Progress in Solid State Chemistry 32 (2004) 131 www.elsevier.

nl/locate/pssc

Calcium phosphates as substitution of bone tissues

Mara Vallet-Reg a,, Jose Mara Gonzalez-Calbet b
a

Departamento de Qumica Inorganica y Bioinorganica, Facultad de Farmacia, Universidad Complutense de Madrid, Pza Ramon y Cajal, 28040 Madrid, Spain b Departamento de Qumica Inorganica, Facultad de Ciencias Qumicas, Universidad Complutense de Madrid, 28040 Madrid, Spain Received 1 December 2003; received in revised form 1 June 2004; accepted 15 July 2004

Abstract Calcium phosphates with clinical applications are used in reconstructive surgery. When dealing with the repairing of a skeletal section, two extremely diverse routes could be initially considered: to replace the damaged part or to substitute it regenerating the bone. This second option is in fact the role played by calcium phosphates, which can be classied among the bioactive ceramics group. Bioceramics, and therefore, calcium phosphates, exhibit very good biocompatibility and bone integration qualities, and constitute the materials showing closest similarity to the mineral component of bone; this fact, together with their bioactivity, make them very good candidates for bone regeneration. This review is focused on calcium phosphate ceramics; therefore, it is important to analyse rstly the biological calcium phosphates as components of natural hard tissues, that is, bone and teeth, and then to look for synthetic methods able to produce calcium decient carbonate apatites with nanometric size, i.e., showing similar features to the biological apatites. In the present work, we describe the synthesis procedures to obtain in the laboratory calcium decient carbonate nanoapatite both in bulk and thin lm forms, as well as the characterization methods applied to these materials, with particular incidence in the electron microscopy. # 2004 Elsevier Ltd. All rights reserved.

Corresponding author. Tel.:+34-91-394-18-61; fax.: +34-91-394-17-86. E-mail address: vallet@farm.ucm.es (M. Vallet-Regi).

0079-6786/$ - see front matter # 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.progsolidstchem.2004.07.001

M. Vallet-Regi, J.M. Gonzalez-Calbet / Progress in Solid State Chemistry 32 (2004) 131

Contents 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Natural hard tissues: bones and teeth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Apatites and other phosphates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Substituted apatites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Calcium phosphate cements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Biphase mixtures of calcium phosphates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fabrications of bulk calcium phosphate pieces . . . . . . . . . . . . . . . . . . . . . . . . . . Ceramic coatings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tissue engineering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 3 5 14 19 20 21 23 26 27

1. Introduction Calcium phosphates with clinical applications constitute an interesting eld of research and development in the production of useful biomaterials for implant fabrication and/or xation. Biomaterials in general, and particularly bioceramics, allow replacing several parts of our body [1,2]. Ceramic materials used in reconstructive surgery can be classied in two large groups: bioinert and bioactive. Bioinert ceramics have almost no inuence in the surrounding living tissue, and their nest example would be alumina. Bioactive ceramics, by contrast, are capable of bonding with living osseous tissues; several calcium phosphates and certain compositions of glasses and ceramic glasses exhibit such feature. The bioactivity phenomenon is another example of the chemical reactivity of ceramic materials with their environment: namely, an articial solution chosen to perform the in vitro assays, or the physiological body uids during in vivo assays. The rst ceramics to be used in clinical applications, alumina and zirconia, are two prototypes of inert ceramics, and this is the main reason why they were chosen for use in implants. Such ceramics feature extremely slow reaction kinetics, to such an extent that they can be considered as almost inert ceramics. Obviously, other ceramics have much faster reaction rates, and even very fast kinetics. As in any other chemical reaction, the product obtained from the reaction of a substance with its environment could yield an undesirable outcome, such as the corrosion of a metal, but it could also lead to a favourable reaction product, through the chemical transformation of the starting material into a sought-after nal product. This is the case with the bioactive ceramics which, when in contact

 M. Vallet-Regi, J.M. Gonzalez-Calbet / Progress in Solid State Chemistry 32 (2004) 131

with physiological uids, react chemically towards the production of newly formed bone. When dealing with the repairing of a skeletal section, two extremely diverse routes could be initially considered: to replace the damaged part, or to substitute it regenerating the bone. This last option is in fact the role played by the bioactive bioceramics, such as calcium phosphates. From a structural point of view, ceramic materials can be classied as crystalline solidsceramics, amorphous solidsglasses, or amorphous solids with crystallization nucleiglass-ceramics, which can in turn be considered as inert, bioactive or resorbable. This review is focused on calcium phosphate bioceramics; therefore, it is important to analyse rstly the biological calcium phosphates as components of natural hard tissues, that is, bones and teeth.

2. Natural hard tissues: bones and teeth The bones and teeth of all vertebrates are natural composite materials, where one of the components is an inorganic solid, carbonate hydroxyapatite. It amounts to 65% of the total bone mass, with the remaining mass formed by organic matter and water. Most of this organic matter is collagen. Its molecules are bonded forming linear chains which are in turn arranged in bres, giving rise to various macroscopic structures (Fig. 1). In between said molecules there are small interstitial empty compartments, regularly spaced, where apatite nanocrystals are deposited, in a controlled biomineralization process involving more than 200 dierent acid proteins. These proteins act as inhibitors, nucleators or templates for the epitaxial growth of nanocrystals, anchored to the collagen [3]. The crystallization of the complex and hardly soluble apatite structures evolves favourably through the kinetically controlled formation of metastable intermediate products. Under in vitro conditions, amorphous calcium phosphate is transformed into octacalcium phosphate (OCP) which, in turn, evolves to carbonate hydroxyapatite; at lower pH values, the intermediate phase seems to be dehydrated dicalcium phosphate (DCPD) [4]. The bones are characterized by their composition, crystalline structure, morphology, particle size and orientation. The carbonate hydroxyapatite of bones ranges between 4% and 8% in carbonate content, which increases with age while the hydrogen phosphate ion decreases. The crystals are nanometer sized, with an average length of 50 nm, 25 nm in width and thicknesses of just 25 nm, scattered in the organic matrix. Their small size is a very important factor related to the solubility of biological apatites when compared with mineral apatites. Small dimensions and low crystallinity are two distinct features of biological apatites which, combined with their non-stoichiometric composition, inner crystalline disorder and presence of carbonate ions in the crystal lattice, allow explaining their special behaviour (Fig. 2). The bones, the body supporting scaold, can exhibit dierent types of integration between organic and inorganic materials, leading to signicant variations

M. Vallet-Regi, J.M. Gonzalez-Calbet / Progress in Solid State Chemistry 32 (2004) 131

Fig. 1. Cortical or compact bone, and trabecular or spongy bone. Arrangement of carbonate hydroxyapatite and collagen in the formation of hard tissues.

in their mechanic properties. The ratio of both components reects the compromise between toughness (high inorganic content) and resiliency or fracture strength (low inorganic content). All attempts to synthesize bone replacement materials for clinical applications featuring physiological tolerance, biocompatibility and long term stability have, up to now, had only relative success; it comes to show the superiority and complexity of the natural structure where, for instance, human femur can withstand loads of up to 1650 kg. The bones of vertebrates, as opposed to the shells of molluscs, can be considered as living biominerals since there are cells inside them under permanent activity. The bone formation process starts by the action of osteoblasts, special cells that synthesize and release the collagen matrix in the form of a jelly substance, the osteoid, which is subsequently mineralized by controlled deposition of calcium phosphate. The osteoblasts remain trapped inside the mineral phase, evolving towards osteocites which continuously maintain the bone formation activity. Meanwhile, another type of cells, the osteoclasts, catabolyse the bone destroying it. This dynamic process of bone formation and destruction accounts for its growth during the development stages of the body, preserving its shape and consistency, and enabling its regeneration in case of fracture. It also constitutes a storage and

 M. Vallet-Regi, J.M. Gonzalez-Calbet / Progress in Solid State Chemistry 32 (2004) 131

Fig. 2. Crystal structure of carbonate hydroxyapatite. Powder X-ray diraction patterns and infrared spectra of enamel, dentine and bone.

hauling mechanism for two essential elements, phosphorus and calcium, which are mainly stored in the bones. The teeth exhibit similar characteristics to the bones, except for their external surface coating, the enamel. Dental enamel has a much larger inorganic content than the bone, up to 90%, and is formed by prismatic crystals, of larger dimensions and strongly oriented. The dierences in crystallinity and carbonate content between bone and dentine (with similar qualities) and enamel are straightforward (Fig. 2). All this explains its dierent mechanical behaviour. In fact, enamel is considered as the most resistant and tough material in the biological world. However, and in contrast with the bone, dental enamel in an adult body does not contain cells, and is therefore unable to regenerate itself; any deterioration that it may suffer becomes irreversible. There is no biological process that repairs degraded or damaged enamel, evidencing the need for enamel-biocompatible materials in the repair of tooth decay. 3. Apatites and other phosphates This group includes not only the phosphates (crystalline solid materials), but also cements and biphasic mixtures with calcium phosphate content. The most used calcium phosphate in implant fabrication is hydroxyapatite, Ca10(PO4)6(OH)2, since it is the most similar material to the mineral component of

M. Vallet-Regi, J.M. Gonzalez-Calbet / Progress in Solid State Chemistry 32 (2004) 131

Fig. 3. Crystal structure of carbonate hydroxyapatite.

bones. Taking into account the hydroxyapatite lattice parameters (a 0:95 nm and c 0:68 nm), and its symmetry (hexagonal, S.G. P63/m) (Fig. 3), most likely its unit cell will be arranged along the c-axis. This would justify a preferred orientation that gives rise to an oriented growth along the c-axis and a needle-like morphology. It exhibits good properties as biomaterial, such as biocompatibility, bioactivity, osteoconductivity, direct bonding to bone, etc. Among the wide range of available calcium phosphates, or with potential formulation, it is important to know the close relation between the Ca/P ratio, acidity and solubility. Thus, the lower the Ca/P ratio is, the larger are the acidity and solubility of the mixture. For Ca=P < 1, both acidity and solubility are extremely high, and both parameters decrease substantially for Ca/P ratios close to 1.67, which is the value of stoichiometric hydroxyapatite, Ca10(PO4)6(OH)2 [5]. The Ca/P ratio is a very useful parameter for scientists working in this eld. Table 1 shows several calcium phosphates arranged according to their Ca/P ratio. The lattice parameters of some of these phosphates are quite similar, leading to overlapping of reection maxima which makes dicult, in some occasions, the adequate interpretation of the powder X-ray diraction pattern. Besides, since synthesis conditions inuence very strongly on crystal size, morphology and structure, a careful characterization by electron microdiraction and high resolution electron transmission microscopy (HRTEM) is very useful to investigate small particles and mixtures of calcium phosphates. For instance, hydroxyapatite (OHAp) and b-tricalcium phosphate (b-TCP), as a function of synthesis conditions, often coexist in v distinct proportions. Moreover, when OHAp is heated at 1050 C it converts partially to b-TCP. When crystal size is small and inhomogeinities appear in the images obtained by HRTEM, electron microdiraction is the most powerful tool to ascertain the real symmetry. Fig. 4a shows the zero-order Laue zone (ZOLZ) microdiraction pattern of an OHAp crystallite along [0001]. The sixfold symmetry

 M. Vallet-Regi, J.M. Gonzalez-Calbet / Progress in Solid State Chemistry 32 (2004) 131 Table 1 Various calcium phosphates with their respective Ca/P atomic ratios Ca/P 2,0 1,67 1,50 1,33 1,0 1,0 1,0 1,0 0,7 0,67 0,5 0,5 Name Tetracalcium phosphate Hydroxyapatite Amorphous calcium phosphate Tricalcium phosphate (a,b,c) Octacalcium phosphate Dicalcium phosphate dihydrate Dicalcium phosphate Calcium pyrophosphate (a,b,c) Calcium pyrophosphate dihydrate Heptacalcium phosphate Tetracalcium dihydrogen phosphate Monocalcium phosphate monohydrate Calcium metaphosphate (a,b,c) Formula Ca4O(PO4)2 Ca10O(PO4)6(OH)2 Ca10-xH2x(PO4)6(OH)2 Ca3(PO4)2 Ca8H2(PO4)65H2O CaHPO42H2O CaHPO4 Ca2P2O7 Ca2P2O72H2O Ca7(P5O16)2 Ca4H2P6O20 Ca(H2PO4)2H2O Ca(PO3)2 Acronym TetCP OHAp ACP TCP OCP DCPD DCPA CPP CPPD HCP TDHP MCPM CMP

characteristic of the OHAp space group is evident. In addition, the rst-order Laue zone (FOLZ) is also visible although with very weak intensity. Using the formula H 2=k R=CL2 , where H stands for the spacing between ZOLZ and FOLZ, R stands for the FOLZ radii and CL is the camera constant, the repeat distance along the c-axis can be approximately calculated [6] giving a value of H 0:7 nm which is very close to the OHAp c-axis (c 0:688 nm). Since crystallites in these two phosphates usually appear as platelets oriented along [0001], this is a nice procedure to distinguish between them. Actually, Fig. 4b shows the ZOLZ microdiraction pattern along [0001] corresponding to b-TCP where a threefold symmetry is observed, in agreement with its space group (R3c). Since the FOLZ is also visible, the c-axis can be calculated, leading to H 3:73 nm in agreement with the bTCP c-axis (3.79 nm). Calcium phosphate ceramics are unstable under the electron beam after prolonged exposure times. Fig. 5 shows the HRTEM image oriented along [0001].

Fig. 4. ZOLZ microdiraction patterns along [0001] of (a) OHAp and (b) b-TCP.

M. Vallet-Regi, J.M. Gonzalez-Calbet / Progress in Solid State Chemistry 32 (2004) 131

Fig. 5. HRTEM image along [0001] corresponding to OHAp. Damaged areas are arrowed.

Irradiation under the electron beam produces small areas (marked with arrows) with fringes dierent from those of OHAp. Areas showing irregular contrast near the edges of the crystal grew with time of exposure and corresponded to density changes related to local variations of Ca/P stoichiometry as also observed by Shulin et al. [7] on enamel apatites as due to a localized demineralization involving loss of calcium and phosphate ions. Fig. 6 shows the eect of the irradiation damage of b-TCP. The HRTEM image oriented along [0001], after exposing the crystal for a long period, shows regions of patches with fringe spacing dierent from those of the pure b-TCP. The most probable product after the decomposition is CaO formed by the diusion of Ca2+ and O2 ions. In order to obtain further information on the nature of the observed structural changes under irradiation, the Moire fringe method has been used. Moire fringe patterns have proven to be not only an excellent tool for the detection of very small local crystal spacing variations (undetectable in the image or diraction pattern) but also to enhance the eects of rotations, defects and strain elds [8]. In the case of OHAp, the study of the evolution of the Moire pattern as a function of the irradiation beam has shown to prove the existence of a secondary phase [9]. Finally, although radiation damage in calcium phosphates tends to blur the images as a result of progressive amorphization, practical results can be obtained with image processing by ltering the Fourier transform around amplitude maxima of dierent reections which enhances the image contrast [10]. Synthetic apatites aimed at emulating the biological scenario should exhibit small particle sizes and presence of CO3 . In this sense, the wet route is the most 2 adequate method of synthesis. There are several methods leading to nanometric

 M. Vallet-Regi, J.M. Gonzalez-Calbet / Progress in Solid State Chemistry 32 (2004) 131

Fig. 6. bTCP HRTEM image along [0001], after irradiation for a long period. Fringes corresponding to CaO are clearly seen.

size apatites. Some examples are, for instance: Aerosol synthesis technique has been used to produce small particles of dierent materials [11,12]. Its main advantage is that this technique has the potential to create particles of unique composition, for which starting materials are mixed in a solution at atomic level. An ulterior thermal treatment can originate important modications on morphology and texture. In consequence, OHAp preparation by this method was deemed of interest [13]. Hydroxyapatite hollow particles have been prepared by pyrolysis of an aerosol produced by ultrahigh frequency of a CaCl2(NH4)H2PO4 solution. Hollow particles were annealed at dierent v temperatures. Thermal treatment at 1050 C produces the growth of nucleated crystallites in the particle surface, with remarkable morphology. The raspberryshaped particles are polycrystalline, the crystallite size being a function of the annealing time (Fig. 7). The mean particle size is 1.2 lm, although there is a wide size range of 0.32.2 lm. The lattice image obtained by electron microscopy along the [0001] zone axis clearly shows the sixfold hexagonal symmetry characteristic of OHAp and is reected in the corresponding microdiraction pattern (left inset). However, dierent contrast is observed in dierent areas of the crystal marked AC. Lattice images in parts A and B are dierent probably due to crystal tilt. Actually, the microdiraction pattern in part B of the crystal (right inset) shows some deviation from the perfect sixfold hexagonal symmetry, which

10

M. Vallet-Regi, J.M. Gonzalez-Calbet / Progress in Solid State Chemistry 32 (2004) 131

Fig. 7. (a) Schematic representation of the pyrosol equipment used to produce powdered materials. (b) SEM micrograph showing that polycrystalline OHAp is constituted by small hollow particles. (c) Magnication of a raspberry-shaped particle. (d) Electron diraction pattern characteristic of polycrystalline OHAp. (e) Electron micrograph of an OHAp crystallite along [0001]. The microdiraction pattern at the right shows some deviation from the perfect hexagonal symmetry observed in the pattern at the left, probably due to crystal tilt.

is probably related to the crystal tilt of area B with respect to area A. Areas of white contrast (marked C) are possibly due to density variations related to local Ca/P stoichiometry. Methods based on precipitation from aqueous solutions are most suitable for preparation of large amounts of apatite, as needed for processing both into ceramic bodies and in association with dierent matrixes. The diculty with most of the conventional precipitation methods used is the synthesis of well-dened and reproducible orthophosphates [14,15]. Problems can arise due to the usual lack of precise control on the factors governing the precipitation, pH, temperature, Ca/P ratio of reagents, etc., which can lead to products with slight dierences in stoichiometry, crystallinity, morphology, etc., that could then contribute to the dierent in vivo/in vitro behaviours described. In this sense, it is important to develop a methodology able to produce massive and reproducible quantities of apatite, optimized for any specic application or processing requirements by controlling composition, impurities, morphology, and crystal and particle size.

 M. Vallet-Regi, J.M. Gonzalez-Calbet / Progress in Solid State Chemistry 32 (2004) 131

11

For quantitative reactions in solutions, the reactants must be calcium and phosphate salts with ions that are unlikely to be incorporated into the apatite lattice. Since it has been claimed that NO and NH are not incorporated into crystal3 4 line apatites, or in the case of NH present a very limited incorporation [16], the 4 chosen reaction for this method was 10CaNO3 2 4H2 O 6NH4 2 HPO4 8NH4 OH ! Ca10 PO4 6 OH2 20NH4 NO3 6H2 O.

Thus, apatites with dierent stoichiometry and morphology have been prepared and the eects of varying synthesis conditions on stoichiometry, crystallinity, and morphology of the powder are analysed. The eects of varying concentration of the reagents, the temperature of the reaction, reaction time, initial pH, aging time, and the atmosphere within the reaction vessel were also studied (Fig. 8). Temperav tures in the range 2537 C are necessary to obtain apatites with crystal sizes in the v range of adult human bone, while 90 C are necessary to obtain apatites with crystal sizes in the range of enamel. Higher reaction times lead to apatites with higher Ca/P ratios. Aging of the precipitated powder can lead to the incorporation of minor quantities of carbonate. It is possible to force the incorporation of carbonate ions into the apatitic structure without introducing monovalent cations [17]. The main results of the studied variations in the reaction conditions are, in short, that higher concentrations of reagents produce higher amounts of products with minor dierences in their characteristics, allowing the production of homogeneous sets of materials. The application of the liquid mix technique, which is based on the Pechini patent [18]. This patent was originally developed for the preparation of multicomponent oxides, allowing the production of massive and reproducible quantities with a precise homogeneity in both composition and particle size. This method is based on the preparation of a liquid solution that retains its homogeneity in the solid state. This method not only allows a precise control of the cation concentration, but also the diusion process is enormously favoured by means of the liquid solution, compared to other classical methods. The solution of metallic salts in polycarboxylic P O3 acid such as citrates is solidied by the addition of a diol which increases the 4 solution viscosity due to the formation of ester-type three-dimensional polymers. When the diol reacts with the citric solution a resin is formed, thus avoiding the partial segregation, which would modify the original homogeneity of the solution [19]. Its application has now extended to the preparation of calcium phosphates. The main diculty of this synthesis lies on the presence of groups that cannot be complexed by citric acid, and may cause its segregation and the formation of separated phosphate phases. The success of this task would suggest the possibility, by modifying the synthesis conditions, of obtaining large amounts of single phases or biphasic mixtures with precise proportions of the calcium phosphates. This method makes it possible to obtain single phase hydroxyapatite, b-TCP and a-TCP and also biphasic materials whose content in b-TCP and OHAp can be precisely predicted from the Ca/P ratio in the precursor solutions [20] (Fig. 9).

12

M. Vallet-Regi, J.M. Gonzalez-Calbet / Progress in Solid State Chemistry 32 (2004) 131

Fig. 8. Schematic diagram of a crystallization setup. Micrographs and electron diraction patterns (insets) of calcium-decient apatites obtained from Ca(NO3)24H2O and (NH4)2HPO4; all the variable production parameters were equal for both materials, except the synthesis temperature (as displayed).

Obviously, several techniques have been utilized for the preparation of hydroxyapatite and other calcium phosphates. The synthetic routes employed can be divided into solid-state reactions and wet methods, which include precipitation, hydrothermal and hydrolysis of other calcium phosphates [17,2126]. Modications of these classical methods (precipitation, hydrolysis or precipitation in the presence of urea, glycine, formamide, hexamethylenetetramine. . .) [2729] or alternative techniques have been employed to prepare hydroxyapatite with morphology, stoichiometry, ion substitution or degree of crystallinity as required for a specic application. Among them, solgel [3034], microwave irradiation [35,36], freezedrying [37], mechanochemical method [3841], emulsion processing [4244], spray pyrolysis [13,45,46], hydrolysis of a-TCP [47], ultrasounds [48,49]. . . can be outlined.

 M. Vallet-Regi, J.M. Gonzalez-Calbet / Progress in Solid State Chemistry 32 (2004) 131

13

Fig. 9. Outline of calcium phosphate synthesis by liquid mix technique. Three dimensional phase diagramCa/P ratio, annealing temperature, OHAp contentof calcium phosphates.

Particular attention must be paid at this point to the essays performed on the absence of gravity. Suvurova and Buat [50] have compared the results obtained when calcium phosphate specimens, in particular, OHAp and triclinic octacalcium phosphate (OCP), are prepared from aqueous solutions under dierent conditions of precipitation. When supersaturated solutions of calcium phosphates are prepared by diusion-controlled mixing in space (EURECA 19921993 ight) several dierences are observed in crystal size, morphology and structural features with respect to those prepared on earth. It is worth stressing that space-grown OCP crystals possess a maximum growth rate in the [001] direction and a minimum rate in the [100] one. Space-grown and terrestrial OHAp crystals dier from each other in size: the former are, at least 11.5 orders of magnitude bigger in length. Diffusion-controlled mixing in space seems to provide a lower supersaturation in the crystallization system comparatively to earth, promoting the crystal growth in the competition between nucleation and growth. These authors conclude that similar processes may most probably arise in the human body (under denite internal conditions) during space ying when quite large OHAp crystals start to grow instead of the small and natural ones. In addition, other modication of OCP crystals with

14

M. Vallet-Regi, J.M. Gonzalez-Calbet / Progress in Solid State Chemistry 32 (2004) 131

huge sizes appears. These elements may disturb the Ca dynamical equilibrium in the body which might lead to possible demineralization of the bone tissue. Biological apatites (mineral component of the bones) are dicult to synthesize at the laboratory with carbonate contents equivalent to those in the bone. Although the carbonate inclusion in itself is very simple (in fact, when producing stoichiometric apatites at the laboratory, a strict control of the synthesis conditions is needed to avoid carbonate inclusion), the carbonate content is always dierent from the fraction of carbonates in the natural bone (48% wt) [21] and/or are located in dierent lattice positions [51]. At this point, it should be mentioned that this carbonate content can be slightly dierent when analysed samples come from other vertebrates [52]. The carbonate easily enters into the apatite structure, but the problem lies in the amount that should be introduced taking into account the carbonate content of biological apatites. When the aim is to obtain carbonate hydroxyapatite and the reaction takes place at high temperatures, the carbonates enter and occupy lattice positions in the OH sublattice (A type apatites). In contrast, the carbonates in biological apatites always occupy positions in the PO3 sublattice 4 (that is, they are B type apatites) [51]. In order to solve this problem, low temperature synthesis routes have to be followed, allowing obtaining carbonate hydroxyapatites with carbonates in phosphate positions [21]. But the amount entered remains to be solved, and it usually is lower than the carbonate content of the mineral component of the bones. These calcium decient and carbonated apatites have been obtained in laboratory by various techniques; nowadays, it is known that apatites with low crystallinity, calcium deciency and carbonate content can be obtained, but with carbonate contents usually unequal to those of the natural bones [17,53,54]. Therefore, the main problem remains in the control of carbonate content and lattice positioning.

4. Substituted apatites Hydroxyapatite is clearly a non-stoichiometric compound, with the ability to accept compositional variations in its three sublattices (Fig. 10). No biological hydroxyapatite shows a stoichiometric Ca/P ratio, but they all move towards tissues, which are linked to an increase in crystallinity. These trends have a remark-

Fig. 10. Generic formulation of apatite minerals, and potential substitutions in the three sublattices.

 M. Vallet-Regi, J.M. Gonzalez-Calbet / Progress in Solid State Chemistry 32 (2004) 131

15

able physiological meaning, since the younger, less crystalline tissue can develop and grow faster, while storing other elements that the body needs during its growth; this is due to the highly non-stoichiometric quality of OHAp, which caters for the substitutional inclusion of dierent amounts of several ions, such as Na+, K+, Mg2+, Sr2+, Cl, F, HPO2 , etc [1]. 4 The more crystalline the OHAp becomes, the more dicult interchanges and growth are. In this sense, it is worth stressing that the bone is probably a very important detoxicating system for heavy metals due to the ease of substitution in OHAp; heavy metals, in the form of insoluble phosphates, can be retained in the hard tissues without important alterations of their structural properties. But the ability to exchange ions in this structure also allows to design, develop and characterize new and better calcium phosphates for certain specic applications. It is known that the bone regeneration rate depends on several factors such as porosity, composition, solubility and presence of certain elements that, released during the resorption of the ceramic material, facilitate the bone regeneration carried out by the osteoblasts. Thus, for instance, small amounts of strontium, zinc or silicates stimulate the action of these osteoblasts and, in consequence, the new bone formation. Carbonate and strontium favour the dissolution, and therefore the resorption of the implant. Silicates increase the mechanical strength, a very important factor in particular for porous ceramics, and also accelerate the bioactivity of apatite [55]. The current trend is, therefore, to obtain calcium phosphate bioceramics partially substituted by these elements. In fact, bone and enamel are some of the most complex biomineralized structures. The attempts to synthesize bone at the laboratory are devoted at obtaining biocompatible prosthetic implants, with the ability to leverage natural bone regeneration when inserted in the human body. Its formation might imply certain temporary structural changes on its components, which demand in turn the presence, at trace levels, of additional ions and molecules in order to enable the mineralization process. This is the case, for instance, with bone growth processes, where the localized concentration of silicon-rich materials coincides precisely with areas of active bone growth. The reason is yet unknown, although the evidences are clear; the possible explanation of this phenomenon would also justify the great activity observed in certain silicon-substituted apatite phases and in some glasses obtained by solgel method, regarding cell proliferation and new bone growth. One way to enhance the bioactive behaviour of hydroxyapatite is to obtain substituted apatites, which resemble the chemical composition and structure of the mineral phase in bones [56,57]. These ionic substitutions can modify the surface structure and electrical charge of hydroxyapatite, with potential inuence on the material in biological environments. In this sense, an interesting way to improve the bioactivity of hydroxyapatite is the addition of silicon to the apatite structure, taking into account the inuence of this element on the bioactivity of bioactive glasses and glass-ceramics [58,59]. In addition, several studies have proposed the remarkable importance of silicon on bone formation and growth [60,61] at in vitro and in vivo conditions.

16

M. Vallet-Regi, J.M. Gonzalez-Calbet / Progress in Solid State Chemistry 32 (2004) 131

Several methods for the synthesis of silicon-substituted hydroxyapatites have been described. Ruys [62] suggested the use of a solgel procedure; however, these materials, besides the hydroxyapatite phase, include other crystalline phases depending on the substitution degree of silicon. Tanizawa and Suzuki [63] tried hydrothermal methods, obtaining materials with a Ca/(P+Si) ratio higher than that of pure calcium hydroxyapatite. Boyer et al. [64] conducted studies on the synthesis of silicon-substituted hydroxyapatites by solid-state reaction, but in these cases the incorporation of a secondary ion, such as La3+ or SO2 , was needed. In 4 these examples, no bioactivity studies were performed on the silicon-containing apatites. Gibson et al. [65] synthesized silicon-containing hydroxyapatite by using a wet method, and its in vitro bioactivity studies gave good results. These authors studied the eects of low substitution levels on the biocompatibility and in vitro bioactivity, determining the ability to form the apatite-like layer by soaking the materials in a simulated body uid (SBF) [66]. Also, Marques et al. [67] synthesized, by wet method, hydroxyapatite with silicon content up to 0.15 wt%, obtainv ing stable materials at 1300 C and noting that the unit cell volume and the a parameter length of the hydroxyapatite decreased as the silicon content increased. Hence, the role of silicon substituting part of the phosphorus atoms present in the hydroxyapatite lattice seems to be an important factor inuencing the bioactive behaviour of the material. However, it is not clearly known whether the silicon present in the material substitutes completely the phosphorus in the hydroxyapatite structure, or whether the replacement is partial, or even if in any of the described synthesis the silicon species remain as an independent phase. In all the cited syntheses, the nal product contains silicon, but its chemical nature is not revealed. A similar work focused on the synthesis and bioactivity study of hydroxyapatites containing orthosilicate anions that isomorphically replace phosphate groups, aimed at improving the bioactivity of the resulting materials as compared with that of pure calcium hydroxyapatite. To accomplish this purpose, two synthesis procedures were used, starting from two dierent calcium and phosphorus precursors and the same silicon reagent in both cases. To assess the proposed substitution, surface chemical and structural characterization of the silicon-substituted hydroxyapatites was performed by means of X-ray diraction (XRD) and X-ray photoelectron spectroscopy (XPS). The in vitro bioactivity of the so-obtained materials was determined by soaking the materials in SBF and monitoring the changes of pH and chemical composition of the solution, whereas the modication at the surface was followed by means of XPS, XRD, and scanning electron microscopy (SEM). Silicon-containing hydroxyapatites were synthesized by the controlled crystallization method. Chemical analysis, N2 adsorption, Hg porosimetry, X-ray diraction, scanning electron microscopy, energy-dispersive X-ray spectroscopy, and Xray photoelectron spectroscopy (XPS) were used to characterize the hydroxyapatite and to monitor the development of a calcium phosphate layer onto the substrate surface immersed in a simulated body uid, that is, in vitro bioactivity tests. The inuence of the silicon content and the nature of the starting calcium and phosphorus sources on the in vitro bioactivity of the resulting materials were studied. A

 M. Vallet-Regi, J.M. Gonzalez-Calbet / Progress in Solid State Chemistry 32 (2004) 131

17

sample of silicocarnotite, whose structure is related to that of hydroxyapatite and contains isolated SiO4 anions that isomorphically substitute some PO3 anions, 4 4 was prepared and used as reference material for XPS studies. An increase of the unit cell parameters with the Si content was observed, which indicated that SiO4 4 units are present in lattice positions, replacing some PO3 groups. By using XPS it 4 was possible to assess the presence of monomeric SiO4 units in the surface of apa4 tite samples containing 0.8 wt% of silicon, regardless the nature of the starting raw materials, either CaNO3 2 =NH4 2 HPO4 =Si OCOCH3 4 or CaOH2 =H3 PO4 =Si OCOCH3 4 . However, an increase of the silicon content up to 1.6 wt% leads to the polymerization of the silicate species at the surface. This technique shows silicon enrichment at the surface of the three samples. The in vitro bioactivity assays showed that the formation of an apatite-like layer onto the surface of siliconcontaining substrates is strongly enhanced as compared with pure silicon-free hydroxyapatite (Fig. 11). The samples containing monomeric silicate species showed higher in vitro bioactivity than that of silicon-rich sample containing polymeric silicate species. The use of calcium and phosphate salts as precursors leads to materials with higher bioactivity [68]. Finally, the results revealed that controlled crystallization is a good procedure to prepare silicon-substituted hydroxyapatites that can be used as a potential material for prosthetic applications. The presence of silicon (Si) in HA has shown an important role on the formation of bone [60]. To study the role of Si, Si-substituted hydroxyapatite (Si-HA) has been synthesized by several methods [6265,67,68] but its structural characteristics and microstructure remain not fully understood. The structural studies carried out until now (mainly by X-ray diraction) have not demonstrated the Si incorporation into the apatite structure. In fact, the very similar scattering factor makes very dicult to determine if Si has replaced some P in the same crystallographic position. Until now, the absence of secondary phases and the dierent

Fig. 11. SEM images of silicon-substituted apatites, before and after soaking in SBF for 6 weeks.

18

M. Vallet-Regi, J.M. Gonzalez-Calbet / Progress in Solid State Chemistry 32 (2004) 131

bioactive behaviour are the best evidences for the Si incorporation. No positive evidence or quantitative study of P substitution by Si has been carried out yet. On the other hand, the hydroxyls groups sited at the 4e position are one of the most important sites for the HA reactivity. The movement of H along the c-axis 2contributes to the HA reactivity. However, XRD is not the optimum tool for the study of light atoms such as H; neutron diraction (ND) is an excellent alternative to solve this problem. The Fermi length for Si and P are dierent enough to be discriminated, whereas neutrons are very sensitive to the H presence. Consequently, XRD and ND have combined to answer one of the most recent subjects in the dentistry and orthopaedic surgery elds. In order to explain the higher bioactivity of the silicon-substituted hydroxyapatite (SiHA), synthetic ceramic hydroxyapatite (HA) and SiHA have been structurally studied by neutron scattering. The Rietveld renements show that the nal compounds are oxy-hydroxyapatites, when obtained by solid-state synthesis under air atmosphere. By using neutron diraction, the substitution of P by Si into

Fig. 12. Thermal ellipsoids schematic drawing for hydroxyapatite and silicon-substituted hydroxyapatite.

 M. Vallet-Regi, J.M. Gonzalez-Calbet / Progress in Solid State Chemistry 32 (2004) 131

19

the apatite structure has been corroborated in these compounds. Moreover, these studies also allow us to explain the superior bioactive behaviour of SiHA, in terms of higher thermal displacement parameters of the H located at the 4e site [69]. The renements of the structure by the Rietveld method indicate that the thermal treatment produces partial decomposition of the OH groups, leading to oxyhydroxy apatites in both samples and the higher reactivity of the Si-substituted HA can be explained in terms of an increasing of the thermal ellipsoid dimension parallel to the c-axis for H atoms (Fig. 12).

5. Calcium phosphate cements Cements based on calcium salts, phosphates or sulphates, have attracted much attention in medicine and dentistry due to their excellent biocompatibility and bone-repair properties [7073]. Moreover, they have the advantage over the bioceramics that they do not need to be delivered in prefabricated forms, because these self-setting cements can be handled by the clinician in paste form and injected into bone cavities. Depending on the cement formulation, or the presence of additives, dierent properties, such as setting time, porosity or mechanical behaviour have been found in these materials [7478]. On the other hand, in the literature on phosphates focused on calcium phosphate cements, the technique employed for obtaining such cements is to mix the dierent components; one of them is responsible for curing the mixture. For instance, in the Constanz cement [79]the rst of its kind to be commercialized the nal product is a carbonateapatite (dahlite) with low crystallinity and a carbonate content reaching 4.6%, in substitution of phosphate groups (B type carbonateapatite) as is the case in bones. Constanz cement is obtained from a dry mixture of a-tricalcium phosphate, a-Ca3(PO4)2, calcium phosphate monohydrate, Ca(H2PO4)H2O, and calcium carbonate, CaCO3. The Ca/P ratio of the rst component is 1.50, and 0.5 for the second one, both values signicantly lower than the Ca/P ratio of 1.67 for hydroxyapatite. A liquid componenta sodium monoacid phosphate solutionis then added to this solid mixture, which allows the formation of an easily injectable paste that will cure over time. The paste curing happens after a very reasonable period of time when considering its use in surgery. In fact, after 5 min it shows a consistency suitable for injection, and upon 10 min it is solid without any exothermal response, exhibiting an initial strength of 10 MPa. Twelve hours later, 90% of its weight has evolved to dahlite, with compression strength of 55 and 2.1 MPa when under stress. This cement is then resorbed and gradually replaced by newly formed bone. Calcium phosphate cements which can be resorbed and injected are being commercialized by various international corporations [80], with slight dierences in their compositions and/or preparation. Research is still under way in order to improve the deciencies still present. These cements cure in eld, are very compatible with the bone and seem to resorb slowly; during this gradual process, the newly formed bone grows and repla-

20

M. Vallet-Regi, J.M. Gonzalez-Calbet / Progress in Solid State Chemistry 32 (2004) 131

ces the cement. However, the properties of the calcium phosphate cements are still insucient for their reliable application. There are problems related to their mechanical toughness, the curing time, the application technique on the osseous defect and the nal biological properties. New improvements in the development of these cements will soon be described, solving at least in part some of these disadvantages. For instance, the curing time will be shortened, even in contact with blood, and the toughness under compression will also improve. Most of the injectable calcium phosphate cements used evolves to an apatitic calcium phosphate during the setting reaction. One of the main drawbacks of these apatitic cements is the slow resorption rate of the apatite. On the other hand, calcium sulphate dihydrate, gypsum, has been used as bone-void ller during many years [70,7981], although it presents a too fast resorption rate to provide a good support for new bone. The combination of both, calcium sulphate and apatite, can overcome the individual drawbacks, and in the last years studies using this biphasic material have been performed [8284]. Despite the advantages, all these implants can act as foreign bodies and become potential sources of infections. Then, the in vivo utilization of these materials requires a preventive therapy and this may be achieved introducing a drug into them, which can be locally released in situ after implantation. In fact, dierent studies using bioceramics and self-setting materials containing active drugs have been performed in the last years [8589]. In this sense, the addition of an antibiotic to calcium sulphate-based cements has also been studied, in order to determine if the presence of the drug aects the physico-chemical behaviour of the cements and to study the release kinetics of the drug from the cement. Two system types were chosen: gypsum and apatite/gypsum. The antibiotic chosen for this study was cephalexin in crystalline form, i.e. cephalexin monohydrate. The presence of cephalexin into the cements does not alter neither the physicochemical behaviour of the cements nor produce structural changes on them. The release of the drug is dierent depending on the composition. For gypsum cements, the cephalexin is quickly released, helped by a dissolution process of the matrix, whereas the drug release is more controlled by the hydroxyapatite presence in hydroxyapatite/gypsum samples. Apatite-containing cements do not only show a dierent drug release process, also the paste viscosity is lower and a faster formation in vitro of an apatite-type layer on their surface is observed [90].

6. Biphase mixtures of calcium phosphates Several attempts have been made to synthesize the mineral component of bones starting from biphase mixtures of calcium phosphates [91]. Hence, bone replacing materials based on mixtures of hydroxyapatite and b-TCP have been prepared; under physiological conditions, such mixtures evolve to carbonate hydroxyapatite. The chemical reactions are based in equilibrium conditions between the more stable phase, hydroxyapatite, and the phase prone to resorption, b-TCP. As a

 M. Vallet-Regi, J.M. Gonzalez-Calbet / Progress in Solid State Chemistry 32 (2004) 131

21

consequence, the mixture is gradually dissolved in the human body, acting as a stem for newly formed bone and releasing Ca2+ and PO3 to the local environ4 ment. This material can be injected, used as coating or in any other form suitable for application as bulk bone replacementforming of bulk pieces, lling of bone defects[92]. At present, a wide range of biphase mixtures are under preparation, using various calcium phosphates, bioactive glasses, calcium sulphates, etc. [55,84,93,94]. Currently, there is an increasing interest on the preparation of mixtures of two or more calcium phosphates. These materials are commonly prepared with hydroxyapatite and a more resorbable material such as tricalcium phosphate (a or b) or calcium carbonate in dierent proportions depending on the characteristics required for a specic application. Some examples of commercial products based on these mixtures are: Triosite2, MBCPTM2, Eurocer1,. . .. The synthesis routes employed commonly in the preparation of these mixtures include the blending of dierent calcium phosphates [95,96], and precipitation [9799]. Other techniques also employed are: solid state [100], treatment of natural bone [101], spray pyrolysis [102], microwave [103], combustion [104]. . . . Some authors have defended the superior properties of the biphasic materials directly prepared over those obtained by mixing two single phases [105]. The promising results obtained with cements and biphase mixtures seem to indicate that it is easier to obtain precursors of synthetic apatites that, when in contact with the biological environment, can evolve towards similar compositions to that of the biological apatite, than to obtain apatites in the laboratory with similar compositional and structural characteristics to those of the biological material, and in adequate quantities, i.e. large, industry-scale amounts with precise composition and easily repeatable batch after batch, for its use in the production of ceramic biomaterials. Bioceramics aimed at the replacement or lling of bones could be obtained by synthesis of apatite precursors through dierent calcium phosphate mixtures, using a wet route. If the information gathered from the calcium cements is put to use, it would be necessary to eliminate the solution added to cure the mixture and search for compositions and ratios that allow to obtain precursors that, when in contact with the body uids, evolve chemically towards the formation of carbonate hydroxyapatite crystals, with small particle size and low crystallinity, calcium decient and with a carbonate content of approximately 4.5% w/w, located in the PO3 sublattice. 4

7. Fabrication of bulk calcium phosphate pieces Traditionally, calcium phosphate ceramics have been processed by high temperature treatments, [5,106]. If the products have previously been synthesized by the wet route at low temperatures, this leads to a very crystalline material, and therefore not similar to biological apatites. Owing to the thermal decomposition of most calcium phosphates at high temperature, this type of process is restricted basically

22

M. Vallet-Regi, J.M. Gonzalez-Calbet / Progress in Solid State Chemistry 32 (2004) 131

to obtaining stoichiometric hydroxyapatite and tricalcium phosphate. It can be employed in the production of both dense and porous pieces. The latter allow the growth of natural tissue inside the pores, helping to mechanically stabilize the implants. Although these treatments are still in use nowadays [107,108], they reduce signicantly the reactivity of the ceramics and the growth kinetics of the bone. Therefore, new forming methods at lower temperatures have been developed [109,110], allowing to obtain pieces without altering the crystallinity of the ceramic starting material. The combination of the synthesis of calcium decient, low crystallinity carbonate hydroxyapatites with processing methods that preserve their chemical and microstructural features is an excellent alternative to the production of bioceramic pieces. Methods such as colloidal processing [107], uniaxial pressing from a precipitated powder [108], cold isostatic pressing [111] and starch consolidation [112] and combination of gel casting and foams out [113] have yielded excellent results. In this sense, porous machinable pieces of OHAp have been prepared from polyurethane foams by combination of gelcasting and burn out methods (Fig. 13). The pieces were constituted by polyhedral-like particles with an average size of 0.451.0 mm that are surrounded by an interconnected network of pores. The porous pieces showed a bimodal distribution of the pores size between 30.8 58.6 and %1.0 lm. The size of the interconnected pores (30.858.6 lm) can be controlled as a function of the cells in the used foam. The composition of the ceramics and the volume of the small pores on the particles surface can be modied as a function of the sintering time. The presence of pores could promote the bone ingrowth and also could be used to insert dierent drugs, which makes these porous pieces a potential candidate to be used as non-load-bearing bone implants and as drug delivery systems. If materials in powder form instead of pieces are used for bone lling applications, the main advantage is their superior adaptability to the prole of each defect, while the clinical insertion is not so adequate. Such materials are dicult to

Fig. 13. SEM micrographs of porous pieces sintered at 1300 C for 3 h, at dierent magnication values.

 M. Vallet-Regi, J.M. Gonzalez-Calbet / Progress in Solid State Chemistry 32 (2004) 131

23

place and secure in the implant region, and their particles retain the potential to migrate for weeks or even months. To avoid this drawback, the powder is mixed with a degradable carrier matrix. At present, the trend is to obtain materials suitable for injection [114]. Both organic and inorganic matrices can be used. Among the organic matrices, non-absorbable polymers such as polymethyl methacrylate (PMMA) [115,116], polyethylene (PE) [117] and polysulphone are being used, but the main problem is that such matrices reduce the bioactivity of the implant. Biodegradable polymers are also in use, such as polylactic acid (PLA) [118] and polyglycolic acid (PGA), as well as natural polymers [119,120] (collagen, cellulose and starch). A inorganic matrix also in use is calcium sulphate (Hapset) [121].

8. Ceramic coatings At present, for all those clinical applications where load-bearing properties are required, most of the implants used are metallic, with subsequent and serious problems due to: (a) the large dierences in mechanical properties between the articial implant and the natural bone, giving rise to ruptures, (b) the presence of ions that, released from the articial implant, could be toxic or harmful and provoke pains, and (c) the impossibility to regenerate natural bone. An alternative option, until a more similar material to bone becomes available, is to coat the metallic implant with ceramics. This technique is being used nowadays, both for dental implants and hip joint prosthesis. There is still a long way to follow, but several metallic implants with ceramic coatings are commercially available already, and the research in problem solving is under way. The ceramic coating process on a metallic substrate is quite complicated, and several methods are available in this sense. A great deal of the clinical success depends on this coating, since the quality and durability of the interface attachment greatly depend on the purity, particle size, chemical composition of the coating, layer thickness and surface morphology of the substrate. An additional advantage when coating a metallic implant with ceramics is the reduction in ion release issues from the metal alloy. The ceramic represents a truly eective barrier that hinders the metallic ion kinetics of release towards the living body. Hydroxyapatite is being specically used for this purpose, in order to improve the attachment of hip joint prostheses, due to its excellent biological properties such as non-toxicity and lack of inammatory response and brous and/or immunitary reactions. Hydroxyapatite (HA), tricalcium phosphate (TCP), and their biphasic combinations are important ceramic materials in the replacement of hard tissues, because they can form a strong bond with the bone and favour bone formation. However, the poor mechanical properties of calcium phosphates limit the use of the bulk material to non-load-bearing implants. For this reason, one of most important uses of these calcium phosphates is to coat inert or biotolerable implants with mechanical properties adequate for orthopaedic substitutions. In this way, the coated implants will not only have the good mechanical properties of the substrate but

24

M. Vallet-Regi, J.M. Gonzalez-Calbet / Progress in Solid State Chemistry 32 (2004) 131

also an enhanced osseointegration and bioactivity due to the calcium phosphate layer. Plasma spray technique is currently the only method commercially available for coating metallic substrates [122124]. The signicant deciencies found in the plasma sprayed HA coatings have promoted the search for new deposition methods, such as ion beam assisted deposition, magnetron sputtering, solgel, pulsed laser deposition. . .[125128]. Moreover, not only the synthesis method, but the crystallinity and the calcium phosphate phases present in the lm inuence the solubility and biological behaviour of these coatings [129,130]. The so-called pyrosol method has been applied for the deposition of a wide variety of ceramic materials with dierent applications [131133]. The ability to control the composition and microstructure of the coatings and to obtain uniform lms onto irregular surfaces, along with the facility for large scale production makes this technique potentially very powerful. Thin lms of calcium phosphates with dierent crystallinity and Ca=P ratios have been prepared by the pyrosol method and dip-coating procedures [134138] and characterized by XRD, FTIR, EDS and SEM (Fig. 14) [139]. The microstructure, crystallinity and composition of the deposited lms can be controlled by modifying the composition of the precursor solution, reactor atmosphere and substrate temperature. By this method, pure hydroxyapatite with varying crystallinity and microstructure can be obtained under more oxidant conditions, whereas tricalcium phosphate (a or b) or biphasic hydroxyapatite/tricalcium phosphate mixtures can be obtained under argon atmosphere. The ability to vary the calcium phosphate phases deposited has signicant implications for the use of these lms in implant devices. The results show that the use of a simple and versatile solution method allows calcium phosphate coatings with dierent crystallinity and phase compo-

Fig. 14. Schematic representation of the pyrosol equipment used to produce coatings (left). SEM micrographs (above, right) of three hydroxyapatite coatings with dierent morphologies, obtained by pyrosol. Powder X-ray diraction pattern (below, right) of one of these coatings.

 M. Vallet-Regi, J.M. Gonzalez-Calbet / Progress in Solid State Chemistry 32 (2004) 131

25

sition to be prepared. Dissolution tests show that these calcium phosphate layers are stable and exhibit dierent behaviours as a function of their composition. The ability to modify the phase composition, crystallinity and microstructure of the coatings should allow designing coatings with tailored in vitro or in vivo behaviour. Although dierent deposition methods have been applied in the last years [139 143], solgel method oers a good alternative since the synthesis temperatures are low, and it can be applied to a great number of substrates, including those which would oxidize at higher temperatures. Several authors have prepared OHAp via solgel technique using dierent precursors [144147], showing that the temperature required to form OHAp depends on the chemical reactivity of the precursors. Livage et al. [148] investigated the solgel synthesis of phosphates and found that alkyl phosphates esters or phosphoric acid were unsuitable precursors, either because the hydrolysis is too slow, or because it reacts so fast that a precipitate is formed, as in the case of phosphoric acid. In the same way, dierent works show the necessity of controlling the ageing eect of the sol, in the synthesis of single phase OHAp powders by the solgel method [149151]. A similar systematic study for the optimization of the lms deposition has been carried out [152]; single phase

Fig. 15. Immersion stage and thermal treatment followed for the preparation of coatings by dip-coating technique. The illustration shows a total joint prosthesis partially coated with hydroxyapatite. Atomic force micrographs of a substrate and two coatings obtained at dierent immersion velocities.

26

M. Vallet-Regi, J.M. Gonzalez-Calbet / Progress in Solid State Chemistry 32 (2004) 131

OHAp coatings were deposited on Ti6Al4V by the solgel dipping technique from aqueous solutions containing triethyl phosphite and calcium nitrate. In order to obtain homogeneous and monophasic hydroxyapatite coatings, the ageing time and temperature of the sol were, fundamentally, the variables studied. The pH measurement was a good tool to evaluate the best conditions of the sols to deposit these coatings. SEM and SFM techniques show that the coatings deposited are dense and homogeneous and with a low roughness (Fig. 15) which depends on the sol viscosity and the lm thickness. The higher the sol temperature, the shorter the ageing time needed to obtain pure OHAp after coating annealing. When the ageing parameters or the annealing temperature are not adequately controlled, additional phases or poor surfaces are obtained. The conditions to obtain the best coatings have been related with the pH decrease on the aqueous sols observed during the ageing, according to the polymerization reaction between calcium and phosphorous. In order to obtain homogeneous, crack-free coatings, the annealing temperature and thickness of the coatings must be controlled. Films roughness is related with the viscosity of the sol-precursor used to do the deposition, as well as with the number of coating layers.

9. Tissue engineering This eld, which started about a decade ago, is now at full research potential and the rst results are currently being developed. The aim is basically to provide an articially made scaold, e.g. a bioceramic, onto which cells are cultured so that the articial piece becomes colonized. This process can be carried out both under in vitro and in vivo conditions. One of the main objectives is to develop materials aimed at the functional repair and reconstruction of biological structures. In this sense, much attention is being devoted to the synthesis and surface study of various substratessuch as calcium phosphates, among othersfor their application in the development of three-dimensional scaolds for tissue engineering. Perhaps the main issue in this subject is the study and tailoring of surface properties of these substrates, in order to control the interaction with biological entities such as macrocells and cells. The repair and replacement strategy for damaged parts of the human body is clearly dierent from the traditional biomedical implants currently in use where, in order to perform a bone replacement, the use of donor tissue is the predominant trend, either from allografts or autografts; there is, however, a certain increase in the application of articial materials. Donor tissues exhibit good biocompatibility, but are also handicapped by clear disadvantages: these tissues are costly, scarce, there are risks of disease transmission, etc., and thus it can be predicted that synthetic materials will conquer a good share of this market in the long run. The current global market for these products has an estimated value of more than 1000 million euro, with an annual growth rate of 7.7%. With the increasing number of complex revision surgery interventions and, as a consequence, the higher use of bone lling materials, it seems reasonable to forecast a market for the next decade

 M. Vallet-Regi, J.M. Gonzalez-Calbet / Progress in Solid State Chemistry 32 (2004) 131

27

in excess of 2000 million euro. There is a general agreement in the fact that this is a eld yet to explore [153]. The feasible production of ceramic pieces with tailored porosity, in order to be used as substrates for tissue engineering, opens up a spectacular future for calcium phosphates.

10. Conclusions In the future, bioceramics could be the ideal biomaterials due to their good biocompatibility and bone integration, and also to their close similarity to the mineral component of bone. However, the use of bioceramics for load-bearing applications is still far from feasible; their inherent rigidity and breakability are the main drawbacks here. Therefore, the applications of bioceramics are currently focused on the production of non-load-bearing implants, such as pieces for middle ear surgery, lling of bone defects in oral or orthopaedic surgery, and coating of dental implants and metallic prosthesis. Nowadays, the synthesis of organicinorganic hybrids is a strong subject of research, as well as the manufacturing of calcium phosphate-based cements, biphasic mixtures that mimic as closely as possible the mineral component of biological apatites, and the production of substrates for cell culture and biochemical factors for tissue engineering. The production of nanostructured materials, similar to the complex hierarchical structures of hard tissues bone and teeth, is also a very attractive eld of research where promising results are already being achieved. Finally, the ability to functionalize surfaces with dierent molecules of varying nature and dimensions, by means of their attachment to the substrate, as well as the potential to physical chemically and topographically nanostructure the surface, will enable in a medium term to act selectively on the biological species, such as proteins, peptides, . . . .

Acknowledgements The authors wish to thank CICYT (Spain) for the nancial support through Research Project MAT2002-0025. The eorts and publications of the research group on calcium phosphates, Departamento de Qumica Inorganica y Bioinorganica, UCM, cited throughout this text, are greatly appreciated. Special thanks to P. Cabanas, F. Conde and J.M. Moreno for their friendship and technical assistance.

References
Vallet-Reg M. Introduction to the world of biomaterials. Ann Quim Int Ed 1997;93:S6S14. Black J, Hastings G, editors. Handbook of biomaterials properties. Chapman and Hall; 1998. Veis A, editor. The chemistry and biology of mineralized connective tissues. Elsevier; 1981, p. 618. Nancollas GH. Biomineralization. In: Mann S, Webb J, Williams RJP, editors. Chemical and biochemical perspectives. VCH; 1989, p. 157. [5] Aoki H. Medical applications of hydroxyapatite. Tokyo, St. Louis: Ishikayu Euro America Inc; 1994. [1] [2] [3] [4]

28

M. Vallet-Regi, J.M. Gonzalez-Calbet / Progress in Solid State Chemistry 32 (2004) 131

[6] Ayer RJ. Electron Microsc Tech 1989;13:16. [7] Shulin W, Jingwei F, Airu L, Tingxing L. Proceedings of the XI International Congress on Electron Microscopy, Kyoto; 1986. p. 30712. [8] Bursill LA, Hutchison JL, Sumida N, Lang AR. Nature 1981;192:518. [9] Nicolopoulos S, Gonzalez-Calbet JM, Alonso MP, Gutierrez-Ros MT, de Frutos MI, Vallet-Reg M. J Solid State Chem 1995;16:265. [10] Nicolopoulos S, Vallet-Reg M, Gonzalez-Calbet JM. Solid State Ionics 1997;175:1013. [11] Cabanas MV, Gonzalez-Calbet JM, Labeau M, Mollard P, Pernet M, Vallet-Reg M. J Solid State Chem 1992;101:265. [12] Vallet-Reg M, Ragel V, Roman J, Martnez JL, Labeau M, Gonzalez-Calbet JM. J Mater Res 1993;8(1):138. [13] Vallet-Reg M, Gutierrez-Ros MT, Alonso MP, de Frutos MI, Nicolopoulos S. J Solid State Chem 1994;112:58. [14] Narasaraju TS, Phebe DE. J Mater Sci 1996;31:1. [15] Suchanec W, Yoshimura M. J Mater Res 1998;13:98. [16] Elliott JC. Studies in inorganic chemistry, vol. 18. Elsevier; 1994. [17] Rodriguez-Lorenzo LM, Vallet-Reg M. Chem Mater 2000;12(8):2460. [18] Pechini MP. US Patent No. 3330697; 1967. [19] Vallet-Reg M. In: Rao KJ, editor. Perspectives in solid state chemistry. India: Narosa Pub. House; 1995, p. 37. [20] Pena J, Vallet-Reg M. J Euro Ceram Soc 2003;23:1687. [21] LeGeros RZ. Monographs in oral science, vol. 15. Basel: Karger; 1991. [22] Elliot JC, editor. Sauramps medical. Montpellier; 1998, p. 2566. [23] Suchanek W, Yoshimura MJ. J Mater Res 1998;13(1):94. [24] Narasaraju TS, Phebe DE. J Mater Sci 1996;31:1. [25] de Groot K. Ceram Int 1993;19:363. [26] Bohner M. J Care Injured 2000;31-S-D37472. [27] Cuneyt Tas A. Biomaterials 2000;21:1429. [28] Andres-Verges M, Fernandez-Gonzalez C, Martnez-Gallego M, Solier I, Cachadina JD, Matijevic EJ. J Mater Res 2000;15(11):2526. [29] Yasukawa A, Matsuura T, Kakajima M, Kandori K, Ishikawa T. Mater Res Bull 1999;24:589. [30] Weng W, Baptista JL. Biomaterials 1998;19:125. [31] Jilavenkatesa A, Condrate RA. J Mater Sci 1998;33:4111. [32] Chai CS, Gross KA, Ben-Nissan B. Biomaterials 1998;19:2291. [33] Layrolle P, Ito A, Tateishi T. J Am Ceram Soc 1998;81(6):1421. [34] Liu DM, Trocynzki T, Tseng W. J Biomater 2001;22:1721. [35] Sampath Kumar TS, Manjubala I, Gunasekaran. J Biomater 2000;21:1623. [36] Fang Y, Agrawal DK, Roy DM, Roy RJ. J Mater Res 1992;7(2):490. [37] Itatani K, Iwafune K, Scott Howellm F, Aizawa M. Mater Res Bull 2000;35:575. [38] Pena J, del Real RP, Rodriguez-Lorenzo LM, Vallet-Reg M. Bioceramics 1999;12:173. [39] Kim W, Zang Q, Saito FJ. J Mater Sci 2000;35:5401. [40] Yeong B, Junmin X, Wang J. J Am Ceram Soc 2001;82:65. [41] Nakano T, Tokumura A, Umakoshi Y, Imazato S, Ehara A, Ebisu S. J Mater Sci Mater Med 2001;12:703. [42] Lim GK, Wang J, Ng SC, Chew CH, Gan LM. Biomaterials 1997;18:1433. [43] Wals D, Mann S. Chem Mater 1996;8:1944. [44] Furuzono T, Walsh D, Sato K, Sonoda K, Tanaka JJ. J Mater Sci Lett 2001;20:111. [45] Aizawa M, Hanazawa T, Itatani K, Howell FS, Kishioka A. J Mater Sci 1999;34:2865. [46] Veilleux D, Barthelemy N, Trombe JC, Verelst M. J Mater Sci 2001;36:2245. [47] Tenhuisen KS, Brown PW. Biomaterials 1998;19:2209. [48] Kim W, Satio F. Ultrasonics Sonochem 2001;8:85. [49] Fang Y, Agrawal DK, Roy DM, Roy R, Brown PW. J Mater Res 1992;7:2294. [50] Suvurova EI, Buat PA. Eur Cells Mater 2001;1:27.

 M. Vallet-Regi, J.M. Gonzalez-Calbet / Progress in Solid State Chemistry 32 (2004) 131

29

[51] Elliot JC, Bond G, Tombe JC. J Appl Crystallogr 1980;13:618. [52] Tadic D, Epple M. Biomaterials 2004;25:987. [53] Okazaki M, Matsumoto T, Taira M, Takakashi J, LeGeros RZ. In: Legeros RZ, LeGeros JP, editors. Bioceramics, vol. 11. New York: World Scientic; 1998, p. 85. [54] Doi Y, Shibutani T, Moriwaki Y, Kajimoto T, Iwayama YJ. J Biomed Mater Res 1998;39:603. [55] Vallet-Reg M, Ramila A, Padilla S, Munoz B. J Biomed Mater 2003. [56] LeGeros RZ. Nature 1965;206:403. [57] Jha LJJ, Best SMJ, Knowles JC, Rehman I, Santos JD, Boneld W. J Mater Sci Mater Med 1997;8:185. [58] Hench LL, Wilson J, Hench LL, Wilson J, editors. An introduction to bioceramics, vol. 20. Boca Raton (FL): World Scientic; 1992. [59] Ohura K, Nakamura T, Yamamuro T, Kokubo T, Ebisawa Y, Kotoura Y, et al. J Biomed Mater Res 1991;25:357. [60] Carlisle EM. Science 1970;167:179. [61] Carlisle EMD. Calcif Tissue Int 1981;33:27. [62] Ruys AJ. J Aust Ceram Soc 1993;29:71. [63] Tanizawa Y, Suzuki T. J Chem Soc Faraday Trans 1995;91:34994350. [64] Boyer L, Carpena J, Lacout JL. Solid State lonics 1997;95:121. [65] Gibson IR, Best SM, Boneld W. J Biomed Mater Res 1999;44:422. [66] Kokubo T, Kushitani H, Sakka S, Kitsugi T, Yamamuro T. J Biomed Mater Res 1990;24:721. [67] Marques PAAP, Magalhaes MCF, Correia RN, Vallet-Reg M. Key Eng Mater 2001;247:1925. [68] Balas F, Perez-Pariente J, Vallet-Reg MJ. Biomed Mat Res 2003;66A:364. [69] Arcos D, Rodriguez-Carvajal J, Vallet-Reg M. Physical B 2004;350:607. [70] Coetzee AS. Arch Otolaryngol 1980;106:405. [71] Lemaitre J, Mirtchi A, Munting E. Sil Ind Ceram Sci Technol 1987;52:141. [72] Chow LC. J.Ceram Soc Jpn 1991;99:954. [73] Sugama T, Allan M. J Am Ceram Soc 1992;75:2076. [74] Mirtchi AA, Lemaitre J, Munting E. Biomaterials 1991;12:505. [75] Otsuka M, Matsuda Y, Suwa Y, Fox JL, Higuchi W. J Biomed Mater Res 1995;29:25. [76] Miyamoto Y, Ishikawa K, Takechi M, Toh T, Yuasa T, Nagayama M, et al. Biomaterials 1998;19:707. [77] del Real RP, Wolke JCC, Vallet-Reg M, Jansen JA. Biomaterials 2002;23:3673. [78] Nilsson M, Fernandez E, Sarda S, Lidgren L, Planell JA. J Biomed Mater Res 2002;61:6007. [79] Constanz BR, Ison IC, Fulmer MT, Fulmer RD, Poser RD, Smith ST, et al. Science 1995;267:1796. [80] Takagi S, Chow LC, Ishikawa K. Biomaterials 1998;9:1593. [81] Pietrzak WS, Ronk R. J Craniofac Surg 2001;11:327. [82] Rawlings III CE, Wilkins RH, Hanker JS, Georgiade NG, Harrelson JM. J Neurosurg 1988;69:269. [83] Sato S, Koshino T, Saito T. Biomaterials 1998;19:1895. [84] Cabanas MV, Rodrguez-Lorenzo LM, Vallet-Reg M. Chem Mater 2002;14:3550. [85] Yu D, Wong J, Matsuda Y, Fox JL, Higuchi WI, Otsuka M. J Pharm Sci 1992;81:529. [86] Hamanishi C, Kitamoto K, Tanaka S, Osuka M, Doi Y, Kitahashi T. J Biomed Mater Res Appl Biomater 1996;33:139. [87] Mousset B, Benoit MA, Delloye C, Bouillet R, Guillard. Int Orthop 1997;21:403. [88] Meseguer L, Ros MJ, Clavel M, Vicente V, Alcaraz M, Lax A, et al. J Biomed Mater Res 2002;61:458. [89] Ratier A, Gibsn IR, Best SM, Freche M, Lacout JL, Rodrguez F. Biomaterials 2001;22:897. [90] Doadrio JC, Arcos D, Cabanas MV, Vallet-Reg M. Biomaterials 2004;25:2629. [91] Daculsi G. Biomaterials 1998;19:1473. [92] Drimandi G, Weiss P, Millot F, Daculsi G. J Biomed Mater Res 1998;39:660. [93] Ragel CV, Vallet-Reg M, Rodriguez-Lorenzo LM. Biomaterials 2002;23:1865. [94] Ramila A, Padilla S, Munoz B, Vallet-Reg M. Chem Mater 2002;14:2439.

30

M. Vallet-Regi, J.M. Gonzalez-Calbet / Progress in Solid State Chemistry 32 (2004) 131

[95] Tancred DC, McCormack BAO, Carr AJ. Biomaterials 1998;19:2303. [96] Bouler JM, Trecant M, Delecrin J, Royer J, Passuti N, Daculsi GJ. J Biomed Mater Res 1996;32:603. [97] Slosarczyk A, Piekarcyk J. Ceram Int 1999;25:561. [98] Kivrak N, Cuneyt Tas A. J Am Ceram Soc 1998;82:2245. [99] Petrov OE, Dyulgerova E, Petrov L, Ropova R. Mater Lett 2001;48:162. [100] Yang X, Wang Z. J Mater Chem 1998;8:2233. [101] Lin FH, Liao CJ, Chen KS, Sun JS, Lin CY. J Biomed Mater Res 2000;51:157. [102] Itatani K, Nishioka T, Seike S, Howell FS, Kishiota A, Kinoshita M. J Am Ceram Soc 1994;77:801. [103] Manjubala I, Sivakimar M. Mater Chem Phys 2001;71:272. [104] Cunneyt Tas A. J Eur Ceram Soc 2000;20:2389. [105] Gauthier O, Bouler JM, Aguado E, LeGeros RZ, Pilet P, Daculsi G. J Mater Sci Mater Med 1999;10:199. [106] Aoki H, Kato K, Ogiso M, Tabata T. J Dent Outlook 1997;49:567. [107] Rodrguez-Lorenzo LM, Vallet-Reg M, Ferreira JMF. Biomaterials 2001;22:583. [108] Rodrguez-Lorenzo LM, Vallet-Reg M, Ferreira JMF. Biomaterials 2001;22:1847. [109] Wan ACA, Khor G, Hasting GW. J Biomed Mater Res 1998;41:541. [110] Du C, Cui FZ, Feng QL, Zhu XD, de Groot K. J Biomed Mater Res 1999;44:407. [111] Tadic D, Beckmann F, Schwarz K, Epple M. Biomaterials 2004;25:3335. [112] Rodrguez-Lorenzo LM, Vallet-Reg M, Ferreira JMF. J Biomed Mater Res 2002;60:232. [113] Padilla S, Roman J, Vallet-Reg M. J Mater Sci: Mater Med 2002;13:1193. [114] Dupraz A, Nguyen TP, Richard M, Daculsi G, Passuti N. Biomaterials 1999;20:663. [115] Pena J, Vallet-Reg M, San Roman J. J Biomed Mat Res 1997;35:129. [116] Rodriguez-Lorenzo LM, Salinas AJ, Vallet-Reg M, San Roman J. J Biomed Mat Res 1996;30:515. [117] Wang M, Kokubo T, Boneld W. In Kokubo T, Nakamura T, Miyaji F, editors. Bioceramics, vol. 9. Japan; 1996. p. 387. [118] Kikuchi M, Sato K, Suetsugu Y, Tanaka J. In: LeGeros RZ, LeGeros JP, editors. Bioceramics, vol. 11. NY: World Scientic; 1998, p. 153. [119] Bakos D, Soldan M, Hernandez-Fuentes I. Biomaterials 1999;20:191. [120] Cherng A, Takagi S, Chow LC. J Biomed Mater 1997;35:273. [121] Constantino PD, Friedman CD. Clin North Am 1996;27:1037. [122] de Groot K, Gessink R, Klein CPAT, Serekain P. J Biomed Mater Res 1987;21:1375. [123] Park E, Condrate RA, Hoelzer DT, Fischman GS. J Mat Sci Mater Med 1998;9:643. [124] MacDonal DE, Betts F, Stranick M, Doty S, Boskey AL. J Biomed Mater Res 2001;54:480. [125] Van Dijk K, Schaehen HG, Wolke JGC, Jansen JA. Biomaterials 1996;17:405. [126] Zeng HT, Laceeld WR. Biomaterials 2000;21:23. [127] Cui FZ, Lou ZS, Feng QL. J Mat Sci Mater Med 1997;8:403. [128] Spoto G, Ciliberto E, Allen GC. J Mater Chem 1994;4:1849. [129] Cao Y, Weng J, Chen J, Fen J, Yang Z, Zhang X. Biomaterials 1996;17:419. [130] Chou L, Marek B, Wagner WR. Biomaterials 1999;20:977. [131] Labeau M, Gautheron B, Cellier F, Vallet-Reg M, Garca E, Gonzalez-Calbet JM. J Solid State Chem 1993;102:434. [132] Martnez A, Pena J, Labeau M, Gonzalez-Calbet JM, Vallet-Reg M. J Mater Res 1995;10:1307. [133] Patil PS. Mater Chem Phys 1999;59:185. [134] Barrere F, Layrolle P, Van Blitterswijk CA, de Groot K, Laydolle P. J Am Ceram Soc 2002;85:517. [135] Choi J, Bogdanski D, Koller M, Esenwein SA, Muller D, Muhr G, et al. Biomaterials 2003;24:3689. [136] Ozawa N, Ideta Y, Yao T, Kokubo T. Bioceramics 2003;15:71. [137] Vallet-Reg M, Izquierdo-Barba I, Gil FJ. J Biomed Mater Res 2003;67:674. [138] Hijon N, Cabanas MV, Izquierdo-Barba I, Vallet-Reg M. Chem Mater 2004;16:1451.

 M. Vallet-Regi, J.M. Gonzalez-Calbet / Progress in Solid State Chemistry 32 (2004) 131 [139] [140] [141] [142] [143] [144] [145] [146] [147] [148] [149] [150] [151] [152] [153]

31

Cabanas MV, Vallet-Reg MJ. Mater Chem 2003;13:1104. Lin S, LeGeros RZ, LeGeros JP. J Biomed Mater Res 2003;66:819. Cotell CM. Appl Surface Sci 1993;69:140. Suchanek W, Yoshimur MJ. J Mater Res 1998;13:94. Tkalcec E, Sauer M, Nonninger R, Schmidt H. J Mater Sci 2001;36:5253. Layrolle P, Lebugle A. Chem Mater 1994;6:1996. Jillavenkatesa A, Condrate Sr RA. J Mater Sci 1998;33:4111. Ben-Nissan B, Green DD, Kannangara GSK, Chai CS, Mile A. J SolGel Sci Tech 2001;21:27. Bezzi A, Celotti G, Landi E, La Torretta TMG, Sopyan I, Tampieri A. Mater Chem Phys 2003;78:816. Livage J, Barboux P, Vandenborre MT, Schmutz C, Taulelle F. J Non Cryst Solids 1992;147 148:18. Chai CS, Gross KA, Ben-Nissan B. Biomaterials 1998;19:2291. Hsieh M, Perng L, Chin T, Perng H. Biomaterials 2001;22:2601. Liu D, Troczynski T, Tseng WJ. Biomaterials 2002;23:1227. Hijon N, Cabanas MV, Izquierdo-Barba I, Vallet-Reg M. Key Eng Mater 2004;254256:363. Vallet-Reg M. Perspective article. J Chem Soc Dalton Trans 2001;2:97.