Influence of temperature and concentration on the sintering by


									                                          Acta Materialia 52 (2004) 5655–5663

         Influence of temperature and concentration on the sintering
            behavior and mechanical properties of hydroxyapatite
                                                             a,**                 a,*
              Chandrasekhar Kothapalli                            , M. Wei              , A. Vasiliev a, M.T. Shaw                b,c

                     Department of Metallurgy and Materials Engineering, Institute of Materials Science, University of Connecticut,
                                            97 North Eagleville Road, U-3136, Storrs, CT 06269, USA
                     Department of Chemical Engineering, University of Connecticut, 191 Auditorium Road, Storrs, CT 06269, USA
                          Polymer Program, Institute of Materials Science, University of Connecticut, Storrs, CT 06269, USA

                            Received 19 February 2004; received in revised form 16 July 2004; accepted 23 August 2004
                                                      Available online 28 September 2004


    Human bone mineral contains calcium-deficient crystalline hydroxyapatite (HA) embedded in collagen fibers. Research over the
past two decades has focused on preparing synthetic HA, which closely resembles bone apatite and exhibits excellent osteoconduc-
tivity. This paper describes the synthesis of nano-HA particles via a wet precipitation method. The concentration of the reactants
(0.5, 1.0 and 2.0 g/dL) and the temperature of the reaction (25, 70 and 100 °C) were varied. FESEM images were used to determine
the size and shape of the resulting nano-particles. The length and breadth of the HA particles were found to increase with the tem-
perature, while the aspect ratio increased with both the concentration and the temperature. The average length of the particles was in
the range 53–165 nm and the average breadth in the range 29–52 nm. Agglomerates of HA precipitates were formed during the syn-
thesis process. HA precipitated at 25 °C and concentration 0.5 g/dL resulted in large agglomerates with a specific surface area of 79.8
m2/g. HA agglomerates synthesized at each condition were pressed into discs and sintered at 1200 °C. It was found that there was a
positive correlation (p = 0.015) between sintered density and biaxial flexural strength. A maximum strength of 57.4 MPa was
observed for the specimens 2.0–70 which also attained the highest density, 92%. XRD results indicated that most of the sintered
discs had slightly decomposed. The decomposition of the specimens and their resulting microstructures also contributed to the
mechanical strength drop of the specimens.
Ó 2004 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

Keywords: Hydroxyapatite; Reactant concentration; Reaction temperature; Density and biaxial flexural strength

1. Introduction                                                               tivity. Many methods of synthesizing HA have been
                                                                              reported, such as solid-state reaction, sol–gel, wet syn-
   Over the past two decades, numerous efforts have                            thesis and hydrothermal methods [1–5]. The sol–gel
been made to prepare hydroxyapatite (HA) for bone tis-                        method involves molecular mixing of the calcium and
sue applications due to its excellent biocompatibility and                    phosphorous resulting in chemical homogeneity, but
bioactivity. Of all the calcium phosphates, HA resem-                         has drawbacks such as the possible hydrolysis of the
bles most closely bone apatite and exhibits osteoconduc-                      phosphates and the high cost of raw materials [6]. Also,
                                                                              the HA prepared by this method resulted in relatively
                                                                              inferior crystallinity and thermal stability. Although
     Corresponding authors. Tel.: +1 860 486 9253; fax: +1 860 486            HA ceramics with high aspect ratios were prepared by
     Tel.: +1 860 486 3543; fax: +1 860 486 4745.
                                                                              hydrothermal and solid-state reactions [1–5], the
    E-mail addresses: (C. Kothapalli),                   procedures employed were relatively complicated and (M. Wei).                                                 the temperatures involved were comparatively high. In

1359-6454/$30.00 Ó 2004 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
5656                                 C. Kothapalli et al. / Acta Materialia 52 (2004) 5655–5663

addition, it has been proven difficult to obtain controlled            ing few drops of strong ammonia (Sigma, ACS rea-
morphology with these two methods.                                   gent). The nitrate solution was either kept at room
   A variety of methods have been explored in attempts               temperature or heated to 70 or 100 °C while stirring
to prepare nano-sized HA particles with relatively high              vigorously. The temperature during the reaction was
aspect ratios [7–9]. Such HA nano-particles have pri-                controlled using a hot plate so that there is a maximum
mary importance for the biomedical field because they                 fluctuation of ±2 °C. The phosphate solution was
can be blended into biodegradable polymers to make                   added into each of the calcium solution in a dropwise
composites with superior mechanical properties. In                   manner. HA precipitates were formed after mixing
addition, it was reported that nano-sized HA exhibited               the solutions, which was continuously stirred and main-
better bioactivity than coarser crystals [10].                       tained at the reaction temperature for 3 h. Finally, the
   Many investigations [5,11,12] studied a wide range of             precipitates were aged for 7 days without heating and
experimental conditions to prepare rod-like HA parti-                stirring at room temperature. They were then washed
cles. However, fibrous HA with relatively low crystallin-             with distilled water until the traces of ammonia were re-
ity and poor purity were formed. The key factors for                 moved, dried at 80 °C, and ground to powders using a
synthesizing HA particles with desired properties such               mortar and pestle. Table 1 describes the notation used
as stability, controllable aspect ratios and bioactivity             for the HA samples prepared at various concentrations
are: reaction temperature, pH and concentration of the               and temperatures.
reactants. The properties of HA, particularly those men-
tioned above, affects the efficiency of the powder in its               2.2. Characterization of HA particles
ultimate applications [13–16]. In this study, a wet precip-
itation method was employed to prepare nano-sized HA                    Hydroxyapatite particles thus synthesized were exam-
particles. This method was advantageous because it re-               ined using XRD (BRUKER AXS D5005) with a copper
sulted in homogenous pure HA precipitates with good                  target. The voltage and current used were 40 kV and
crystallinity.                                                       40 mA, respectively. A step size of 0.02° and a scan
   The effects of temperature and concentration of the                speed of 1.0° minÀ1 were used. The analysis was con-
reacting solutions were examined, along with their                   ducted for both the as-prepared and calcined specimens
influence on HA properties such as morphology, sinter-                at 1200 °C. The morphological features of the HA pre-
ing behavior and mechanical strength. HA was synthe-                 cipitates prepared at different conditions were studied
sized by setting the initial concentration of the reactant           using JEOL JSM-6340 (15 kV) type high-resolution
solutions at 0.5, 1.0 and 2.0 g/dL, while the tempera-               FESEM. The specific surface area of the HA powder
tures of the reaction were maintained at 25, 70 and                  was calculated using the BET method from a N2 adsorp-
100 °C at each concentration. Hence nine independent                 tion isotherm obtained from a Micromeritics ASAP
batches of HA particles were synthesized. The samples                2010 (Micrometrics) surface area analyzer operating be-
were characterized using X-ray diffraction (XRD),                     tween 10 and 127 kPa.
Brunauer–Emmett–Teller (BET) specific surface area,
scanning electron microscope (SEM) and field emission                 2.3. Mechanical testing and characterization of sintered
scanning electron microscope (FESEM). In the present                 HA
paper, we choose to sinter the pellets at 1200 °C as our
rationale was to have consistent sintering temperature                  Hydroxyapatite powders prepared at different condi-
with other researchers in the field to compare the                    tions were pressed into half-inch diameter discs at a
mechanical strength results we obtained. This paper is               pressure of 150 MPa for 1 min and sintered in air at
the first known report which systematically studied                   1200 °C for 1 h with a heating rate of 5 °C/min. Nine
the effect of reactant concentration on HA particle
morphology, sintering behavior and mechanical
properties.                                                          Table 1
                                                                     Preparation conditions of the samples studied in this work
                                                                     Sample ID           Concentration (g/dL)           Temperature (°C)
2. Materials and methods                                             0.5–25              0.5                             25
                                                                     0.5–70                                              70
2.1. Preparation of HA                                               0.5–100                                            100
                                                                     1.0–25              1.0                             25
   Fifty milliliters of calcium nitrate (99% pure, Sigma)            1.0–70                                              70
and 82 mL of ammonium hydrogen orthophosphate                        1.0–100                                            100
(99% pure, Sigma) solutions in distilled water were pre-             2.0–25              2.0                             25
pared at concentrations of 0.5, 1.0 and 2.0 g/dL, and                2.0–70                                              70
                                                                     2.0–100                                            100
the pH of each solution was brought up to 12 by add-
                                                                                                C. Kothapalli et al. / Acta Materialia 52 (2004) 5655–5663                             5657

discs in each group were prepared, and the average bulk                                                                         on the losses incurred during washing of precipitates.
density of each specimen before and after sintering was                                                                         HA particles prepared at different temperatures and
calculated by measuring its weight and size. XRD was                                                                            concentrations were observed using FESEM as shown
used to examine the decomposition of the sintered                                                                               in Fig. 2, which provided the basis for subsequent par-
specimens. The biaxial flexural strength of sintered spec-                                                                       ticle size distribution analysis. The particle size distri-
imens was measured using Instron testing machine                                                                                bution studies were not done before grinding the HA
according to ASTM C 1499-03 [24]. The surface mor-                                                                              powders using mortar and pestle. Five different FES-
phology of the sintered specimens was studied by scan-                                                                          EM images of each sample were considered for meas-
ning electron microscopy (SEM). The sintered pellets                                                                            uring the length, breadth and aspect ratio of the HA
were polished using various grades of silicon carbide pa-                                                                       particles; though we had included only one representa-
pers (grade 400–1200) and 1.0 lm diamond paste was                                                                              tive image of each sample in this paper. Results indi-
used for the final polishing. The pellets were then coated                                                                       cated that the length and breadth of the particles of
with a gold film and their surface morphologies were ob-                                                                         all samples follow normal distribution. It was observed
served using an environmental scanning electron micro-                                                                          that the morphology of HA crystals was strongly
scope (ESEM).                                                                                                                   dependant on the concentration of the reactants and
                                                                                                                                the temperature of the reaction. Table 2 summarizes
                                                                                                                                the particle size distribution of HA particles synthe-
3. Results                                                                                                                      sized at various conditions. At a constant reactant
                                                                                                                                concentration, as the reaction temperature increased,
3.1. Before calcination                                                                                                         the particle size and aspect ratio also increased
                                                                                                                                (p = 0.006); at a constant reaction temperature, with
3.1.1. XRD patterns of as-prepared HA precipitates                                                                              the increase of the reactant concentration, the aspect
   Fig. 1 shows the typical XRD pattern of the HA pow-                                                                          ratio of the precipitates increased (p = 0.033) with a
der synthesized at different concentrations and tempera-                                                                         predicted maximum of 3.6 at reaction conditions of
tures (Table 1). All peaks perfectly matched with the                                                                           2.0 g/dL and 100 °C.
JCPDS pattern 9-432 for HA, which suggested that pure
HA was obtained for all the samples prepared at varied                                                                          3.2. Formation of HA agglomerates
conditions. The lattice parameters of the HA were found
                     ˚            ˚                 ˚
to be a = b = 9.426 A, c = 6.887 A and c = 119.97 A. It                                                                         3.2.1. Agglomerate size and distribution
was observed that as the reaction temperature increased                                                                            Most of the methods used for synthesizing HA
at a constant concentration, the diffraction peaks be-                                                                           produced heterogeneous agglomerates either during
came more intense, indicating an increase in the crystal-                                                                       the chemical reaction or subsequent drying process.
linity of the HA powders (p = 0.0055).                                                                                          The average agglomerate size (by weight) obtained
                                                                                                                                from the light scattering technique is shown in
3.1.2. Particle size distribution based on FESEM images                                                                         Fig. 3. There was not enough evidence to draw con-
   At each concentration and temperature, the typical                                                                           clusions regarding the effect of concentration on
yield of HA powder was around 5.5–5.8 g depending                                                                               agglomerate size. However, with the increase of the
                                                                                                                                reactant temperature, the HA agglomerate size de-
                                                                                                                                creased (p = 0.046).

                                                                                                                                3.2.2. Specific surface area

                                                                                                                                   Fig. 4 shows the specific surface area of as-precipi-
                                                                                                                                tated HA synthesized at various conditions. Increasing

                                                                                                                                the reaction temperature resulted in large HA crystals
 Arbitrary units

                                                                                                                                with a reduction in their specific surface areas. It was

                                                                                                                                found from Fig. 4 that the smallest HA particles syn-
                                                                                                                                thesized at the concentration of 0.5 g/dL and the tem-
                                                                                                                                perature of 25 °C had the surface area of 79.8 m2/g,
                                                                                                    (402) (410)


                                                                                                                                while the particles obtained at higher reaction temper-



                                                                                                                                atures such as 2.0–100, had lower surface areas. The






                                                                                                                                specific surface area values obtained were in the range

                                                                                                                                reported by few other research groups [25–28] who also
                   10           20                   30                         40                50          60     70         employed the wet method to synthesize HA. The slight
                                                                                                                                difference between the observed and reported values
                                                                                                                                might be due to the different synthesizing conditions
 Fig. 1. XRD powder patterns for the samples of HA as-prepared.                                                                 used.
5658                                C. Kothapalli et al. / Acta Materialia 52 (2004) 5655–5663

                             Fig. 2. FESEM images of HA particles synthesized at various conditions.

3.3. Characterization of sintered HA discs                          for the samples 0.5–100 and 1.0–100, all the remaining
                                                                    samples were partially decomposed: a low peak of
3.3.1. XRD analyses                                                 monetite (CaHPO4) was observed at 31° 2h in the
   Fig. 5(a)–(c) shows the XRD patterns of HA discs                 specimens (0.5–25, 1.0–25 and 2.0–25) as shown in
sintered at 1200 °C for 2 h. It was found that except               Fig. 5(a), while CaO was detected in the specimens
Table 2
Statistics of particle size and distribution of HA synthesized at different
Sample ID                                 Number           Average          Average           Average                                                       2.1 Tm 0567 672.985 l 3.843
                                          of particles     length (nm)      breadth (nm)      aspect ratio

                                                                                                              Relative intensity
0.5–25                                     57               53              29                1.8
0.5–70                                    110               76              32                2.3
0.5–100                                    49              157              58                2.7
1.0–25                                     40               64              29                2.2
1.0–70                                     76              104              37                2.8
1.0–100                                    18              141              52                2.7
2.0–25                                     10               56              29                1.9
2.0–70                                     23              132              40                3.3
2.0–100                                    12              165              42                3.9
                                                                                                                                   20   30   40        50            60

                                  1.8           0.5 g/dL
 Log (agglomerate size, µm)




                                  1.0                                              2.0 g/dL
                                          1.0 g/dL



                                     20         40          60         80        100                 120
                                                         Reaction temperature,˚C

Fig. 3. Average volume-weighted agglomerate size of particles syn-
thesized at various conditions.

   Specific surface area, m2 /g



                                                                                 0.5 g/dL

                                  20                             2.0 g/dL

                                                                                       1.0 g/dL
                                    20          40            60            80         100           120
                                                     Reaction temperature,˚C

Fig. 4. Specific surface area results of as-precipitated HA powder
synthesized at various conditions.                                                                           Ca10 ðPO4 Þ6 ðOHÞ2 !2Ca3 ðPO4 Þ2 þCa3 P2 O8 þCaOþH2 O
prepared at 70 °C (0.5–70, 1.0–70 and 2.0–70) as shown in
Fig. 5(b). In comparison, a large proportion of sample
2.0–100 was decomposed into a-tricalcium phosphate                                                           3.3.2. Bulk density
(Ca3(PO4)2), b-tricalcium phosphate (Ca3P2O8) and                                                               The bulk density of HA discs before and after sinter-
calcium oxide according to the following equation [19]:                                                      ing are plotted in Fig. 6. It was found that the density of
5660                                              C. Kothapalli et al. / Acta Materialia 52 (2004) 5655–5663

                                                                                  as the temperature increased, the flexural strength in-
                                                                                  creased proportionally (p = 0.004). The observed maxi-
                                                                                  mum flexural strength of 57.4 MPa was close to the
                                                                                  values reported by Jarcho et al. [22] and Rodriguez
                                                               Conc, g/dL
                                                                                  et al. [29].
 ρ, g/cm3

            2.5                                                                   3.4. Surface morphology

            2.0                                                                       Fig. 9 shows the representative SEM images of HA
                                                                  1.0             ceramics sintered at 1200 °C for 1 h. Nearly pore-free
                              pre-sintered                        2.0
            1.5                                                                   surfaces were observed for the sample 2.0–70, which
                                                                                  incidentally demonstrated maximum flexural strength.
            1.0                                                                   Especially, the sample 2.0–100, which decomposed into
               20        40          60         80          100             120   tricalcium phosphates, after sintering, showed a highly
                                Reaction temperature, ˚C
                                                                                  porous surface with pores distributed all over the sur-
Fig. 6. Bulk densities of the samples before and after calcination as a           face. Open round pores were also observed for the sam-
function of synthesis temperature.                                                ples synthesized at 25 °C and concentrations of 0.5, 1.0
                                                                                  and 2.0 g/dL (the images of which are not shown here).
                                                                                      Fig. 10 shows the variation of agglomerate size of HA
the pre-sintered HA discs increased with synthesis tem-
                                                                                  precipitates and their specific surface area with the par-
perature (p = 0.01) in spite of the concentration of the
                                                                                  ticle surface area. The surface area of the particles was
reactants. The highest density observed was for the discs
                                                                                  calculated using the relation:
pressed from the HA powders precipitated at 100 °C,
and concentrations of 0.5, 1.0 and 2.0 g/dL. After sinter-                        4à ½1 þ 0:5=ðL=Dފ=qà D                            ðm2 =gÞ;
ing at 1200 °C, the density of the discs pressed from
                                                                                  where D is the average breadth of the particles, L is the
samples 0.5–25, 1.0–25 and 2.0–25 reached 80–85% of
                                                                                  average length of the particles and q is the theoretical
the theoretical density, i.e., 2.91 g/cm3. The samples
                                                                                  density of HA (3.16 g/cm3). It was observed that the
2.0–70 exhibited the highest sintered density of 92%
                                                                                  agglomerate size of the HA precipitates increased with
while the samples 2.0–100 showed the lowest density
                                                                                  the surface area of the particles (p = 0.001). The HA pre-
among all the samples tested.
                                                                                  cipitates having small particle sizes and thereby high
                                                                                  agglomerate sizes possessed high specific surface areas.
3.3.3. Biaxial flexural strength
                                                                                     It was observed that a strong correlation exists be-
   Fig. 7 shows the biaxial flexural strength of the HA
                                                                                  tween the observed flexural strength of the pellets and
discs sintered at 1200 °C. Except for the sample 2.0–
                                                                                  the crack dimension observed from the SEM images of
100 which was severely decomposed, it was found that
                                                                                  the surface-polished pellets. Random cracks were se-
temperature had more impact on the strength of discs
                                                                                  lected on the surface polished pellets and the crack size
than the concentration. At a constant concentration,
                                                                                  was calculated by measuring the length (l) and width
                                                                                  (w) of the crack. The flexural strength was correlated

                                                                                                                      HA discs, sintered
                                                                                                                      p = 0.015
                                                                                    Log (flex strength, Pa)




                                                                                                                    0.38      0.40       0.42       0.44   0.46   0.48
                                                                                                                                     Log (ρ, g/cm3)

                                                                                  Fig. 8. Correlation between density and flexural strength of sintered
              Fig. 7. Flexural strength of samples after calcination.             pellets.
                                                                                C. Kothapalli et al. / Acta Materialia 52 (2004) 5655–5663                                                   5661

                                                           Fig. 9. SEM micrographs of the top surface morphology of samples sintered at 1200 °C for 1 h.

                                  100                                                             80                                 4. Discussion
                                                                                                       Specific surface area, m /g
                                                                                                                                        Apparently, the preparation conditions had signifi-

    Agglomerate size, µm

                                                                                                                                     cant impact on the resulting HA precipitates. The exper-
                                   60                                                             50                                 imental design was a three-level full factorial, with no
                                                                                                  40                                 repetitions of the HA batch preparations. The observa-
                                   40                                                                                                tions were handled in terms of linear models that used
                                                                                                                                     only first-order (trend) terms significant at the 95% con-
                                                                                                  20                                 fidence level and a curvature term with temperature thus
                                                                                                  10                                 leaving five degrees of freedom for error. In this fashion,
                                    0                                                              0                                 we reported with good certainty on trends with concen-
                                    20         25      30    35      40    45     50      55     60                                  tration and temperature, and the curvature with temper-
                                                        Particle surface area, m2/g                                                  ature without repetitions. The HA particles were
Fig. 10. Variation of agglomerate size and specific surface area with
                                                                                                                                     precipitated by a wet process and aged for 7 days. It
particle surface area of HA precipitates.                                                                                            was found that aging was a crucial process during HA
                                                                                                                                     synthesis which allowed the precipitates ripening [33].
                                                                                                                                     With the increase of the reaction temperature, the size
to the average crack size for each sample according to                                                                               of the precipitates increased. At a high reaction temper-
the relation r = r0/(1 + 2(l/w)), where r0 is a constant.                                                                            ature, the rate of particle ripening was high and the pre-
Fig. 11 shows the positive correlation (p = 0.04) between                                                                            cipitates grew large during the allotted time. As a result,
the biaxial flexural strength and the crack size. It was                                                                              large precipitates were observed for the HA particles
observed that the biaxial flexural strength rose exponen-                                                                             precipitated at high temperatures, i.e., 70 and 100 °C
tially with the decrease in the crack size.                                                                                          (as shown in Table 2). It was observed that when the
                                                                                                                                     reactant concentration was high, the precipitates
                                  70                                                                                                 selectively grew quicker along the c-axis compared to
                                            p = 0.04
                                                                                                                                     the a-axis resulting in high-aspect-ratio HA crystals
 Biaxial flexural strength, MPa

                                  60                                                                                                 (p = 0.033). Reactant concentration also had an effect
                                                                                                                                     on the resulting morphology of the precipitates. Increas-
                                                                                                                                     ing the reactant concentration while maintaining the
                                  40                                                                                                 reacting temperature constant, resulted in HA precipi-
                                                                                                                                     tates with increased agglomerate size and aspect ratio.
                                                                                                                                        Fig. 3 shows the average agglomerate size (by weight)
                                  20                                                                                                 obtained at various concentrations and temperatures
                                                                                                                                     studied. It was observed that larger agglomerates were
                                                                                                                                     formed for relatively smaller particles, such as those syn-
                                   0                                                                                                 thesized at 25 °C (refer to Fig. 10). Rodriguez et al. [20]
                                        0       0.02     0.04    0.06    0.08      0.1    0.12    0.14                               showed that the precipitation of HA was a nucleation–
                                                                   1/[1+2(l/w)]                                                      aggregation–agglomeration growth process where the
Fig. 11. Correlation between biaxial flexural strength of the pellets and                                                             aggregation process was ruled by surface energy minimi-
the crack size observed on the SEM images of the pellets.                                                                            zation. Due to their large surface areas, small precipitates
5662                                 C. Kothapalli et al. / Acta Materialia 52 (2004) 5655–5663

were more likely to form large agglomerates so as to min-            sition was accompanied by a dramatic microstructure
imize their surface energy [17]. This was in agreement               change of the specimens. In contrast, a smooth, dense
with Gibson et al. [30] who also reported that smaller               surface morphology almost free of pores was observed
particles appeared to form into large agglomerates with              for the specimens 2.0–70 (Fig. 9), which were prepared
rough surfaces, which is indicative of high surface areas.           under reaction conditions that produced the highest ob-
    The HA agglomerates were pressed into discs, and                 served density and strength among the entire specimens
sintered at 1200 °C for 1 h. The variation of density with           tested. While it would be more intuitive to think that the
reaction conditions for all the HA discs before and after            micro structural defects inside the sintered pellets would
sintering is shown in Fig. 6. The HA discs attained a                have a profound effect on the dynamics of the fracture
maximum sintered density of 92% for the samples syn-                 propagation and ultimately dictate the mechanical
thesized at 70 °C; this density was slightly lower than              strength of the pellet, the images of the polished samples
the values reported by both Ruys et al. [21] and Jarcho              gave us a better view of the porosity (or density) of the
et al. [22] who both used smaller HA particles. The phys-            samples than those of the fractural sections. It would of
ical basis of the Herring law of sintering [31], suggested           interest to note that the fractured surface of the sintered
that the sintering rate at a given temperature was inver-            pellets did not reveal micro-structural features more evi-
sely proportional to the square of the powder particle               dently than the polished surface of the pellet, the images
size. Thus, the smaller the agglomerate size, the easier             of which are not included in this paper.
for the powder to achieve high density at the same sin-
tering temperature. As shown in Figs. 3 and 6, with
the increase of the reaction temperature, the powder                 5. Conclusions
agglomerate size decreased, and their sintering density
also improved in spite of the decrease of the surface area              Different HA powders were synthesized using a wet
with agglomerate size. However, it was reported that the             precipitation method wherein the concentration of the
high sinterability of the synthetic HA was due to the                reactants and the temperature of the chemical reaction
high surface area and chemical purity of the HA                      were varied. The results indicate that these two factors
[18,30,32]. In both reports, single-phase HA powders                 were critical in controlling the size and shape of the
were studied and they did not decompose to secondary                 resulting HA particles. The length and breadth of the
phases on sintering. In contrast, most of the HA used                HA particles were found to increase with the reaction
in this study were decomposed when sintered at                       temperature. HA agglomerates were formed during the
1200 °C, which was one of the reasons resulting in rela-             synthesis process. Relatively large agglomerates were
tively low sintering density in this study.                          found for small HA precipitates (such as those precipi-
    A low sintering density was observed for the speci-              tated at 25 °C), due to their large surface areas. These
mens 2.0–100, which had severely decomposed during                   HA agglomerates were pressed into discs and sintered
the sintering process and caused their structure failure.            at 1200 °C for 1 h. It was discovered that there was a
It was reported that HA decomposed within the temper-                positive correlation between the sintered density and
ature range 1200–1450 °C, and the decomposition tem-                 biaxial flexural strength. With the increase of the sinter-
perature was strongly dependent on the characteristics               ing density, the biaxial flexural strength increased. How-
of the HA powders [23]. During the decomposition, a                  ever, due to severe decomposition, a significant drop in
large amount of water was formed, as indicated in                    strength as well as the density was found in specimens
Eq. (1). It was hypothesized that as the water forced                2.0–100. This was also accompanied by a porous micro-
its way out, it created many pores in the sintered discs,            structure, which was witnessed by the SEM observation.
as observed in the SEM images of the polished speci-
mens (Fig. 9). As a result, the sintering density of the
specimens 2.0–100 dramatically decreased.
    It was found that there was also a positive correlation
(p = 0.015) between the sintered density and biaxial flex-
                                                                        The authors acknowledge the support of University
ural strength of the specimens tested, as shown in Fig. 8.
                                                                     of Connecticut Research Foundation for sponsoring
With the increase of the sintering density, the biaxial
                                                                     the project.
flexural strength increased. The microstructure change
also reflected the density and strength of the specimens.
A surface full of micro-pores was observed for the pol-
ished 2.0–100 specimens as shown in Fig. 9. These spec-              References
imens also had the lowest observed sintering density and
                                                                      [1] Asaoka N, Suda H, Yoshimura M. Preparation of hydroxyapatite
biaxial flexural strength in the current study. This was in                whiskers by hydrothermal method. Chem Soc Jpn 1995;1:25–9.
agreement with the observation by Ruys et al. [21] that               [2] Ota Y, Iwashita T, Kasuga T. Novel preparation method of
the catastrophic strength drop caused by the decompo-                     hydroxyapatite fibers. J Am Ceram Soc 1998;81:1665–8.
                                           C. Kothapalli et al. / Acta Materialia 52 (2004) 5655–5663                                       5663

 [3] Kamiya K, Yoko T, Tanaka K. Growth of fibrous hydroxyapatite           [19] Liu HS, Chin TS, Lai LS, Chiu SY, Chung KH, Chang CS, et al.
     in the gel system. J Mater Sci 1989;24:827–32.                             Ceram Int 1997;23:19–25.
 [4] Fujishiro Y, Yabuki H, Kawamura K. Preparation of needle-like         [20] Rodriguez-Clemente R, Lopez-Macipe A, Gomez-Morales J,
     hydroxyapatite by homogeneous precipitation under hydrother-               Torrent-Burgues J, Castano VM. Hydroxyapatite precipitation:
     mal conditions. J Chem Tech Biotechnol 1993;57:349–53.                     a case of nucleation–aggregation–agglomeration-growth mecha-
 [5] Suzuki S, Ohgaki M, Ichiyanagi M. Preparation of needle-like               nism. J Eur Ceram Soc 1998;18:1351–6.
     hydroxyapatite. J Mater Sci Lett 1998;17:381–3.                       [21] Ruys AJ, Wei M, Sorrell CC, Dickson MR, Brandwood A,
 [6] Jillavenkatesa A, Condrate RA. Sol–gel processing of hydroxy-              Milthorpe BK. Sintering effects on the strength of hydroxyapatite.
     apatite. J Mater Sci 1998;33(16):4111–9.                                   Biomaterials 1995;16(5):409–15.
 [7] Liu J, Ye X, Wang H, Zhu M, Wang B, Yan H. The influence of            [22] Jarcho M, Bolen CH, Thomas MB, Bobick J, Kay JF, Doremus
     pH and temperature on the morphology of hydroxyapatite                     RH. Hydroxyapatite synthesis and characterization in dense
     synthesized     by    hydrothermal     method.      Ceram     Int          polycrystalline form. J Mater Sci 1976;11(11):2027–35.
     2003;29(6):629–33.                                                    [23] Ruys AJ, Zeigler KA, Standard OC, Brandwood A, Milthorpe
 [8] Ioku K, Yamauchi S, Fujimori S, Goto S, Yoshimura M.                       BK, Sorrell CC. Hydroxyapatite sintering phenomena: densifi-
     Hydrothermal preparation of fibrous apatite and apatite sheet.              cation and dehydration behavior. In: Bannister MJ, editor.
     Solid State Ionics 2002;151(1–4):147–50.                                   Ceramics: adding the value. Proceedings for the international
 [9] Liou SZC, Chen SY, Liu DM. Synthesis and characterization of               ceramic conference, AUSTCERAM-92, 2. Melbourne: CSIRO;
     needlelike apatitic nanocomposite with controlled aspect ratios.           1992. p. 605–10.
     Biomaterials 2003;24(22):3981–8.                                      [24] ASTM C 1499-03, Test method for monotonic equi-biaxial
[10] Webster TJ, Ergun C, Doremus RH, Siegel RW, Bizios R.                      flexural strength of advanced ceramics at ambient temperature.
     Enhanced osteoclast-like cell functions on nano-phase ceramics.            Annual book of ASTM standards: electronics, vol. 10.04,
     Biomaterials 2001;22(11):1327–33.                                          1991.
[11] Kandori K, Horigami N, Yasukawa A. Texture and formation              [25] Brook RJ. Pore-grain boundary interactions and grain growth. J
     mechanism of fibrous calcium hydroxyapatite particles prepared              Am Ceram Soc 1969;52:339–40.
     by decomposition of calcium–EDTA chelates. J Am Ceram Soc             [26] Saeri MR, Afshar A, Ghorbani M, Ehsani N, Sorrell CC. The wet
     1997;80:1157–64.                                                           precipitation process of hydroxyapatite. Mater Lett
[12] Yasukara A, Takase H, Kandori K. Preparation of calcium                    2003;57:4064–9.
     hydroxyapatite using amides. Polyhedron 1994;13:3071–8.               [27] Honda T, Takagi M, Uchida N, Saito K, Uematsu K. J Mater Sci
[13] Aoki H. Science and medical applications of hydroxyapa-                    1990;11:114.
     tite. Tokyo (Japan): Japanese Association of Apatite Science;         [28] Berndt CC, Haddad GN, Farmer AJD, Gross KA. Thermal
     1991.                                                                      spraying for bioceramic applications. Mater Forum 1990;14:161.
[14] Hench LL. Bioceramics: from concept to clinic. J Am Ceram Soc         [29] Rodriguez-Lorenzo LM, Vallet-Regi M, Ferreira JMF. Colloidal
     1991;74(7):1487–510.                                                       processing of hydroxylapatite. Biomaterials 2001;22:1847–52.
[15] Suchanek W, Yoshimura M. Processing and properties of                 [30] Gibson IR, Ke S, Best SM, Bonfield W. Effect of powder
     hydroxyapatite-based materials for use as hard tissue replacement          characteristics on the sinterability of hydroxyapatite powders. J
     implants. J Mater Res 1998;13:94–117.                                      Mater Sci 2001;12:163–71.
[16] LeGeros RZ. Calcium phosphates in oral biology and medi-              [31] Kingery WD, Bowen HK, Uhlmann DR. Introduction to
     cine Basel: Switzerland Karger; 1991.                                      ceramics. New York: Wiley; 1976.
[17] Bouyer E, Gitzhofer F, Boulos MI. Morphological study of              [32] Patel N, Gibson IR, Ke S, Best SM, Bonfield W. Calcining
     hydroxyapatite nanocrystal suspension. J Mater Sci                         influence on the powder properties of hydroxyapatite. J Mater Sci
     2000;11:523–31.                                                            2001;12:181–8.
[18] Landi E, Tampieri A, Celotti G, Sprio S. Densification behavior        [33] Wei M, Ruys AJ, Milthorpe BK, Sorrell CC. Solution ripening of
     and mechanisms of synthetic hydroxyapatites. J Eur Ceram Soc               hydroxyapatite nanoparticles: effects on electrophoretic deposi-
     2000;20:2377–87.                                                           tion. J Biomed Mater Res 1999;45(1):11–9.

To top