Characterization, thermal stability and sintering of hydroxyapatite powders prepared by different routes

Document Sample
Characterization, thermal stability and sintering of hydroxyapatite powders prepared by different routes Powered By Docstoc
					                                                Materials Chemistry and Physics 94 (2005) 333–341

         Characterization, thermal stability and sintering of hydroxyapatite
                       powders prepared by different routes
                                                             Nasser Y. Mostafa ∗
                                 Chemistry Department, Faculty of Science, Suez Canal University, Ismailia 41522, Egypt

                                 Received 4 October 2004; received in revised form 18 April 2005; accepted 10 May 2005


   Hydroxyapatite (HAP) powder precursors have been used as starting material for biomedical applications, such as synthetic bone graft
materials and scaffold for hard tissue engineering. Considering the numerous applications of hydroxyapatite, three different routes for HAP
powders preparations was investigated. Two powders were prepared by chemical precipitation reactions at 100 ◦ C and one by mechanochemical
reaction. The powders were characterized using chemical analysis, surface area measurements, laser diffraction, X-ray diffraction (XRD) and
SEM. The Ca/P ratios were varied from 1.67 to 1.58. The chemical composition, the crystallinity and the agglomeration characters depend on
the preparation route. The effect of powder characteristics on the sinterability was investigated. Although, the thermal stability and hence the
start of sintering dependents on the Ca/P ratio, the final sintering density and hence the mechanical properties depends on the agglomeration
characteristics and the particle size distribution. Hydroxyapatite powder prepared by mechanochemical route have nano-sized crystallites with
a uniform smaller agglomerated particle size distribution and have a butter sinterability.
© 2005 Elsevier B.V. All rights reserved.

Keywords: Hydroxyapatite; Mechanochemical synthesis; Agglomeration; Crystallinity; Sintering

1. Introduction                                                                 of phosphate ions PO4 3− by hydrogen phosphate HPO4 2−
                                                                                allows a continuous variation of the Ca/P atomic ratio
   Hydroxyapatite [Ca10 (PO4 )6 (OH)2 ] is well known as                        between 9/6 and 10/6. This leads to calcium-deficient hydrox-
the mineral component of bones and teeth. Thus, it has a                        yapatites, Ca10−x (PO4 )6−x (HPO4 )x (OH)2−x . Calcium-defi-
considerable interest in dental and medical research [1–4].                     cient hydroxyapatite powders can be precipitated from
Hydroxyapatite ceramics and hydroxyapatite–polymer                              conventional wet chemical methods and decomposed
composites are very promising for biomedical applications,                      into a mixture of HAP and tricalcium phosphate,
such as synthetic bone graft [5–7] and scaffold for hard tissue                 Ca3 (PO4 )2 (TCP) by thermal treatment above 700 ◦ C.
engineering [8,9]. Considering the numerous applications                        This allows a direct processing of biphasic calcium phos-
of hydroxyapatite in biomedical fields, numerous HAP                             phate ceramics HAP/TCP without the step of powder
synthesis techniques have been developed [2,10]. The most                       blending.
commonly used technique for the formation of HAP powder                             The main interest in the last years is not only control-
is the wet methods in aqueous solutions both by simple                          ling the stoichiometry of synthetic HAP but also controlling
preciptation method [11,12] or hydrolysis of acidic calcium                     the crystal size and shape and the agglomeration character-
phosphate salts [13–15]. Synthesis of HAP using wet method                      istics of the powder. Since these synthetic HAPs are further
is very complicated and needs a special attention to control                    processed to produce ceramics or incorporated with other
the Ca/P ratio as well as the crystallinity. The substitution                   materials (polymers) to produce biocomposites. Therefore,
                                                                                a number of novel processing routes have been developed
 ∗   Tel.: +20 64 382216; fax: +20 64 322381.                                   for preparing fine hydroxyapatite powders, including sol–gel
     E-mail address:                                         syntheses [16,17], hydrothermal reactions [18,19], emulsion

0254-0584/$ – see front matter © 2005 Elsevier B.V. All rights reserved.
334                                  N.Y. Mostafa / Materials Chemistry and Physics 94 (2005) 333–341

and microemulsion syntheses [20,21] and mechanochemical                 freshly boiled distilled water. H3 PO4 was dropped slowly to
syntheses [22–25].                                                      the stirred suspension until a Ca/P ratio of 1.70 was reached.
   Mechanochemical powder synthesis is a solid-state                    Boiling and stirring was continued for 2 h. The supernatant
synthesis method that takes advantage of the perturbation of            liquid was removed. Two liters of water was added to the ves-
surface-bonded species by pressure to enhance thermody-                 sel and the suspension was boiled and stirred for 2 h to remove
namic and kinetic reactions between solids [26]. The main               the Ca(OH)2 . The resulting hydroxyapatite was washed
advantages of mechanochemical synthesis of ceramic pow-                 three times with water and finally with 0.001 M H3 PO4 .
ders are simplicity and low cost, which make it a valuable                  Calcium-deficient hydroxyapatite (d-HAP) was prepared
method for industrial production of HAP powder. Equip-                  by dropping slowly and simultaneously with stirring aqueous
ments used for mechanochemcal synthesis are conventional                solutions of (CH3 COO)2 Ca (0.1 M), and K2 HPO4 (0.06 M),
milling equipment, such as ball mills and vibratory mills.              with a stoichiometric Ca/P molar ratio of 1.67, into 1 l of
   Since the mechanochemical synthesis involves only                    boiling water. The precipitate was aged for 2 h at 100 ◦ C after
solid-state reaction, Suchanek et al. [27] distinguished                dropping solutions, then left to cool down for 24 h. The final
mechanochemical from the mechanochemical–hydrothermal                   pH was find to be weak acidic (pH ∼5.0).
synthesis (wet mechanochemical). In the latter case, an                     The third hydroxyapatite was prepared through wet
aqueous phase incorporates in the system. The aqueous                   mechanochemical route (m-HAP) with Ca(OH)2 and
phase accelerates kinetic processes that commonly rate limits           (NH4 )2 HPO4 as starting materials. Appropriate amounts of
a process, such as dissolution, diffusion, adsorption, reaction         the two starting materials at a molar ratio of 1.67 were mixed
rate, and crystallization (nucleation and growth) [28]. The             together in a conventional ball mill with zirconia balls as
mechanochemical activation of slurries can generate local               the milling medium and 100% distilled water. Millings were
zones of high temperatures (up to 450–700 ◦ C) and high                 carried out for 8 h with the rotating speed of 170 rpm in a
pressures due to friction effects and adiabatic heating of              high-density polyethylene bottle. After milling, the slurry
gas bubbles (if present in the slurry), while the overall               was filtered under suction and washed with freshly boiled
temperature is close to the room temperature [29]. The                  distilled water.
mechanochemical method has many advantages over the                         All the preparations of hydroxyapatite were dried under
hydrothermal methods, especially, on the industrial scale.              N2 , at 105 ◦ C for 24 h then calcined at 600 ◦ C for 2 h. Pellets
Mechanochemical method can produce large amounts of                     specimens were molded into a stainless steel mold by hand
HAP powder at lower temperature, i.e. room temperature, as              to produce specimens 13 mm in diameter and about 2 mm
compared to 90–200 ◦ C for the hydrothermal process. Thus,              high, at a pressure of 200 MPa. The green pellets of about
for the mechanochemical–hydrothermal process, there is no               50–55% of the theoretical density were sintered in air for 1 h
need for a pressure vessel and external heating.                        at temperatures from 850 to 1200 ◦ C, with 50 ◦ C intervals.
   The aim of this research is to characterize several HAP              Specimens 13 mm in diameter and about 18 mm high were
powders prepared by different routes and comparing their                molded, for compressive strength measurements. The heating
sintering behaviors. This is the primary step in a project to           and cooling rates were limited to 300 ◦ C h−1 to prevent crack
investigate the utilization of these hydroxyapatites with poly-         formation.
mers to produce scaffold for tissue engineering with different              The green and sintered densities of compacts were
mechanical properties and different resorbation rates. Com-             obtained from the measurement of dimensions and sample
positional, microstructural, morphological and mechanical               mass. Relative densities were given as a percentage of the
characterizations were carried out on the powder as well as             theoretical density of HAP, 3.16 g cm−3 .
on dense sintered bodies.                                                   X-ray diffraction (XRD) analyses were performed using
                                                                        an automated diffractometer (Scintag Inc., Sunnyvale, CA),
                                                                        at a step size of 0.02◦ , scan rate of 2◦ min−1 , and a scan range
2. Materials and methods                                                from 4◦ to 60◦ 2θ. The average crystallite size of the precip-
                                                                        itates was estimated by using the simple Scherrer equation:
   Pure CaCO3 powder was heated at 1100 ◦ C for 4 h, and
the resultant CaO powder was used as it is or hydrated under                       kλ
nitrogen with a third equivalent quantity of distilled water to                [β1/2 cos θ]
produce Ca(OH)2 . All the other chemicals were of analytical
grades. Two hydroxyapatite powders were prepared by                                               ˚
                                                                        where D is the size in A measured using reflection (0 0 2),
aqueous precipitation method. The precipitation was carried             k the shape factor equal to 0.9, λ the wavelength of X-rays
out in a three-necked bottle equipped with a reflux condenser            equal to 1.5418 A, θ the diffraction angle equal to 12.92 for
protected from the atmosphere by a CO2 -absorbing trap                  the reflection (0 0 2), and β1/2 is defined as the diffraction peak
and ports for introducing N2 and the titrant. Stoichiometric            width at half height, expressed in radians. The particle size of
hydroxyapatite (s-HAP) was prepared by titrating a boiling              the powders was analyzed by using a laser diffraction particle
suspension of Ca(OH)2 with 0.5 M H3 PO4 . The Ca(OH)2                   size analyzer (SHIMADZU), by suspending the powder in
suspension was prepared by adding 56 g of CaO to 3 l of                 water using sonication.
                                            N.Y. Mostafa / Materials Chemistry and Physics 94 (2005) 333–341                                              335

   The specific surface area measurements were performed
with BET method employing a Quantachrome (Monosorb)
Surface Area Analyzer. The Ca/P ratio was measured
by Inductively Coupled Plasma Spectrometry (ICPMS)
(JOBINYBON, mod. JY-38). A universal test machine
(Instron) was used for compression tests.
   The powders and the fracture surfaces of the sintered
samples were investigated using SEM (HITACHI S-3500N)
equipped with secondary electron detector and EDX. All
samples were coated with gold.

3. Results and discussions

3.1. Characterization
                                                                                    Fig. 1. XRD patterns of hydroxyapatite powders dried at 100 ◦ C for 24 h.
    XRD patterns of HAP powders as shown in Fig. 1, revealed
no secondary phases besides hydroxyapatite. The Ca/P ratio
of the HAP powder is one of the most crucial parame-                               (57 nm) and d-HAP has the largest one. These results agree
ters in determining properties and thermal stabilities [30].                       well with the after-mentioned surface area measurements.
Chemical analysis results are presented in Table 1, it is                          d-HAP have the largest crystallites, since it was prepared
clear that the Ca/P ratio of s-HAP powder coincide well                            under acidic conditions. Many authors [31,32] found that the
with the theoretical values for stoichiometric hydroxyap-                          crystallite size increases with increasing the acidity of the
atite (1.67). Hydroxyapatite prepared at low pH (d-HAP) is                         precipitation medium.
highly Ca-deficient (Ca/P = 1.58), and wet mechanochemical                              Fig. 2 shows plots of volume percent of particles sus-
route produces hydroxyapatite (m-HAP) with Ca/P ratio of                           pended in water over a range of particle diameter as deter-
1.65. Table 1 also shows the chemical formula corresponds to                       mined by laser diffraction particle size analyzer. Samples
the after-mentioned chemical analysis, with considering the                        s-HAP and m-HAP have a clear bimodal particle size distribu-
charge compensation of the deficient calcium by substituting                        tion, while d-HAP shows a unimodal particle size distribution
PO4 3− with HPO4 2− and creating a vacancy on OH− site,                            (single maximum at about 15 m). The absence of particles
Ca10−x (PO4 )6−x (HPO4 )x (OH)2−x [2,13,14].                                       with diameters less than 0.2 m (Fig. 3) suggests there is no
    The BET surface of HAP powders measured using N2                               suspended single crystallites of HAP in all samples.
adsorption are given in Table 1. Mechanochemical route                                 To test the effect of ultrasonic energy on the agglomer-
produces powder (m-HAP) with the highest surface area                              ation, the particle sizes were measured for each sample at
(86 m2 g−1 ). Hydroxyapatite prepared in acidic media d-HAP                        different sonication time up to 30 min. This gives no changes
have the lowest surface area (19 m2 g−1 ).                                         in the particle size distribution of Fig. 2 for samples sonicated
    Particle size in agglomerated or aggregated systems like                       for only 5 min. There also no particle sizes were determined
hydroxyapatite powders is one of the most important param-                         corresponding to the results of the XRD, i.e. agglomerates
eters controlling ceramic processing. In this study, particle                      are not broken by ultrasonic energy to the primary particles
size of the powders were investigated using XRD (Scherrer                          (crystallites).
equation) and laser diffraction. Each method of measure-                               Critical parameters describing all curves are shown in
ment senses a particular aspect of particle size. The profile                       Table 2 and include particle-size range, particle size at
broadening by powder X-ray diffraction senses monocrys-                            maximum volume percent, and total volume percent under
talline domains. Laser diffraction method determines only                          a given peak. The two peaks of s-HAP powder bimodal
the agglomerated or aggregated particles in the suspended                          distribution were in the range of 0.6–0.9 and 0.9–45 m. It
solvent.                                                                           was also calculated from the distribution that around 57% of
    The mean crystallite sizes estimated by using the Scherrer                     the powder full in the micrometer size range (0.9–45 m).
equation for the (0 0 2) planes are shown in Table 1. The                          For m-HAP, around 86% of the powder consisted of particles
m-HAP has the smallest crystallites in the nanoscale size                          within 0.2–1.0 m and the other 14% of the mass was

Table 1
Characteristics of hydroxyapatite powders
Sample                Ca/P ratio               Composition                                          Crystallite size (nm)              Surface area (m2 g−1 )
s-HAP                 1.67                     Ca10 (PO4 )6 (OH)2                                   141.3                              48
m-HAP                 1.65                     Ca9.9 (HPO4 )0.1 (PO4 )5.9 (OH)1.9                    57.4                              86
d-HAP                 1.58                     Ca9.48 (HPO4 )0.48 (PO4 )5.48 (OH)1.48               211.9                              19.5
336                                           N.Y. Mostafa / Materials Chemistry and Physics 94 (2005) 333–341

                                                                                 was 0.20 m (Fig. 3(a)). m-HAP powder agglomerates are
                                                                                 small in size than s-HAP agglomerates. It is obvious that the
                                                                                 agglomerate particles in case of m-HAP have lower particle
                                                                                 size than those of s-HAP. The typical volume of primary
                                                                                 m-HAP agglomerates, at 0.1–1.0 m, which agrees with the
                                                                                 laser diffraction results (Fig. 2).
                                                                                     Fig. 3(c) shows that d-HAP, which formed in acidic
                                                                                 medium had a rod-like shape with mean particle lengths of
                                                                                 2 m and diameters of 0.25 m. This figure shows clearly
                                                                                 the agglomeration of the primary crystallites with interparti-
                                                                                 cle bonds glow between them.
                                                                                     In general, the formation of agglomerated HAP particles
                                                                                 can be explained by nucleation–aggregation–agglomeration–
                                                                                 growth mechanism proposed by Randolph and Larson [33]
                                                                                 and Rodriguez et al. [34]. According to this mechanism,
                                                                                 particle formation goes in the following steps: (a) nucleation
                                                                                 and growth to form crystallites in nanometric size range; (b)
                                                                                 aggregation of elemental nanocrystals by physical attrac-
                                                                                 tions; (c) further crystal growth, at a constant residual super-
                                                                                 saturation, acting as cementing agent inside the aggregates
                                                                                 to form stable agglomerate. These agglomerates continue
                                                                                 to grow as distinct particles and appear as submicrometer
                                                                                 particles in case of m-HAP and s-HAP. Finally, the increase
                                                                                 in particle size takes place by aggregation of these sub-
                                                                                 micrometer size agglomerated particles to form secondary
                                                                                 agglomerate in the micrometric size range (second bimodal
                                                                                     Gomez-Morales et al. [35] found that the aggregation
                                                                                 processes directed by different forces, in the nanomet-
                                                                                 ric/colloidal scale, surface free energy minimization would
                                                                                 be responsible for the aggregation, while electrical surface
                                                                                 charges would direct the micrometric scale particles aggre-
           Fig. 2. Particle size distribution by laser diffraction.              gation.

                                                                                 3.2. Sintering and thermal stability
particles of 1.0–25.0 m (Table 2). The agglomerate of m-
HAP is homogenous and most of it full in the submicrometer                           Understanding the sintering behavior of hydroxyapatite
range. This was conformed by SEM.                                                powders is important, because this allows to design ceramics
   The presence of two particle size ranges in case of m-HAP                     with controlled grain growth, microstructure and mechanical
and s-HAP powders possibly reflects the aggregation in the                        properties. In the present study, the effect of powder charac-
submicrometer sizes range, yielding the so-called primary                        teristics on densification, microstructural development and
agglomerates (the first peak). These agglomerates continue                        mechanical properties of HAP compacts were studied.
to aggregate by forming interparticle bonds until they reach                         Fig. 4 shows the effect of sintering temperature on the
the micrometer size range, the second peak (>1 m).                               densification (percent of sintered densities with respect to
   Morphological results from SEM investigation are shown                        theoretical density of HAP; 3.16 g cm−3 ) after cooling to
in Fig. 3(a–c). s-HAP and m-HAP powder consists of hard                          room temperature. The green density for s-HAP and m-
agglomerates which are composed of fine crystallites. In case                     HAP are comparable and smaller than that of d-HAP. This
of s-HAP, the dominant observed agglomerated particle size                       is due to the nature of agglomerated particles of m-HAP

Table 2
Mean parameters of particle size for hydroxyapatite powders
Sample               First peak                                                               Second peak

                     Range ( m)                 Peak ( m)             Volume (%)              Range ( m)          Peak ( m)           Volume (%)
s-HAP                0.6–0.9                     0.6                   43                     0.9–45              2.0                 57
m-HAP                0.2–1.0                     0.3                   86                     1.0–25              2.0                 14
d-HAP                  7–100                    15                    100                     –                   –                   –
                                        N.Y. Mostafa / Materials Chemistry and Physics 94 (2005) 333–341                                         337

                                                                           Fig. 4. Densification of hydroxyapatite compacts sintered at different tem-

                                                                           900 and 1000 ◦ C, compared to between 950 and 1000 ◦ C for
                                                                           m-HAP and between 1000 and 1050 ◦ C for stoichiometric
                                                                              This is clearly shown in SEM micrographs of the fracture
                                                                           surfaces of powder compacts sintered at 1000 ◦ C (see Fig. 5).
                                                                           Compared with the as synthesized powder (Fig. 3) evidence
                                                                           exists for a particle coalescence which associated with the
                                                                           surface reduction [36]. The particle coarsening occurred more
                                                                           significantly with samples with low Ca/P ratio. The necking
                                                                           among the particles was apparent in Ca-deficient samples (d-
                                                                           HAP and m-HAP) due to localized sintering at 1000 ◦ C. At
                                                                           this sintering temperature the particle size decreases as the
                                                                           Ca/P ratio increases.
                                                                              It is remarkable to notice that at 1000 ◦ C, sample s-
                                                                           HAP shows a collapse of the large agglomerate (compare
                                                                           Figs. 3(a) and 5(a)). In another investigation [37], the authors
                                                                           find, the calcination reduces the particle size and narrows the
                                                                           size distribution.
                                                                              The second stage of sintering corresponds to densifica-
                                                                           tion and the removal of most of the specimen porosity. The
                                                                           onset of densification as indicated by a sharp increase in the
                                                                           sintered density, for d-HAP and m-HAP were started nearly
                                                                           at 1000 ◦ C, whereas stoichiometric hydroxyapatite (s-HAP)
                                                                           required higher temperature 1050 ◦ C.
                                                                              For d-HAP, the maximum rate of densification was
                                                                           between 1050 and 1150 ◦ C. Comparing this behavior with
                                                                           that of stoichiometric hydroxyapatite, calcium-deficient HAP
                                                                           sinters at much lower temperature, and the temperature of
                                                                           maximum densification decreases as the calcium deficiency
                                                                              m-HAP achieves the highest final sintered density of
Fig. 3. SEM micrographs of hydroxyapatite powders: (a) s-HAP; (b) m-       93.5% of the theoretical density at 1200 ◦ C, whereas d-HAP
HAP; (c) d-HAP.
                                                                           and s-HAP reach only 89.4 and 76.8% respectively, at the
                                                                           same sintering temperature. In conclusion, the start of densifi-
and s-HAP, which are composed of nano-sized crystallites.                  cation depends on the chemical composition of the precursor
A small increase in density is observed before the onset of                powder, i.e. Ca/P ratio. The final density of the sintered bod-
densification and this corresponds to the first stage of sinter-             ies mainly depends on the homogeneity and the particle size
ing, where necks are forming between powder particles. For                 distribution of the precursor powder. This conclusion is sup-
highly calcium-deficient d-HAP, this process occurs between                               u
                                                                           ported by G¨ ltekin and Oktar [39]; they found that the final
338                                        N.Y. Mostafa / Materials Chemistry and Physics 94 (2005) 333–341

Fig. 5. SEM of hydroxyapatite sintered at 1000 ◦ C: (a) s-HAP; (b) m-HAP;     Fig. 6. SEM of hydroxyapatite sintered at 1200 ◦ C: (a) s-HAP; (b) m-HAP;
(c) d-HAP.                                                                    (c) d-HAP.

                                                                              increase of sintering temperature improves the homogene-
limiting densification level depend predominantly on the sur-                  ity and favors grain growth, resulting in an increase in mean
face area of the HAP powder and to a lesser extent on Ca/P                    grain size. The increase in grain size is also depending on the
ratio [39].                                                                   homogeneity of the precursor powder.
    SEM micrographs of the fracture surfaces of powder com-                       Sample m-HAP shows the isolated closed pores in the
pacts sintered at 1200 ◦ C are given in Fig. 6. In general, the               matrix of uniform closed packed HAP grains. Samples
                                          N.Y. Mostafa / Materials Chemistry and Physics 94 (2005) 333–341                                     339

Table 3
Chemical composition and physicomechanical properties of sintered ceramics
Sample              Ca/P molar ratio            TCP (wt.%)              HAP (wt.%)                Porosity (%)           Compressive strength (MPa)
s-HAP               1.67                         0.0                    100                       23.2                    58
m-HAP               1.65                         9.3                     90.7                      6.5                   108
d-HAP               1.58                        50.1                     49.9                     10.6                    87

s-HAP and d-HAP show larger pores in the range from 0.3
to 1 m size. These pores seem to be connected to form a
continuous network separated by thick walls. This may be
an advantage for the circulation of the physiological fluid
when it is used for biomedical purposes but it is not useful
for the intergrowth of cells [38].
    Comparing the final density of sample m-HAP and s-HAP,
which have the same green density, leads to the conclusion
improved packing homogeneity leads to improved sinterabil-
ity in the form of higher sintered density at a given sintering
temperature and the ability to achieve high density with fine
particle size.

3.3. Compressive strength

    Table 3 shows porosity and compressive strength of HAP
sintered compacts at 1200 ◦ C. It can be seen that the decrease
in porosity is accompanied by an increase in the mechanical
strength. Sample m-HAP, with improved packing homogene-
ity attains the highest compressive strength (1080 MPa) and                     Fig. 7. XRD patterns of hydroxyapatite powders compacts sintered at
the lowest porosity (6.5%). Stoichiometric HAP, s-HAP have                      1100 ◦ C.
the lowest compressive strength (578 MPa) and the highest
porosity (23.2%).
                                                                                where x is the calcium deficiency, and thus, Ca/P = (10 − x)/6.
                                                                                The mole fraction of -TCP to the mole fraction of HAP
3.4. Thermal stability                                                          in the sintered bodies is given by: XTCP /XHAP = 3x/(1 − x).
                                                                                Since, XHAP = 1 − XTCP, and XTCP /(1 − XTCP ) = 3x/(1 − x),
    XRD patterns of sintered materials at 1100 ◦ C, as shown                    the percent of -TCP can be calculated by knowing
in Fig. 7, highlighted the difference in the thermal stability                  calcium deficiency (x) from the chemical analysis.
of HAP powders. HAP synthesized by mechanochemical                              Table 2 shows the Ca/P ratio and weight percent of -
method and from Ca(CH3 COO)2 precursor (in acidic                               TCP and HAP in the sintered compacts. Results agree
media) transformed partially into -tricalcium phosphate, -                      well with the semiquantitative analysis using XRD in
Ca3 (PO4 )2 , which related to the Ca-deficiency in the lattice                  Fig. 7.
of HAP powders. This agreed with the research of Yoshimura
et al. [19], in which he indicated that Ca-deficient HAP tends
to transform to -Ca3 (PO4 )2 on heating at 900 ◦ C depending                    4. Conclusion
on the deficiency of calcium. Stoichiometric hydroxyapatite
(s-HAP) with Ca/P ratio of 1.67 is stable up to 1200 ◦ C.                          The present study compares the powder properties of
    Raynaud et al. [41] proved the reliability of the Ca/P                      hydroxyapatite prepared by wet mechanochemical method
atomic ratio of the initial HAP powder calculated by                            with those prepared at 100 ◦ C using wet solution precipita-
determining the phase proportions in calcined biphasic                          tion methods. The wet mechanochemical method produces
calcium phosphates. Thus, the percent of -TCP and HAP in                        nanocrystallites of hydroxyapatite with homogenous fine
the sintered bodies can be determined from the Ca/P ratio of                    particle size distribution with most of the agglomerated
the precursor HAP powder. Considering calcium-deficient                          particles full on the submicrometer size range. This alters the
HAP change with thermal treatment to stoichiometric HAP                         sintering characteristics of hydroxyapatite powder compacts
and -TCP according to the following equation:                                   by attaining high sintering density at lower temperature in
                                                                                comparison with those prepared by wet solution precipita-
Ca10−x (HPO4 )x (PO4 )6−x (OH)2−x
                                                                                tion method at 100 ◦ C. Hence, this preparation route could
  → (1 − x)Ca10 (PO4 )6 (OH)2 + 3x -Ca3 (PO4 )2 + xH2 O                         emerge as a simple synthetic route to produce HAP ceramics
340                                  N.Y. Mostafa / Materials Chemistry and Physics 94 (2005) 333–341

with an enhanced characteristics, especially, for industrial            References
                                                                         [1] L.L. Hench, J. Am. Ceram. Soc. 81 (7) (1998) 1705.
    The sintering behavior of HAP powders with different
                                                                         [2] J.C. Elliott, Structure chemistry of the apatites and other calcium
characteristics, allows the identification of deferent factors                orthophosphates, in: Studies in Inorganic Chemistry, Elsevier, Ams-
effecting sintering behavior and density of the final ceramics.               terdam, 1994.
At the beginning of the sintering, below 1000 ◦ C, particle              [3] F.H. Jones, Surf. Sci. Rep. 42 (2001) 75.
coalescence occurs without densification or with little densi-            [4] C. Rey, Calcium phosphate for medical applications, in: Z. Amjad
                                                                             (Ed.), Calcium Phosphate in Biological and Industrial Systems,
fication. The particle coalescence usually associated with a
                                                                             Kluwer Academic Publishers, Boston, 1998.
reduction of the specific surface area [36]. The temperature, at          [5] Y. Shikinami, M. Okuno, Biomaterials 20 (1998) 859.
which start of this coalescence occurs, decreases as the Ca/P            [6] K.S. TenHuisen, R.I. Martin, M. Klimkiewicz, P.W. Brown, J.
ratio of the powder decreases. Nearly at the same temperature,               Biomed. Mater. Res. 29 (1995) 803.
the calcium-deficient hydroxyapatites dissociate into a mix-              [7] K.S. TenHuisen, P.W. Brown, J. Biomed. Mater. Res. 28 (1994)
ture of stoichiometric hydroxyapatite and -tricalcium phos-
                                                                         [8] D.W. Hutmacher, Biomaterials 21 (2000) 2529.
phate. Stoichiometric; s-HAP (Ca/P = 1.67) is stable upon all            [9] S.H. Li, K. deGroot, P. Layrolle, Key Eng. Mater. 218–220 (2002)
the temperature range studied. This agrees with fact that the                35.
transfer of matter by superficial diffusion is enhanced by the           [10] A.D. Papargyris, A.I. Botis, S.A. Papargyri, Key Eng. Mater.
increase of vacancies and defects in the structure (i.e. by a                206–213 (2002) 83.
                                                                        [11] Y. Liua, W. Wanga, Y. Zhana, C. Zhenga, G. Wanga, Mater. Lett.
decrease of the Ca/P ratio), i.e. the dissociation reaction helps
                                                                             56 (2002) 496.
in activation the transfer of matter by superficial diffusion            [12] R.E. Rimana, W.L. Suchaneka, K. Byrappaa, C.W. Chena, P. Shuka,
[36].                                                                        C.S. Oakesa, Solid State Ionics 151 (2002) 393.
    The densification stage is seemed to be also assisted by             [13] K.S. TenHuisen, P.W. Brown, Biomaterials 19 (1998) 2209.
the dissociation of calcium-deficient HAP, since the start               [14] R. Martin, P.W. Brown, J. Biomed. Mater. Res. 35 (1997)
of densification for Ca-deficient hydroxyapatite (d-HAP
                                                                        [15] P.W. Brown, R.I. Martin, K.S. TenHuisen, Factors influencing the
and m-HAP) start at lower temperatures than stoichiometric                   formation of monolithic hydroxyapatite at physiological temperature,
hydroxyapatite (s-HAP). In this stage, the densification is                   in: Biomedical and Biological Applications of Glass and Ceramics,
known to result from volume diffusion and grain boundary                     Am. Ceram. Soc. (1996) 37–48.
diffusion [36].                                                         [16] G. Bezzi, G. Celotti, E. Landi, T.M.G. La Torretta, I. Sopyan, A.
                                                                             Tampieri, Mater. Chem. Phys. 78 (2003) 816.
    Although, the start of densification depends mainly on the
                                                                        [17] T. Anee Kuriakose, S. Narayana Kalkura, M. Palanichamy, D.
Ca-deficiency (Ca/P ratio), the final sintering density depends                Arivuoli, G. Karsten Dierks, C. Bocelli, Betzel, J. Cryst. Growth
on the particle size, the homogeneity and the agglomeration                  263 (2004) 517.
character of the powder precursor. Thus, the final densifi-               [18] H. Hattori, Y. Iwadate, J. Am. Ceram. Soc. 73 (1990) 1803.
cation ratio of sintered materials was maximum for m-HAP                [19] M. Yoshimura, H. Suda, K. Okamoto, K. Ioku, J. Mater. Sci. 29
                                                                             (1994) 3399.
at 93.5% of the theoretical density. It was slightly lower, at
                                                                        [20] M.G. Murray, J. Wang, C.B. Ponton, P.M. Marquis, J. Mater. Sci.
89.4% in case of stoichiometric hydroxyapatite and very low;                 30 (1995) 3061.
76.8% in case of highly deficient hydroxyapatite (d-HAP),                [21] G.K. Lim, J. Wang, L.M. Gan, Mater. Lett. 30 (1996)
i.e. m-HAP, which was produced by wet mechanochemical                        431.
route and has a nanocrystallites with homogenous agglom-                [22] W. Kim, Q. Zhang, F. Saito, J. Mater. Sci. 35 (2000)
erated particles, attains the highest densification. These
                                                                        [23] K.C.B. Yeong, J. Wang, S.C. Ng, Biomaterials 22 (2001)
results are in partial disagreement with the result of Raynaud               2705.
et al. [40,41], who found the densification decrease with                [24] S. Rhee, Biomaterials 23 (2002) 1147.
increasing the -TCP contents in HAP/TCP sintered bodies.                [25] B. Yeong, X. Junmin, J. Wang, J. Am. Ceram. Soc. 84 (2001)
Thus, they concluded, the lower sinterability of calcium-                    465.
                                                                        [26] E. Gutman, Mechanochemistry of Materials, Cambridge International
deficient HAP in comparison with the stoichiometric
                                                                             Science Publishing, Cambridge, UK, 1997.
HAP.                                                                    [27] W.L. Suchanek, P. Shuk, K. Byrappa, R.E. Riman, K.S. TenHuisen,
    In conclusion, HAP ceramics prepared from mechano-                       V.F. Janas, Biomaterials 23 (2002) 699.
chemical synthesis powders were found to exhibit better den-            [28] M. Yoshimura, W. Suchanek, Solid State Ionics 98 (1997) 197.
sification, uniform microstructure and improved mechanical               [29] N.V. Kosova, A.K. Khabibullin, V.V. Boldyrev, Solid State Ionics
                                                                             101–103 (1997) 53.
                                                                        [30] M.A. Fanovich, J.M. Porto LoPez, J. Mater. Sci. Mater. Med. 9
                                                                             (1998) 53.
                                                                        [31] S. Suzuki, M. Ohgaki, M. Ichiyanagi, M. Ozawa, J. Mater. Sci. Lett.
                                                                             17 (1998) 381.
Acknowledgements                                                        [32] K. Kandori, N. Horigami, A. Yasukawa, T. Ishikawa, J. Am. Ceram.
                                                                             Soc. 80 (1997) 1157.
                                                                        [33] A.D. Randolph, M.A. Larson, Theory of Particulate Processes, sec-
   The author acknowledges the financial support from HEC                     ond ed., Academic Press, NY, 1986.
(Egypt) and Dr. P. W. Brown (Pennsylvania State University)                           ı                      o                  o
                                                                        [34] R. Rodr´guez-Clemente, A. L´ pez-Macipe, J. G´ mez-Morales, J.
for useful discussions.                                                                     e            n
                                                                             Torrent-Burgu´ s, V.M. Casta˜ o, J. Eur. Ceram. Soc. 18 (1998) 1351.
                                     N.Y. Mostafa / Materials Chemistry and Physics 94 (2005) 333–341                                    341

[35] J. Gomez-Morales, J. Torrent-Burgues, R. Rodriguez-Clemente,       [38] B. Chang, C. Lee, K. Hong, H. Youn, H. Ryu, S. Chung, K. Park,
     Cryst. Res. Technol. 36 (2001) 1065.                                    Biomaterials 21 (2000) 1291.
[36] D. Bernache-Assollanta, A. Ababoua, E. Championa, M. Heughe-                u
                                                                        [39] G. G¨ ltekin, F.N. Oktar, Mater. Lett. 56 (2002) 142.
     baertb, J. Eur. Ceram. Soc. 23 (2003) 229.                         [40] S. Raynaud, E. Champion, D. Bernache-Assollant, J. Laval, J. Am.
[37] N. Thangamania, K. Chinnakalib, F.D. Gnanama, Ceram. Int. 28            Ceram. Soc. 84 (2001) 359.
     (2002) 355.                                                        [41] S. Raynaud, E. Champion, D. Bernache-Assollant, Biomaterials 23
                                                                             (2002) 1073.

materialresearch materialresearch