A THERMODYNAMIC APPROACH TO THE HYDROTHERMAL SYNHTESIS OF HYDROXYAPATITE- BASED NANOCOMPOSITE

MATERIALS

Roxana M. Piticescu1, Radu Robert Piticescu1, Gabrielle C. Chitanu2, Madalina L. Popescu1

1 Institute for Non-ferrous and Rare Metals, Bucharest,102 Biruintei Blvd., Pantelimon,

Ilfov, Romania 2 Petru Poni Institute of Macromolecular Chemistry of the Romanian Academy, Iasi,

Romania

Abstract The synthesis of HAP by hydrothermal procedure has been studied to establish the most suitable conditions for addition of tetragonal zirconia or polymer to HAP as reinforcing agent in order to obtain biological active nanomaterials. Thermodynamic prediction has been used to predict the synthesis pH and temperature for the complex system. Validation of the synthesis route has been done on the basis of the structural and compositional characterization of composites. Densities values were very close of theoretical density for HAP (2.76 g/cm3 for the composite and 3.02 g/cm3 for pure HAP). Hydroxyapatite crystallite sizes are in the nanometer range in all cases. Nanocrystalline HAP is expected to present high chemical and biological activity and high sorption capacity.

1. Introduction

Ceramic-based composite nanomaterials play an increasing role in the high-tech fields from structural (replacing traditional materials) to functional (from electronics to biomaterials) applications. Different physical, mechanical, chemical or mixed methods have been proposed to obtain the initial nanostructured powders with desirable characteristics for producing dense sintered materials (bulk, thick or thin films) with required properties. Hydroxyapatite-Ca10(PO4)6(OH)2-is one of the most well known biocompatible materials due to its similar composition with the human bounds (natural composite consisting of calcium phosphates like HAP and collagen) and high bioactivity. Its utilisation in orthopedic implants and dentistry is unfortunately limited by its low mechanical strength. As a consequence many efforts have been drawn toward synthesis of HAP-composite nanomaterials with enhanced bioactivity and mechanical properties [1-16]. Hydrothermal method is a wet chemical process presenting some advantages like: one-step process, low synthesis temperatures, obtaining of materials with nanocrystalline structure, controlled nucleation, homogeneous composition, versatility (oxides, sulphides, carbon nanotubes) and low pollution [17]. Main problems are related to hydroxyl bonding producing powder agglomeration. To solve this problems different approaches have been proposed, like surface modification using adequate polymers or using external driving forces like electric field including deposition of nanocrystalline thin films. A convenient approach is to predict first the thermodynamic equilibrium in the complex water-ions system and then to verify the conditions to obtain powders with the desired characteristics. The present paper gives some results on the thermodynamic predictions and their applications in the synthesis of HAP-composite powders from some definite systems.

2. Thermodynamic considerations regarding hydrothermal synthesis of ceramic oxides

Generally the hydrothermal processes may take place by two ways: hydrothermal reactions (reactions between species in hydrothermal solutions under the influence of temperature and pressure) and hydrothermal crystallization (transformation of amorphous species in crystalline ones under the influence of temperature and pressure). Figures 1a and 1b shows that increasing temperature and electronegativity the equilibrium goes toward formation of oxides, the Gibbs free energy being higher in the case of hydrothermal reaction than for the hydrothermal crystallization. Calculations have been done using a special soft [18].

Fig. 1.a. Gibbs free energies for the synthesis of some oxides by hydrothermal reactions.

Fig. 1.b. Gibbs free energy for the synthesis of some oxides by hydrothermal crystallisation.

1 1 1 1 1 1 1.11.11.11.11.21.21.51.61.61.61.71.71.8

-350

-300

-250

-200

-150

-100

-50

0

50

100

150

200

5000C

3000C

00C

Sn, Ni

Mn, Fe, Pb, Zn

AlZr

Mg

Ce

Y, La, PrCa, Sr, Ba

Li, Na, K

Gib

bs fr

ee E

nerg

y [k

J/m

ol]

Electronegativity

1 1 1 1 1 1 1.11.11.11.21.21.31.51.51.61.61.71.71.71.8

-600

-400

-200

0

200

400

600

5000C

3000C

00C

Ni

Zn, Si, Sn

Fe, Pb

Be, Al

Zr, Ti

Mg

Y, La, PrCa, Sr, Ba

Li, Na, K

Gib

ss fr

ee E

nerg

y [k

J/m

ol]

Electronegativity

As a consequence, in most of our works we synthesized different oxides starting from water soluble compounds using alkaline hydroxides as mineralizing agents. It comes out that role of solution pH should also be taken into account. A simple and reasonable prediction of species formed vs. solution pH at different temperatures consists in calculation the potential-pH (known as Pourbaix) diagrams.

3. Experimental procedures

The precursor solutions were prepared by dissolution of pure soluble salts (calcium nitrate Ca(NO3)2*4H2O, calcium phosphate Ca5(PO4)3OH, zirconium tetrachloride ZrCl4, yttrium nitrate Y(NO3)3 and sodium maleate-vinyl acetate copolymer) under vigorous stirring into distilled water in the appropriate amount corresponding to the programmed ratio. The pH of the solution was adjusted to the desired value by mixing it with ammonium hydroxide solution. The composite powders were then obtained by hydrothermal treatment of the suspension in a 2L Teflon autoclave (CORTEST, USA) for two hours in the temperature range 100-2500C. The precipitates were separated by filtering, washed with distilled water to remove the soluble chlorides or nitrates and with ethanol to reduce agglomeration and finally dried for several hours in air at 1100C. Phase composition of powders was investigated by XR diffraction. The phase composition has been calculated according to Bragg-Brentano method using CuK radiation. Specific surface area has been measured according to BET method. The picnometric densities were also measured using the inert gas adsorption method. Microstructural characterisation of powders was done using FT-IR analysis and the SEM method (Gemini-LEO 1530).

4. Results and discussions 4.1. Synthesis of Yttria-doped zirconia nanopowders

Yttria-doped zirconia materials with tetragonal symmetry are known to increase the mechanical properties of many ceramic matrix composites including HAP due to its intrinsic high toughness. A comparison of mechanical properties of alumina and YTZP mechanical characteristics are presented in table 1. Synthesis of YTZP powders was consequently first addressed.

Table 1: Some characteristics of 96% Al2O3 and Y-TZP (tetragonal ZrO2)

Al2O3 (96%) Y-TZP Young’s modulus E (GPa) 250-330 200 Bending strength (MPa) 300 800-1200 Fracture toughness (MPa.m1/2) 3-3.5 6-8 Poisson’s ratio 0.22 0.28 Hardness (VH) (GPa) 1600 1250

The calculated Pourbaix diagram of the system at normal temperature (fig. 2) shows that complete precipitation of both ZrO2 and Y2O3 takes place in the pH range about 7-10. Increasing temperatures up to 3000C and corresponding vapour pressure of the solution the binary system is stable in a larger pH domain (aprox. 4.5-10.5). The experimental verification of the chemical composition shows an increase of the Yttria molar ratio with increasing pH and temperature, confirming the thermodynamic predictions. Consequently an increase of the tetragonal phase has been observed in the XRD pattern presented in fig. 3.

Fig. 2. Potential-pH diagram of the system Zr-Y3-H2O vs. hydrothermal temperature

Fig. 3. Yttria molar content (right) and corresponding XRD pattern of YTZP powders obtained at 2500C (left) vs. hydrothermal temperature and pH

4.2. In-situ hydrothermal synthesis of HAP-YTZP-polymer nanopowders

Sodium maleate-vinyl acetate copolymer (MAc-VA) was used as polymer in order to control the crystallization process of HAP composites. The selection was based on the previous experiments showing its good hydrophilic properties.

0.00

1.00

2.00

3.00

4.00

1 2 3 4 5 6 7 8

pH

Y2O

3, m

ol%

T=150

T=200T=250

20 30 40 50 60

T=2500C C, T

M

M

C, T

pH 9

pH 7

pH 5

pH 4

pH 3

pH 2

2, deg.

14121086420

2.0

1.5

1.0

0.5

0.0

-0.5

-1.0

-1.5

-2.0

25

25

100

100

200

200

300

300

2525 100 100100200 200200300 30025 100 20025 100 100200 200300 300

Zr - Y - H2O - System at 25.00, 100.00, 200.00 and 300.00 C

C:\My pH

Eh (Volts)

ZrO2HZrO3(-a)

ZrOH(+3a) Y2O3*2ZrO2ZrO2

ELEMENTS Molality PressureZr 4.700E-02 Variable Y 5.200E-03 Variable

Thermodynamic predictions were made using the same soft. Due to the fact that the existing database does not contain any information regarding thermodynamic values for polymeric compounds, an approximation has been done introducing the values of the monomers. The main possible reactions in the proposed systems that have been considered are: C4H6O2(VACg) + NH4OH = C2H5NO + CH3CHO(l) + H2O (1) 5Ca(NO3)2*4H2O + 3NH4*H2PO4 + 7NH4OH = Ca5(PO4)3OH + 10NH4NO3 + 26H2O (2) 5Ca(NO3)2*4H2O + 3NH4*H2PO4 +C4H6O2(VACg)+ 8NH4OH = Ca5(PO4)3OH + C2H5NO +CH3CHO(l)+10NH4NO3 +27 H2O (3) 5Ca(NO3)2*4H2O + 3NH4*H2PO4 + C4H6O2 (VACg)+ 8NH4OH = Ca5(PO4)3OH +C2H5NO+C2H4O2(ACAg)+10NH4NO3 +26 H2O+H2(g) (4) 5Ca(NO3)2*4H2O+3NH4*H2PO4+C4H6O2(VACg)+8NH4OH= Ca5(PO4)3OH +CH3CHO(l)+ C2H4O2 (ACAg)+10NH4NO3 +26 H2O+NH3(g) (5) The calculated Gibbs energies of the reactions assumed before are presented in table 2. It may be observed that reaction (1) is very close to the equilibrium while polymer addition in the system slightly increase the standard Gibbs thermodynamic potential of the HAP formation reactions.

Table 2. Calculated Gibbs free energies

Reaction 1 Reaction 2 Reaction 3 Reaction 4 Reaction 5 T, 0 C ΔrG, kJ ΔrG, kJ ΔrG, kJ ΔrG, kJ ΔrG, kJ

0 -24.349 -140.470 -164.820 -166.470 -157.102 10 -23.992 -140.899 -164.892 -166.993 -157.880 20 -23.636 -141.387 -165.023 -167.569 -158.712 30 -23.281 -141.932 -165.213 -168.196 -159.596 40 -22.926 -142.644 -165.570 -168.985 -160.643 50 -22.571 -143.440 -166.011 -169.852 -161.769 60 -22.217 -144.292 -166.509 -170.770 -162.947 70 -21.862 -145.200 -167.062 -171.738 -164.176 80 -21.508 -146.162 -167.670 -172.756 -165.456 90 -21.153 -147.234 -168.387 -173.879 -166.841 100 -20.798 -148.397 -169.195 -175.087 -168.313 110 -20.442 -149.615 -170.057 -176.345 -169.836 120 -20.086 -150.887 -170.973 -177.654 -171.411 130 -19.729 -152.346 -172.075 -179.144 -173.168 140 -19.371 -153.982 -173.353 -180.807 -175.098 150 -19.012 -155.673 -174.685 -182.520 -177.079 160 -18.653 -157.419 -176.072 -184.285 -179.113 170 -18.292 -159.251 -177.543 -186.131 -181.229 180 -17.930 -163.213 -181.143 -190.102 -185.471 190 -17.568 -167.239 -184.807 -194.134 -189.775 200 -17.204 -171.330 -188.534 -198.225 -194.139

The influence of the polymer addition on the species formed in hydrothermal solutions at 1500C is clearly observed in the Pourbaix diagrams (fig. 4). Addition of 10 mol% VACg strongly increases the stability domain of the precipitated Ca5(PO4)3OH.

Fig. 4. Potential-pH diagrams of HAP (left) and HAP-polymer system at 1500C

The influence of the MAc-VA/HAP ratio onto the precipitation behaviour was done on the basis of turbidimetric titration. It is observed (fig. 5) that polymer/HAP ratios less then 20% are required to obtain stable colloidal suspensions during titration with ammonia.

Fig. 5. Turbidimetric titration of MAc-VA/HAP solutions

The evolution of crystalline phases of composite HAP-ZrO2 and HAP- MAc-VA powders (fig. 6) evidenced that crystalline powders containing apatite (A) as single phase were

Turbidimetric titration of MAc-VA, Na with hydroxyapatite

0

20

40

60

80

100

0 1 2 3 4 5VNH3, mL

t, %

without polymerc=6.67%c=33.35%c=66.7%

14 12 10 8 6 4 2 0

2.0

1.5

1.0

0.5

0.0

-0.5

-1.0

-1.5

-2.0

pH

Eh (Volts)

Ca(NO3)2*4H2O

Ca5(PO4)3OH

Ca(+2a)

14 12 10 8 6 4 2 0

2.0

1.5

1.0

0.5

0.0

-0.5

-1.0

-1.5

-2.0

pH

Eh (Volts)

Ca5(PO4)3OH H3PO4

NH4*H2PO4

obtained directly from the hydrothermal process. Both pure ZrO2 and YTZP substituted the Ca (II) ions in the crystalline lattice. Increasing polymer content had a very slight influence on the crystallization degree of apatite phase.

Fig. 6. XRD pattern of HAP-ZrO2 (left) and HAP- MAc-VA (right) hydrothermal powders

The nature of HAP- MAc-VA interactions was evidenced using the FT-IR analysis (fig. 7). Based on the assignment of bonding observed (table 3) it is proposed that physical interactions between pure or YTZP-doped HAP inorganic matter and polymeric MAc-VA matter are formed in-situ during hydrothermal reaction.

Fig. 7. Ft-IR spectra of HAP- MAc-VA (5% and 10%) powders (left) compared with pure

commercial and hydrothermal HAP powders (right).

10 20 30 40 50 60 70 80

AA

AAA

AA

AA

Inte

nsity

(a.u

.)

2

HAP HAP Z3 HAP ZY3

0 10 20 30 40 50 60 70 80In

tens

ity (a

.u.)

AA

A

AA

A

A

AA

2

HAP-5% polymer HAP-10% polymer HAP

0

65

20

40

60

4000 400100020003000

%T

Wavenumber[cm-1]

500 1000 1500 2000 2500 3000 3500 4000 45000

10

20

30

40

50

60

T %

Wavenumber [cm -1 ]

HAP HAP commercial

Table 3. Nature of bonding in the FT-IR spectra

Wavenumber [cm- 1] 5% 10% HAP commercial

HAP assignment

3572 3572 3570 3565 OH stretching 3443 3443 3450 3424 OH streching 1723 1723 - - Esther carbonyl stretching

1577 1577 - - C=O or C=O-Ca - - - - CO2 trace 1103 1103 - 1090 HPO4 (1135) and asymmetric

HAP (1109) 1033 1033 1030 1038 P=O 873 873 - - PO43- substitution with

CO32-

Microstructural characterization of composites powders using SEM is presented in figure 8. It may be observed that powders microstructure is strongly influenced by the presence of the MAc-VA polymer. Pure and YTZP-doped HAP powders consist of small agglomerated powders while MAc-VA powders present prismatic non-agglomerated crystals.

a) b) c)

Fig. 8. SEM of pure HAP (a), HAP doped with 10% YTZP (b) and HAP-10% MAc-VA

Some physical characteristics of the powders are presented in table 4. It is observed that MAc-VA/HAP composite powders obtained in-situ by hydrothermal reactions have a larger surface area compared to both pure and YTZP-doped HAP powders.

Table 4. Physical characteristics of HAP powders obtained by hydrothermal route

HAP HAP-YTZP HAP-5%MAc-VA

HAP- 10% MAc-VA

SBET, m2/g 63.66 119.67 166.30 156.81 Density g/cm3 3.02 3.09 2.76 2.73

Grain size nm

31.20 16.19 14.01 13.08

5. CONCLUSIONS During hydrothermal in-situ treatment of HAP-maleic acid copolymer mixtures cross-linked polymer-calcium phosphate network are formed. Maleic acid copolymers seem to be active crystal growth regulators for calcium phosphate, leading to powders with nanometric crystallite sizes, as calculated from the BET specific area. Densities values were very close of theoretical density for HAP (2.76 g/cm3 for the composite and 3.02 g/cm3 for pure HAP. Hydroxiapatite dBET is in the nanometer range in all cases. Nanocrystalline HAP is expected to present high chemical and biological activity and high sorption capacity. The results will be used in the near future to validate the biocompatibility of the system compared with testing various types of polymers for the hydrothermal synthesis of HAP-polymer nanocomposites as well as for obtaining HAP-polymer films using electrodepositing method.

Acknowledgements

National Programme for New and Advanced Materials, Micro and Nanotechnologies MATNANTECH for financial support of the research Dr. Maria Giurginca from National Consultancy Center for Environmental Protection of University POLITEHNICA of Bucharest for performing FT-IR measurements Mr.Viorel Badilita, IMNR –Bucharest, for performing XRD analysis The Nanomaterials Laboratory from UNIPRESS Warsaw for performing BET, picnometric densities and SEM analysis

Reference 1. T. Nakamura, Bioceramics 9 (1996) 31-34. 2. K. Kaneda, S. Assano, T. Hashimoto, S. Satoh, M. Fujiya, Spine 17 (1992) s295-s303. 3. M. Wang, Biomaterials 24 (2003) 2133-2151. 4. M. C. Chang, J. Tanaka, Biomaterials 23 (2002) 4811-4818. 5. N. Spanos, V. Deimede, P. G. Koutsoukos, Biomaterials 23 (2002) 947-953. 6. R. Joseph, W. J. McGregor, M. T. Martyn, K. E. Tanner, P. D. Coates, Biomaterials 23

(2002) 4295-4302. 7. W. Bonfield, Bioceramics 9 (1996) 11-13. 8. R. N. Downes, S. Vardy, K. E. Tanner, W. Bonfield, Bioceramics 4 (1991) 9. M. C. Chang, J. Tanaka, Biomaterials 23 (2002) 3879-3885. 10. M. C. Chang, T. Ikoma, M. Kikuchi, J. Tanaka, J. Mater. Sci. Lett. 20(13) (2001) 1129-

1201. 11. M. Kikuchi, Y. Suetsugu, J. Tanaka, S. Itoh, S. Ichinose, K. Shinoyama, Y.

Hiraoka, Y. Mandai, S. Nakatani, Bioceramics 12 (1999) 393-396. 12. M. C. Chang, C. C. Ko, W. H. Douglas, Biomaterials 24 (2003) 2853-2862. 13. L. L. Hench, Bioceramics (1988) 54-71. 14. A. P. Marques, R. L. Reis, J. A. Hunt, Biomaterials 23 (2002) 1471-1478. 15. R. L. Reis, A. M. Cunha, P. S. Allan, M. J. Bevis, Plastics in medicine and surgery - PIMS

'96, Glasgow, UK: University of Strathclyde (1996) 195-202. 16. A. L. Oliveira, P. B. Malafaya, R. L. Reis, Biomaterials 24 (2003) 2575-2584. 17. D.Segal, Chemical Synthesis of Advanced Ceramic Materials, Cambridge Univ. Press

(1989) 18. A. Roine, Outokumpu HSC Chemistry for Windows, version 4.0, ISBN 952-950007-05-4

Pori, Finland (1999)