Carbonate substitution in precipitated hydroxyapatite: An

investigation into the effects of reaction temperature and

bicarbonate ion concentration

J. Barralet,1,2 S. Best,2 W. Bonfield2

1
Smith and Nephew Group Research Centre, York Science Park, Heslington, York, YO1 5DF, United Kingdom
2
IRC in Biomedical Materials, Queen Mary and Westfield College, Mile End Road, London E1 4NS, United Kingdom

Received 5 July 1996; accepted 1 October 1997

Abstract: Carbonate substitution in the apatite crystal lat- copy indicated the presence of carbonate in the A site for low
tice can occur in either the hydroxyl or the phosphate sites, carbonate contents (< 4 wt%), and at higher carbonate con-
designated as A or B type, respectively, and previous inves- tents (> 4 wt%), the carbonate was located predominantly in
tigations generally have described precipitated carbonate the B site. On the basis of these observations and chemical
hydroxyapatite as being B type on the basis of infra red and analyses, a new AB carbonate substitution mechanism is
X-ray data. This paper documents the effects of two precipi- proposed that better describes the experimental data than
tation variables, namely temperature and bicarbonate ion the B-type models used previously. © 1998 John Wiley &
concentration, on the morphology, phase composition, and Sons, Inc. J Biomed Mater Res, 41, 79–86, 1998.
calcium, phosphorus, and carbon contents of precipitated
carbonate hydroxyapatite. Variations in both temperature
and bicarbonate concentration could yield either acicular or Key words: carbonate; hydroxyapatite; substitution; precipi-
spheroidal crystals. X-ray diffraction and infra red spectros- tation

INTRODUCTION crystallinity and precipitate size,4–9 making it more

difficult to characterize the nature of carbonate substi-
Human bone mineral differs in composition from stoi- tution. It also previously has been observed experi-
chiometric hydroxyapatite (HA) [Ca10(PO4)6(OH)2] in mentally that precipitated CHA containing low levels
that it contains additional ions, of which carbonate is of carbonate can be A type.10
the most abundant species (∼ 8 wt%).1 The mechanism Various models have been devised in the past to
of in vivo mineralization and the nature of the carbon- describe carbonate substitution in precipitated apa-
ate substitution in apatite have been two of the major tites.5,7,9–13 Some of these models assume that a
driving forces behind in vitro carbonate hydroxyapa- coupled substitution occurs whereby the negative
tite (CHA) precipitation studies. charge caused by replacing PO43− with CO32− is bal-
It generally is agreed that the carbonate ion can sub- anced by the substitution of a monovalent ion (typi-
stitute at two sites in the apatite structure, namely the cally sodium) for calcium. Other models effect the
hydroxyl and the phosphate ion positions, giving A- charge balance required by the substitution of carbon-
and B-type CHA, respectively.2 A-type CHA can be ate for phosphate by calcium vacancy formation or by
formed by high-temperature reaction of HA with car- a mixture of these two mechanisms.14
bon dioxide gas, yielding a highly crystalline material3 This paper reports the findings that have resulted
that lends itself to characterization techniques, such as from a study of the precipitation of CHA prepared,
X-ray diffraction (XRD), infrared spectroscopy, and using a method based on that of Nelson and Feather-
transmission electron microscopy (TEM). On the other stone,5 over a range of temperatures and bicarbonate
hand, B-type CHA often is formed by precipitation, ion concentrations. It was found that the conventional
and carbonate in the B site has been found to reduce carbonate for phosphate model gave a good descrip-
tion of the experimentally determined change of phos-
phorus content with carbon. However, the model
Correspondence to: J. Barralet, Smith and Nephew Group failed to describe accurately the observed change in
Research Center, York Science Park, Heslington, York, Y01
5DF, United Kingdom calcium with carbon, implying that the substitution
mechanism upon which the conventional model is
© 1998 John Wiley & Sons, Inc. CCC 0021-9304/98/010079-08 based is not the mechanism that actually occurs in this
80 BARRALET, BEST, AND BONFIELD

system. On the basis of this finding and from our ex- RESULTS
perimental evidence for an A-type substitution at low
carbonate contents, a new model was devised TEM micrographs (Fig. 1) revealed that the mor-
whereby simultaneous A- and B-type substitutions are phology of the apatite crystals ranged from acicular
accounted for. This new model, termed here the AB (500–700 nm long) at reaction temperatures of 90° and
model, accurately described the change in composi- 70°C to smaller (> 20 nm) and spheroidal below 70°C.
tion of precipitated CHA with carbonate content that The effect of increasing bicarbonate concentration was
was observed experimentally in this study. Therefore similar to that of decreasing reaction temperature in
it is proposed that a combined AB substitution, and that precipitate size decreased and more spheroidal-
not a simple B-type substitution, occurs in this CHA shaped crystals were formed. Figure 2 illustrates this
precipitation system. effect with CHA crystals at 70°C, at bicarbonate con-
centrations of 10, 40, and 160 mM. For reaction tem-
peratures at which hydroxyapatite precipitates were
spheroidal, a slight decrease in size was noted with
MATERIALS AND METHODS
increasing bicarbonate concentration. However, acic-
ular crystallites were formed only for reactions where
Precipitation the bicarbonate ion concentration was low and the
reaction temperature was high. The effect of reaction
A precipitation reaction, based on that developed by Nel- conditions on the morphology of the crystallites is
son and Featherstone,5 was used for the preparation of hy- summarized in Figure 3.
droxyapatite and CHA with varying degrees of carbonate The carbon content of the apatites precipitated un-
substitution. A 130 mM solution of analytical grade (AR) der the various conditions are shown in Figure 4. It
tri-ammonium orthophosphate at pH >11 was dripped into appeared that the products obtained from 37°C and
a continuously stirred 210 mM solution of AR calcium ni- 45°C reactions had higher carbon contents than those
trate 4-hydrate over a period of approximately 2 h. Six re-
prepared in the 3°C and 70°C reactions, where bicar-
action temperatures were selected: 3°, 25°, 37°, 45°, 60°, 70°,
bonate ion concentrations of 160 mM were used in the
and 90°C. The 3°C temperature was achieved by immersing
the reaction vessel in an ice bath while other temperatures reaction. However, Figure 4 also indicates that the ma-
were maintained by use of a thermostatically controlled hot terial produced in the 3°C and 37°C reactions could
plate. Up to nine concentrations of AR sodium bicarbonate accommodate a larger maximum carbonate substitu-
were added to the phosphate solution at each temperature, tion than the CHA precipitated at 45°C and 70°C. The
between 0 mM and 640 mM. The resulting precipitates were maximum carbon content observed was 2.8 wt% (14
aged for 24 h before being washed and filtered. wt% carbonate), and there was a rapid increase in car-
bon content at low bicarbonate concentrations (0–10
mM ) for CHA at all precipitation temperatures inves-
tigated.
Characterization
It was found that the sodium content of a 4 wt%
CHA was < 0.1 wt%, that is, below the detection level
The crystallite morphology resulting from the precipita- of the apparatus, whereas an 11.5 wt% CHA contained
tion reaction was determined by transmission electron mi- 1.5 wt% sodium. The calcium and phosphorus weight
croscopy (TEM) with a JEOL 1200 EX2 (JEOL, Tokyo, Japan).
percent as a function of carbon weight percent of apa-
The precipitates were examined in bright-field mode and
using electron diffraction at magnifications typically up to
tites precipitated at 70°, 45°, 37°, 25°, and 3°C are
50,000X, using an accelerating voltage of 100kV. Carbon con- shown in Figure 5(a, b). The calcium and phosphorus
tent was determined using a Control Equipment Corpora- content appeared to decrease with increasing carbon
tion Model 240 XA CHN elemental analyzer. Calcium, phos- content, and this was unaffected by precipitation tem-
phorus, and sodium measurements were made with a Perki- perature. A least-squares fit of all the data [Fig. 5(a)]
n–Elmer Plasma 40 emission spectrometer. The except the outlying point at Ca = 32.5 wt%, circled,
reproducibility of this method was found to be better than gives a gradient of 0.8. The intercept of the line at x =
2% of the measured value. Infrared spectra were recorded 0 indicates that the precipitated HA is calcium defi-
on a Nicolet 800 FTIR spectrophotometer using a photo- cient. A definite relationship exists between the de-
acoustic sampling technique (MTEC Photoacoustic PAC200 crease in phosphorus content with the increase in car-
system). Spectra were obtained between wave numbers 400–
bon content that appears to be independent of precipi-
4000 cm−1. XRD patterns of CHA powders were collected
using a Siemens D5000 diffractometer with the data ana-
tation temperature [Fig. 5(b)].
lyzed using the Diffrac AT program. The phases present Infrared spectroscopy results indicated the presence
were determined by comparing the patterns with JCPDS of the major carbonate and phosphate peaks expected
standards.15 The lattice parameters of the CHAs were deter- in CHA for the samples precipitated at 3°, 37°, 70°, and
mined by Rietveld refinement of the diffraction profiles with 90°C, and these are summarized in Table I. The HA
a least-squares fitting program (G.S.A.S.).16 precipitated in the absence of bicarbonate ions con-
CARBONATE SUBSTITUTION IN PRECIPITATED HA 81

Figure 1. The effect of precipitation temperature on the size and morphology of HA precipitates at (a) 25°C, (b) 45°C, (c)
70°C, and (d) 90°C.

tained a small trace of carbonate, possibly from dis- temperature increased. The phosphate y1 band of the
solved atmospheric carbon dioxide. As the precipita- HAs prepared in bicarbonate-free conditions ap-
tion temperature increased, the carbonate y3 bands at peared to be unaffected by precipitation temperature.
higher wave numbers became less prominent, and at The y2 carbonate bands are observed at 875 cm−1 in
70° and 90°C, only those at 1455 and 1420 cm−1 were bicarbonate-free conditions at all temperatures.
observed. Two phosphate y3 bands were present and The effect of bicarbonate ion concentration on the
became increasingly separated as the precipitation infrared spectra for reactions performed at 3°C is dem-
82 BARRALET, BEST, AND BONFIELD

Figure 2. The effect of bicarbonate ion concentration on the size and morphology of CHA precipitates at 70°C: (a) 10, (b) 40,
(c) 160 mM, and (d) 320 mM.
CARBONATE SUBSTITUTION IN PRECIPITATED HA 83

Figure 3. Effect of bicarbonate concentration and precipi-

tation temperature on the morphology and phase of apatite
crystals.

onstrated in Figure 6. As might be expected, there was

a general increase in the ratio of carbonate/phosphate
peak relative intensity with increasing bicarbonate ion
addition. There also was evidence of both A- and B-
type substitution.5,17 Table I indicates that for a given
bicarbonate ion concentration, increasing reaction
temperature resulted in the disappearance of peaks at
higher wave numbers. The carbonate y2 band was lo-
cated at wave number 875 cm−1 at all reaction tem-
peratures in bicarbonate-free conditions. As the bicar-
bonate ion concentration increased, the peak position
appeared to move slightly to a lower wave number
(873 cm−1 ). Although the difference was small, it was
observed in all samples.
TEM studies indicated that at temperatures below
45°C and at high bicarbonate concentrations, the crys-
tal dimensions of the precipitate were small (in the
order of tens of nanometers). As a consequence, the
Figure 5. (a) Calcium and (b) phosphorus contents of
X-ray diffraction patterns of apatites formed under CHAs precipitated at 70°, 45°, 37°, 25°, and 3°C as a function
of carbon content.

these conditions were typified by broad peaks. There-

fore it was not possible to determine the lattice param-
eters of these apatites with a high level of accuracy.
The apatites precipitated at 70°C were of large enough
crystal dimensions to give XRD patterns suitable for
lattice parameter determination from structure refine-
ment. Figure 7(a,b) shows the c and a axis dimensions
as a function of carbonate content. It can be seen that
the c axis decreases slightly and then increases with
increasing carbonate substitution. The a axis shows
the reverse trend, increasing in size as the carbonate
content increases up to 4 wt%, thereafter it decreases.

DISCUSSION

Figure 4. Carbonate content of apatites precipitated at 3°, The results indicate that both the size and shape of
37°, 45°, and 70°C at various bicarbonate ion concentrations. the HA crystals can be controlled by altering both the
84 BARRALET, BEST, AND BONFIELD

TABLE I
Peak Positions of the FTIR Spectra of Some CHAs
Precipitated at Different Temperatures and Bicarbonate
Concentrations Compared with Reported Peak
Assignments
Species CO32− PO43− PO43− CO32−
Vibration y3 y3 y1 y2
mode
Sample cm−1 cm−1 cm−1 cm−1

3C0M 1574, 1557, 1534 1086, 1059 961 875

1488, 1454, 1420
3C01M 1557, 1456, 1418 1084, 1041 960 875
3C09M 1557, 1539 1419 1063, 1048 960 873
3C32M 1505, 1489, 1473, 1050, 1008 961 871
1457, 1434, 1419
3C48M 1505, 1488 < 1473 1052 960 872
< 1457, 1434, 1419
37C0M 1488, 1456, 1419 1090, 1061 961 875
37C04M 1488, 1456, 1419 1042 961 873 Figure 7. (a) Change in the c axis of CHA precipitated at
37C16M 1473, 1456, 1419 1045 961 872 70°C with increasing carbonate substitution; (b) change in a
37C32M 1473, 1457, 1419 1046 961 872 axis.
70C0M 1455, 1419 1090, 1049 962 875
70C01M 1455, 1419 1087, 1040 961 874
70C04M 1454, 1417 1042 961 873
70C16M 1456, 1418 1052 961 873 been observed here that the carbonate concentration of
70C24M 1467, 1450 < 1413 1039 961 873 the precipitate depends upon the precipitation tem-
90C0M 1455, 1420 1093, 1035 962 875 perature. However, variations in phosphate and cal-
Key: XCYM; X indicates the reaction temperature; Y indi- cium weight percent with carbon content were not
cates the bicarbonate concentration in mM/10. temperature dependent, suggesting the mechanism of
carbonate substitution was unchanged at precipitation
precipitation temperature and the bicarbonate ion temperatures of 3°–90°C.
concentration. The effects of temperature and carbon- On the basis of IR and XRD studies, changes in crys-
ate content on crystal size have been observed previ- tal lattice parameters have been attributed to different
ously by several investigators, most studies reporting types of substitution location: the a axis expands and
a broadening of the XRD patterns at temperatures be- the c axis contracts as a result of an A-type carbonate
low 80°C or at high carbonate contents.4,6,7,18 It has substitution, and the reverse trend is observed for a
B-type substitution. In this study, FTIR spectra
showed that at low carbonate contents and at low pre-
cipitation temperatures there are y3 carbonate bands in
the range of 1550–1500 cm−1. Bands in this region pre-
viously have been assigned to carbonate in the hy-
droxyl or A-site,10,18,19 Nelson and Featherstone5 pre-
pared B-type apatites at 90°C and found that low car-
bonate content apatites (2 wt%) had y3 carbonate
bands at 1550, 1500, 1469, 1454, and 1415 cm−1. These
additional bands were not observed at high precipita-
tion temperatures in this study.
The XRD refinement data clearly showed an in-
crease in a axis at carbonate contents up to 4 wt% (Fig.
7) while a decrease was observed at higher contents.
The opposite trend was observed in the c axis mea-
surements. Nelson and Featherstone,5 who used the
same precipitation method as reported here, observed
a similar change in lattice parameters with carbonate
content and attributed this effect to the incorporation
of water in the structure. Nelson and Featherstone5
hypothesized that sodium substituted for calcium in
CHA and reported sodium contents of between 0.4
Figure 6. Effect of bicarbonate concentration on IR carbon- and 0.7 wt% for two samples of CHA with similar
ate and phosphate bands of CHAs precipitated at 3°C. carbonate contents. A large variation in sodium con-
CARBONATE SUBSTITUTION IN PRECIPITATED HA 85

tent was reported at other carbonate contents, and this for three phosphate ions with the formation of one
would tend to cast doubt on a definite relationship hydroxyl and two calcium vacancies.
between carbonate and sodium in their samples. By The results obtained in this study were compared
comparison, LeGeros,7 who used a different reaction with all the current models by plotting the predicted
and showed a clear relation between sodium and car- weight percent calcium and phosphorus [Fig. 5 (a, b)].
bonate content of CHA, reported sodium contents of It was found that none of these models accurately de-
2.4 wt% for a 4 wt% precipitated CHA. In this study at scribed the data from this study. As a result of this, a
low carbonate content (4 wt%), the sodium content new model has been devised that could describe the
was < 0.1 wt% in the CHA precipitated at 3°C whereas simultaneous substitution of the A and B sites. First,
a more substituted CHA (11.5 wt% carbonate) precipi- the A and B sites were filled in equal proportions as
tated at the same temperature was found to contain described by:
1.5 wt% sodium. These results imply that either so-
dium is included as an impurity second phase or it Ca10−p(PO4)6−p(CO3)p(OH)2−p(CO3)p (5)
does not play a major role in the mechanism of sub- where 0 < p < 1.
stitution of carbonate in this system. This substitution was thought to cease before all of
There are a number of models for the substitution of the A sites were full since evidence for hydroxyl
carbonate in hydroxyapatite. They have been derived groups was observed on the IR spectra of high car-
variously on the basis of IR, XRD, and chemical analy- bonate content CHA.
sis of carbonate apatites and on the fundamental re- When p = 1, further substitution then could occur by
quirement for charge balance within a crystal lattice. the Kühl and Nebergal model11:
One of the first was that of Kühl and Nebergal,11 who
proposed the formula for precipitated carbonate apa- Ca9−y(PO4)5−y(CO3)1+y(OH)1−y (6)
tite as: where y < 0.8.
Ca10−y+u(PO4)6−y(CO3)y(OH)2−y+2u (1) The graphs of Ca and P wt%, with C wt% predicted
by this and previous models, were plotted and com-
where 0 < y < 2, and 0 < 2u < y. This model assumes
pared with our experimental data. Table II. shows the
that when one phosphate ion is replaced by one car-
slopes of linear fits of the experimental data and pre-
bonate ion, generating one calcium and one hydroxyl
vious carbonate substitution models and the new AB
vacancy, a degree (u) of these vacancies may be filled.
model. It was found that most models gave a reason-
This formula generally is accepted for precipitated
able fit of the P wt% as a function of C wt%, but none
CHA.13
adequately described the change in calcium content
The formula was later modified slightly18 when it
with carbon except the AB model devised in this
was proposed that u ion pairs of [CO3, OH] could be
study.
substituted without the need for vacancy formation:
Ca10−y(PO4)6−2y(CO3OH)2y(OH)(2−y) (2)
17
This was refined slightly by Vignoles, who took into CONCLUSIONS
account the effect of ionic concentration:
Ca10−y+u hy−u(PO4)6−y(CO3 h)y − u It has been shown that the size and morphology of
(CO3 OH)u(OH)2−y+u hy−u (3) precipitated carbonate hydroxyapatites can be af-
where 0 ø y ø 2 and 0 ø 2u ø y; h is a vacancy, and
u depends on the preparation conditions. TABLE II
Comparison of the Gradients of Composition Graphs
Nelson and Featherstone,5 who investigated the
Determined Experimentally By the AB Mechanism and
same CHA precipitation system as that studied in this From Previously Published Mechansims
report, concluded on the basis of their results that the
Gradient Ca/C P/C Ca/P
mechanism of substitution could be described by:
Experimentally observed −0.8 −1.3 0.7
Ca10−yNay{(PO4)6−y(CO3)4y/3}(OH)(2−2y/3)) (4)
AB mechanism −0.9 −1.1 0.8
All these workers observed that carbonate could be Vignoles (1984)
u=0 −0.4 −1.1 0.3
included in the A site at low carbonate levels but did
u = y/2 POSITIVE −2.5 NEGATIVE
not include this observation in their substitution mod- Nelson and −1.4 −1.4 1.0
els. One of the few reports of a mixed AB mechanism Featherstone (1982)
was that of Driessens et al.,12 who reported that high Bonel (1972) POSITIVE −1.1 NEGATIVE
temperature AB-type CHA was substituted by a Kühl and Nebergal (1963)
u=0 −0.3 −1.1 0.2
mechanism whereby a small quantity of carbonate
u = y/2 POSITIVE −1.7 NEGATIVE
filled the A site and three carbonate ions substituted
86 BARRALET, BEST, AND BONFIELD

fected by control of the precipitation temperature and tion in the Apatite Structure, Ph.D. Thesis, New York University,
bicarbonate ion concentration. The maximum carbon- New York, 1967.
ate content at a given bicarbonate concentration de- 8. N. C. Blumenthal, F. Betts, and A. S. Posner, ‘‘Effect of carbon-
ate and biological macromolecules on formation and proper-
pends on the temperature of the precipitation reaction.
ties of hydroxyapatite,’’ Calcif. Tiss. Res., 18, 81–90 (1975).
The FTIR and XRD data indicated that the carbonate
9. G. Bonel, ‘‘Contribution à 1 étude de la carbonation des apa-
was substituted into the A and B sites of the hydroxy- tites. II. Synthèse et étude des propriétés physico-chimiques
apatite lattice in products where carbonate contents des apatites carbonatées de type B. III. Synthèse et étude des
were low. At higher carbonate contents, B-type sub- propriétés physico-chimiques des apatites carbonatés dans
stitution appeared to be more prevalent. On the basis deux types de sites. Evolution des spectres infrarouge en fonc-
of this observation, a new AB carbonate substitution tion de la composition des apatites,’’ Ann. Chim., 147, 127–144
(1972).
mechanism has been proposed that better fits our ex-
10. M. Vignoles, G. Bonel, D. W. Holcomb, and R. A. Young, ‘‘In-
perimental results than previous models. This model
fluence of preparation conditions on the composition of type B
may describe the mechanism of carbonate incorpora- carbonated hydroxyapatite and on the localisation of the car-
tion into hydroxyapatite in vivo. bonate ions,’’ Calcif. Tiss. Int., 43, 33–40 (1988).
11. G. Kühl and W. H. Nebergal, ‘‘Hydrogenphosphat- und car-
The authors acknowledge the continuing support of bonat apatat,’’ Z. Anorg. Allgem. Chem., 324, 313–320 (1963).
EPSRC for the IRC program and for the provision of a re- 12. F. C. M. Driessens, R. M. H. Veerbeck, and P. Kiekens, ‘‘Mecha-
search studentship (JB). We also gratefully acknowledge the nism of substitution in carbonate apatites,’’ Z. Anorg. allgem.
experimental assistance of Dr. J. Knowles and I. Abrahams Chem., 504, 195–200 (1983).
and helpful discussions with Professor J. Elliott. 13. J. C. Elliott, Structure and Chemistry of the Apatites and Other
Calcium Orthophosphates, Studies in Inorganic Chemistry, Vol. 18,
Elsevier, Amsterdam, 1994, pp. 234–239.
14. E. A. P. De Maeyer, R. M. H. Verbeeck, and D. E. Nassens,
‘‘Stoichiometry of Na+- and CO32-containing hydroxyapatite
References
obtained by the hydrolysis of monetite,’’ Inorg. Chem., 32, 5709–
5714 (1993).
1. F. C. M. Driessens, ‘‘Formation and stability of calcium phos-
15. ‘‘Powder diffraction file (inorganic phases),’’ JCPDS Card No.
phates in relation to the phase composition of the mineral in
9-432, JCPDS International Centre of Diffraction Data, Newton
calcified tissue,’’ in Bioceramics of Calcium Phosphate, K. De-
Square, Pennsylvania, USA 1980.
Groot (ed.), CRC Press, Inc., Bocca Raton, Florida, 1983, pp.
1–32. 16. A. C. Larson, R. B. von Dreele, and M. Lujan, Jr., G.S.A.S.—
2. R. Z. LeGeros, O. R. Trautz, E. Klein, and J. P. LeGeros, ‘‘Two Generalized Crystal Structure Analysis System, Neutron Scat-
types of carbonate substitution in apatite substitution,’’ Specia- tering Centre, Los Alamos National Laboratory, California,
lia Experimentia, 25, 5–7 (1969). 1990.
3. J. C. Elliott, The Crystallographic Structure of Dental Enamel and 17. G. Bonel, ‘‘Contribution à l’étude de la carbonation des apa-
Related Apatites, Ph.D. Thesis, University of London, 1964. tites. I. Synthèse et étude des propriétés physico–chimiques
4. N. C. Blumenthal and A. S. Posner, ‘‘Hydroxyapatite: Mecha- des apatites carbonatées du type A,’’ Ann. Chim., 147, 65–88
nism of formation and properties,’’ Calc. Tiss. Res., 13, 235 (1972).
(1973). 18. C. Vignoles, Contribution à L’étude de L’influence des Ions Alcalins
5. D. G. A. Nelson and J. D. B. Featherstone, ‘‘Preparation, analy- sur la Carbonation dans les Sites de Type B, Thèse, L’Institut Na-
sis and characterization of carbonated apatites,’’ Calcif. Tiss. tional Polytechnique de Toulouse, Toulouse, France, 1984.
Int., 34, 569–581 (1982). 19. Y. Doi, Y. Moriwaki, T. Aoba, M. Okazaki, J. Takahashi, and K.
6. J. -C. Labarthe, G. Bonel, and G. Montel, ‘‘Sur la structure et les Joshin, ‘‘Carbonate apatites from aqueous and nonaqueous
propriétés des apatites carbonatées de Type B phospho- media studied by esr, IR, and X-ray diffraction: Effect of NH4+
calciques,’’ Ann. Chim., 7, 289–301 (1973). ions on crystallographic parameters,’’ J. Dent. Res., 61, 429–34
7. R. Z. LeGeros, Crystallographic Studies of the Carbonate Substitu- (1982).