Effect of phase transformations during synthesis on the chemical composition and structure of calcium-deficient hydroxyapatite

Article abstract of DOI:10.1134/S0020168507100160

The chemical and phase changes during the precipitation of calcium-deficient hydroxyapatite from a solution of natural chalk in a mixture of phosphoric and nitric acids, with the use of aqueous ammonia as the precipitant, have been studied by potentiometric titration and physicochemical analysis. The initial solution concentration and precipitation time are shown to have a significant effect on the composition and structure of the precipitate. The processes in the solution and precipitate are interpreted in terms of thedegree of protonation of phosphate ions under various conditions.

Full text of DOI:10.1134/S0020168507100160

ISSN 0020-1685, Inorganic Materials, 2007, Vol. 43, No. 10, pp. 1119–1127. © Pleiades Publishing, Inc., 2007.
Original Russian Text © N.V. Kitikova, I.L. Shashkova, Yu.G. Zonov, O.A. Sycheva, A.I. Rat’ko, 2007, published in Neorganicheskie Materialy, 2007, Vol. 43, No. 10, pp. 1246–
1
255.
Effect of Phase Transformations during Synthesis
on the Chemical Composition and Structure of Calcium-Deficient
Hydroxyapatite
N. V. Kitikova, I. L. Shashkova, Yu. G. Zonov,
O. A. Sycheva, and A. I. Rat’ko
Institute of General and Inorganic Chemistry, Belarussian Academy
of Sciences, ul. Surganova 9, Minsk, 220072 Belarus
e-mail: kitikova@igic.bas-net.by
Received October 3, 2006; in final form, March 14, 2007
Abstract—The chemical and phase changes during the precipitation of calcium-deficient hydroxyapatite from
a solution of natural chalk in a mixture of phosphoric and nitric acids, with the use of aqueous ammonia as the
precipitant, have been studied by potentiometric titration and physicochemical analysis. The initial solution
concentration and precipitation time are shown to have a significant effect on the composition and structure of
the precipitate. The processes in the solution and precipitate are interpreted in terms of the degree of protonation
of phosphate ions under various conditions.
DOI: 10.1134/S0020168507100160
INTRODUCTION
metastable intermediates determines the composition
and properties of the final reaction product. Therefore,
the key to understanding the process of HA formation
under various experimental conditions, in particular for
the targeted synthesis of HA with controlled properties,
lies in studying phase formation and its effect on the
stoichiometry of the material [10].
The crystallization of hydroxyapatite (HA),
Ca (êé ) (éH) , is of considerable scientific and
10
4 6
2
technological interest because HA is widely used in
biomedical and other technological applications both as
an individual compound and in composites. Special
attention has been given to nonstoichiometric HA,
whose composition can be represented by the general
HA is commonly precipitated by pouring together
calcium nitrate and ammonium hydrogen phosphate
solutions having the same pH, from 6 to 12 [5, 7], or by
neutralizing a calcium hydroxide suspension with a
phosphoric acid solution [6, 11]. Calcium carbonate is
often used as the calcium precursor, especially in the
synthesis of carbonated HA [12]. Use can also be made
formula Ca1 (Hêé ) (êé ) (éH)2 – x with 0 ≤ x ≤ 2.
0 – x
4 x
4 6 – x
Calcium-deficient HA is an analog of biological HA,
has a higher solubility [1], bioactivity, and biocompati-
bility [2] in comparison with stoichiometric HA, and
offers good sorptive properties in chromatographic sep-
arations of bioorganic polymers [3] and considerable
catalytic activity [4, 5].
of ëa(éH) suspensions prepared from CaO after firing
2
The best approach for the synthesis of nonstoichio- calcium carbonate at 1000°ë for 24 h [13]. In contrast
metric HA is chemical precipitation from solution [5, to conventional processes, we prepared calcium-defi-
6
]. The mechanism of HA formation under various
cient HA by precipitation from a pH 1 reaction mixture
based on natural chalk, using aqueous ammonia as the
precipitant. The chalk was dissolved in a mixture of
phosphoric and nitric acids. Nitric acid was added in
order to maintain the pH value at 1, thereby preventing
early formation of calcium hydrogen phosphates,
which otherwise would have impeded further chalk dis-
experimental conditions has been the subject of exten-
sive studies because the processes involved are rather
complex: HA formation in supersaturated solutions is
always preceded by the precipitation of one or more
intermediates [7, 8]. These include crystalline anhy-
drous calcium hydrogen phosphate (CHP), ëaHêé4;
calcium hydrogen phosphate dihydrate (CHPD), solution and HA crystallization [14]. Phase formation
ëaHêé · 2H é; and ëa H(êé ) · 2.5H é; also pos- in different stages of the reaction was studied by ana-
4
2
4
4 3
2
sible is the formation of amorphous tricalcium phos- lyzing the products formed at the equivalence points in
phate (TCP), ëa (êé ) · nH é [8, 9]. The precipita- the pH curves obtained in the course of precipitation.
3
4 2
2
tion sequence of these phases as a function of pH and As shown in many studies, the factors determining the
solution composition depends not only on the solubility crystallization behavior of HA include the pH and con-
product of the corresponding phosphate but also on centration of the starting solutions and the reaction time
kinetic factors [7, 8]. The transformation sequence of [7]. The purpose of this work was to study the effect of
1
119

1
120
KITIKOVA et al.
Table 1. Synthesis conditions and characteristics of intermediate phases in calcium-deficient HA precipitation
Solution parameters
Characteristics of phases
at the equivalence point
Synthesis conditions
at the equivalence points
Intermediate
sample no.
–
2+
[
Ca], [H PO ], NH · H O introduc-
éç /ë‡
molar ratio
Ca/P
2
3
4
3
2
ç
S, m /g
mol/l
mol/l
tion rate, ml/min
molar ratio
1
2
3
4
5
6
0.11
0.42
0.70
0.42
0.42
0.42
0.06
2.7*
1-1
4.2
7.4
9.5
1.95
2.84
5.29
1.74
1.56
1.45
47
76
88
1
1
-2
-3
0.25
0.42
0.25
0.25
0.25
2-1
2-2
3.6
7.1
9.5
0.77
1.42
2.84
1.11
1.19
1.40
16
58
85
2
2
.7
.7
2
-3
3-1
3-2
3.2
6.7
9.5
0.57
1.20
2.46
1.22
1.06
1.36
17
33
87
3
-3
4.0**
2.7**
1.3**
–
–
–
3.8
0.97
1.57
2.76
–
–
–
–
–
–
6
9
.6
.5
3.8
0.86
1.54
2.76
–
–
–
–
–
–
6
9
.9
.5
3.4
0.72
1.36
2.53
–
–
–
–
–
–
7
9
.1
.5
*
10.5 M aqueous ammonia.
*
*5.3 M aqueous ammonia.
the initial solution concentrations and precipitation rate
Dissolution and precipitation were conducted at a
of the conversion of metastable intermediates to cal- temperature of 20°ë, which was maintained by a water
cium-deficient HA.
EXPERIMENTAL
thermostat. The synthesized samples were stored in the
mother solution for 24 h, followed by washing with hot
distilled water, filtration, and drying, first in air and then
at 150°ë for 3 h (the reaction intermediates were dried
at 65°ë).
The calcium source used was natural chalk (96%-
1
pure calcite, ëaëé ). In our preparations, we used the
We carried out several series of experiments, of
which two were selected to illustrate the effects of the
initial solution concentrations (series 1) and precipita-
tion rate (series 2). The initial calcium and phosphate-
ion concentrations in the reaction mixtures and the pre-
cipitant (aqueous ammonia) concentration and intro-
duction rate are given in Table 1. To study the phase for-
mation process, the reaction intermediates in series 1
were taken at pH values corresponding to equivalence
points.
3
size fraction 0.09–0.25 mm.
The chalk was dissolved in a reaction vessel con-
taining a stirrer and the electrodes of a pH meter. Cal-
cium carbonate (10 g) was gradually added
(
ꢀ0.1 g/min) to an appropriate amount of phosphoric
acid, while maintaining the pH at 1 by adding 30%
nitric acid. The initial Ca/P molar ratio in the solutions
corresponded to the HA stoichiometry, 1.67. To the
resultant solution was added aqueous ammonia at a
constant rate (deviation within ±0.1 ml/min), using a
The structure and composition of the synthesized
micropump, until the solution pH reached 9.5. During materials were determined by x-ray diffraction (XRD),
precipitation, the solution pH was visualized in the IR spectroscopy, thermal analysis (DTA + TG), and
form of a potentiometric titration curve using an XY chemical analysis. XRD patterns were collected on a
recorder.
DRON-3 powder diffractometer (CuK radiation). Lat-
tice parameters were refined using the 002, 211, 222,
213, 321, 410, and 004 reflections. IR spectra were
α
1
Vitebsk oblast, Belarus.
INORGANIC MATERIALS Vol. 43 No. 10 2007

EFFECT OF PHASE TRANSFORMATIONS DURING SYNTHESIS
1121
pH
d(pH)/d(OH/Ca) pH
d(pH)/d(OH/Ca)
200
1
0
8
6
4
2
(
a)
(b)
4
1
1
5
0
50
00
0
3
2
1
6
5
6
'
5
'
3
'
4
'
2
'
1
'
0
1
2
3
4 0
1
2
3
–
2+
OH /Ca molar ratio
Fig. 1. Variation in pH and derivative curves (primed numbers) for the synthesis of calcium-deficient HA (a) from solutions of dif-
ferent concentrations and (b) at different precipitation rates; the numbers at the curves identify the samples as referenced in Table 1; [NH3 ·
H O] = (a) 10.5 and (b) 5.3 mol/l; [H êé ] = (1) 0.06, (2, 4–6) 0.25, and (3) 0.42 mol/l; precipitation rate of (4) 4.0, (1–3, 5) 2.7,
2
3
4
and (6) 1.3 ml/min.
–1
measured from 400 to 4000 cm on a Midac Fourier ing phase and the mother solution. The resulting precip-
transform IR spectrometer, using 2-g samples prepared itate has a variable composition [17]. The formation of
as KBr pellets.
reaction intermediates depends on the reaction condi-
tions, as evidenced by chemical analysis, XRD, IR
spectroscopy, and DTA data.
Thermal analysis was performed in a Paulik–Pau-
lik–Erdey thermoanalytical system during heating in
air from 20 to 1000°C at a rate of 10°C/min. The spe-
HA crystallization in dilute solutions (Table 1, sam-
ple 1; Fig. 1a, curve 1) is preceded by the formation of
poorly crystallized TCP (Fig. 2) with a large Ca/P
molar ratio (Table 1). In the IR spectra of the reaction
2
cific surface area S (m /g) was determined by low-tem-
perature nitrogen adsorption measurements. The chem-
ical composition of solid products was determined by
elemental analysis after dissolving a sample in nitric
acid. Calcium was determined by back titration with
–1
intermediates (Fig. 3), the bands at 3580 and 630 cm ,
–
due to constitutional éH groups, are missing, but the
Trilon B and ZnSO solutions [15], and phosphorus, by
4
–1
band around 875 cm attests to the presence of trace
the phosphovanadomolybdate spectrophotometric
method [16]. The uncertainty in our determinations was
estimated at ±0.02 mmol/g.
levels of the hydrogen phosphate ion. This impurity
accounts for the weight loss at 330 and 720°ë, due to
the polymerization of the hydrogen phosphate ion [18]
(Fig. 4). Raising the solution pH to 9.5 leads to the
RESULTS AND DISCUSSION
topochemical transformation of TCP to HA, as evi-
denced by the XRD pattern of the as-precipitated final
product (Fig. 2).
Figure 1 shows potentiometric titration curves illus-
trating the variation in the pH of the mother solution in
pseudoequilibrium with the precipitate during the intro-
Increasing the initial solution concentration causes
duction of the precipitant for a varied initial solution precipitation to begin at pH 2–3 (Fig. 1a, sample 2,
concentration and a constant rate of precipitation from
solutions of various concentrations (Fig. 1a) and for a
varied precipitant introduction rate at a constant solu-
tion concentration (Fig. 1b). The parameters of the
solutions at the equivalence points in their titration
curves and the chemical compositions of the corre-
sponding intermediates are listed in Table 1.
curve 2). The resultant well-crystallized phase had a
reduced Ca/P molar ratio (Table 1) and was identified
by a combination of physicochemical characterization
techniques (Figs. 2–4) as CHP [19–21]. In the pH range
3
.7–4.5, the plateau is broader than that in curve 1.
Starting above pH 9, HA begins to form, as evidenced
by extra lines in the XRD pattern (Fig. 2) and the
The presence of a family of curves representing the increase in Ca/P molar ratio (Table 1). This behavior is
precipitation process as a function of reaction time associated with the fact that, in alkaline solutions, the
(
Fig. 1b) is characteristic of systems in which the for- hydrogen phosphate solubility is higher [22], and the
mation of primary products is followed by secondary dissolution of the intermediate is followed by HA crys-
chemical processes at the interface between the form- tallization.
INORGANIC MATERIALS Vol. 43 No. 10 2007

1
122
KITIKOVA et al.
8
0
200
200
I
(a)
(b)
(c)
II
III
1
1
-1
-2
2-1
3
3
-1
-2
2
2
-2
-3
1
-3
3-3
4
0
30
20
10
40
30
2
20
θ, deg
10
40
30
20
10
Fig. 2. XRD patterns of reaction intermediates in samples 1–3 at the (1-1, 2-1, 3-1) first and (1-2, 2-2, 3-2) second equivalence points
and (1-3, 2-3, 3-3) those of the as-precipitated final product; the numbers at the curves identify the samples as referenced in Table
1
; (I) HA, (II) ëaHêé · 2H é, (III) ëaHêé .
4 2 4
Curve 3 in Fig. 1, representing the variation in pH the precipitant leads to washing out of the nitrate ion,
during HA synthesis at the highest solution concentra- which is then missing in the final product. The presence
tion, [H PO ] = 0.42 mol/l, has a different shape. In the of the nitrate ion and quantitative changes in its content
3
4
first step, we observe the formation of an amorphous are also demonstrated by the TG curves, which attest to
phase (Fig. 2) with a trace amount of the dihydrogen NO
removal at temperatures from 500 to 600°ë.
2
–
phosphate ion, H PO . The presence of this ion is evi-
2
4
In addition, the precipitation process is influenced
denced by the IR spectrum, which shows, in addition to significantly by the synthesis duration, which is deter-
–
1
the band at 875 cm , several features in the range mined by the precipitant introduction rate. The varia-
–
1
3
000–3500 cm . These latter arise from the stretching tion in pH has then some features in common with the
mode ν(éH) and are characteristic of the spectrum of curves obtained at different solution concentrations.
ëa(H êé ) · H é [20]. The presence of the dihydro- Curve 4 in Fig. 1b represents the variation in the pH of
2
4 2
2
the reaction mixture at the highest precipitation rate and
is similar in shape to curve 1 in Fig. 1a, which was
gen phosphate ion is also evidenced by the endotherms
at 150, ꢀ270°ë, and 335°ë in the DTA curve, which
–
2+
correspond to three-step water removal [23]. The first obtained in dilute solutions. Accordingly, the éH /ëa
jump in pH is followed by a decrease, rather than by a molar ratio at the equivalence points in curve 4 is larger
plateau, and then by a second jump in pH, due to the compared to curves 5 and 6, which were obtained at
formation of CHP (Fig. 2), which has a less perfect slower precipitation rates (Table 1). Moreover, with
crystal structure in comparison with sample 2. Like in decreasing precipitation rate, curves 5 and 6 show a
solutions of medium concentrations, the final step of characteristic drop in pH after the first step of precipi-
HA formation occurs at pH above 9.0, where the inter- tation, like at high solution concentrations (Fig. 1a,
mediate, less thermodynamically stable phases dis- curve 3).
solve, and the most stable phase, HA, is formed.
Reducing the precipitant concentration and increas-
A noteworthy feature of the primary phases that ing the reaction time have nearly the same effect, as fol-
form in the first step at phosphoric acid concentrations lows from comparison of samples 2 and 5 (Fig. 1a,
from 0.25 to 0.42 mol/l is the high concentration of curve 2; Fig. 1b, curve 5).
–
NO ions, which seem to be adsorbed by the precipi-
3
This difference between the effects of initial solu-
tate surface [10] and can be identified from the IR band tion concentration and pH on the phase formation pro-
–
1
at 1380 cm , which becomes stronger with increasing cess can be accounted for by the presence of a diversity
initial solution concentration. Increasing the amount of of phosphate ions with different degrees of dissociation
INORGANIC MATERIALS Vol. 43 No. 10 2007

EFFECT OF PHASE TRANSFORMATIONS DURING SYNTHESIS
1123
1
-1
1
2
-2
-1
2
3
3
-2
-1
-2
4
000
3500
3000
2500
2000
1500
1000
500
–
1
Wavenumber, cm
Fig. 3. IR spectra of reaction intermediates in samples 1–3 at the (1-1, 2-1, 3-1) first and (1-2, 2-2, 3-2) second equivalence points;
the numbers at the curves identify the samples as referenced in Table 1.
and protonation, which were studied in detail earlier
24, 25].
As pointed out by Abbona et al. [9], HA formation
in uncontrolled-pH processes is due to the reaction
between the dihydrogen phosphate and calcium ions,
accompanied by a decrease in pH and the predominant
[
In the first step of H êé dissociation, at pH 1–4, the
3
4
solution contains, in addition to the dihydrogen phos- formation of CHPD, CHP, ëa H(êé ) · 2.5H é, and
4
4 3
2
–
TCP, which crystallize more rapidly:
phate ion (H PO ), undissociated H êé molecules
2
4
3
4
–
4
–
2–
2+
+
ëa + H PO
ëaHêé + H ,
(1)
(2)
and H(H PO ) and (H PO ) dimers. In the second
2
4
2
4
2
2
4
2
step of dissociation, the solution contains both mono-
–
4
2
+
+
3
ëa + 2H PO
ëa (êé ) + 4H .
3 4 2
2
2
4
–
–
4
2–
2
mers, HPO , and H PO , and dimers, H (HPO )
2
2
4
That reaction (1) prevails in the first step of precipita-
tion was pointed out in many reports (see, e.g., [7, 8]).
Moreover, according to Heughebaert et al. [10], the
forming solid phase maintains electroneutrality owing
3
–
and H(HPO ) . As follows from earlier data, at the
4
2
lowest initial H êé concentration in this study,
3
4
0
.06 mol/l, the protonated form H êé prevails up to
2+
–
3
4
to Ca and éH removal, whereas the hydrogen phos-
phate ion is incorporated into the crystalline phase,
forming an apatite structure. For this reason, the rela-
tionship between the calcium and dihydrogen phos-
pH 2.5, whereas at the highest concentration,
.42 mol/l, the concentrations of the dihydrogen phos-
phate ion and protonated molecules differ very little. At
0
pH > 2.5, the degree of protonation decreases markedly phate ion concentrations plays a central role in deter-
with decreasing initial solution concentration.
mining the reaction path.
INORGANIC MATERIALS Vol. 43 No. 10 2007

1
124
KITIKOVA et al.
(
a)
(b)
(c)
2
2
-1
-2
3
3
-1
-2
1
1
-1
-2
DTA
0
0
0
0
1
1
-1
-2
2
4
6
TG
2-2
-1
3
3
-1
-2
2
0
200 400 600 800
0 200 400 600 800
0 200 400 600 800
t, °C
Fig. 4. TG and DTA curves of reaction intermediates; the numbers at the curves identify the samples as referenced in Table 1.
Given the degree of protonation of phosphate ions, The minimum in the potentiometric titration curve
an important feature of the reaction medium in HA syn- (Fig. 1a, curve 3) after the first jump in pH is then, most
thesis in the dilute solutions (Fig. 1a, curve 1) is the likely, due to the insufficient precipitant introduction
higher content of the acid in predominantly low-reac- rate, so that the amount of released protons increases
tivity, molecular form, which results in a significant more rapidly than that in the dilute solutions. The
excess of calcium ions compared to high-reactivity increase in the degree of protonation of phosphate ions
phosphate ions. Under such conditions, the reaction after dissociation in more concentrated solutions in the
–
only begins after a significant éH excess is introduced second step also contributes to the considerable rise in
and the pH reaches a level needed for reaction (2) to the concentration of dihydrogen phosphate ions, which
2
+
react with calcium by reaction (2). This accelerates the
reaction compared to the precipitation from the dilute
solutions, as evidenced by the relative heights of the
peaks in the derivative curves (Fig. 1a). If the precipita-
tion and neutralization processes cancel one another, as
at the medium solution concentrations (Fig. 1a, curve 2),
the curve has a plateau. The minimum in the potentio-
metric titration curves at reduced precipitant concentra-
tions (Fig. 1b, curve 5) or precipitant introduction rates
take place. The unreacted Ca ions are actively sorbed
by the surface of the forming precipitate, which fea-
tures a high adsorption capacity because of the small
size of the primary particles and, accordingly, the rela-
tively large specific surface [10]. This accounts for the
high calcium content of the solid product in the first
step (Table 1). Moreover, the HA formation through
amorphous TCP by reaction (2) is favored by the more
–
2+
rapid rise in éH /ëa molar ratio, which ensures neu-
tralization of the released protons. At a solution pH near 7,
the acid dissociation in the first step is followed by a
second jump in pH (Fig. 1a), accompanied by reac- controlled pH conditions is governed by the relation-
tion (1), which reduces the Ca/P molar ratio in the ship between the initial calcium concentration and the
(
Fig. 1b, curves 5, 6) lends support to this assumption.
Thus, the formation of reaction intermediates under
forming material (Table 1).
hydroxyl concentration in the precipitant solution and
also by the nature and amount of phosphate ions result-
ing from dissociation. Because of the dissimilarity
between the processes involved in crystallization, the
final reaction products differ markedly not only in
The same reaction underlies the formation of cal-
cium hydrogen phosphates in the second step in more
concentrated solutions, containing a larger fraction of
–
H PO ions (Fig. 1a, curve 2). At the highest solution chemical and phase compositions but also in crystal
2
4
structure and morphology (Table 2, Figs. 5, 6).
–
concentration in this study, the presence of the H PO4
2
According to XRD and chemical analysis data, the
synthesized HA is a compound of variable composi-
tion, with a Ca/P molar ratio from 1.34 to 1.60, and has
the same hexagonal structure as pure HA, even at the
largest deviation from stoichiometry and in the pres-
ence of calcium hydrogen phosphate impurity (Fig. 5).
ion in the reaction intermediate, evidenced by its IR
spectrum, indicates that the dimeric phosphate ions
react with calcium:
–
2
+
+
ëa + H(H PO )
ëa(H êé ) + H .
(3)
2
4
2
2
4 2
INORGANIC MATERIALS Vol. 43 No. 10 2007

EFFECT OF PHASE TRANSFORMATIONS DURING SYNTHESIS
1125
Table 2. Characteristics of final products in the synthesis of calcium-deficient HA
Lattice parameters
2
Final sample no.
Ca/P molar ratio
S, m /g
a, nm
c, nm
V, nm3
1
2
3
4
5
6
1.60
1.51
1.40
1.55
1.50
1.34
0.939 ± 0.002
0.947 ± 0.003
0.948 ± 0.002
0.947 ± 0.002
0.950 ± 0.003
0.951 ± 0.003
0.690 ± 0.001
0.693 ± 0.001
0.695 ± 0.001
0.693 ± 0.001
0.693 ± 0.001
0.693 ± 0.001
0.528 ± 0.002
0.539 ± 0.003
0.541 ± 0.002
0.538 ± 0.002
0.542 ± 0.003
0.543 ± 0.003
143
93
80
99
90
84
–1
The present XRD and IR spectroscopy data lead us to known IR absorption bands at 875 and 1640 cm , by a
assume that Ca-deficient HA may form by different
processes, which can be represented, according to Rod-
riguez-Lorenzo [26], by three models:
–1
band at 1215 cm , which was missing in our spectra
Fig. 6).
(
The present IR spectroscopy data attest to partial
surface adsorption of hydrogen phosphate groups,
substitutions in the crystal lattice, and
2
–
substitution of hydrogen phosphate ions, HPO , by
4
3
4
–
–1
the formation of an intercrystalline mixture of HA PO
groups, as evidenced by the band at 875 cm ,
due to stretches of the hydrogen phosphate group, and
the band at 1640 cm , corresponding to the bending
mode of water [27]. These bands were present in all of
the spectra. At the same time, only in sample 1 was this
and ëa H(êé ) · 2.5H é.
4
4 3
2
–1
Heughebaert et al. [10] identified adsorbed hydro-
gen phosphate groups on the surface of samples which
were not washed thoroughly enough to remove the
residual mother solution. According to their results, substitution not accompanied by changes in the bands
such groups are characterized, in addition to the well near 3500 cm , due to vibrations of water molecules.
–1
(
a)
(b)
4
1
2
5
6
3
4
0
30
20
1040
θ, deg
30
20
10
2
Fig. 5. XRD patterns of the final products; the numbers at the curves identify the samples as referenced in Table 1; the phases are
designated as in Fig. 2.
INORGANIC MATERIALS Vol. 43 No. 10 2007

1
126
KITIKOVA et al.
eral tendency for similar potentiometric titration curves
1
and, accordingly, that the processes involved are identi-
cal. During precipitation from dilute solutions, which
2+
contain an excess of Ca ions relative to the high-reac-
2
3
–
tivity phosphate ions, H PO , because of the dissocia-
2
4
tion in the first step and the high degree of protonation
of phosphate ions, or precipitation at high rates, which
produce an excess of hydroxyls, the primary phase is
poorly crystallized TCP. As a result of topochemical
reactions, this phase then converts to Ca-deficient HA,
containing hydrogen phosphate groups instead of
3–
4
5
6
PO . The samples synthesized under such conditions
4
have a small particle size and, accordingly, large spe-
cific surface.
Increasing the initial solution concentration or reac-
–
tion time increases the H PO concentration in the
2
4
–
2
solution and leads to the formation of H(H PO )
2
4
dimers, which is favorable for the formation of hydro-
gen and dihydrogen phosphates as primary phases.
Subsequent HA formation in the final step, at pH > 9,
involves, in contrast to the processes in dilute solutions,
dissolution of reaction intermediates and formation of
either an intercrystalline mixture of HA and calcium
hydrogen phosphate or a mechanical mixture of these
phases. A distinguishing feature of the final product is
the substantially increased calcium deficiency and
small specific surface, which may be due to aggregation
of primary particles.
4
000 3500 3000 2500 2000 1500 1000 500
–
1
Wavenumber, cm
Fig. 6. IR spectra of the final products; the numbers at the
curves identify the samples as referenced in Table 1.
This sample has the largest Ca/P ratio, 1.60, and the
smallest lattice parameters, which leads us to conclude
that the Ca deficiency in this sample is indeed due to
substitutions in the crystal lattice. In contrast, in all of
the other samples, both the a parameter and unit-cell
volume increase markedly as the Ca/P molar ratio
CONCLUSIONS
We prepared calcium-deficient HA with a Ca/P
decreases to 1.34, and their IR spectra undergo signifi- molar ratio in the range 1.34 to 1.60 via precipitation
–1
cant changes in the range 3165–3536 cm . In addition, from an acid calcium phosphate solution by aqueous
the XRD patterns from some of the samples, e.g., 2 and ammonia under controlled pH conditions.
4
, show only peaks from HA, whereas the samples with
a = 0.948 nm and greater contain an impurity phase, as
evidenced by XRD data and the band at 896 cm , char-
acteristic of CHP [20]. These results suggest that, dur-
ing HA crystallization in more concentrated solutions,
through the formation of calcium hydrogen phosphates
The formation of Ca-deficient HA in dilute solu-
tions (below 0.1 M H êé ) at high precipitation rates
3
4
–1
(
short reaction times) is preceded by the formation of
amorphous TCP, which then converts to Ca-deficient
2
–
HA containing HPO groups through topochemical
4
as reaction intermediates, the intermediate is dissolved reactions.
only partially in the final step, at pH > 9, because of the
small éH /ëa molar ratio, ≤1.5. As a result, the new
phase grows on residual primary particles as on crystal-
Raising the initial solution concentration to above
.25 mol/l or reducing the precipitation rate leads to the
–
2+
0
formation of calcium hydrogen phosphate as a primary
lization centers, forming either an intercrystalline mix- phase, which then dissolves and recrystallizes into Ca-
ture (samples 2, 4) or a two-phase system (samples 3, 5, 6), deficient HA. As the calcium deficiency increases,
2
–
incomplete dissolution of the primary phase leads to the
formation of either an intercrystalline mixture or a two-
phase system of HA and calcium hydrogen phosphate
crystals.
depending on the HPO content. In contrast to what
4
was reported by Rodriguez-Lorenzo [26], the impurity
phases in our samples are CHP and CHPD, instead of
ëa H(êé ) · 2.5H é.
4
4 3
2
In contrast to conventional processes in pH ≥ 7 solu-
Analysis of the potentiometric titration curves, tions, the precipitation of Ca-deficient HA from pH 1–2
chemical analysis data, and phase compositions of the solutions of a calcium salt in phosphoric acid enables
final products leads us to conclude that there is a gen- an effective process control owing to the variable
INORGANIC MATERIALS Vol. 43 No. 10 2007

EFFECT OF PHASE TRANSFORMATIONS DURING SYNTHESIS
1127
degree of protonation of the phosphoric acid and the 13. Lazic, S., Zec, S., Miljevic, N.D., and Milonjic, S., The
Effect of Temperature on the Properties of Hydroxyapa-
tite Precipitated from Calcium Hydroxide and Phospho-
ric Acid, Thermochim. Acta, 2001, vol. 374, no. 1,
pp. 13–22.
presence of different phosphate ions in the solution.
REFERENCES
1
4. Shashkova, I.L., Kitikova, N.V., Rat’ko, A.I., and
D’yachenko, A.G., Preparation of Calcium and Magne-
sium Hydrogen Phosphates from Natural Dolomite and
Their Sorptive Properties, Neorg. Mater., 2000, vol. 36,
no. 8, pp. 990–994 [Inorg. Mater. (Engl. Transl.), vol. 36,
no. 8, pp. 826–829].
5. Schwarzenbach, G. and Flaschka, H., Die komplexome-
trische Titration, Stuttgart: Ferdinand Enke, 1965.
6. Charlot, G., Les méthodes de la chimie analytique: Anal-
yse quantitative minérale, Paris: Masson, 1961, 4th ed,
part 2.
1
2
. Mavropoulos, E., Rossi, A.M., Rocha, N.C.C., et al.,
Dissolution of Calcium-Deficient Hydroxyapatite Syn-
thesized at Different Conditions, Mater. Charact., 2003,
vol. 50, nos. 2–3, pp. 203–207.
. Raynaud, S., Champion, E., Lafon, J.P., and Bernache-
Assollant, D., Calcium Phosphate Apatites with Variable
Ca/P Atomic Ratio: III. Mechanical Properties and Deg-
radation in Solution of Hot Pressed Ceramics, Biomate-
rials, 2002, vol. 23, no. 4, pp. 1081–1089.
1
1
3
. Tanaka, H., Watanabe, T., and Chikazawa, M., FTIR and
TPD Studies on the Adsorption of Pyridine, n-Buty-
lamine, and Acetic Acid on Calcium Hydroxyapatite,
J. Chem. Soc., Faraday Trans., 1997, vol. 93, no. 24,
pp. 4377–4381.
1
1
1
7. Vasserman, I.M., Khimicheskoe osazhdenie iz rastvorov
(
Chemical Precipitation from Solutions), Moscow:
Khimiya, 1980.
8. Perdok, W.G., Christoffersen, J., and Arends, J., The
Thermal Lattice Expansion of Calcium Hydroxyapatite,
J. Cryst. Growth, 1987, vol. 80, no. 1, pp. 149–154.
9. ASTM Standard X-ray Diffraction Powder Patterns,
Inorganic, Philadelphia: ASTM, 1967.
4
5
. Dykman, A.S., Origin of the Catalytic Activity of Imper-
fect Hydroxyapatite, Zh. Prikl. Khim. (S.-Peterburg),
2
003, vol. 76, no. 2, pp. 234–236.
. Matsumura, Y., Kanai, H., and Moffat, J.B., Catalytic
Oxidation of Carbon Monoxide over Stoichiometric and
Non-stoichiometric Hydroxyapatites, J. Chem. Soc., 20. Atlas infrakrasnykh spektrov fosfatov: Ortofosfaty
Faraday Trans., 1997, vol. 93, no. 24, pp. 4383–4387.
(Atlas of Infrared Spectra of Phosphates: Orthophos-
phates), Pechkovskii, V.V. et al., Eds., Moscow: Nauka,
6
7
. Gibson, I.R., Rehman, I., Best, S.M., and Bonfield, W.,
Characterization of the Transformation from Calcium-
1
981.
Deficient Apatite to β-Tricalciumphosphate, J. Mater. 21. Lepilina, R.G. and Smirnova, N.M., Termogrammy
Sci.: Mater. Med., 2000, vol. 11, no. 12, pp. 799–804.
. Van Kemenade, M.J.J.M. and De Bruyn, P.L., A Kinetic
neorganicheskikh fosfatnykh soedinenii (Thermal Anal-
ysis of Inorganic Phosphates), Leningrad: Nauka, 1984.
Study of Precipitation from Supersaturated Calcium 22. Shashkova, I.L., Rat’ko, A.I., and Kitikova, N.V.,
Phosphate Solutions, J. Colloid Interface Sci., 1987,
vol. 118, no. 2, pp. 564–585.
Removal of Heavy Metal Ions from Aqueous Solutions
by Alkaline-Earth Metal Phosphates, Colloids Surf., A,
1
999, vol. 160, pp. 207–215.
8
9
. Boistelle, R. and Lopez-Valero, I., Growth Units and
Nucleation: The Case of Calcium Phosphates, J. Cryst. 23. Pechkovskii, V.V., Salonets, G.I., Dzyuba, E.D., and
Growth, 1990, vol. 102, no. 3, pp. 609–617.
Chubarov, A.V., Electrical Conductivity of Calcium and
Barium Monophosphates during Dehydration, Zh.
Neorg. Khim., 1980, vol. 25, no. 2, pp. 387–390.
. Abbona, F., Lundager Madsen, H.E., and Boistelle, R.,
The Final Phases of Calcium and Magnesium Phos-
phates Precipitated from Solution of High to Medium
Concentration, J. Cryst. Growth, 1988, vol. 89, no. 4,
pp. 592–602.
2
4. Ivanenko, V.I., Udalova, I.A., Lokshin, E.P., and Kalin-
nikov, V.T., State of Phosphate Ions in Aqueous Solu-
tions in the Presence of NaCl, KCl, NaNO , and KNO ,
3
3
Zh. Neorg. Khim., 2002, vol. 47, no. 6, pp. 1026–1032.
The Growth of NonstoichiometricApatite fromAqueous 25. Prodan, I.E., Eshchenko, L.S., and Pechkovskii, V.V.,
1
0. Heughebaert, J.C., Zawacki, S.J., and Nancollas, G.H.,
Solution at 37°C: I. Methodology and Growth at pH 7.4,
J. Colloid Interface Sci., 1990, vol. 135, no. 1, pp. 20–
Role of Complexation in the Crystallization of Iron
Phosphates, Zh. Neorg. Khim., 1990, vol. 35, no. 4,
pp. 843–848.
3
2.
1
1
1. Altunina, L.I., Pechkovskii, V.V., Eshchenko, L.S., and 26. Rodriguez-Lorenzo, L.M., Studies on Calcium Deficient
Gafarova, A.F., Formation Conditions and Properties of
Hydroxyapatite, Zh. Neorg. Khim., 1986, vol. 31, no. 5,
pp. 1107–1112.
Apatites Structure by Means of MAS-NMR Spectros-
copy, J. Mater. Sci.: Mater. Med., 2005, vol. 16, no. 5,
pp. 393–398.
2. Sampath Kumar, T.S., Manjubala, I., and Gunaseka- 27. Cheng, Z.H., Yasukawa, A., Kandori, K., and Ish-
ran, J., Synthesis of Carbonated Calcium Phosphate
ikawa, T., FTIR Study on Incorporation of CO2 into Cal-
cium Hydroxyapatite, J. Chem. Soc., Faraday Trans.,
1998, vol. 94, no. 10, pp. 1501–1505.
Ceramics Using Microwave Irradiation, Biomaterials,
2
000, vol. 21, no. 16, pp. 1623–1629.
INORGANIC MATERIALS Vol. 43 No. 10 2007