Thermoanalytical and structural characterization of fluoridated calcium phosphates prepared in anhydrous alcohols

Article abstract of DOI:10.1007/s10973-009-0654-7

A thermoanalytical, morphological, and structural study of fluoridated calcium phosphates that were prepared by different variants of a synthesis in anhydrous alcohols is reported. The obtained materials were neither fully amorphous nor single-phased crystalline, and their nature considerably depended on the synthesis conditions. In all cases, the retention of significant amounts of solvent in the solid product was observed. A complete removal of the solvent was only possible by heating to temperatures above ～573-673 K which resulted in variations in the elemental composition, phase changes, and an increase of the crystallinity. Consequently, this synthesis in anhydrous alcohols is not a viable route to obtain materials with a defined crystallinity and stoichiometry.

Full text of DOI:10.1007/s10973-009-0654-7

J Therm Anal Calorim (2010) 100:509–517
DOI 10.1007/s10973-009-0654-7
Thermoanalytical and structural characterization of fluoridated
calcium phosphates prepared in anhydrous alcohols
•
•
Ricardo G. Simo˜es Ana I. Aleixo
•
Ana L. C. Lagoa Manuel E. Minas da Piedade
•
•
•
Joa˜o P. Leal Thorsten Peitsch Matthias Epple
Portuguese Special Chapter
Received: 17 February 2009 / Accepted: 20 March 2009 / Published online: 12 January 2010
´
´
ꢀ Akademiai Kiado, Budapest, Hungary 2010
Abstract A thermoanalytical, morphological, and struc-
tural study of fluoridated calcium phosphates that were
prepared by different variants of a synthesis in anhydrous
alcohols is reported. The obtained materials were neither
fully amorphous nor single-phased crystalline, and their
nature considerably depended on the synthesis conditions.
In all cases, the retention of significant amounts of solvent
in the solid product was observed. A complete removal of
the solvent was only possible by heating to temperatures
above *573–673 K which resulted in variations in the
elemental composition, phase changes, and an increase of
the crystallinity. Consequently, this synthesis in anhydrous
alcohols is not a viable route to obtain materials with a
defined crystallinity and stoichiometry.
Introduction
Calcium phosphates are the inorganic component of human
hard tissues, e.g. bone and teeth [1–4]. This has fostered
many research efforts to develop biomimetic calcium
phosphate materials to be used in medical applications,
such as the repair of bone fractures, tooth restoration, or the
surface coating of metallic implants to improve their bio-
compatibility [5–8].
The mineral phase of human hard tissues is a non-
stoichiometric hydroxyapatite, whose deviation from the
stoichiometric formula, Ca10(PO4)6(OH)2, and the crystal-
lite size depend on the biological function [1, 2, 5, 9–12].
The bone mineral, for example, amounts to about half of
the total mass of the bone and essentially consists of
nanocrystals of a carbonated hydroxyapatite (the so-called
‘‘biological apatite’’), with the approximate formula Ca10–
x(PO4,CO3)6–x(HPO4)x(OH)2–x [1, 2, 5, 9–13]. Dental
enamel, on the other hand, has a larger inorganic content
than bone (about 97%) and is formed by larger and strongly
oriented prismatic crystals of biological apatite [5].
Keywords Calcium phosphates ꢀ Apatites ꢀ
Thermal analysis ꢀ Thermogravimetry ꢀ
Differential scanning calorimetry
Fluoridated calcium phosphates are also important bi-
ominerals. Shark teeth, for instance, contain fluoroapatite
Ca10(PO4)6F2 [14, 15], and the replacement of OH- by
F- in the hydroxyapatite phase of human tooth enamel by
fluoridated dental products readily occurs to yield a mate-
rial similar to hydroxyapatite/fluoroapatite mixed crystals,
Ca10(PO4)6(F,OH)2, formed in vitro [2, 16–21]. This sub-
stitution contributes to the prevention of tooth decay
because fluoroapatite is less soluble in the acidic solutions
produced by oral bacteria than the original hydroxyapatite
present in teeth [22]. It has been noted, however, that the
main caries-preventing effect of fluoride is related to its
ability to promote the remineralization of teeth using
R. G. Simo˜es ꢀ A. I. Aleixo ꢀ A. L. C. Lagoa ꢀ
M. E. Minas da Piedade (&) ꢀ J. P. Leal
´
´
ˆ
Departamento de Quımica e Bioquımica, Faculdade de Ciencias,
Universidade de Lisboa, 1649-016 Lisboa, Portugal
e-mail: memp@fc.ul.pt
J. P. Leal
Unidade de Ciencias Quımicas e Radiofarmaceuticas, Instituto
ˆ
´
Tecnologico e Nuclear, 2686-953 Sacavem, Portugal
ˆ
´
´
T. Peitsch ꢀ M. Epple
Inorganic Chemistry and Center for Nanointegration Duisburg-
¨
Essen, University of Duisburg-Essen, Universitatsstrasse 5-7,
45117 Essen, Germany
123

510
R. G. Simo˜es et al.
calcium dissolved in saliva, rather than the formation of
fluoridated apatitic phases that are more resistant to acidic
attacks [23–25].
under nitrogen at 353 K until all the calcium had reacted to
calcium ethanolate (*2 h). The solvent was removed in
vacuum and Ca(OEt)2ꢀ0.36EtOH (9.1 g, 62.0 mmol) was
obtained in the form of a fine white powder. Then,
7.0020 g (47.7 mmol) of this sample were dissolved in
200 cm3 of ethanol inside the glovebox to yield a
0.24 mol dm-3 solution of Ca(OEt)2 (S1). A solution of
hydrofluoric acid, HF, (S2) was prepared by refluxing
0.4041 g of NH4F (10.9 mmol) in 200 cm3 of ethanol
inside a 2 dm3 three necked round bottom flask. The
decomposition of NH4F generates HF according to the
reaction [29]:
Synthetic calcium phosphates are conventionally pre-
pared either by high-temperature solid state reactions or by
aqueous precipitation methods [2]. Calcination techniques
lead to highly crystalline materials with an increased sta-
bility towards dissolution due to a high crystallinity and
large crystallites. Hence, in spite of their good biocom-
patibility, they exhibit low biodegradation rates in vivo
compared to the nanosized biological apatites [26, 27]. The
drawback of aqueous precipitation routes is the difficult
control over stoichiometry caused by the phosphate/
hydrogenphosphate equilibrium and, in the case of fluoro-
apatite, the impossibility to obtain a phase-pure material
that contains fluoride only and no hydroxide. Such prob-
lems are directly related to the precipitation from the
medium water which participates in the reaction. A syn-
thesis in anhydrous alcohols using calcium ethanolate,
phosphoric acid, and ammonium fluoride was proposed by
Layrolle and Lebugle, which, in principle, allows a better
adjustment of the stoichiometry [28, 29]. Here, we report a
systematic thermoanalytical and structural study of fluori-
dated calcium phosphates that were prepared using differ-
ent variations of this synthesis to investigate if it provides a
viable route to produce calcium phosphates with a defined
composition and crystallinity.
NH4F ðslnÞ ! NH3 ðgÞ þ HF ðslnÞ
ð1Þ
A 0.215 mol dm-3 solution of orthophosphoric acid,
H3PO4, (S3), produced by dissolving 3.1673 g (32.3 mmol)
of solid H3PO4 in 150 cm3 of ethanol inside the glovebox,
was transferred to an addition funnel and added dropwise to
solution S2 under reflux. Finally, solution S1 was added
dropwise to this reaction mixture from a second addition
funnel. The formation of a white gelatinous precipitate was
immediately observed. The reaction mixture was refluxed
under nitrogen atmosphere for 3 h. After cooling, the sus-
pension was transferred to a Schlenk tube and the solvent was
removed in vacuum. The obtained solid was dried in vacuum
(1.3 Pa), first for 3 days at room temperature and then for
another day at 373 K until a constant mass was achieved. The
product was divided into three portions: sample A1 was used
without further treatment; samples A2 and A3 were heated
for 10 h under nitrogen flux at 573 and 873 K, respectively.
Experimental
Synthesis
Synthesis of samples B1, B2 and B3
All syntheses were performed using modified versions of the
methodproposedbyLayrolleandLebugle[29]. Thesyntheses
were carried out under an oxygen- and water-free (\5 ppm)
nitrogen atmosphere inside a glove-box or using Schlenk
techniques. Metallic calcium (Sigma-Aldrich, 99%, gran-
ules), phosphoric acid crystals (Sigma-Aldrich, 99.999%),
and ammonium fluoride (Panreac, 95%) were used without
The synthesis of these samples was carried out to investigate
the effect of excess calcium on the nature of the produced
materials. Metallic calcium (3.0983 g, 77.3 mmol) and
200 cm3 of ethanol were placed into a two-necked round
bottom flask inside a glovebox. The reaction system was
taken out of the glovebox and connected to a reflux con-
denser. The mixture was refluxed under nitrogen atmosphere
at 353 K until all the calcium had reacted (*10 h) to yield a
0.39 mol dm-3 solution of Ca(EtO)2 (S4). A HF solution
(S5) was prepared according to reaction (1) by refluxing
0.4138 g of NH4F (11.2 mmol) in 200 cm3 of ethanol inside
a three-necked 2 dm3 round bottom flask. A solution of
orthophosphoric acid (S6) was prepared by dissolving
3.2546 g (33.2 mmol) of H3PO4 in 150 cm3 of ethanol inside
the glovebox. This solution was added to solution S5 under
reflux. Finally, solution S4 was added to the reaction mixture
with the immediate formation of a white gelatinous precip-
itate. The resulting suspension was refluxed under nitrogen
atmosphere for 1 h. After cooling, the suspension was
transferred to a Schlenk tube, the solvent was removed in
¨
further purification. Ethanol (Riedel-de-Haen, p.a.) and iso-
propanol (Merck, 99.8%) were pre-dried with CaH2, refluxed
under nitrogen with iodine-activated magnesium, and finally
distilled. Four families of calcium phosphate samples denoted
A to D were prepared and characterized.
Synthesis of samples A1, A2, and A3
In total, 3.0202 g (75.4 mmol) of metallic Ca and 150 cm3
of ethanol were placed in a three-necked 2 dm3 round
bottom flask inside a glovebox. The setup was removed
from the glovebox, connected to a nitrogen line and
adapted to a reflux condenser. The mixture was refluxed
123

Thermoanalytical and structural characterization of fluoridated calcium phosphates
511
vacuum (1.3 Pa), and the obtained solid was dried at room
temperature, first for 3 days at 1.3 Pa and subsequently for
7 days at 13 mPa. The product was divided into three por-
tions: sample B1 was used without further processing, and
samples B2 and B3 were heated for 17 h under nitrogen flux
at 573 and 773 K, respectively.
under nitrogen flux for 10 h at 573, 873 and 1173 K,
respectively.
Characterization
Elemental analyses (C, H, N) were made with a CE instru-
ments EA-1110 apparatus. The calcium contents of samples
A and D were determined with a Pye Unicam SP9 atomic
absorption spectrometer (AAS) with an air/acetylene flame.
The detection was set to 422 nm and the samples were
previously reacted with La2O3/HCl. The calcium contents of
samples B and C, and the phosphorus contents of all the
studied samples were obtained with a Perkin Elmer Optima
2000 DV ion coupled plasma atomic emission spectrometry
(ICP) instrument. The fluoride content of all the samples was
determined by the potentiometric method reported by Duff
and co-workers [30]. A Metrohm 632 pH-Meter, a fluoride
selective electrode (Mettler Toledo), and a calomel refer-
ence electrode (Metrohm CH-9101 Herisau) were used.
Fourier-transform infrared spectra (FTIR) were determined
in KBr plates with a Nicolet 6700 instrument. X-ray powder
diffractograms (XRD) were recorded on a Philips PW1730
diffractometer (Cu Ka radiation, graphite monochromator).
Scanning electron microscopy (SEM) with energy disper-
sive X-ray spectroscopy (EDX) analyses were performed in
high vacuum on Au/Pd-sputtered samples with a FEI ESEM
Quanta 400 FEG instrument. Simultaneous thermogravi-
metric (TG) and differential scanning calorimetry (DSC)
Synthesis of samples C1, C2 and C3
Metallic calcium (5.0354 g, 125.6 mmol) and 500 cm3 of
ethanol were placed in a three-necked round bottom flask
inside a glovebox. The reaction system was removed from
the glovebox and connected to a reflux condenser. The
mixture was refluxed under nitrogen atmosphere at 353 K
until all the calcium had reacted (solution S7). A solution
of orthophosphoric acid (S8) was prepared by dissolving
7.5525 g (77.1 mmol) of H3PO4 in 300 cm3 of ethanol
inside the glovebox. The vessel containing the solution was
taken out of the glovebox and connected to a nitrogen line
and to a bubbler containing a concentrated NaOH solution.
Gaseous HF was produced in an independent vessel by
reacting 5 cm3 of H2SO4 (Panreac, 95%) with 1.0132 g
(27.4 mmol) of NH4F according to
H2SO4 ðlÞ þ 2NH4F ðsÞ ! ðNH4Þ2SO4 ðsÞ þ 2HF ðg) ð2Þ
The gas was bubbled into solution S8 to yield solution S9.
Solution S7 was then added to solution S9, giving a white
precipitate. The resulting suspension was refluxed under
nitrogen atmosphere for 2 h, transferred to a Schlenk tube
and dried in vacuum at room temperature, first for 3 days at
1.3 Pa and then for 7 days at 13 mPa. The product was
divided in three portions: sample C1 was used without
further treatment, and samples C2 and C3 were heated for
17 h under nitrogen flux at 573 and 773 K, respectively.
Table 1 Temperatures of the annealing after precipitation, the results
of the elemental analysis (mass/%), the molar calcium/phosphorus
ratio, n(Ca)/n(P), and the molar calcium/fluoride ratio, n(Ca)/n(F), of
all samples prepared in this study
Sample T/K
C
H
N
F
Ca
P
n(Ca)/ n(Ca)/
n(P)
n(F)
Synthesis of samples D1, D2, D3, and D4
A/theoretical
1.48
4.38
12.7
A1
A2
A3
373 0.36 1.55 1.59 1.02 27.3 21.1 1.00
573 0.29 0.40 0.42 0.96 26.3 22.7 0.90
873 0.10 0.02 0.05 0.28 27.7 25.0 0.86
The procedure followed in the preparation of the D samples
was essentially the same as described for samples C, except
for the use of iso-propanol instead of ethanol as solvent.
The calcium iso-propoxide solution was obtained from
5.0376 g (125.7 mmol) of metallic calcium and 500 cm3 of
iso-propanol. An orthophosphoric/hydrofluoric acid mix-
ture was prepared by dissolving 7.5604 g (77.2 mmol) of
H3PO4 in 300 cm3 of iso-propanol, and bubbling gaseous
HF in the obtained solution. The hydrofluoric acid was
produced by the reaction of 5 cm3 of H2SO4 (Panreac,
95%) with 1.0197 g (27.5 mmol) of NH4F. The final
reaction mixture was taken to dryness in vacuum at room
temperature, and the obtained solid was further dried in
vacuum (1.3 Pa) and at 373 K during 4 days. The product
was divided into four parts: sample D1 was used without
further treatment, and samples D2, D3, and D4 were heated
13.0
46.9
B/theoretical
2.33
6.90
B1
B2
B3
298 15.33 3.75 0.51 0.03 19.5 6.5 2.32
573 5.82 0.56 0.21 0.03 18.2 5.8 2.43
308.1
287.6
233.5
773 1.14
0
0.12 0.04 19.7 6.7 2.27
1.69
C/theoretical
4.58
C1
C2
C3
298 8.92 1.89
573 1.65 0.97
773 1.04 0.37
0
0
0
0
0
0
29.8 17.6 1.31
32.4 19.5 1.28
35.1 20.9 1.30
1.63
–
–
–
D/theoretical
4.57
D1
D2
D3
D4
373 1.93 1.55
0
0
0
0
0.53 31.3 15.5 1.56
0.63 29.6 16.3 1.40
28.0
573 1.83 0.67
873 0.37 0.09
22.3
23.8
28.0
0.70 35.1 18
1.51
1173 0.18
0
0.70 41.4 19.2 1.67
123

512
R. G. Simo˜es et al.
were performed with a Setaram TG-DSC 111 apparatus.
Samples were placed in SiO2 crucibles and heated from 298
to 873 K at a rate of 3 K min-1 under dynamic nitrogen
atmosphere (2 dm3 h-1). Combined thermogravimetric-
infrared analysis (TG-IR) and differential thermal analysis
(DTA) were carried out on a Netzsch STA 409/Bruker
carbon and hydrogen contents of the samples heated above
373 K under nitrogen atmosphere [31]. The nitrogen
present in samples A and B originates from residual NH4þ
Vertex 70 system (298–1273 K; heating rate 3 K min-1
;
dynamic nitrogen or oxygen atmosphere, 2 dm3 h-1; Al2O3
crucibles).
Results and discussion
The temperatures used in the final thermal treatment during
preparation and the results of the elemental analysis of
samples A–D are summarized in Table 1. The results show
that the obtained products have a different calcium to
phosphorus molar ratio, n(Ca)/n(P), than expected from the
amounts of calcium and phosphorus used in the preparations.
The presence of significant amounts of carbon and
hydrogen in samples A1, B1, C1, and D1 indicates that the
solvent (ethanol or iso-propanol) remains partially incor-
porated in the sample lattice, even after drying in vacuum
at a temperature ranging from 298 to 373 K. The carbon
and hydrogen contents of the samples decrease with the
increase of the annealing temperature, indicating a loss of
solvent by evaporation and/or thermal decomposition. The
catalytic thermal decomposition of ethanol and iso-propa-
nol over calcium phosphates in the range of 523–773 K is
well documented, and it is probably the source of the
Fig. 2 SEM images of samples a A1, heated to 373 K, b A2, heated
to 573 K, c A3, heated to 873 K, d C1, heated to 298 K, e C2 heated
to 573 K, and f C3, heated to 773 K
D4 (T = 1173 K)
D3 (T = 873 K)
D2 (T = 573 K)
D1 (T = 373 K)
C3 (T = 773 K)
C2 (T = 573 K)
C1 (T = 298 K)
P
(a)
(b)
P
Ca
B3 (T = 773 K)
Ca
Au
B2 (T = 573 K)
B1 (T = 298 K)
O
C
O
C
Ca
Au
Pd
Ca
Au
F
1.00 2.00 3.00 4.00
E/keV
0.80 1.60 2.40 3.20 4.00 4.80
E/keV
A3 (T = 873 K)
A2 (T = 573 K)
P
Ca
(c)
(d)
P
Ca
A1 (T = 373 K)
O
5
10
15
20
25
30
35
40
45
50
55
60
2θ/°
O
Ca
Ca
Fig. 3 X-ray powder diffractograms of samples A–D with corre-
sponding preparation temperatures in parenthesis. The marked peaks
are assigned to the following phases: hydroxy/fluoroapatite ( );
monetite (•); calcium oxide (|); c-calcium pyrophosphate (m);
a-calcium pyrophosphate ( ); a-tricalcium phosphate (r), and
b-tricalcium phosphate (j)
Pd
C
Si
C
F
1.00 2.00 3.00 4.00
E/keV
0.90 1.80 2.70 3.60 4.50
E/keV
Fig. 1 EDX spectra for samples A1 (a), B1 (b), C1 (c), and D1 (d)
123

Thermoanalytical and structural characterization of fluoridated calcium phosphates
513
used in the in situ production of HF. This had also been
observed by Layrolle and Lebugle in their original syn-
thesis [29]. The nitrogen incorporation in the products was
eliminated in the synthesis of samples C and D by the
addition of gaseous HF produced externally by reaction (2).
The results in Table 1 also indicate that all preparations led
to samples with considerably lower fluoride content than
those expected from the syntheses. Therefore, the products
are highly fluoride-deficient in comparison to stoichiome-
tric fluoroapatite, Ca10(PO4)6F2. The fact that the fluoride
content also decreased upon heating of the samples A
indicates that fluoride was probably not incorporated into
the lattice but rather adsorbed on the surface. The samples
B showed a n(Ca)/n(P) ratio very close to that used in the
synthesis but with almost no incorporation of fluoride. In
samples C, no fluoride was incorporated at all, which is
unexpected because the difference between the prepara-
tions of samples C and D only consisted in the change of
the alcoholic solvent.
which did not significantly change upon heating. Similar
morphologies were observed for samples B and D. In
contrast, sample C1 showed a plate-like morphology which
did not substantially change upon the heat treatment that
led to samples C2 and C3 (Fig. 2d–f).
The X-ray powder diffractograms obtained for all sam-
ples are shown in Fig. 3. Note that it is not possible to
distinguish between hydroxyapatite and fluoroapatite by
conventional X-ray powder diffraction due to the very
similar diffraction patterns and also due to the fact that they
form a complete substitution row. Therefore, we will
denote these phases as hydroxy/fluoroapatite in the fol-
lowing. Except for the samples C, a clear increase of the
crystallinity of the products is observed with increasing
temperature during the thermal treatment. Sample A1 is
practically amorphous. All other samples are more or less
crystalline, but they all contain more than one phase. The
results of the phase identification through peak assignments
using the DIFFRACplus EVA 11.0.0.3 program [32] and
the FIZ/NIST database [33] are summarized in Table 2.
Analysis of Table 2 suggests that upon heating the initially
obtained amorphous A1 material to 573 K, a mixture of
monetite and fluoroapatite is formed. Further heating to
873 K leads to pyrophosphate formation by the well-
known reaction [34]:
The results of the EDX analysis (Fig. 1) corroborate the
general conclusions drawn from the elemental analysis.
Samples A1, B1, C1, and D1 show evidence of carbon
from solvent retention in the lattice, and no fluoride is
detectable in the case of samples B and C. The trace of Si
in Fig. 1d probably results from the HF attack to the glass
containers used in the synthesis, and the presence of Au
and Pd originates from the sputtering of the samples prior
to the SEM analysis.
2CaHPO4 ðsÞ ! Ca2P2O7 ðsÞ þ H2O ðg)
ð3Þ
A clear signature of fluoroapatite could not, however, be
identified in the diffractogram of sample A3. Samples B1
and B2 are almost amorphous with only traces of
hydroxyapatite, and sample B3 corresponds to a mixture of
The SEM micrographs of samples A are shown in
Fig. 2a–c. The initially obtained A1 product consisted of
compact crystallites with a diameter of about 100 nm,
Table 2 Phase composition of samples A–D as derived from X-ray powder diffraction analysis
Sample Phase assignment
A1
A2
A3
Amorphous
Mixture of monetite (CaHPO4; JCPDS 00-009-0080) and hydroxy/fluoroapatite (JCPDS 00-034-0011)
a- and c-calcium pyrophosphate (Ca2P2O7; JCPDS 00-009-0345 and JCPDS 00-015-0197) and traces of hydroxy/fluoroapatite (JCPDS
00-034-0011)
B1
B2
B3
C1
C2
C3
Almost amorphous; traces of hydroxy/fluoroapatite (JCPDS 00-034-0011)
Almost amorphous; traces of hydroxy/fluoroapatite (JCPDS 00-034-0011)
Hydroxy/fluoroapatite (JCPDS 01-072-1243) and traces of CaO (JCPDS 37-1497)
Monetite (CaHPO4; JCPDS 00-009-0080)
Monetite (CaHPO4; JCPDS 00-009-0080)
c-Calcium pyrophosphate (Ca2P2O7; JCPDS 00-009-0345); traces of CaO (JCPDS 37-1497), and possibly of hydroxy/fluoroapatite
(JCPDS 01-072-1243)
D1
D2
D3
Unknown phase
Mixture of hydroxy/fluoroapatite (JCPDS 00-034-0011), CaO (JCPDS 37-1497), and an unidentified phase
Mixture of hydroxy/fluoroapatite (JCPDS 00-034-0011), a-tricalcium phosphate (a-TCP; a-Ca3(PO4)2; JCPDS 00-029-0359), and CaO
(JCPDS 37-1497)
D4
Mixture of hydroxy/fluoroapatite (JCPDS 00-034-0011), b-tricalcium phosphate (b-TCP; b-Ca3(PO4)2, JCPDS 01-070-2065), and
traces of CaO (JCPDS 37-1497)
123

514
R. G. Simo˜es et al.
hydroxyapatite and possibly CaO as indicated by the peak
at 2h = 37.4ꢁ. Calcium oxide could result from the thermal
decomposition of excess Ca(OEt)2 used in the synthesis
[35]. The hydroxylation source necessary for the formation
of hydroxyapatite may originate from the thermal decom-
position of Ca(OEt)2 and ethanol [31, 35]. Note finally that
in this case the elemental analysis indicates practically no
fluoride in the samples, thus implying the absence of a
fluoroapatite phase. Sample C1 corresponds to monetite.
Hence, the patterns obtained by successively heating C1 to
573 and 773 K are dominated by the formation of pyro-
phosphates according to reaction (3) as in the case of
samples A and to the thermal decomposition of Ca(OEt)2
leading to CaO. The sample D1 prepared in iso-propanol is
compatible with a nanocrystalline material, as indicated by
the broad peaks observed in the diffraction pattern. The
corresponding phase could not, however, be identified.
Subsequent heating from 373 to 1173 K leads to a mixture
of hydroxy/fluoroapatite, calcium oxide, and b-tricalcium
phosphate, Ca3(PO4)2.
D4 (T = 1173 K)
D3 (T = 873 K)
PO4
PO4
OH
OH
PO4
PO4
PO4
PO4
D2 (T = 573 K)
D1 (T = 373 K)
PO4
OH
PO4
C3 (T = 773 K)
C2 (T = 573 K)
C1 (T = 298 K)
P2O7
HPO4
HPO4
P2O7
OH
OH
HPO4
HPO4
B3 (T = 773 K)
B2 (T = 573 K)
B1 (T = 298 K)
OH
OH
PO4
PO4
PO4
PO4
PO4
OH
PO4
PO4
The FTIR spectra of samples A–D are shown in Fig. 4.
The presence of surface-bound and occluded solvent (eth-
anol or iso-propanol) in the samples A1, B1, C1, and D1 is
confirmed by the observation of the broad bands in the
range 3700–2600 cm-1 and at *1650 cm-1, assigned to
O–H stretching and O–H bending modes, respectively
[36–38]. Note that water cannot be present due to the fully
water-free syntheses. In the case of samples A and D, the
observed decrease of the intensity of those bands on
heating is consistent with the removal of the solvent and
the formation of phases that do not contain hydroxide
groups. This is in agreement with the results of X-ray
A3 (T = 873 K)
A2 (T = 573 K)
P2O7
PO4
P2O7
PO4
PO4
OH
HPO4
OH
A1 (T = 373 K)
PO4
HPO4
4000 3600 3200 2800 2400 2000 1600 1200 800
400
~
/cm–1
Fig. 4 FTIR spectra of the samples A–D
Table 3 Results of the TG-DSC and TG-IR experiments carried out on samples A1, B1, C1, and D1 in the temperature ranges 298–873 K and
298–1273 K
Method
Sample
A1
B1
C1
D1
T/K
Dm/%
T/K
Dm/%
T/K
Dm/%
T/K
Dm/%
TG-DSC (N2)
TG-IR (N2)
TG-IR (O2)
298–873
10.2
298–657
657–707
12.9
3.1
298–549
549–751
4.7
7.7
298–660
660–709
709–873
298–655
660–707
707–1173
298–660
660–707
840–947
7.3
4.7
1.1
8.7
5.0
2.5
8.1
3.8
0.6
298–902
902–910
10.1
0.5
298–626
626–818
818–1273
9.8
8.1
6.5
298–578
578–761
815–947
1.7
8.2
2.5
298–902
902–910
10.7
0.6
Note: The mass losses are given as mass percentage (%Dm) and the nature of the purging gas is indicated in parentheses
Dm/% = 100 9 (m0- m)/m0 where m0 is the initial mass of sample and m is the mass of sample at the given temperature in the TG scan
123

Thermoanalytical and structural characterization of fluoridated calcium phosphates
515
102
diffraction (Table 2). In contrast, heating samples B1 and
C1 to 573 and 773 K did not lead to the elimination of the
OH bands. The sharp OH bands at *3600 cm-1 belong to
2
100
98
96
94
92
90
1
TG
hydroxide groups in hydroxyapatite. The spectra of sam-
3ꢁ
3
DTA
~
ples A1 and A2 show the PO4 (m ¼ 1102, 1043, 607, and
570 cm-1) and HPO4 (m ¼ 904 cm-1) bands expected
2ꢁ
~
for the apatite and monetite phases detected by X-ray dif-
4
DSC
5
6
2ꢁ
~
fraction. The disappearance of the HPO4 peak at m ¼ 904
cm-1 and the occurrence of peaks characteristic for P2O7
4ꢁ
(m ¼ 1160, 985, 720, 550, and 499 cm-1) in sample A3 are
~
also consistent with the X-ray diffraction analysis. The
290
490
690
890
1090
1290
3ꢁ
~
observation of PO4
peaks at m ¼ 1102, 607, and
T/ K
570 cm-1 in the FTIR spectrum of sample A3 suggests the
presence of fluoroapatite, a conclusion that could not be
drawn from the X-ray diffraction data. It is, however,
unlikely that the fluoroapatite present in sample A2 would
100.04
100.0
99.5
99.0
100.00
99.96
99.92
%T
%
T
98.5
98.0
97.5
decompose upon heating to 873 K. In the case of the
2ꢁ
1000
1000
4000 3000 2000
4000 3000 2000
(T = 467 K)
(T = 298 K)
~
samples B, the absence of HPO4 bands at m ¼ 900 and
/cm–1
(1)
~
/cm–1
~
1400 cm-1 and the presence of PO4 peaks (Fig. 4) are
3ꢁ
(2)
100.0
80.0
60.0
100.0
80.0
60.0
also consistent with the X-ray diffraction results. These
indicate the formation of a hydroxyapatite phase upon
heating from 298 to 773 K and show no evidence of
hydrogenphosphate-containing phases (e.g. monetite) or
pyrophosphate originating from their decomposition
through reaction (3). The FTIR spectra of the samples C1
40.0
20.0
0.0
40.0
20.0
0.0
1000
4000 3000 2000
1000
4000
2000
3000
(T = 537 K)
(T = 670 K)
~
/cm–1
~
/cm–1
(3)
(4)
2ꢁ
100.0
99.5
100.0
99.5
and C2 show the HPO4 peaks (Fig. 4) expected for the
monetite phase indicated by the X-ray diffraction analysis.
The decomposition of this phase on heating according to
reaction (3) is also evidenced by the presence of peaks
99.0
98.5
98.0
99.0
98.5
97.5
4000 3000 2000 1000
4ꢁ
1000
4000
2000
3000
(T = 918 K)
(T = 768 K)
attributed to P2O7 in the FTIR spectrum of sample C3.
~
/cm–1
~
/cm–1
(5)
The FTIR spectra of samples D show the presence of
3ꢁ
(6)
PO4 peaks (Fig. 4) consistent with the presence of flu-
Fig. 5 Results of TG-IR, DSC and DTA experiments on sample A1,
using nitrogen as the purging gas. The scale on the left corresponds to
the sample mass percentage 100 9 (m/m0) determined in the TG
experiments. The numbered IR spectra refer to the given tempera-
tures. In the DSC and DTA measuring curves, peaks corresponding to
endothermic events point downwards
oroapatite and tricalcium phosphate phases observed by
X-ray diffraction.
Table 3 summarizes the TG-DSC (298–873 K) and TG-
IR (298–1273 K) results obtained for samples A1, B1, C1,
and D1. Figure 5 shows an overlay of three measuring
curves corresponding to sample A1: TG in the range of
298–1273 K, DSC in the range of 298–873 K, and DTA in
the range of 873–923 K. The DTA experiment was run to
complement the DSC results which (due to instrumental
limitations) were only performed up to 873 K. It should
also be noted that the relative intensities of the DSC and
DTA peaks in Fig. 5 are not comparable in absolute terms,
as they were scaled relative to their minimum values in the
temperature ranges covered by the experiments.
observation of a band at m ꢂ 1000 cm-1 in the IR spectrum
~
2 at 467 K suggests ethanol evaporation up to *473 K.
This is followed by its thermal decomposition as shown by
the formation of H2O and CO2 between *533 K and
*673 K (IR spectra 3 and 4 in Fig. 5). The thermal
2ꢁ
decomposition of HPO4 accompanied by the release of
water according to reaction (3) may also contribute to the
mass loss observed in this range. This reaction was indeed
reported to be endothermic and occur in a wide range
(*573–973 K) without a resolvable mass loss step
[34, 39]. Finally, the nature of the small mass loss (0.5–
0.6%) at *903 K with H2O and CO2 release, also detected
by DTA, could not be determined.
Figure 5 also shows several IR spectra of the gases
evolved during the thermogravimetric experiment. A mass
loss of about 10% up to 873 K with two or three unre-
solvable steps is observed. The accompanying decrease in
heat flow observed in the DSC measurements where at
least three small unresolved peaks are evident indicates that
it corresponds to a complex endothermic process. The
As shown by the overlay of the TG data for samples
A1–D1 in Fig. 6, the initial and pronounced mass loss due
123

516
R. G. Simo˜es et al.
þ
105
100
95
incorporation of NH4 in the structure. The retention of
significant amounts of solvent in the solid product even
after drying in vacuum at temperatures between *298 and
373 K is a severe drawback of this type of synthesis to
produce materials with a defined composition and crystal-
linity. A more complete removal of the solvent is achieved
by heating the materials to temperatures above *573 K
but this leads to a significant increase of crystallinity and to
structural changes due to decomposition reactions involv-
ing, e.g. hydrogenphosphate.
(a)
A1
90
C1
D1
85
80
B1
Acknowledgements This work was supported by Fundac¸a˜o para a
75
ˆ
Ciencia e a Tecnologia (FCT, Portugal) and the CRUP (Portugal;
300
500
700
900
1100
1300
A-31/09)/DAAD (Germany) bilateral exchange program. PhD and
Post Doctoral grants from FCT are also gratefully acknowledged by
R.G.S. (SFRH/BD/48410/2008) and A.L.C.L. (SFRH/BPD/35053/
2007), respectively.
T/K
0
-2
(b)
Endothermic
References
-4
B1
D1
1. LeGeros RZ. Biological and synthetic apatites. In: Brown PW,
Constantz B, editors. Hydroxyapatite and related materials. Boca
Raton: CRC Press; 1994.
-6
Φ
2. Elliot JC. Structure and chemistry of the apatites and other cal-
cium orthophosphates. Amsterdam: Elsevier; 1994. p. 208.
3. Baeuerlein E. Biomineralization. Progress in biology, molecular
biology and application. Weinheim: Wiley-VCH; 2004.
4. Epple M, Baeuerlein E. Biomineralisation: medical and clinical
aspects. Weinheim: Wiley-VCH; 2007.
5. Dorozhkin SV, Epple M. Biological and medical significance of
calcium phosphates. Angew Chem Int Ed. 2002;41:3130–46.
6. Salinas AJ, Vallet-Regi M. Evolution of ceramics with medical
applications. Z Anorg Allg Chem. 2007;633:1762–73.
7. Vallet-Regi M, Gonzalez-Calbet JM. Calcium phosphates as sub-
stitution of bone tissues. Prog Solid State Chem. 2004;32:1–31.
8. Kenny SM, Buggy M. Bone cements and fillers: a review. J Mater
Sci Mater Med. 2003;14:923–38.
-8
A1
C1
800
-10
-12
300
400
500
600
700
900
T/K
Fig. 6 Overlays of the TG curves (a) and the DSC curves (b)
obtained for the samples A1–D1
to solvent evaporation/thermal decomposition and eventu-
ally reaction (3) detected for sample A1 was also observed
for samples B1, C1, and D1. Note that all samples show a
clear endothermic DSC peak at *673 K. This peak is
9. Weiner S, Wagner HD. The material bone: structure-mechanical
function relations. Annu Rev Mater Sci. 1998;28:271–98.
10. Wopenka B, Pasteris JD. A mineralogical perspective on the
apatite in bone. Mater Sci Eng C. 2005;25:131–43.
11. Rey C, Combes C, Drouet C, Sfihi H, Barroug A. Physico-
chemical properties of nanocrystalline apatites: implications for
biominerals and biomaterials. Mater Sci Eng C. 2007;27:198–205.
12. Horvath AL. Solubility of structurally complicated materials: II.
bone. J Phys Chem Ref Data. 2006;35:1653–69.
13. Peters F, Schwarz K, Epple M. The structure of bone studied with
synchrotron X-ray diffraction, X-ray absorption spectroscopy and
thermal analysis. Thermochim Acta. 2000;361:131–8.
14. Moller IJ, Melsen B, Jensen SJ, Kirkegaard E. A histological,
chemical and X-ray diffraction study on contemporary (Car-
charias glaucus) and fossilized (Macrota odontaspis) shark teeth.
Arch Oral Biol. 1975;20:797–802.
15. Daclusi G, Kerebel LM. Ultrastructural study and comparative
analysis of fluoride content of enameloid in sea-water and fresh-
water sharks. Arch Oral Biol. 1980;25:145–51.
16. Jha LJ, Best SM, Knowles JC, Rehman I, Santos JD, Bonfield W.
Preparation and characterization of fluoride-substituted apatites.
J Mater Sci Mater Med. 1997;8:185–91.
within the range typical of the thermal decomposition of
2ꢁ
HPO4
in accordance with reaction (3) [34, 39] (see
above). It cannot, however, be unequivocally assigned to
this process because the presence of hydrogenphosphate or
pyrophosphate is evident by X-ray diffraction (Fig. 3) and
FT-IR analysis (Fig. 4) of samples A and C, but it is not
clear in the corresponding results for samples B and D.
Conclusions
The nature of fluoridated calcium phosphates obtained by
water-free precipitation from alcoholic media studied in
this work is very sensitive to variations of the synthetic
conditions. The fluoridation of the products by in situ
NH4F decomposition during synthesis seems to be more
efficient than the use of an HF solution, but it leads to the
17. Manjubala I, Sivakumar M, Nikkath SN. Synthesis and charac-
terisation of hydroxy/fluoroapatite solid solution. J Mater Sci.
2001;36:5481–6.
123

Thermoanalytical and structural characterization of fluoridated calcium phosphates
517
18. Rodriguez-Lorenzo LM, Hart JN, Gross KA. Influence of fluorine
in the synthesis of apatites. Synthesis of solid solutions of
hydroxy-fluorapatite. Biomaterials. 2003;24:3777–85.
19. Gross KA, Rodriguez-Lorenzo LM. Sintered hydroxyfluorapa-
tites. Part I: sintering ability of precipitated solid solution pow-
ders. Biomaterials. 2004;25:1375–84.
29. Layrolle P, Lebugle A. Synthesis in pure ethanol and character-
ization of nanosized calcium phosphate fluoroapatite. Chem
Mater. 1996;8:134–44.
30. Duff EJ, Stuart JL. Determination of fluoride in calcium phos-
phates with a fluoride-selective electrode. Anal Chim Acta. 1970;
52:155–7.
20. Busch S, Schwarz U, Kniep R. Morphogenesis and structure of
human teeth in relation to biomimetically grown fluorapatite–
gelatine composites. Chem Mater. 2001;13:3260–71.
21. Prymak O, Sokolova V, Peitsch T, Epple M. The crystallization
of fluoroapatite dumbbells from supersaturated aqueous solution.
Cryst Growth Des. 2006;6:498–506.
22. Bartlett JD, Dwyer SE, Beniash E, Skobe Z, Payne-Ferreira TL.
Fluorosis: a new model and new insights. J Dent Res. 2005;84:
832–6.
31. Monma H. Catalytic behavior of calcium phosphates for decom-
positions of 2-propanol and ethanol. J Catal. 1982;75:200–3.
32. DIFFRACplus EVA 11.0.0.3. Karlsruhe: Bruker AXS; 1996–
2005.
33. FIZ/NIST Inorganic Crystal Structure Database (ICSD), NIST
Standard Reference Database 84. Gaithersburg: National Institute
of Standards and Technology; 2006.
34. Wikholm NW, Beebe RA, Kittelberger JS. Kinetics of the con-
version of monetite to calcium pyrophosphate. J Phys Chem.
1975;79:853–6.
35. Dongare MK, Sinha APB. Thermal analysis of some metal alk-
oxides. Thermochim Acta. 1982;57:37–45.
23. ten Cate JM, Duijsters PPE. Influence of Fluoride in solution on
tooth demineralization. 2. Microradiographic data. Chemical-
data. Caries Res. 1983;17:513.
24. Cate JM, Duijsters PPE. Influence of Fluoride in solution on tooth
demineralization. 1. Chemical-data. Caries Res. 1983;17:193–9.
36. Klee WE, Engel G. I.R. spectra of the phosphate ions in various
apatites. J Inorg Nucl Chem. 1970;32:1837–43.
¨
25. Ogaard B, Rolla G, Ruben J, Dijkman T, Arends J. Microradio-
37. Fowler BO. Infrared studies of apatites. I. Vibrational assign-
ments for calcium, strontium, and barium hydroxyapatites uti-
lizing isotopic substitution. Inorg Chem. 1974;13:194–207.
38. Penel G, Leroy G, Rey C, Sombret B, Huvenne JP, Bres E.
Infrared and Raman microspectrometry study of fluor-fluor-
hydroxy and hydroxy-apatite powders. J Mater Sci Mater Med.
1997;8:271–6.
graphic study of demineralization of shark enamel in human
caries model. Scand J Dent Res. 1988;96:209–11.
26. Tadic D, Peters F, Epple M. Continuous synthesis of amorphous
carbonated apatites. Biomaterials. 2002;23:2553–9.
27. Tadic D, Epple M. A thorough physicochemical characterisation
of 14 calcium phosphate-based bone substitution materials in
comparison to natural bone. Biomaterials. 2004;25:987–94.
28. Layrolle P, Lebugle A. Characterization and reactivity of nano-
sized calcium phosphates prepared in anhydrous ethanol. Chem
Mater. 1994;6:1996–2004.
39. Berry EE. The structure and composition of some calcium-defi-
cient apatites. J Inorg Nucl Chem. 1967;29:317–27.
123