You are on page 1of 6

Indian Journal of Chemistry

Vol. 47A, November 2008, pp. 1626-1631

Synthesis of nano sized hydroxyapatite powder using sol-gel technique and its
conversion to dense and porous bodies
Iis Sopyana,*, Ramesh Singhb & Mohammed Hamdic
a

Department of Manufacturing & Materials Engineering, Faculty of Engineering, International Islamic University Malaysia,
P.O. Box 10, 50728 Kuala Lumpur, Malaysia
Email: sopyan@iiu.edu.my
b
Department of Mechanical Engineering, College of Engineering, University Tenaga Nasional, Malaysia
c
Department of Engineering Design and Manufacture, Faculty of Engineering, University Malaya, Malaysia
Received 13 May 2008; revised 20 October 2008

Hydroxyapatite powder has been prepared via sol-gel procedure using calcium nitrate tetrahydrate and diammonium
hydrogen phosphate as the precursors for calcium and phosphorus, respectively. XRD measurement shows that the powder
contains hydroxyapatite crystals with -TCP and calcium oxide as secondary phases. Hydroxyapatite powder of higher
purity, i. e., the correct Ca/P ratio, has been obtained by adding an appropriate amount of diammonium hydrogen phosphate
and heating with stirring. Morphological evaluation by SEM measurement shows that the particles of the HA are tightly
agglomerated and globular in shape with an average size of 1-2 m. The primary particulates have average diameters of
50-200 nm, as detected by SEM and nanoparticle sizer. Purity (almost 100%) of the obtained hydroxyapatite has been
confirmed by XRD analysis. Its performance has been tested by making dense and porous samples.
IPC Code: Int Cl.8 B82B3/00; C01B25/45; C01F11/00

Hydroxyapatites (HA) are particularly attractive


materials for bone and tooth implants since they
closely resembles human tooth and bone mineral and
have proved to be biologically compatible with these
tissues.1 Since the use of hydroxyapatite for the first
time in 1981 for periodontal lesion filling, its use in
the medical field has extended to solid blocks, solid
components, and films for dental implants. Many
studies have shown that HA ceramics show no
toxicity,
inflammatory
response,
pyrogenetic
response, or fibrous tissue formation between implant
and bone. Also, these materials have the ability to
bond directly to the host bone.2
The main limitation of HA ceramics as well as all
other bioactive materials is that they have poor
mechanical properties. Basically, all bioceramics
which have good mechanical properties and suitable
for load bearing applications should be bioinert.
Hydroxyapatite, on the other hand, has high
bioactivity, with many medical applications in the
form of porous, dense, granules, and, as coatings.2,3
Several research groups have developed preparative procedures for hydroxyapatite. Traditionally, two
main methods are employed for preparation of HA
powders: wet (chemical) method (including
precipitation method,4 hydrothermal technique,5 and

hydrolysis6) and dry (solid state reaction) method


(refs cited in Ref. 7). Differences in the preparative
routes leads to variations in morphology,
stoichiometry, and level of crystallinity. Other
methods, such as sol-gel,8,9 spray pyrolysis,10,11
mechano-chemical method,12 have also been
developed recently and are well documented.7
Sol-gel procedure was first employed for the
preparation of HA by Sakka and co-workers.8 They
used calcium diethoxide and phosphorus triethoxide
as starting materials. Hydrolysis polycondensation
of the monomers in neutral and acidic solutions gave
HA powders of high purity. Extraordinarily fine
amorphous particulates with less than 10 nm in
diameter were obtained from precipitation of the
solutions, which increased to only ca. 100 nm after
calcination at 900C.
In this present study, we have developed sol-gel
procedure for preparing HA powder. It is well known
that sol-gel techniques have several advantages for
producing ceramic particulates of high purity, high
crystallinity, nano sizes, and high reactivity. Sol-gel
process, however, has some drawbacks such as
expensive raw materials and low homogenity of the
final product. We report herein a novel sol-gel method
for preparing extraordinarily fine hydroxyapatite

SOPYAN et al.: SYNTHESIS OF NANO SIZED HYDROXYAPATITE POWDER USING SOL-GEL TECHNIQUE

powder, utilizing easily obtainable raw materials of


relatively low cost. Simplicity of experimental
execution, in respect of methods employing wet
chemical reaction, is one of the most important
advantages offered by this method. Physico-chemical
characterization of the hydroxyapatite powder
obtained from the sol-gel procedure has been carried
out. Morphology as well as the mechanical properties
of dense and porous bodies prepared using the powder
have also been investigated.
Materials and Methods
For preparation of hydroxyapatite powder, calcium
nitrate tetrahydrate and diammonium hydrogen
phosphate (reagent grade) were used as calcium and
phosphorus precursors, respectively. Both reagents
were purchased from Merck KGaA, Germany. Urea
(R&M Chemicals, UK) was used as gelling and
ammonium donor agent. EDTA (Merck KGaA) was
used as chelating agent to prevent an immediate
precipitate formation calcium ions in the course of gel
formation. The reaction was conducted in basic
solution using ammonium solution (R&M Chemicals,
UK) as solvent.
Preparation of the stoichiometric hydroxyapatite powder

Ammoniun solution was heated at 60C, and


EDTA (181 g) was added while stirring until it
dissolved. Into this, 200 mL aqueous solution of 129 g
calcium
nitrate
tetrahydrate
was
poured.
Diammonium hydrogen phosphate (39.83 g) and urea
(45.20 g) were subsequently added. The mixture was
heated at 100C for 3-4 h. The obtained gel was dried
at 350C under ambient static air and subsequently
subjected to an 820C calcination under flowing air.
The powder was examined by X-ray diffraction
techniques to determine the phases formed. It was
observed that Ca/P molar ratio is ca. 1.8.
Accordingly, to compensate the upward deviation
from the stoichiometric ratio (1.667), the powder was
mixed with an appropriate amount of diammonium
hydrogen phosphate, followed by suspending in water
and heating at 90C with rigorous stirring. This
procedure restored the Ca/P ratio of hydroxyapatite
powder to 1.67. After drying, pure hydroxyapatite
powder was obtained.
Preparation of dense and porous samples

For preparation of dense bodies, the as-prepared


hydroxyapatite powder (50 g), was mixed with
poly(vinyl alcohol) of 15.000 MW(2.5 g) and 100 ml

1627

distilled water. The suspension was homogenized


using a magnetic stirrer followed by spray-drying on a
Buchi mini spray dryer (B-290 type). Well dispersed
powders obtained were compressed uniaxially using
cold pressing technique at 800 kg/cm2. The pellets
thus obtained were sintered in air at 1250C for 1h.
The same procedure was applied for a commercial
hydroxyapatite powder (Sigma Aldrich).
For preparation of porous bodies, slurries were
prepared with the as-prepared HA powder, and
adjusted loading of HA (ca. 25 wt%), using Duramax
of D-3021 or D-3005 type (Rohm and Haas, USA) as
the dispersant. Commercial cellulosic sponges were
used and an initial impregnation procedure with a
considerable fluidity slurry was employed. After
soaking into the slurry, the sponges were dried in
ambient air for 72 h and then subjected to heat
treatment at 600C for 1 h to eliminate organic
matrix. Sintering was carried out at 1250C for 1h.
Characterization

Scanning electron microscopic measurement for


morphology evaluation of the powder was performed
on a Jeol FESEM instrument (model JSM 6700F).
The particle size distribution and mean particle size
were measured using a back scattering method on a
Malvern Nanosizer (model Nano-S). Differential and
thermogravimetric analyses were performed on the
as-prepared of HA powders and dried gel in ambient
air using Perkin Elmer instrument (model PYRIS
Diamond) with a 10C/min heating rate. The
crystalline phase compositions of the powders and of
the dense samples were evaluated in a Rigaku
diffractometer with copper K radiation and a scan
rate of 2 in 2 min-1. XRD patterns obtained were
utilized for quantitative phase analysis according
to literature.13
Density of porous bodies was measured as apparent
density (geometrical weight/volume measurement).
The compressive strength was measured on
cylindrical specimens (10 mm ht 10 mm dia.) using
an Instron 1195 apparatus. The compressive strength
was calculated from the maximum load registered
during the test divided by the original area. Several
specimens were used for the testing.
Density of cylindrical dense bodies was measured
as apparent density (geometrical weight/volume
measurement).
Bars
of
the
polycrystalline
hydroxyapatite of dimension 25 mm 2.0 mm
2.5 mm were cut from cylindrical plates with a
diamond saw. No further chemical treatment was

1628

INDIAN J CHEM, SEC A, NOVEMBER 2008

performed on bars before testing. The flexural tests


were conducted in four-point bending, using a Lloyd
LR10K plus mechanical tester.
Results and Discussion
After refluxing the reaction mixture at 100C for
ca. 4 h, concentrated sol was formed. Subsequently it
was converted to white gel through in situ solvent
evaporation.
Reaction of hydroxyapatite formation can be
expressed as follows:
5Ca(NO3)2 + 3(NH4)2HPO4 + 4NH4OH
Ca5(PO4)3OH + 10NH4NO3 + 3H2O

(1)

The black dried gel obtained was then subjected to


TG-DTA for thermal characterization. DTA-TG
curves of the gel dried at 350C shows the first weight
drop of 10% at 100C due to water evaporation. A
subsequent decrease in weight of ca. 50% occurs until
700C which is attributed to decomposition and
elimination of ammonia, nitrate, urea, organic
compounds, and carbon dioxide. There are two
exothermic peaks in the curve; the first ranging
from 350 420C, is attributed to decomposition of
ammonia and organic compounds. The second one,
from 420 500C is a large exothermic peak which
may be due to decomposition of urea and carbon
dioxide. Subsequently, calcination of dried gel at
820C for 2h under flowing air converted it into
hydroxyapatite powder. The yields of the HA
powders were 90-95%.
XRD pattern of crystalline phase of the powder
after the heat treatment at 820C (Fig. 1) shows that
hydroxyapatite is the main component in the powder
(ca. 75%). Calcium oxide (5%) and -tricalcium
phosphate (20%) were present as secondary phases.
Generally, the powder mixture obtained at this stage
contained 75-85%, 15-20%, and 4-6% of HA, -TCP,
and CaO, respectively. At this composition, the
Ca/P molar ratio is about 1.8. Since optimum
stoichiometric Ca/P molar ratio is 1.667, to
compensate for the upward deviation of Ca/P,
diammonium hydrogen phosphate was added to the
suspension of the mixture HA, and heated at 90C
with rigorous stirring until the solvent was completely
removed.
In the XRD pattern of the pure powder obtained
after this treatment the peaks of -TCP and CaO
disappear, proving 100% purity of the hydroxyapatite

powder. We also checked the possiblility of the


presence of calcium hydroxide in the powder. A
phenolphtalein test shows that no the hydroxide is
present in the powder.
SEM photograph of the as-prepared HA powder
(Fig. 2) shows that individual hydroxyapatite particles
formed in globular shape with an average size of
~ 50-200 nm in diameter. The nanometric primary
particles agglomerated tightly into micrometric
aggregates of various shape and size. On the other
hand, the particle size distribution of the
hydroxyapatite powder as measured by nanoparticle
sizer shows two separate size distributions; the lower
distribution ranging from ca. 50 500 nm may be
attributed to individual particles and the higher
distribution from 2000 7000 nm may be attributed
to tightly bonded particle agglomerates. Quite likely it

Fig. 1 XRD pattern of HA powder mixed with impurities CaO


and -TCP.

Fig. 2 SEM photograph of the sol-gel derived HA powder.

SOPYAN et al.: SYNTHESIS OF NANO SIZED HYDROXYAPATITE POWDER USING SOL-GEL TECHNIQUE

is difficult to disperse all agglomerates even after


rigorous stirring for hours. The specific surface area
measured by BET method gave a low value of 7 m2/g.
This value is unusual for particles as fine as in the
present case at nano levels, hence it is considered that
the surface area measured by BET is for agglomerates
and not of the particles. The particle size of the HA
powder obtained in this study is considerably fine, as
confirmed by SEM measurement, in respect of the
HA powder prepared by sol-gel technique. Earlier
studies have reported that sol-gel derived HA powders
have particle size of about 100 nm in diameter8.
Figure 3 shows the IR spectra of the dried gel
(a) and hydroxyapatite powder (b). The dried gels
spectrum clearly shows broad peaks, a characteristic
of amorphous products. Bands of carboxylic and
carbonate (1400-1600 cm-1), phosphate (9001100 cm-1), ammino (1400, 1600 and 3200 cm-1) and
acetate (2800, 2300 cm-1) groups were detected.
Obviously, CO32- and HPO42- groups are present,
partially substituting groups of PO43- and/or OH- in the
HA structure. In Fig. 3(b), FTIR spectrum of the final

1629

powder shows the characteristic peaks corresponding


to OH- (630, 3560 cm-1) and PO43- (960, 1050,
1090 cm-1) vibrations, together with weak bands of
CO32- group at 870 and 1540 cm-1, which indicates
that the sol-gel HA powder is partially carbonated
hydroxyapatite, as commonly observed in synthesis
involving organic reagents.
Dense and porous samples

Dense and porous bodies were prepared to evaluate


the performance of the obtained sol-gel HA powder.
Dense samples were also prepared using commercial
HA powder. Table 1 lists the physical properties of
dense samples prepared using the sol-gel HA and
commercial HA powders after sintering at 1250C.
The sintered dense bodies for the sol-gel derived HA
powder showed flexural strength of 57.7 MPa and an
apparent density of 2.855 g/cm3 at 90% density (i. e.,
10% porosity). Although the flexural strength is not
so high, it is still in the range of the flexural strength
of human cortical bones. The value is in fact much
better than that obtained using commercial powder
(36.0 MPa), at 89% relative density.
Figure 4 shows XRD pattern of the cylindrical
dense bodies. It is clear that -TCP appeared at
2 = 30.7 as the secondary phase after the sintering,
Table 1 Physical properties of dense samples prepared using
the sol-gel HA and commercial HA powders.

Fig. 3 FTIR spectra of (a) dried gel HA and (b) sol-gel derived
HA powder.

Type of HA
powder

Sintering
temp. (C)

Flexural
strength
(MPa)

Apparent
density
(g/cm3)

Rel.
density
(%)

Sol gel
Commercial

1250
1250

57.7
36.0

2.855
2.810

90
89

Fig. 4 XRD pattern of the HA dense body after 1250C


sintering.

1630

INDIAN J CHEM, SEC A, NOVEMBER 2008

although with an acceptable quantity (less than 3%).


The formation of -TCP itself is unusual since
normally -TCP will form first at temperatures
<1250C, then -TCP and tetracalcium phosphate
appear at higher temperature.14 This phenomenon was
observed in HA powders prepared by precipitation
method. It is deduced that the phenomena was
attributed to the high reactivity of the obtained HA
powder. However, this is an advantage since sintering
process at lower temperatures suffice to obtain
comparable particle densification level as indeed
proven by the much higher mechanical strengths of
the present HA dense bodies.
Figure 5 shows SEM pictures of the fractured
surface of dense samples. Micropores of 1-2 m in
diameter are present on the surface, which are
responsible for high porosity (10%). In the dense
samples. Besides micropores, a macropore of ca.
20 m in size was found, representing a less-

compacted region which was probably the origin of


the fracture.
Porous bodies were also prepared by the polymeric
sponge method as reported in literature.15,16 Figure 6
illustrates the resulting macrostructure of the porous
hydroxyapatite. It is clear that the sample has circular
open-cell pores of 500 microns - 2 mm in diameter.
Presumably, the pore structures are similar to that of
the original matrix.
A 44% sintering shrinkage was obtained after
1250C sintering, with an apparent density of
1.290 g/cm3 and a relative porosity of 59%. On the
other hand, measurement of mechanical properties of
porous bodies provided a compressive strength of
1.96 MPa. It is important to note that conditions for
slurry preparation have not been optimised asyet. In
fact, large cavities were found in several samples as
the result of incomplete slurry penetration into the
pores. The slurry itself had to be ultrasonically treated
to prevent coagulation. So, the compressive strength
obtained, 1.96 MPa, was considered low for such
value of apparent density (1.290 g/cm3). It is also
reasonable to presume that such low compressive
strength is partly attributed to high macroporosity of
the sample.
Figure 7 shows the microporos structure of the wall
of the open pores in which micropores of 1-2 m in
diameter were observed. Hing et al. reported that they
succceded in preparing porous HA with compressive
stress of 1 11 MPa for the increased apparent
density from 0.38 1.25 g/cm3.17 Even though it is
compared in respect of porosity, the value of
1.96 MPa is still low if we refer to the value reported

Fig. 5 SEM picture of the fractured surface of dense samples.

Fig. 6 Macrostructure of the porous hydroxyapatite.

Fig. 7 Microporosity structure of the wall of the open pores.

SOPYAN et al.: SYNTHESIS OF NANO SIZED HYDROXYAPATITE POWDER USING SOL-GEL TECHNIQUE

by Guicciardi et al18. (16.1 MPa) for the speciment


of 56% total porosity. The preliminary test on powder
performance shows that the sol-gel powders have
excellent physical properties which may be used to
produce dense and porous bioactive bone implants
with desired properties.
Conclusions
Hydroxyapatite (HA) powders with particles
having nano- to submicro-metric size were
succesfully prepared via sol-gel procedure using
calcium nitrate tetrahydrate and diammonium
hydrogen phosphate as the precursors. The primary
particulates have globular shape with an average
diameter of 50-200 nm, as detected by SEM. The
results of the the nanoparticle sizer measurements are
in good agreement with those obtained by
morphological investigation. All the powders
obtained have good purity (nearly 100%).
Dense samples were prepared successfully from the
HA powder using cold pressing technique. Sintering
at 1250C led to a total porosity of 10%, with
apparent density of 2.855 g/cm3. The mechanical test
of the dense samples provided a higher flexural
strength (57.7 MPa) than that using commercial HA
powder (36.0 MPa), thus showing its suitability for
load bearing bone implant applications.
Porous bodies produced via sponge impregnation
technique using the sol-gel derived HA powder have
apparent density of 1.290 g/cm3, with 59% total
porosity and 44% sintering shrinkage. The porous
bodies contained open macropores of 0.5-2 mm
diameter as the result of removal of the polymeric
matrix as well as micropores of 1-2 m diameter. The
compressive strength measurement yielded a value of

1631

1.96 MPa, proving that porous HA samples can be


used for human spongy bone substitutes. Thus,
physical characterization of the sol-gel HA powder
followed by the preliminary test on powder
performance in making porous and dense samples
shows that the powder has excellent physical
properties and may be to produce dense and porous
bioactive bone implants with desired properties.
References
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18

Willmann G, Interceram, 42 (1993) 206.


Willmann G, British Ceram Trans, 95 (1996) 212.
Levernia C & Schoenung, J M, Ceram Bull, 70 (1991) 95.
Tas A C, Biomaterials, 21 (2000) 1429.
Fujishiro Y, Sato T & Okuwaki, A., J Mater Sci Mater Med,
6 (1995) 172.
Monma H & Kamiya T, J Mater Sci, 22 (1987) 4247.
Suchanek W & Yohsimura M, J Mater Res, 13 (1998) 94.
Masuda Y, Matsubara K & Sakka S, Nippon Seramikkusu
Kyokai Gakujitsu Ronbunshi, 98 (1990) 1255.
Liu D M, Troczynski T & Tseng W J, Biomaterials, 22
(2001) 1721.
Luo P & Nieh T G, Mater Sci Eng, C3 (1995) 75.
Aizawa M, Itatani K, Howell F S & Kishioka A, J Ceram
Soc Japan, 103 (1995) 1214.
Toriyama M, Ravglioli A, Krajewki A, Galassi C, Roncari E
& Piancastelli A, J Mater Sci, 30 (1995) 3216.
Tampieri A, Celotti G, Sprio S & Mingazzini C, Mater Chem
Phys, 64 (2000) 54.
Bertoluzza A, Cacciari S, Tinti A, Vasina M & Morelli M A,
J Mater Sci Mater Med, 6 (1995) 76.
Galassi C, Landi E, Tampieri A & Celotti G, Am Ceram Soc
Ceram Trans, 112 (2001) 215.
Saggio Woyansky J, Scott C E & Minnear W P, Am Ceram
Soc Bull, 71 (1992) 1674.
Hing K A, Best S M & Bonfield W, J Mater Sci Mater Med,
10 (1999) 135.
Guicciardi S, Galassi C, Landi E, Tampieri A, Satou K &
Pezzoti G, J Mater Res, 16 (2001) 163.