Eiomaterials 18 (1997) 737-741

0 1997 Elsevier Science Limited

Printed in Great Britain. All rights reserved
PII SOl42-9612 (96) 00203-7 014s9612/97/$17.00

Hydrolysis and phase transition of

alpha-tricalcium phosphate
Li Yubao, Zhang Xingdong and K. de Groat*
Engineering Research Ceitre in B~omater~als, Sichuan Union University, Chengdu 610064, China; ‘Bjomaterja/s
Research Group, University of Leiden, The Netherlands

Increasing attention has been paid in recent years to alpha-tricalcium phosphate (a-TCP), which is
used as the main constituent of calcium phosphate bioactive bone cements and biphasic calcium
phosphate ceramics. Its hydrolysis and conversion into apatite phase may play an important role in
new bone formation in vivo. In this experiment, c(-TCP powder was made and immersed in deionized
water. The morphology change and phase composition were analysed before and after immersion.
Based on the results of hydrolysis, a formula for the converted apatite-TCP phase containing lattice
water was put forward and a partial structural model along the hydroxyl column was proposed. 0
1997 Elsevier Science Limited. All rights reserved

Keywords: wTCP, apatite, hydrolysis, phase conversion, structure model

Received 23 May 1996; accepted 20 November 1996

Hydroxyapatite (HA) forms the main mineral apatite containing lattice water along the hydroxyl
constituent of human hard tissues’.2. HA ceramics column was proposed.
have proved to be biocompatible and bioactive
materials which can chemically bond with bone3,
and have been successfully used clinically for repair MATERIALS AND METHODS
of bone defects and augmentation of osseous
tissues4.“. Calcium phosphate bioactive bone Laboratory wet-synthesized apatite powder with a Ca/P
cements based on alpha-tricalcium phosphate (r- molar ratio of 1.50 was prepared according to a method
TCP) have attracted increasing attention since the reported previously’3. The apatite powder was sintered
1980sfi.7. These bone cements can be easily moulded at 1250°C for 3 h followed by subsequent quenching in
during operation and simply implanted or injected air to obtain x-TCP. The quenched a-TCP powder was
into the bone defects, and will turn into apatite after ground, passed through a 200-mesh sieve and then
setting and hardening’.“. Bonding with bone will immersed into deionized water in a ratio of lg to
finally be realized through the converted apatite 25 ml at room temperature. After 30 days the immersed
phase. This distinguishing feature of the bone powder was filtered and dried at 80°C for 24 h. Some
cements can be attributed to the hydrolysis of x-TCP immersed powder was re-sintered at 750 and 1100°C
to a large extent. In recent years, biphasic calcium for 1 h separately to further determine the phase
phosphate ceramics consisting of HA//GTCP or HA/ composition and Ca/P molar ratio of the powder after
r-TCP have been evaluated in both osseous and non- immersion.
osseous tissues’“-“. The results showed that these The powders were analysed by X-ray diffraction
biphasic ceramics were biologically more active than (XRD), scanning electron microscopy (SEM) and
pure HA ceramics alone, and that the biological infrared spectroscopy (IR) before and after immersion
behaviour of the biphasic ceramics containing x-TCP and after sintering.
was superior in new bone formation. This
phenomenon can also be closely related to the
hydrolysis of the z-TCP phase; in other words, to the RESULTS
change of its phase structure and morphology during
hydrolysis. In this experiment, r-TCP powder was Figure ZA shows the XRD spectrum of starting a-TCP
made and immersed in deionized water. Its powder. It matches the ASTM data for r-TCP,
morphology change, phase transition and Ca/P molar exhibiting a pure r-TCP crystal structure. After
ratio were analysed and determined. Based on the immersion in deionized water for 30 days, most of the
results of hydrolysis, a partial structural model of x-TCP phase has been converted into a poorly
crystallized apatite structure, except that some GI-TCP
Correspondence to Dr Li Yubao, Analytical and Testing phase is still present, as shown in Figure zB. When re-
Center, Sichuan Union IJniversity, Chengdu 610064, China. sintered at 750°C for 1 h, the relative intensity of the

737 Biomaterials 1997, Vol. 18 No. 10

738
a-TCP
I I I I
30
2G(degtee)
apatite
apatite
II
Figure1 XRD spectra of c(-TCP before and after immersion and after re-sintering.
powder for 1 month, C, re-sintering at 750°C for 1 h and D, re-sintering at 1100°C for 1 h.
converted apatite phase is increased (Figure IC). At
llOO”C, the XRD spectrum of the immersed powder
again gives a pure TCP structure (P-TCP at this
temperature). No apatite spectral peaks are present at
this time.
Figure 2 shows the SEM photos of a-TCP powder
before and after immersion. It can be seen that the
starting a-TCP powder is composed of well-crystallized,
small ceramic granules. After immersion in water, the
clear, dense and irregular morphology of the granules
becomes indistinct, loose (microporous) and ball-like,
with tiny crystallites on the surfaces. The size of the
granules looks larger after immersion.
Figure 3A shows the IR spectrum of a-TCP powder.
After immersion in water, the powder showed an
apatite structure without hydroxyl (OH-) groups, as
shown in Figure 3B, and the increase of water peaks or
bands in the spectrum is obvious. However, the 750°C
IR spectrum in Figure 3C clearly shows the presence of
Biomaterials 1997, Vol. 18 No. 10
Hydrolysis and phase transition of M-TCP: Li Yubao et al.
apatite
apatite
L I I I 1
40
A, Starting cc-TCP powder, B, immersed
OH- groups at about 3570 and 630cm-‘. The 1100°C
IR spectrum shows a pure TCP structure without OH-
groups.
DISCUSSION
The purpose of re-sintering at 750°C is to confirm the
presence of OH- groups in the converted apatite
phase. From the spectra in Figures ZB and 3B, we may
conclude that the converted apatite phase is oxyapatite,
without OH- groups but not HA, because no OH-
peaks can be observed in Figure 3B. However, at
750”C, with the increase of the relative intensity of the
converted apatite phase, i.e. the increase of its crystalli-
nity, OH- peaks appear in the IR spectrum, as shown in
Figure 3C. This means that the OH- group does exist in
the converted apatite phase. The OH- peak that is not
seen in the IR spectrum in Figure 3B could be
Hydrolysis and phase transition of cr-TCP: Li Yubao et al. 739

a
HE--- 3600 3200 2600 2400 2600 i 600 Ii00 Boo00
WAVENUNEER

Figure 3 The IR spectra of cr-TCP powder before and after

immersion and after re-sintering. A, Starting wTCP
powder, 8, immersed powder for 1 month, C, re-sintering
at 750°C for 1 h and D, re-sintering at 1100°C for 1 h.

Ca,,_,(OH), _,(HP0,J,(P0,)6_, (0 < x < 1) will give an

apatite crystal structure in its XFXI spectrum, despite its
Ca/P molar ratio being 1.67 (x = 0), 1.5 (x = 1) or
between 1.67 and 1.5 (0 < x < 1). Hence, from the
spectrum in Figure IB, it is difficult to determine the Ca/
P ratio of the converted apatite phase. Sintering at
1100°C for 1 h is a simple and suitable way to determine
the Ca/P ratio of the converted apatite phase. From
Figure ID, it is easy to find if the converted apatite still
has a Ca/P ratio of 1.50, corresponding to x = 1 in the
formula mentioned above, i.e. the formula for the
converted apatite phase is CaJOH) (HPO,) (PO,),. This is
a special non-stoichiometric apatite called apatite-
TCP14, which has an apatite crystal structure but
possesses the Ca/P ratio of TCP (1.50). Based on the
discussion above, the hydrolysis and phase conversion
of cr-TCP can be described as:

3Ca3E’0412 + Hz0 ---f CadOH)(HPO,)(PO,),

(Ca/P = 1.5)

b After re-sintering at llOO”C, the apatite-TCP phase

Figure 2 SEM photos of the a-TCP powder: a, before and b, decomposes to the TCP phase (/I-TCP at this
after immersion. temperature), i.e.

Ca9(OH)(HP04)(P04), + 3Ca3(P04), + Hz0

concealed by the poor crystallinity of the converted (Ca/P = 1.5)
apatite phase and by the orientation disorder of the
OH- group along the c-axis at room temperature. The SEM photos in Figure 2 show that the hydrolysis
Usually, non-stoichjometric apatite with formula of a-TCP is not only a process of phase transition but

Biomaterials 1997, Vol. 18 No. 10

740 Hydrolysis and phase transition of LX-TCP:Li Yubao et al.

also a process of morphology change. The phase 1 Hydroxyl column

transition and morphology change happened at the
same time. The IR spectrum in Figure 3B shows an (c-axis)
obvious increase of water peaks after immersion.
These facts indicated that water, not only in the form
of ionized H+ and OH- but also in the molecular form,
destroyed the crystal structure of c(-TCP and/or
incorporated into its crystal lattice to recompose an
apatite crystal structure.
During the hydrolysis and conversion of a-TCP into
apatite-TCP, two water molecules combined with three
3
TCP molecules to form one apatite-TCP molecule
containing lattice water, i.e. 4

(a) 3Cas(PO& -+ 9Ca2+ + 6POi-

(b) Hz0 + H+ + OH- (located at one OH- site of
apatite-TCP)
H+ +POi- -+I-IPO:- (located at one PO:- site of
apatite-TCP)
Combining (a) and (h), we obtain

3Cas(PO,) + Hz0 + 9CaZ+ + OH- + HPO:- + 5POi-

--$ CaS(OH)(HP04)(P04)s
(apatite-TCP)

However, comparison of the apatite-TCP molecule with 1

the HA molecule [CalD(OH),(P04),), it can be noted that
the apatite-TCP lacks one Ca2+ and one OH- in the unit 4
cell of the apatite crystal structure. That is, vacancies
appear in both the cationic Ca2+ sites and the anionic
OH- sites of the apatite crystal structure. This provides \ Ca J
the possibility for a water molecule to occupy these \ -A
vacant sites in unionized form. The incorporation of vacancy
I
lattice water in these vacant sites also increases the
stability of the apatite-TCP crystal structure. In view of Figure 4 Partial structural model of apatite-TCP containing
this, the apatite-TCP formula should be rewritten as lattice water along the hydroxyl column.

(CasH)(OH I OH) (HPO,)(PO,), (Ca/P = 1.50)

According to this formula, a structural model of apatite-
TCP containing lattice water along the hydroxyl
column is proposed as shown in Figure 4.
As we know, non-stoichiometric apatite is a calcium
phosphate material with a Ca/P molar ratio between
1.67 and 1.33, but it has the crystal structure of
stoichiometric HA. HA is known to have a hexagonal
structure with a P63/m space group15-17. In its unit
cell, hydroxyl ions lie at the corners of the projected
rhombic basal plane, forming columns of hydroxyls
with a spacing of half the unit cell height. Six of the 10
calcium ions are associated with these hydroxyls,
forming equilateral triangles centred on and perpendi-
cular to the hydroxyl columns. Successive calcium
triangles are rotated 60”. The hydroxyls are oriented
such that the oxygen-hydrogen bond is along the
column axis but does not cross the plane of the
calcium triangle. It is noteworthy0 that the hydroxyl
oxygen atoms are displaced by 6.3 A from the plane of
the calcium triangles, thus the bond angle between the
hydroxyl hydrogen-oxygen bond and the hydroxyl
oxygen-calcium (of the nearest calcium triangle) bond
is near to that between two hydrogens of a water
molecule (104.5”). On the other hand, in non-stoichio-
metric apatite the substitution of I-IPO:- for PO:- in
the crystal structure occurs, with a loss of Cazf and
OH- ions and, simultaneously, vacancies appear in
both the cationic Ca2+ sites and the anionic OH- sites
around the hydroxyl column14.
On the l/4 plane of the structure model in Figure 4,
one hydrogen ion of the water molecule is situated at a
calcium vacancy site, while the other H+ ion or OH-
group of the same water molecule is arranged at the
hydroxyl site of the hydroxyl column, in which the H+
ion or OH- group is oriented to the c-axis by two
calcium ions of the calcium triangle. In such a manner,
water molecules are placed in the crystal lattice.
CONCLUSIONS
From this experiment, it can be seen that a-TCP will be
converted into apatite in aqueous solution with
immersion time. It is not enough to say the converted
apatite is oxyapatite or HA only from the XRD and IR
spectra at room temperature, and to say that the
converted apatite phase is stoichiometric or non-
stoichiometric. Re-sintering at an appropriate
temperature, such as 750 and llOO”C, is necessary
(helpful) to further determine the composition,
structure and Ca/P ratio of the converted apatite phase.
The conversion of LX-TCP into poorly crystallized
non-stoichiometric apatite containing lattice water in
aqueous solution is important in bone formation. We
chose a special immersion time (30 days); this is

Biomaterials 1997, Vol. 18 No. 10

Hydrolysis and phase transition of cz-TCP:Li Yubao et al. 741

because new bone formation on bioceramics starts 9. Constantz, B. R., Ison, I. C., Fulmer, M. T. et al., Skeletal
several weeks after implantation, which is accordant repair by in situ formation of the mineral phase of
with this immersion time. The apatite in human bone bone. Science, 1995,267,1796-1799.
is also poorly crystallized, non-stoichiometric and
10. Frayssinet, P., Trouillet, J. L., Rouquet, N., Azimus, E.
and Autefage, A., Osseointegration of macroporous
contains lattice water’s-20. This means that the
calcium phosphate ceramics having a different
converted apatite is more analogous to the apatite in
chemical composition. Biomaterials, 1993, 14, 423-
human bone. This may be the reason why biphasic 429.
ceramics containing a-TCP phase give superior 11. Daculsi, G., Passuti, N., Martin, S., Deudon, C., LeGeros,
performance in new bone formation. R.Z. and Raher, S., Macroporous calcium phosphate
ceramic for long bone surgery in humans and dogs:
clinical and histological study. I. Biomed. Mater. Res.,
1990,24,379-396.
REFERENCES 12. Klein, C.P.A.T., de Groot, K., Weiqun, C., Yubao, L.
and Xingdong, Z., Osseous substance formation
1. Groot, K. de, Bioceramics of Calcium Phosphate. CRC induced in porous calcium phosphate ceramics in soft
Press, Boca Raton, FL, 1983. tissues. Biomaterials, 1994, 15, 31-34.
2. Aoki, H., Science and Medical Applications of 13. Yubao, L., Klein, C.P.A. T., Xingdong Z. and de Groot,
Hydroxyapatite. JAAS Press, Tokyo, 1991. K., Formation of a bone apatite-like layer on the
3. Ducheyne, P., Kolkubo, T. and van Blitterswijk, C.A., surface of porous hydroxyapatite ceramics. Biomater-
Bone-Bonding Biomaterials. Reed Healthcare ials,1994, 15,835-841.
Communications Press, Leiderdorp, 1992. 14. Lacout, J.L., Calcium phosphates as bioceramics. In
4. Jarcho, M., Calcium phosphate ceramics as hard tissue Biomaterials- Hard Tissue Repair and Replacement,
prosthetics. Clin. 0rthop. Rel. Res., 1981, 157,259-278. ed. D. Muster. Elsevier Science, Amsterdam, 1992, p.
5. LeGeros, R. Z., Calcium phosphate materials in restora- 81.
tive dentistry: a review. Adv. Dent. Res., 1988, 2(l), 15. Katz, J. L. and Harper, R. A., In Encyclopedia of Materi-
164-180. als Science and Engineering. ed. M.B. Bever.
6. Chow, L. C., Development of self-setting calcium Pergamon, Oxford, 1986, p. 475.
phosphate cements. The Centennial Memorial Issue 16. Kay, M.I., Young, R.A. and Posner, A.S., Crystal
of the Ceramic Society of Japan, 1991, 99(10), 954- structure of hydroxyapatite. Nature, 1964, 204, 1050-
964. 1052.
7. Ginebra, M.P., Boltong, M. G., Femandez, E., Planell, 17. Park, J. B. and Lakes, R. S., Biomaterials: An Zntroduc-
J. A. and Driessens, F.C.M., Properties profile of tion. Plenum Press, New York, 1992, p. 112.
Biocement (H). In Trans. 21st Annual Meeting in 18. Posner, A.S., The mineral of bone. Clin. Orthop. Rel.
Conjunction with ithe 27th Int. Biomaterials Symp., San Res.,1985, 200,87-99.
Francisco, CA, 1995, p. 99. 19. Ducheyne, P., Bioceramics: material characteristics
8. Kurashina, K., Ogiso, A., Kotani, A., Takeuchi, H. and versus in vivo behaviour. I. Biomed. Mater. Res., 1987,
Hirano, M., Histological and microradiographic evalua- 21,219-236.
tion of hydrated a:nd hardened a-tricalcium phosphate/ 20. LeGeros, R. Z., Bonel, G. and Legros, R., Types of “HzO”
calcium phosphate dibasic mixtures. Biomaterials, in human enamel and in precipitated apatite. Calcif
1994,15,429432. Tiss. Res., 1978,26,111-118.

Biomaterials 1997, Vol. 18 No. 10