 and ~1420 cm–1) agrees with
the assumption that carbonate groups in our samples
occupy two structural positions (e.g., Brophy & Nash
1968, Elliott 2002). The corresponding bending seems
unsplit, marked by a band centered at ~873 cm–1. In fact,
the out-of-plane bending of carbonate must in turn be
doubly degenerate owing to the presence of carbonate
groups in two different structural environments, but a
second band, normally recorded at ~880 cm–1, is too
weak to be observed (Elliott 2002). The n2 bending of
carbonate obscures the P–O(H) symmetric stretching
of the (HPO4)2– groups, which must be expressed by
a band centered at practically the same wavenumber
(Berry 1967, Ishikawa et al. 1993). In turn, the (P)O–H
antisymmetric stretching at ~2930 cm–1 is too weak to
be observed as a shoulder of the broad band due to the
stretching of the molecular H2O.
(4) The tetrahedral phosphate anion seems to
generate nine vibrational modes, whose corresponding
bands are tentatively assigned in Table 6. It is unclear
if the band at ~335 cm–1 corresponds to the n2 out-of-
plane bending of the phosphate group (Badiel & Berry
1966, Bhatnagar 1968) or is due to a Ca–OH motion
(Fowler 1973), or if the band at ~261 cm–1 corresponds
to the n2’ out-of-plane bending of the phosphate group
(Badiel & Berry 1966, Bhatnagar 1968) or to a lattice
mode (Fowler 1973), but the band multiplicity (3n3 +
1n1 + 3n4 + 2n2) is consistent with a C6 point symmetry
of the phosphate anion.
(5) Shoulders toward the high-wavenumber side
of the large complexes of bands at 900–1200 cm–1
(the phosphate-stretching region) and 500–650 cm–1
fiG . 5. Infrared spectra of two selected samples of apatite-(CaOH) from Cioclovina: FTIR spectrum (top) and IR spectrum
(bottom).
463_vol_46-2_art_15.indd 474 08/04/08 12:05:22

a P at i t e -(Cao H ) i n fossil B a t GUano DePosit, roMania 475
(phosphate in-plane bending region), potentially due to
the presence of sulfate groups in the structure, are too
poorly resolved. Such bands were reported by Baumer
et al. (1990) in sulfate-substituted uorapatite samples,
but are not evident in our case.
Ge n e t i C Co n s i D e r at i o n s
Textural evidence, as well as parallels with other
similar deposits, show that apatite-(CaOH) from
Cioclovina formed by two complementary processes,
namely: (1) by authigenesis, as reaction products
between carbonate (basement, boulders or moonmilk
ows) and acid solutions derived from guano, and (2)
by sedimentation followed by diagenesis of organic
relics such as bones, teeth and other tissues, in one
of the richest deposits of fossil fragments in Europe.
Both authigenic and biogenic apatite-(CaOH) from
Cioclovina are carbonate- and sulfate-bearing and
Ca-decient. The lack of major physical, crystallo-
graphic and chemical differences between the two kinds
of apatite-(CaOH) from Cioclovina is an argument in
favor of re-equilibration during diagenesis.
An XRD analysis coupled with SEM–EDS, as well
as textural evidence, show that the authigenic apatite-
(CaOH) had a gel-like, X-ray-amorphous precursor.
The nature of this precursor (“ACP phase”) is unclear,
but we presume that it is identical to that observed
during the growth of synthetic apatite-(CaOH), having
a composition like Ca3(PO4)2•nH2O (Francis & Webb
1971, Abbona & Franchini-Angela 1990). This phase is
unstable and normally transforms into the more stable
phases octacalcium phosphate [Ca4H(PO4)3•2.5H2O]
and apatite-(CaOH) (Abbona & Franchini-Angela
1990). As shown by Christoffersen et al. (1990), the
octacalcium phosphate is the intermediate phase in the
formation of apatite-(CaOH) via heterogeneous nucle-
ation at increased supersaturation, which could explain
the presence at Cioclovina of the spherical aggregates
of densely matted material apparently crystallized from
gels. The authigenic crystallization of apatite-(CaOH)
by direct precipitation, as well as the recrystallization of
biogenic fragments during diagenesis, were controlled
by values of pH higher than 6.93 (Elliott et al. 1959) or
6.2 (Simpson 1964), this pH value essentially depending
on the saturation of the mother solution. At low ranges
of pH, the calcium phosphate expecxted to precipitate is
brushite (Elliott et al. 1959, Simpson 1964). As proved
by the frequency of the inclusions of quartz and illite in
apatite-(CaOH) aggregates, the nucleation of Ca phos-
phates and the subsequent growth of apatite-(CaOH)
seed crystals were stimulated by the pre-existence of
silicate nuclei.
The contents of REE + Y reported previously (Table
3) are in the range of 2.95 to 47.11 ppm (average
12.79 ppm), even lower than those recorded by us in
ten selected samples of carbonates from the limestone
support and from the solidified flows of moonmilk
on the walls, which range from 21.00 to 34.19 ppm
(average 25.28 ppm). The classical hypothesis of the
formation of authigenic apatite-(CaOH) by direct reac-
tion between the phosphate solutions derived from the
guano mass and the carbonates in the cave is conse-
quently fully conrmed.
The coexistence of apatite-(CaOH) with brushite
is essentially a function of the local variations of pH.
The action of magnesium as local inhibitor of apatite-
(CaOH) crystallization (Abbona & Franchini-Angela
1990) is excluded since the magnesium phosphates
are lacking in the fossil guano, and in the surrounding
limestones the level of concentration of magnesium
is low. The inhibitor of the crystal growth of apatite-
(CaOH) must in fact be the local excess of carbonate
ions resulting from the breakdown of carbonates due to
the acidity of solutions derived from the guano mass.
This may explain the extensive development of brushite
in the vicinity of the gel-like aggregates of Ca phosphate
preceding apatite-(CaOH). The excess of sulfate in the
system and the low values of pH (up to 5.5) destabi-
lize both brushite and apatite-(CaOH), favoring the
crystallization of ardealite or even gypsum (Rinaudo
& Abbona 1988). This environment could explain the
local occurrence of sprays and crusts of ardealite and
gypsum directly on apatite-(CaOH).
aC k n o w l e D G e M e n t s
This research was supported by the cooperative
ANCS – DRI research program under project 8/2005
between the Romanian and the Walloon governments,
by two MIRA grants awarded by the Rhône-Alpes
Region in France to the rst two authors, and by the
CERES project No. 4–153/04.11.2004 of the Roma-
nian Ministry of Education and Research. Support
from Ecole Nationale Supérieure des Mines de Saint-
Etienne (ENSMSE, France) is gratefully appreciated.
F.H. acknowledges the FNRS (Belgium) for the award
of a position of Chargé de Recherches in 2004–2006
and for grant 1.5.112.02. The authors are indebted to
Dr. Jacques Moutte (ENSMSE) for assistance with
ICP–AES analysis, to Mr. Olivier Valfort (ENSMSE)
for technical assistance during the thermally assisted
X-ray powder work, to Mrs. Gabriela Stelea (Geological
Institute of Romania, Bucharest) and Mr. Pierre Lefèvre
(University of Liège, Belgium) for recording the
infrared absorption spectra, to Miss Véronique Bourgier
for recording the thermal curves, and to Mr. Jean-Pierre
Poyet (ENSMSE) for the ion-chromatographic analysis.
Field assistance by Mr. Bogdan Tomuş is gratefully
acknowledged. The quality of this contribution was
improved after reviews by Dr. Clive Trueman and an
anonymous referee. Robert F. Martin and Associate
Editor Louis Raimbault are gratefully acknowledged for
handling the manuscript.
463_vol_46-2_art_15.indd 475 08/04/08 12:05:22

476 tHe CanaDian MineraloGist
re f e r e n C e s
aB B o n a , f. & fr a n C H i n i -an G e l a , M. (1990): Crystallization
of calcium and magnesium phosphates from solutions of
low concentration. J. Crystal Growth 104, 661-671.
al B e r i U s -He n n i n G , P., la n D a -Ca n o va s , a., la r s s o n , a.-k.
& li D in , s. (1999): Elucidation of the crystal structure of
oxyapatite by high-resolution electron microscopy. Acta
Crystallogr. B 55, 170-176.
aPPleMan, D.e. & ev a n s , H.T., Jr. (1973): Indexing and
least-squares renement of powder diffraction data. U.S.
Geol. Surv., Comput. Contrib. 20 (NTIS Doc. PB-216).
Ba D D i e l , C.B. & Be r r y , e.e. (1966): Spectra structure cor-
relations in hydroxy and uorapatite. Spectrochim. Acta
22, 1407-1416.
Balenzano, f., Dell’anna, l. & Di Piero, M. (1974):
Ricerche mineralogiche su alcuni fosfati rinvenuti nelle
grotte di Castellana (Bari): strengite aluminifera, vivianite,
taranakite, brushite e idrossiapatite. Rend. Soc. Ital. Min-
eral. Petrogr. 30, 543-573.
BaUMer, a., Ca r U B a , r. & GanteaUMe, M. (1990): Car-
bonate-uorapatite: mise en évidence de la substitution
2PO43– ! SiO44– + SO42– par spectrométrie infrarouge.
Eur. J. Mineral. 2, 297-304.
Benoit, P. H. (1987): Adaptation to microcomputer of the
Appleman–Evans program for indexing and least-squares
renement of powder-diffraction data for unit-cell dimen-
sions. Am. Mineral. 72, 1018-1019.
Be r r y , e.e. (1967): The structure and composition of some
calcium-deficient apatites. J. Inorg. Nucl. Chem. 29,
317-327.
BH a t n a G a r , v.M. (1968): Infrared spectra of hydroxyapatite
and uorapatite. Bull. Soc. Chim. France, 1771-1773.
Br o P H y , G.P. & na s H , t.J. (1968): Compositional, infrared
and X-ray analysis of fossil bone. Am. Mineral. 53,
445-454.
BrUnet, f., allan, D.r., reDfern, a.t.s., anGel, r.J.,
Mi l e t i C H , r., re i C H M a n n , H.J., se r G e n t , J. & Ha n f l a n D ,
M. (1999): Compressibility and thermal expansivity of
synthetic apatites, Ca5(PO4)3X with X = OH, F and Cl. Eur.
J. Mineral. 11, 1023-1035.
BU r k e , e.a.J. (2008): Tidying up mineral names: an IMA–
CNMNC scheme for suffixes, hyphens and diacritical
marks. Mineral. Rec. 39, 131-135.
CH r i s t o f f e r s e n , M.r., CH r i s t o f f e r s e n , M.r. & ki B a l C z y C ,
w. (1990): Apparent solubilities of two amorphous calcium
phosphates and of octacalcium phosphate in the tempera-
ture range 30–42°C. J. Crystal Growth 106, 349-354.
ConstantinesCU, e., MarinCea, Ş. & CrăCiun, C. (1999):
Crandallite in the phosphate association from Cioclovina
cave (Şureanu Mts., Romania). In Scientic works by Emil
Constantinescu. Mineralogy in the system of Earth sci-
ences. Imperial College Press, London, U.K. (1-5).
Di a C o n u , G. & Me D e ş a n , a. (1975): Spéléothèmes de dahllite
dans la grotte “Peştera Muierii”, Baia de Fier – Roumanie.
Trav. Inst. Spéol. “Emile Racovitza” 14, 149-156.
DorozHkin, s.v. & ePPle M. (2002): Die biologische und
medizinische Bedeutung von Calciumphosphaten. Angew.
Chem. 114, 3260-3277.
Du M i t r a ş D.G. & Ma r i n C e a , Ş. (2004):
DuMitraş D.G., MarinCea, Ş. & fransolet, a.M. (2004):
Brushite in the bat guano deposit from the “dry” Cioclo-
vina Cave (Şureanu Mountains, Romania). Neues Jahrb.
Mineral., Monatsh., 45-64.
el l i o t , J.s., sH a r P , r.f. & le w i s , l. (1959): The effect of the
molar Ca/P ratio upon the crystallization of brushite and
apatite. J. Phys. Chem. 63, 725-726.
elliott, J.C. (1994): Structure and Chemistry of the Apa-
tites and the Other Calcium Orthophosphates. Elsevier,
Amsterdam, The Netherlands.
elliott, J.C. (2002): Calcium phosphate biominerals. In
Phosphates – Geochemical, Geobiological and Materials
Importance (M.L. Kohn, J. Rakovan & J.M. Hughes, eds.).
Rev. Mineral. Geochem. 48, 427-454.
fiore, s. & laviano, r. (1991): Brushite, hydroxylapatite,
and taranakite from Apulian caves (southern Italy): new
mineralogical data. Am. Mineral. 76, 1722-1727.
fowler, B.o. (1973): Infrared studies of apatites. I. Vibra-
tional assignments for calcium, strontium and barium
hydroxyapatites utilizing isotopic substitution. Inorg.
Chem. 13, 194-207.
fr a n C i s , M.D. & we B B , n.C. (1971): Hydroxyapatite forma-
tion from a hydrated calcium monohydrogen phosphate
precursor. Calc. Tiss. Res. 6, 335-342.
HU G H e s , J.M., Ca M e r o n , M. & Crowley, k.D. (1989):
Structural variations in natural F, OH and Cl apatites. Am.
Mineral. 74, 870-876.
io s o f , V. & ne a C ş u , V. (1980): Analysis of silicate rocks by
atomic absorption spectrometry. Rev. Roum. Chim. 25,
589-597.
is H i k a wa , k., DU C H e y n e , P. & ra D i n , s. (1993): Determina-
tion of the Ca/P ratio in calcium-decient hydroxyapatite
using X-ray diffraction analysis. J. Mater. Sci., Mater.
Med. 4, 165-168.
iva n o v a , t.i., frank-kaMenetskaya o.v., kaltsov, v. &
UG o l k o v , l. (2001): Crystal structure of calcium-decient
carbonated hydroxylapatite. Thermal decomposition. J.
Solid State Chem. 160, 340-349.
ka y , M.i., yo U n G , r.a. & Po s n e r , a.s. (1964): Crystal struc-
ture of hydroxylapatite. Nature 204, 1050-1052.
463_vol_46-2_art_15.indd 476 08/04/08 12:05:23

a P at i t e -(Cao H ) i n fossil B a t GUano DePosit, roMania 477
leGeros, r.z., leGeros, J.P., traUtz, o.r. & klein, e.
(1970): Spectral properties of carbonate in carbonate-
containing apatites. Develop. Appl. Spectrosc. 7 B, 3-12.
liBowitzky, e. (1999): Correlation of O–H stretching fre-
quencies and O–H...O hydrogen bond lengths in minerals.
Monatsh. Chem. 130, 1047-1059.
ManDarino, J.a. (1981): The Gladstone–Dale relationship.
IV. The compatibility concept and its application. Can.
Mineral. 19, 441-450.
MarinCea, Ş. & DuMitraş, D. (2003): The occurrence
of taranakite in the “dry” Cioclovina Cave (Sureanu
Mountains, Romania). Neues Jahrb. Mineral., Monatsh.,
127-144.
Ma r i n C e a , Ş. & Du M i t r a ş , D. (2005): First reported sedimen-
tary occurrence of berlinite (AlPO4) in phosphate-bearing
sediments from Cioclovina Cave, Romania – Comment.
Am. Mineral. 90, 1203-1208.
Ma r i n C e a , Ş., Du M i t r a ş , D. & Gi b e r t , r. (2002): Tinsley-
ite in the “dry” Cioclovina Cave (Sureanu Mountains,
Romania): the second world occurrence. Eur. J. Mineral.
14, 157-164.
Mo r t i e r , a., leMaitre, J. & ro U X H e t , P.G. (1989):
Temperature-programmed characterization of synthetic
calcium-decient phosphate apatites. Thermochim. Acta
143, 265-282.
ne w e s e l e y , H. (1963): Kristallchemische und micromorphol-
ogische Untersuchungen an Carbonat-Apatiten. Monatsh.
Chem. 94, 270-280.
on a C , B.P., BreBan, r., kearns, J. & taMas, t. (2002):
Unusual minerals related to phosphate deposits in Cioclo-
vina Cave, Sureanu Mountains (Romania). Theor. Appl.
Karst. 15, 27-34.
on a C , B.P., et t i n G e r , k., ke a r n s , J & Ba l a s z , i.i. (2005): A
modern, guano-related occurrence of foggite, CaAl(PO4)
(OH)2•H2O and churchite-(Y), YPO4•2H2O in Cioclovina
Cave, Romania. Mineral. Petrol. 85, 291-302.
on a C , B.P. & wH i t e , w.B. (2003): First reported sedimentary
occurrence of berlinite (AlPO4) in phosphate-bearing sedi-
ments from Cioclovina Cave, Romania. Am. Mineral. 88,
1395-1397.
ri n a U D o , C. & aB B o n a , f. (1988): A contribution to the study
of the crystal chemistry of calcium sulfate phosphate
hydrate. Mineral. Petrogr. Acta 31, 95-105.
sCHaDler, J. (1929): Mineralogische-petrographische Char-
acteristik der Phosphat-Ablagerung in the Cioclovinahöhle
bei Pui. Pub. Muz. Hunedoara 5(27), 1-3.
sCHaDler, J. (1932): Ardealit, ein neues Mineral
CaHPO4•CaSO4+4H2O. Zb. Mineral. A, 40-41.
sHaPiro, l. & BrannoCk, w. w. (1962): Rapid analysis of
silicate, carbonate and phosphate rocks. U. S. Geol. Surv.
Bull. 1144-A (A 14-15, A 49-51).
si M P s o n , D.r. (1964): The nature of alkali carbonate apatites.
Am. Mineral. 49, 363-376.
to M U ş, r.B. (1999): The Karst Complex Ciclovina. The Basin
2063. Proteus, Hunedoara, Romania (in Romanian).
tr U e M a n , C.n. & tU r o s s , n. (2002): Trace elements in recent
and fossil bone apatite. In Phosphates – Geochemical,
Geobiological and Materials Importance (M.L. Kohn, J.
Rakovan & J.M. Hughes, eds.). Rev. Mineral. Geochem.
48, 489-522.
Received February 27, 2007, revised manuscript accepted
January 15, 2008.
463_vol_46-2_art_15.indd 477 08/04/08 12:05:23

463_vol_46-2_art_15.indd 478 08/04/08 12:05:23