PHYSICOCHEMICAL CHARACTERIZATION OF SIX COMMERCIAL HYDROXYAPATITES FOR MEDICAL-DENTAL APPLICATIONS AS BONE GRAFT

1DDS, MS, PhD student of the Metallurgical and Materials Engineering Department of the Federal University of Rio de Janeiro Rio de Janeiro, RJ, Brazil. 2DDS, MS, PhD Associate Professor of the Biological Science Department, Bauru Dental School, University of Sao Paulo, SP, Brazil. 3Eng, MS, PhD, Associate Professor of the Metallurgical and Materials Engineering Department of the Federal University of Rio de Janeiro, Rio de Janeiro, RJ, Brazil.


INTRODUCTION
The use of biomaterials for hard tissue replacement has grown as a consequence of the increase in reconstructive surgeries that require, in most cases, adjuvant grafting 8 .Several applications of bone graft materials in medicine and dentistry, such as bone repair, augmentation and substitution can be highlighted 11 .According to data from the Brazilian Ministry of Health, the number of surgical procedures using hydroxyapatites as bone graft has more than doubled between 2002 and 2003.Bone graft may be obtained from different origins: autogenous (from the same individual, highly osteogenic), allogeneic graft (banked freeze-dried bones of human cadavers), xenogenous (usually bovine bone-derived) or synthetic calcium phosphate materials 11,13 .Autografts are considered the "gold standard" for bone repair and substitutions 12 , but they could only be taken in limited amounts.On the other hand, allografts and xenografts are sometimes avoided due to the potential transmission of infectious diseases 15 .
Calcium phosphate (Ca-P) materials can be found in nature (coralline hydroxyapatite) or synthesized by precipitation methods using chemical reagents 1 .Hydroxyapatite, Ca 10 (PO 4 ) 6 (OH) 2 -HA, is the most well known and studied calcium phosphate.In medical and dental fields, the term "hydroxyapatite" is sometimes used to describe any calcium phosphate material.Synthetic Ca-P grafts may be the chosen material, especially when large defects need to be filled.It is generally accepted that these bioceramics are only osteoconductive (have the ability to support tissue ingrowth and bone formation) and non-osteoinductive, which means ability to form bone when implanted in non-osseous sites 3 .In several cases, synthetic materials can be mixed with autogenous bone in order to increase the osteogenic behavior.
The bone graft material may be either resorbable or nonresorbable, with this classification being related to the extent of dissolution of Ca-P materials.The factors affecting the dissolution properties were similar to those affecting biodegradation or bioresorption 9,11 .According to Ducheyne and Qiu 8 , the larger the solubility rate of the ceramic, the more pronounced the enhancement effect of bone tissue growth.Among several factors, chemical composition, particles size and cristallinity are likely to affect the ceramics solubility, which may be adjusted for the desired purpose.Different applications require materials with different resorption rates, which can be regulated by the mixture of several calcium phosphate phases.
The extent of calcium phosphate ceramics (CPC) dissolution in acidic buffer increases from crystalline hydroxyapatite to amorphous apatite, with tricalcium phosphate exhibiting medium solubility 12 .Magnesium or carbonate substitutions were shown to cause reduction in crystal size and an increase of dissolution of synthetic apatites 12 .For the same composition, the resorption increases with the decrease in crystallinity and with the increase in surface area 5 .Crystallinity is highly dependent on sintering temperature: the higher the sintering temperature, the more perfect the crystal and thus the lower the degradation rate.Resorbable calcium phosphate materials are usually unsintered Ca-P materials.
Changes in composition and cristallinity of calcium phosphate-based materials from one manufacturer to another or even from different batches of the same manufacturer were observed.These variations can be a consequence of lack of control in the manufacturing process and directly affect the clinical result, becoming a limiting factor for the use of this material 11 .
The purpose of this study was to determine the physicochemical properties of six commercial hydroxyapatite granules available in the Brazilian market, comparing the results obtained with information from manufacturers' specifications.

MATERIALS AND METHODS
A total of six hydroxyapatite samples in granular form were used in this study: two were imported (HA-4 and HA-6), although they are easily available in the Brazilian market, and four of them were made in Brazil.Concerning the degradation rate, four samples were indicated by the producer as resorbable materials (from HA-3 to HA-6) and the other two -HA-1 and HA-2 -as non-resorbable.According to the products' specifications, the hydroxyapatite granules were either of bovine origin (named "natural") -HA-5 and HA-6) or by using synthetic routes (HA-1, HA-2, HA-3 and HA-4).Information following packing was quite variable.The manufacturers' names were not disclosed to protect their identities.
The granules were bought from or, alternatively, donated by manufacturers and characterized without further treatment.The following characteristics were determined: granules morphology, size range and surface area.The presence of other phases besides hydroxyapatite, cristallinity and chemical groups were also investigated.Size range was determined by vibratory sieving using 75, 125, 250, 350, 420 and 600mm aperture screen size (ABNT).Scanning electron microscopy (SEM -ZEISS, DSM 940A model) was used to investigate granules morphology and granules porosity.The specific surface area of each hydroxyapatite was determined by BET analysis which estimates of surface area by nitrogen adsorption at 77K 7 .
X-ray diffraction (XRD) allows the identification of other crystalline phases besides hydroxyapatite and the unit cell parameters.Cristallinity was determined by the methodology used by Landi, et al. 10 , which results in the following formula: Xc = 1-(V 112/300 / I 300 ) x 100, where Xc = the sample cristallinity in %, I 300 is the intensity of (300) reflection and V 112/300 is the intensity of the hollow between (112) and (300) reflections.XRD analysis was carried out in a Rigaku MINIFLEX XRD operating at 30 kV, 15mA and CuKa radiation.Data was obtained in the range of 5-100 0 2q.
Fourier-transform infrared spectroscopy (FTIR, ABB Bomem Inc., MB series, Quebec, Canada) equipped with reflectance attachment was also used in order to confirm XRD results and to detect some groups, such as CO 3 2-.The spectra were collected at room temperature at a nominal resolution of 4.00 and number of sample scans equal to 1000.The FTIR spectra were recorded in the 400 -4000 cm -1 range using specular reflection.
One gram of each granule was randomly separated and crushed into a fine powder for XRD and FTIR analysis.All procedures were conducted in duplicates and run in separated experiments.

RESULTS
Table 1 shows the particle size obtained and the range size indicated in the manufacturer's specifications.Despite some error associated with the methodology used, the only sample that clearly is in accordance with these specifications was sample HA-6. Figure 1 shows SEM image (50x magnification) and the BET specific surface area for each condition.It is possible to observe that sample HA-3 exhibits the smallest particle size, which corresponds to a high surface area (49.6 m 2 /g) while sample HA-6 exhibits large particles (250 -> 600 mm) but also large pores.Thus, the highest specific surface area of HA-6 (84.5 m 2 /g) is probably a consequence of the porosity observed on these granules, which is compatible with the bovine bone origin.Samples HA-3 and HA-4 exhibited intermediate values of surface area, and, for samples HA-1, HA-2 and HA-5, the BET specific surface areas were relatively low.
Figure 2 shows the XRD pattern (displayed range in 20-60°) of all samples.Only hydroxyapatite was identified (9-432 JCPDS card) with the three most intense peaks corresponding to (211), (300) and (002) planes, respectively.The crystallinity percentage was also indicated in Figure 2 and shows that HA-1, HA-2 and HA-5 have highest, HA-3 intermediate and HA-4 and HA-6 the lowest crystallinity.
Figure 3 shows the FTIR spectra of the samples examined.The bands at 640 and 3575 cm -1 are characteristic OH - absorption bands of HA.The band at 1650 cm -1 was assigned to adsorbed H 2 O.The absorption bands at 1092, 1044, 1036, 960, 602, 573, and 475 cm -1 were assigned to the vibration in the PO 4 3-group of the hydroxyapatite (HA).The C-O vibration in the CO 3 2-group was located in the 1414-1545 cm -1 range and at 875 cm -1 , corresponding to A-type and B-type of carbonated apatite, respectively, and were largely variable from one sample to another.The triplet around 2000 cm -1 was not identified.All granules examined show a composition compatible with that of hydroxyapatite containing variable CO 3 2-substitution.The samples were assumed to have high, medium or low CO 3 2-substitution, based on the comparison of the relative intensity of CO 3 2-bands to the phosphate band.For example, HA-1, HA-5 and HA-6 were classified as high CO 3 2-content, which may improve "in vivo" degradation.HA-3 and HA-4 were considered as containing intermediate CO 3 2-content, while carbonate substitution was almost absent on sample HA-2.
All granules that have exhibited non-expected results were re-characterized in order to confirm previous results.In the case of samples HA-1 and HA-5, new granules were bought and all methodology was again applied.No discrepancies were observed between samples from different batches.

TABLE 1 -
Comparison between particle size determined by sieving and the manufacturers' specifications.HAsamples in granular form FIGURE 1-Scanning electron micrography (50x magnification) and the BET specific surface area for each hydroxyapatite examined