Effect of water vapor pressure and temperature on the

amorphous-to-crystalline HA conversion during heat

treatment of HA coatings

Weidong Tong,* Jiyong Chen, Yang Cao, Lin Lu, Jiaming Feng, and Xingdong Zhang
Institute of Materials Science and Technology, Sichuan University, Chengdu 610064, China

X-ray diffraction was used to characterize the increment of
crystallinity of HA coatings after heat treatment. Coatings
were heated over the temperature (T) interval of 300°–460°C
with a partial water vapor pressure of 0.01 MPa and 0.001
MPa. Heat treatment also was done in air, as a contrast. It
was found that the ratio (n) of the increment of crystallinity
to the crystallinity of the as-received HA coatings was more
significant for the coatings heated in atmosphere with water
vapor than for those heated in air. This ratio also increased
with water vapor pressure. The logarithm of the ratio in-
creased linearly with 1/T, indicating that the ratio is expo-
nential to T. The reason might be that recrystallization of the
amorphous phase is a diffusion controlled process; the
nucleation rate and growth velocity of the crystallites are in
proportion to the diffusion coefficient, which is exponential
to the temperature (T). Incorporation of water vapor in the
atmosphere during heat treatment may decrease the activa-
tion energy for diffusion, which helps raise the diffusion
coefficient of the atoms. Thus recrystallization of the amor-
phous phase can be accelerated. © 1997 John Wiley & Sons, Inc.

INTRODUCTION ration of water vapor may significantly promote the

recrystallization of HA when the coatings are heated
to 630°C whereas the intrinsic properties of the atmo-
Plasma-sprayed hydroxyapatite (HA) coatings have
sphere, such as nitrogen (N2), oxygen (O2), and air
been widely used in dental and hip joint replacements
have no effect on the recrystallization.9 Water vapor
due to the excellent biocompatibility of hydroxyapa-
pressure may affect the recrystallization of HA, but
tite.1,2 However, the as-received HA coatings are ac-
the effect of water vapor on the recrystallization is not
tually composed of multiphases of calcium phos-
significant unless both temperature and water vapor
phates, including crystalline HA, amorphous phase
pressure increase to certain values, indicating that
(or amorphous-like substances), and the decomposi-
temperature and water vapor are two important fac-
tion products that consist of a,b-Ca3(PO4)2 (a,b-TCP)
tors affecting the recrystallization process during post-
Ca4O(PO4)2 (TTCP) and calcium oxide (CaO). The
treatment.10
number of these phases may vary with the spraying
In this study, we investigated the effect of tempera-
parameters, such as starting powder sizes, intrinsic
ture and water vapor pressure on the recrystallization
properties of the powders, power input, thickness of
of an amorphous phase during post-heat-treatment
the coatings, and so forth.3–5 The degradation of the
over the temperature interval of 300°–460°C with par-
coatings initially is dependent on the phase contents of
tial water vapor pressure of 0.01 MPa and 0.001 MPa.
the coatings: the greater the percentage of crystalline
HA in the coating, the lower the degradation.6,7 It
should be mentioned herein that the microstructure of
the coatings, such as porosity, is another independent MATERIALS AND METHODS
factor affecting the stability of the coatings.3,8
Many methods have been tried to raise crystallinity
(percentage of crystalline HA in the coating), among The starting powder was pure and well recrystal-
which post-treatment is the most promising. Incorpo- lized HA with a Ca/P ratio of 1.67. A Metco MN
(USA) system was used to coat HA on titanium sub-
strates (sized 12 × 8 × 2 mm3) that had been polished,
*To whom correspondence should be addressed. sandblasted by SiC, and cleaned before spraying ac-

Journal of Biomedical Materials Research, Vol. 36, 242–245 (1997)

© 1997 John Wiley & Sons, Inc. CCC 0021-9304/97/020242-04
AMORPHOUS-TO-CRYSTALLINE HA CONVERSION
cording to ASTM B600-74.11 All the coatings were
sprayed with the same parameters. The thicknesses of
the coatings were about 50 mm. Coatings were heated
at room temperature intervals of 300°–460°C with par-
tial water vapor pressure of 0.01 MPa or 0.001 MPa for
2 h. Heat treatment in air also was done at tempera-
ture points of 300°, 400°, and 460°C. X-ray diffraction
(XRD) was employed to detect the structure of the
coatings. Increments of crystallinity were evaluated by
the variation of 211 peak area on XRD patterns of the
coatings before and after heat treatment, according to
the following formula: C = Sb − Sa, where C is the
increment of crystallinity, Sb is the 211 peak area on
XRD patterns of the coatings after heating, and Sa is
the 211 peak area on XRD patterns of the as-received
coatings. The ratio (n) of the increment of crystallinity
(C) to the 211 peak area of the as-received coatings (Sa)
was calculated as n = C/Sa.
RESULTS AND DISCUSSION
Figure 1 shows a typical XRD pattern of the as-
received coating, which consists of a series of sharp
peaks, the diffusion background, and additional
phases (arrows). The diffusion background represents
the amorphous phase, and the sharp peaks represent
the crystalline HA. There were only traces of addi-
tional phases, which were the decomposition products
of HA during plasma spraying. The dependence of the
ratio (n) of C to Sa on 1/T is shown in Figure 2. It can
Figure 1. XRD pattern of the as-received coating.
243
be observed that n increased steeply with the increase
of temperature regardless of the atmosphere of the
heat treatment, but n was much higher for the coatings
heated in an atmosphere with water vapor than it was
for those heated in air, and n after heat treatment in
different atmospheres varied as nair << n0.001 MPa <
n0.01 MPa. Figure 3 shows the dependence of Ln(n) on
1/T. Logarithms of nair, n0.001 MPa, and n0.01 MPa are
straight lines, indicating that nair, n0.001 MPa, and
n0.01 MPa were exponential to temperature. It also can
be observed that there was a minor difference in the
slopes of the straight lines, which can be described as:
kair > k0.001 MPa > k0.01 MPa (k = absolute value of the
slope).
In addition to crystalline HA, the as-received
HA coatings actually were composed of an amor-
phous phase and decomposition products of the
starting powder [additional phases, such as a,b-
Ca3(PO4)2(a,b-TCP) Ca4O(PO4)2 (TTCP)] (see Fig. 1).
Each particle of starting powder injected into the
plasma flame usually was partially molten, consisting
of the molten portion and the unmelted cores. An
amorphous phase in the coating was formed because
of the fast cooling of the molten portion by the cold
substrate.12 Therefore, the possible phase transitions
in the coatings during post-heat-treatment consist of
transformations of additional phases and the amor-
phous-to-crystalline HA conversion. The increment of
crystallinity after heat treatment is dependent on the
efficiency of the transformations of the additional
phases and amorphous-to-crystalline HA conversion.
However, there were only minor fractions of addi-
tional phases in the coating compared to the amount
of the amorphous phase; thus the increment of crys-
tallinity was affected mainly by the recrystallization of
the amorphous phase.
Figure 2. Dependence of n (the ratio of C to Sa) on 1/T. n
increased steeply with 1/T, and it is more when water vapor
pressure in atmosphere is higher. T1 = 573K, T2 = 623K, T3
= 673K, T4 = 703k, T5 = 733K.
244
Figure 3. Dependence of logarithm of n on 1/T. Ln(n) var-
ied linearly with 1/T, indicating that n is exponential to T.
The differences of the slope of the straight lines indicate the
activation energy for diffusion varies with water vapor pres-
sure in atmosphere, which is fair > f0.001 MPa > f0.01 MPa. T1
= 573K, T2 = 623K, T3 = 673K, T4 = 703k, T5 = 733K.
It is known that the amorphous phase is a meta-
stable phase and tends to recrystallize.13 The critical
temperature of recrystallization of the amorphous
phase in vacuum (<10−5 torr) is above 600°C (the re-
ported value ranged from 600°–630°C).14,15 Heat treat-
ment in air showed that high crystallinity can be
achieved at temperatures over 630°C for 1 h.16 In our
previous study we had shown that heat treatment (at
630°C) in air with water vapor raised crystallinity
more significantly than that in air without water va-
por.9 In this study we showed that at comparatively
low temperatures (300°–460°C) recrystallization of the
amorphous phase is accelerated by the incorporation
of water vapor in the atmosphere, and the ratio of the
increment of crystallinity to the crystallinity of the as-
received coating is higher if water vapor pressure is
higher (Fig. 2). The recrystallization of the amorphous
phase during heat treatment usually consists of nucle-
ation and growth of the nucleus. The nucleation rate
and growth velocity (u) of crystal nucleases are in pro-
portion to the diffusion coefficient, which is D =
Doexp(-f/kT), where T, f, Do, and k represent abso-
lute temperature, activation energy for the diffusion,
the diffusion constant, and Boltzman’s constant (k =
1.38066 × 10−23), respectively.
It was reported that the newly formed crystallites
during heat treatment grow directly on the surface of
the unmelted cores of the starting powder in the coat-
ing, and recrystallization of the amorphous phase is
related to the number and the distribution of the un-
melted cores. The as-received coatings possessed the
same number and distribution of the unmelted cores
(thus Sa is the same) because the spraying parameters
were the same. During heat treatment, volume (V) of
a newly formed crystallite can be calculated accord-
TONG ET AL.
ingly: V = gu3(t-t)3, where V, g, u, and t are (V) vol-
umes of the newly formed crystallite at time t, (g)
shape factor (for example, g of the global crystallite is
equal to 4p/3), (u) growth velocity, and (t) time to
form a stable nucleus. The values of t, g, and t in this
analysis don’t vary with experimental conditions.
Briefly, V = au3, a = g(t − t)3, and the relationship
between u, n, and D can be described as u = bD, C = fV
= fau3 = fab3D3, n = C/Sa = [ab3f(Do)3exp(-3f/kT)]/Sa
= mexp(-3f/kT), m = ab3fD3o/Sa, where a, b, f, and m
are constants. Logarithms of n should be in proportion
to 1/T: Ln(n) = Ln(m) − 3f/kT (Eq. 1), which should
be a straight line on the diagram of Ln(n) via 1/T.
Equation (1) can be alternatively represented by:
Ln1000n = Ln(m) + Ln1000−[3f/(100000k)] ×
100000/T (Eq. 2).
Experimental results of heat treatment at the tem-
perature interval of 300°–460°C were in good agree-
ment with the theory of phase transformation of amor-
phous solids (Fig. 3). Lines A, B, and C show clearly
that Ln(n) is proportional to 1/T. Statistical analysis
showed that lines A, B, and C can be represented by
the following equation: line A, ln1000n = 22.87 − 0.11
× (100000/T), RTN = 0.9987; line B, ln1000n = 25.04 −
0.13 × (100000/T), RTN = 0.9903; line C, ln10000n =
25.64 − 0.14 × (100000/T), RTN = 0.9926. It can be
observed that cross points of lines A, B, and C on the
vertical axis are satisfactorily equal. This agrees with
our analysis that ln(m)+ln1000 should be a constant. In
addition, absolute values of the slope of the lines in
Figure 3 decreased with the increasing of water vapor
pressure: kair > k0.001 MPa > k0.01 MPa (kk = absolute
value of the slope). According to Equation (2), this
decrease is directly related to the decreased activation
energy for diffusion owing to the incorporation of wa-
ter vapor in the atmosphere. The calculated values of
the activation energy for diffusion in different atmo-
spheres were listed as below: Nofair = 9.24 kcal,
Nof0.001 MPa = 8.58 kcal, Nof0.01 MPa = 7.26 kcal, where
No = 6.02204 × 1023/mol. It can be seen clearly that
activation energy for diffusion decreased with the in-
crease of water vapor pressure.
The effect of water vapor on the recrystallization of
an amorphous phase functioned on the activation en-
ergy for diffusion, which is related directly to the dif-
fusion coefficient. The lower the activation energy, the
higher the diffusion coefficient. It also can be observed
that significant increases in water vapor pressure
(from 0.001 MPa to 0.01 MPa) resulted in decreases of
the activation energy of only about 15 per cent. This
may explain why increments of crystallinity of the
coatings water-vapor-treated for 6 h at 127°C with wa-
ter vapor pressure of 0.15 MPa were comparable only
to those of coatings heated at 490°C for 1 h in atmo-
sphere with partial water vapor pressure of 0.01
MPa.10
It was determined previously that completely crys-
AMORPHOUS-TO-CRYSTALLINE HA CONVERSION
talline HA coatings can be obtained at 490°C in atmo-
sphere with partial water vapor pressure of 0.01 MPa
for 2 h.10 Incorporation of water vapor in atmosphere
during heat treatment indeed accelerated the amor-
phous to crystalline conversion by decreasing the ac-
tivation energy for the diffusion. However, the mecha-
nism by which the water vapor decreases activation
energy is still unclear and needs further investigation.
Heat treatment with water vapor, which is superior to
heat treatment in vacuum and in air in terms of phase
transitions, may be the most promising way to raise
crystallinity of the coatings and decrease the tempera-
ture of heat treatment, and it is beneficial to decreasing
both degradation of the coatings and internal stress at
the ceramic/metal interface. Only these phasically and
mechanically stable coatings are suitable for long-term
implantation.
CONCLUSION
Recrystallization of an amorphous phase is a diffu-
sion-controlled process. Incorporation of water vapor
in atmosphere accelerated significantly the amor-
phous-to-crystalline conversion. The higher the water
vapor pressure, the higher the increment of crystallin-
ity. Increase in water vapor pressure resulted in the
decrease of the activation energy for diffusion, thus
raising the diffusion coefficient, which is beneficial to
the recrystallization of the amorphous phase. Both
temperature and water vapor pressure are important
factors affecting the amorphous-to-crystalline conver-
sion.
References
1. K. Hayshi, T. Inadome, T. Mashima, and Y. Sugioka,
‘‘Comparison of bone–implant interface shear strength
of solid hydroxyapatite and hydroxyapatite-coated ti-
tanium implants,’’ J. Biomed. Res., 27, 557–563 (1993).
2. B. C. Wang, E. Chang, and C. Y. Yang, ‘‘A histomor-
phometric study on osteoconduction and osseointegra-
tion of titanium alloy with and without plasma-
sprayed hydroxyapatite coating using back-scattered
electron images,’’ J. Mater. Sci. Mater. Med., 4, 394–403
(1993).
3. W. Tong, J. Chen, X. Li, Y. Cao, J. Feng, Z. Yang, and X.
Zhang, ‘‘Effect of particle size on molten states of start-
245
ing powder and degradation of the relevant plasma-
sprayed hydroxyapatite coatings,’’ Biomaterials, 17,
1507–1513 (1996).
4. J. Wen, X. Liu, and X. Zhang, ‘‘Intrinsic factors of apa-
tite influencing its amorphorization during plasma-
spray coating,’’ Biomaterials, 16, 39–44 (1995).
5. W. Tong, J. Chen, and X. Zhang, ‘‘Amorphorization
and recrystallization during plasma spraying of hy-
droxyapatite coatings,’’ Biomaterials, 16, 829–832 (1995).
6. P. Ducheyne, S. Radin, and L. King, ‘‘The effect of cal-
cium phosphate ceramic composition and structure on
in vitro behavior. I. Dissolution,’’ J. Biomed. Res., 7, 25–34
(1993).
7. R. Z. LeGeros, R. Kijkowska, J. P. LeGeros, R. Zheng,
and D. Fan, ‘‘Calcium phosphate coatings on implants:
Dissolution characteristics,’’ in Transactions of the 19th
Annual Meeting of the Society for Biomaterials, April 28–
May 2, 1993, Birmingham, Alabama, Society for Bioma-
terials, Minneapolis, 1993, p. 221.
8. J. E. Dalton and S. D. Cook, ‘‘In vivo mechanical and
histological characteristics of HA-coated implants vary
with coating vendor,’’ J. Biomed. Mater. Res., 29, 239–
245, 1995.
9. J. Chen, W. Tong, Y. Cao, J. Feng, and X. Zhang, ‘‘Effect
of atmosphere on the phase transformation in plasma-
sprayed HA coatings during heat treatment,’’ J. Biomed.
Mater. Res., 34, 15–20 (1997).
10. W. Tong, J. Chen, J. Feng, Y. Cao, L. Lu, and X. Zhang,
‘‘Further study on post-treatment of plasma-sprayed
hydroxyapatite coatings with water vapor,’’ in 15th
Southern Biomedical Conference, March 28–31, 1996, Day-
ton, Ohio, pp. 160–164.
11. ‘‘Standard recommendation practice for descaling and
cleaning titanium and titanium alloy surfaces,’’ ASTM
B600–74, American Society for Testing and Materials,
Philadelphia, 1974, pp. 669–671.
12. Z. Zyman, J. Weng, X. Liu, X. Zhang, and Z. Ma,
‘‘Amorphous phase and morphological structure of hy-
droxyapatite plasma coatings,’’ Biomaterials, 14, 225–
228 (1993).
13. R. Zallen, The Physics of Amorphous Solids, Wiley , New
York, 1983.
14. J. Chen, J. Wolke, and K. de Groot, ‘‘Microstructure and
crystallinity in hydroxyapatite coatings,’’ Biomaterials,
15, 396–399 (1994).
15. Z. Zyman, J. Weng, X. Liu, and X. Zhang, ‘‘Phase and
structural changes in hydroxyapatite coatings under
heat treatment,’’ Biomaterials, 15, 151–155 (1994).
16. J. Chen, Y. Cao, J. Feng, and X. Zhang, ‘‘The recrystal-
lization of amorphous phase in HA coatings during
heat treatment,’’ in 1st Far-Eastern Symposium on Bio-
medical Materials, October 6–8, 1993, Beijing, China, X.
Zhang and Y. Ikada (eds.), Kobunshi Kankokai, Inc.,
Kyoto, Japan, 1993, pp. 17–18.
Received March 26, 1996
Accepted July 30, 1996