3.2. Polymers

Polymers are a class of materials consisting of large molecules, often containing many thousands of small units, or monomers, joined together chemically to form one giant chain. In this respect, polymers are comparable with major functional components of the biological environment: lipids, proteins and polysaccharides. They differ from each other in chemical composition, molecular weight, polydispersity, crystallinity, hydrophobicity, solubility and thermal transitions. Besides, their properties can be fine-tuned over a wide range by varying the type of polymer, chain length, as well as by copolymerization or blending of two or more polymers [79,80]. Opposite to ceramics, polymers exhibit substantial viscoelastic properties and easily can be fabricated into complex structures, such as sponge-like sheets, gels or complex structures with intricate porous networks and channels [81]. Being X-ray transparent and non-magnetic polymeric materials are fully compatible with the modern diagnostic methods such as computed tomography and magnetic resonance imaging. Unfortunately, most of them are unable to meet the strict demands of the in vivo physiological environment. Namely, the main requirements to polymers suitable for biomedical applications are that they must be biocompatible, not eliciting an excessive or chronic inflammatory response upon implantation and, for those that degrade, that they breakdown into non-toxic products only. Unfortunately, polymers, for the most part, lack rigidity, ductility and ultimate mechanical properties required in load bearing applications. Thus, despite their good biocompatibility, many of the polymeric materials are mainly used for soft tissue replacements (such as skin, blood vessel, cartilage, ligament replacement, etc). Moreover, the sterilization processes (autoclave, ethylene oxide and ⁶°Co irradiation) may affect the polymer properties [82].

There is a variety of biocompatible polymers suitable for biomedical applications [83,84,85]. For example, polyacrylates, poly(acrylonitrile-co-vinylchloride) and polylysine have been investigated for cell encapsulation and immunoisolation [86,87]. Polyorthoesters and PCL have been investigated as drug delivery devices, the latter for long-term sustained release because of their slow degradation rates [88]. PCL is a hydrolytic polyester having appropriate resorption period and releases nontoxic byproducts upon degradation [89,90]. PU is in use in engineering of both hard and soft tissues, as well as in nanomedicine [91]. Polymers considered for orthopedic purposes include polyanhydrides, which have also been investigated as delivery devices (due to their rapid and well-defined surface erosion), for bone augmentation or replacement since they can be photopolymerized in situ [88,92,93]. To overcome their poor mechanical properties, they have been co-polymerized with imides or formulated to be crosslinkable in situ [93]. Other polymers, such as polyphosphazenes, can have their properties (e.g., degradation rate) easily modified by varying the nature of their side groups and have been shown to support osteoblast adhesion, which makes them candidate materials for skeletal tissue regeneration [93]. PPF has emerged as a good bone replacement material, exhibiting good mechanical properties (comparable to trabecular bone), possessing the capability to crosslink in vivo through the C=C bond and being hydrolytically degradable. It has also been examined as a material for drug delivery devices [88,92,93,94,95]. Polycarbonates have been suggested as suitable materials to make scaffolds for bone replacement and have been modified with tyrosine-derived amino acids to render them biodegradable [88,96]. Polydioxanone has been also tested for biomedical applications [97]. PMMA is widely used in orthopedics, as a bone cement for implant fixation, as well as to repair certain fractures and bone defects, for example, osteoporotic vertebral bodies [98,99]. However, PMMA sets by a polymerization of toxic monomers, which also evolves significant amounts of heat that damages tissues. Moreover, it is neither degradable nor bioactive, does not bond chemically to bones and might generate particulate debris leading to an inflammatory foreign body response [92,100]. A number of other non-degradable polymers applied in orthopedic surgery include PE in its different modifications such as low density PE, HDPE and ultrahigh molecular weight PE (used as the articular surface of total hip replacement implants [101,102]), polyethylene terepthalate and PP, which are applied to repair knee ligaments [103]. Polyactive™, a block copolymer of PEG and PBT, was also considered for biomedical application [104,105,106]. Cellulose [107,108] and its esters [109,110] are also popular. Finally yet importantly, polyethylene oxide, PHB and blends thereof have also been tested for biomedical applications [29].

Nonetheless, the most popular synthetic polymers used in medicine are the linear aliphatic poly(α-hydroxyesters) such as PLA, PGA and their copolymers—PLGA (Table 4). These materials have been extensively studied; they appear to be the only synthetic and biodegradable polymers with an extensive FDA approval history [29,93,111,112]. They are biocompatible, mostly non-inflammatory, as well as degrade in vivo through hydrolysis and possible enzymatic action into products that are removed from the body by regular metabolic pathways [88,93,113]. Besides, they might be used for drug delivery purposes [114]. Poly(α-hydroxyesters) have been investigated as scaffolds for replacement and regeneration of a variety of tissues, cell carriers, controlled delivery devices for drugs or proteins (e.g., growth factors), membranes or films, screws, pins and plates for orthopedic applications [88,93,115,116]. Additionally, the degradation rate of PLGA can be adjusted by varying the amounts of the two component monomers (Table 4), which in orthopedic applications can be exploited to create materials that degrade in concert with bone ingrowth [117]. Furthermore, PLGA is known to support osteoblast migration and proliferation [93,118], which is a necessity for bone tissue regeneration. Unfortunately, such polymers on their own, though they reduce the effect of stress-shielding, are too weak to be used in load bearing situations and are only recommended in certain clinical indications, such as ankle and elbow fractures [113]. In addition, they exhibit bulk degradation, leading to both a loss in mechanical properties and lowering of the local solution pH that accelerates further degradation in an autocatalytic manner. As the body is unable to cope with the vast amounts of implant degradation products, this might lead to an inflammatory foreign body response. Finally, poly(α-hydroxyesters) do not possess the bioactive and osteoconductive properties [93,119].

Several classifications of the biomedically relevant polymers are possible. For example, some authors distinguish between synthetic polymers like PE, PMMA, PLA, PGA, PCL, etc., and polymers of biological origin, which comprise polysaccharides (starch, alginate, chitin/chitosan [120,121], gellan gum, cellulose, hyaluronic acid and its derivatives), proteins (soy, collagen, gelatin, fibrin, silk) and a variety of biofibers, such as lignocellulosic natural fibers [122,123]. Among them, natural polymers often posses highly organized structures and may contain an extracellular substance, called ligand, which is necessary to bind with cell receptors. However, they always contain various impurities, which should be removed prior use. As synthetic polymers can be produced under the controlled conditions, they in general exhibit predictable and reproducible mechanical and physical properties such as tensile strength, elastic modulus and degradation rate. Control of impurities is a further advantage of synthetic polymers. Other authors differentiate between resorbable or biodegradable (e.g., poly(α-hydroxyesters), polysaccharides and proteins) and non-resorbable (e.g., PE, PP, PMMA and cellulose) polymers [123]. Furthermore, polymeric materials can be broadly classified as thermoplastics and thermosets. For example, HDPE and PEEK are the examples of thermoplastics, while polydimethylsiloxane and PMMA are the examples of thermosets [82]. The list of synthetic biodegradable polymers used for biomedical application as scaffold materials is available as Table 1 in Ref. [123], while further details on polymers suitable for biomedical applications are available in literature [82,116,124,125,126,127,128,129] where the interested readers are referred to. Good reviews on the synthesis of different biodegradable polymers [130], as well as on the experimental trends in polymer composites [131] are available elsewhere.

4.1. Biocomposites with Polymers

Typically, the polymeric components of biocomposites and hybrid biomaterials comprise polymers that both have shown a good biocompatibility and are routinely used in surgical applications. In general, since polymers have a low modulus (2–7 GPa, as the maximum) as compared to that of bone (3–30 GPa), CaPO₄ bioceramics need to be loaded at a high weight % ratio. Besides, general knowledge on composite mechanics suggests that any high aspect ratio particles, such as whiskers or fibers, significantly improve the modulus at a lower loading. Thus, some attempts have been already performed to prepare biocomposites containing whisker-like [167,168,169,170,171,172,173] or needle-like [174,175,176,177] CaPO₄, as well as CaPO₄ fibers [178].

The history of implantable CaPO₄/polymer formulations started in 1981 (however, a more general topic “ceramic-plastic material as a bone substitute” is, at least, 18 years older [179]) from the pioneering study by Prof. William Bonfield and colleagues at Queen Mary College, University of London, performed on HA/PE blends [180,181]. That initial study introduced a bone-analogue concept, when proposed biocomposites comprised a polymer ductile matrix of PE and a ceramic stiff phase of HA, and was substantially extended and developed in further investigations by that research group [66,182,183,184,185,186,187,188,189,190,191,192]. More recent studies included investigations on the influence of surface topography of HA/PE composites on cell proliferation and attachment [193,194,195,196]. The material is composed of a particular combination of HA particles at a volume loading of ~40% uniformly dispensed in a HDPE matrix. The idea was to mimic bones by using a polymeric matrix that can develop a considerable anisotropic character through adequate orientation techniques reinforced with a bone-like bioceramics that assures both a mechanical reinforcement and a bioactive character of the composite. Following FDA approval in 1994, in 1995 this material has become commercially available under the trade-name HAPEX™ (Smith and Nephew Richards, Bartlett, TN, USA), and to date it has been implanted in over 300,000 patients with the successful results. It remains the only clinically successful bioactive composite, which was a major step in the implant field [148,197]. The major production stages of HAPEX™ include blending, compounding and centrifugal milling. A bulk material or device is then created from this powder by compression and injection molding [37]. Besides, HA/HDPE biocomposites might be prepared by a hot rolling technique that facilitated uniform dispersion and blending of the reinforcements in the matrix [198]. In addition, PP might be used instead of PE [199,200,201,202].

A mechanical interlock between the both phases of HAPEX™ is formed by shrinkage of HDPE onto the HA particles during cooling [66,67,203]. Both HA particle size and their distribution in the HDPE matrix were recognized as important parameters affecting the mechanical behavior of HAPEX™. Namely, smaller HA particles were found to lead to stiffer composites due to general increasing of interfaces between the polymer and the ceramics; furthermore, rigidity of HAPEX™ was found to be proportional to HA volume fraction [187]. Furthermore, coupling agents, e.g., 3-trimethoxysiyl propylmethacrylate for HA and acrylic acid for HDPE might be used to improve bonding (by both chemical adhesion and mechanical coupling) between HA and HDPE [204,205]. Obviously, other types of CaPO₄ might be used instead of HA in biocomposites with PE [206]. Furthermore, attempts were performed to improve the mechanical properties of HAPEX™ by incorporating other ceramic phases into the polymer matrix, such as PSZ [207] and alumina [208]. For example, a partial replacement of HA filler particles by PSZ particles was found to lead to an increase in the strength and fracture toughness of HA/HDPE biocomposites. The compressive stress, set up by the volume expansion associated with tetragonal to monoclinic phase transformation of PSZ, inhibits or retards the crack propagation within the composite. This results in an enhanced fracture toughness of the HA/ZrO₂/HDPE biocomposite [207].

Various studies revealed that HAPEX™ attached directly to bones by chemical bonding (a bioactive fixation), rather than forming fibrous encapsulation (a morphological fixation). Initial clinical applications of HAPEX™ came in orbital reconstruction [209] but since 1995, the main uses of this composite have been in the shafts of middle ear implants for the treatment of conductive hearing loss [210,211]. In both applications, HAPEX™ offers the advantage of in situ shaping, so a surgeon can make final alterations to optimize the fit of the prosthesis to the bone of a patient and subsequent activity requires only limited mechanical loading with virtually no risk of failure from insufficient tensile strength [66,67,149]. As compared to cortical bones, HA/PE composites have a superior fracture toughness for HA concentrations below ~40% and similar fracture toughness in the 45%–50% range. Their Young’s modulus is in the range of 1–8 GPa, which is quite close to that of bone. The examination of the fracture surfaces revealed that only mechanical bond occurs between HA and PE. Unfortunately, the HA/PE composites are not biodegradable, the available surface area of HA is low and the presence of bioinert PE decreases the ability to bond to bones. Furthermore, HAPEX™ has been designed with a maximized density to increase its strength but the resulting lack of porosity limits the ingrowth of osteoblasts when the implant is placed into the body [23,150]. Further details on HAPEX™ are available elsewhere [66,67]. Except of HAPEX™, other types of both HA/PE [212,213,214,215,216,217,218,219,220] and HA/undisclosed polymer (HA^nano Bone, Promimic AB, Göteborg, Sweden) biocomposites are also known.

Both linear and branched PE was used as a matrix and the biocomposites with the former were found to give a higher modulus [213]. The reinforcing mechanisms in CaPO₄/polymer formulations have yet to be convincingly disclosed. Generally, if a poor filler choice is made, the polymeric matrix might be affected by the filler through reduction of molecular weight during composite processing, formation of an immobilized shell of polymer around the particles (transcrystallization, surface-induced crystallization or epitaxial growth) and changes in conformation of the polymer due to particle surfaces and inter-particle spacing [66,67]. On the other hand, the reinforcing effect of CaPO₄ particles might depend on the molding technique employed: a higher orientation of the polymeric matrix was found to result in a higher mechanical performance of the composite [218,219].

Many other blends of CaPO₄ with various polymers are possible, including rather unusual formulations with dendrimers [221]. Even light-curable CaPO₄/polymer formulations are known [222]. The list of the appropriate CaPO₄ is shown in Table 3 (except of MCPM and MCPA – both are too acidic and, therefore, are not biocompatible [16]; nevertheless, to overcome this drawback, they might be mixed with basic compounds, such as HA, TTCP, CaCO₃, CaO, etc.) many biomedically suitable polymers have been listed above. The combination of CaPO₄ and polymers into biocomposites has a twofold purpose. The desirable mechanical properties of polymers compensate for a poor mechanical behavior of CaPO₄ bioceramics, while in turn the desirable bioactive properties of CaPO₄ improve those of polymers, expanding the possible uses of each material within the body [223,224,225,226]. Namely, polymers have been added to CaPO₄ in order to improve their mechanical strength [223], while CaPO₄ fillers have been blended with polymers to improve their compressive strength and modulus, in addition to increasing their osteoconductive properties [119,227,228,229,230]. In 1990-s, it was established that with increasing of CaPO₄ content, both Young’s modulus and bioactivity of the biocomposites generally increased, while the ductility decreased [23]. However, the later investigations revealed that the mechanical properties of CaPO₄/polymer biocomposites were not so straightforward: the strength was found to decrease with increasing the CaPO₄ content in such biocomposites [231]. Nevertheless, biocompatibility of such biocomposites is enhanced because CaPO₄ fillers induce an increased initial flash spread of serum proteins compared to the more hydrophobic polymer surfaces [232]. What’s more, experimental results of these biocomposites indicated favorable cell-material interactions with increased cell activities as compared to each polymer alone [225]. Furthermore, such formulations can provide a sustained release of calcium and orthophosphate ions into the milieus, which is important for mineralized tissue regeneration [224]. Indeed, a combination of two different materials draws on the advantages of each one to create a superior biocomposite with respect to the materials on their own.

It is logical to assume that the proper biocomposite of a CaPO₄ (for instance, CDHA) with a bioorganic polymer (for instance, collagen) would yield the physical, chemical and mechanical properties similar to those of human bones. Different ways have been already realized to bring these two components together into biocomposites, like mechanical blending, compounding, ball milling, dispersion of ceramic fillers into a polymer-solvent solution, a melt extrusion of a ceramic/polymer powder mixture, co-precipitation and electrochemical co-deposition [22,37,233,234,235]. Three methods for preparing a homogeneous blend of HA with PLLA were compared [233]. A dry process, consisting in mixing ceramic powder and polymer pellets before a compression-molding step, was used. The second technique was based on the dispersion of ceramic fillers into a polymer-solvent solution. The third method was a melt extrusion of a ceramic/polymer powder mixture. Mixing dry powders led to a ceramic particle network around the polymer pellets, whereas the solvent and melt methods also produced a homogeneous dispersion of HA in the matrix. The main drawback of the solvent casting method is the risk of potentially toxic organic solvent residues. The melt extrusion method was shown to be a good way to prepare homogeneous ceramic/polymer blends [233].

Besides, there is in situ formation, which involves either synthesizing the reinforcement inside a preformed matrix material or synthesizing the matrix material around the reinforcement [37,236,237]. This is one of the most attractive routes, since it avoids extensive particle agglomeration. For example, several papers have reported in situ formation technique to produce various composites of CaPO₄ with carbon nanotubes [238,239,240,241]. Other examples comprise using amino acid-capped gold nano-sized particles as scaffolds to grow CDHA [242] and preparation of nano-sized HA/PA biocomposites [243,244]. In certain cases, a mechano-chemical route [245,246], emulsions [247,248,249,250,251,252], freeze-drying [253,254] and freeze-thawing techniques [255] or gel-templated mineralization [256] might be applied to produce CaPO₄-based biocomposites. Various fabrication procedures are well described elsewhere [22,37,233], where the interested readers are referred.

The interfacial bonding between the phases is an important issue of any CaPO₄/polymer biocomposite. Four types of mutual arrangements of nanodimensional particles to polymer chains have been classified by Kickelbick (Figure 1): (1) inorganic particles embedded in inorganic polymer, (2) incorporation of particles by bonding to the polymer backbone, (3) an interpenetrating network with chemical bonds, (4) an inorganic-organic hybrid polymer [257]. If adhesion among the phases is poor, the mechanical properties of a biocomposite suffer. To solve the problem, various approaches have been already introduced. For example, a diisocyanate coupling agent was used to bind PEG/PBT (Polyactive™) block copolymers to HA filler particles. Using surface-modified HA particles as a filler in a PEG/PBT matrix significantly improved the elastic modulus and strength of the polymer as compared to the polymers filled with ungrafted HA [228,258]. Another group used processing conditions to achieve a better adhesion of the filler to the matrix by pressing blends of varying PLLA and HA content at different temperatures and pressures [259]. The researchers found that maximum compressive strength was achieved at ~15 wt % of PLLA. By using blends with 20 wt % of PLLA, the authors also established that increasing the pressing temperature and pressure improved the mechanical properties. The former was explained by decrease in viscosity of the PLLA associated with a temperature increase, hence leading to improved wettability of HA particles. The latter was explained by increased compaction and penetration of pores at higher pressure, in conjunction with a greater fluidity of the polymer at higher temperatures. The combination of high pressures and temperatures was found to decrease porosity and guarantee a close apposition of a polymer to the particles, thereby improving the compressive strength [223] and fracture energy [260] of the biocomposites. The PLLA/HA biocomposites scaffolds were found to improve cell survival over plain PLLA scaffolds [261].

Open in a separate window

Figure 1

Four types of mutual arrangements of nano-sized particles to a polymer chain: (1) inorganic particles embedded in an inorganic polymer, (2) incorporation of particles by bonding to the polymer backbone, (3) interpenetrating network with chemical bonds, (4) inorganic-organic hybrid polymer. Reprinted from Ref. [257] with permission.

It is also possible to introduce porosity into CaPO₄-based biocomposites, which is advantageous for most applications as bone substitution material. The porosity facilitates migration of osteoblasts from surrounding bones to the implant site. Various material processing strategies to prepare composite scaffolds with interconnected porosity comprise thermally induced phase separation, solvent casting and particle leaching, solid freeform fabrication techniques, microsphere sintering and coating [123,262,263,264,265]. A supercritical gas foaming technique might be used as well [233,266,267].

4.1.1. Apatite-Based Formulations

A biological apatite is known to be the major inorganic phase of mammalian calcified tissues [14,15]. Consequently, CDHA, HA, carbonateapatite (both with and without dopants) and, occasionally, FA have been applied to prepare biocomposites with other compounds, usually with the aim to improve the bioactivity. For example, polysulfone composed with HA can be used as a starting material for long-term implants [268,269,270]. Retrieved in vivo, HA/polysulfone biocomposite coated samples from rabbit distal femurs demonstrated direct bone apposition to the coatings, as compared to the fibrous encapsulation that occurred when uncoated samples were used [268]. The resorption time of such biocomposites is a very important factor, which depends on polymer’s microstructure and the presence of modifying phases [269].

Various apatite-containing biocomposites with PVA [255,271,272,273,274,275], PVAP [276] and several other polymeric components [277,278,279,280,281,282,283,284,285,286] have been already developed. Namely, PVA/CDHA biocomposite blocks were prepared by precipitation of CDHA in aqueous solutions of PVA [255]. An artificial cornea consisted of a porous nano-sized HA/PVA hydrogel skirt and a transparent center of PVA hydrogel has been prepared as well. The results displayed a good biocompatibility and interlocking between artificial cornea and host tissues [271,272]. PVAP has been chosen as a polymer matrix, because its phosphate groups can act as a coupling/anchoring agent, which has a higher affinity toward the HA surface [276]. Greish and Brown developed HA/Ca poly(vinyl phosphonate) biocomposites [280,281,282]. A template-driven nucleation and mineral growth process for the high-affinity integration of CDHA with PHEMA hydrogel scaffold has been developed as well [285].

PEEK [167,168,287,288,289,290,291,292,293] and high impact polystyrene [294,295] were also applied to create biocomposites with HA having a potential for clinical use in load bearing applications. The study on reinforcing PEEK with thermally sprayed HA particles revealed that the mechanical properties increased monotonically with the reinforcement concentration, with a maximum value in the study of ~40% volume fraction of HA particles [287,288,289]. The reported ranges of stiffness within 2.8–16.0 GPa and strength within 45.5–69 MPa exceeded the lower values for human bone (7–30 GPa and 50–150 MPa, respectively) [288]. Modeling of the mechanical behavior of HA/PEEK biocomposites is available elsewhere [290].

Biodegradable poly(α-hydroxyesters) are well established in clinical medicine. Currently, they provide with a good choice when a suitable polymeric filler material is sought. For example, HA/PLGA formulations were developed which appeared to possess a cellular-compatibility suitable for bone tissue regeneration [296,297,298,299,300,301,302,303,304]. Zhang and Ma seeded highly porous PLLA foams with HA particles in order to improve the osteoconductivity of polymer scaffolds for bone tissue engineering [227]. They pointed out that hydration of the foams prior to incubation in simulated body fluid increased the amount of carbonated CDHA material due to an increase of COOH and OH groups on the polymer surface, which apparently acted as nucleation sites for apatite. The mechanical properties of PLA/CaPO₄ biocomposites fabricated using different techniques, as well as the results of in vitro and in vivo experiments with them are available in literature [300].

On their own, poly(α-hydroxyesters), such as PGA and PLA, are known to degrade to acidic products (glycolic and lactic acids, respectively) that both catalyze polymer degradation and cause inflammatory reactions of the surrounding tissues [305]. Thus, in biocomposites of poly(α-hydroxyesters) with CaPO₄, the presence of slightly basic compounds (HA, TTCP) to some extent neutralizes the acid molecules, provides with a weak pH-buffering effect at the polymer surface and, therefore, more or less compensates these drawbacks [119,300,306,307,308]. However, additives of even more basic chemicals (e.g., CaO, CaCO₃) might be necessary [123,308,309,310]. Extensive cell culture experiments on pH-stabilized composites of PGA and carbonateapatite were reported, which afterwards were supported by extensive in vitro pH-studies [311]. A consequent development of this approach has led to designing of functionally graded composite skull implants consisting of polylactides, carbonateapatite and CaCO₃ [312,313]. Besides the pH-buffering effect, inclusion of CaPO₄ was found to modify both surface and bulk properties of the biodegradable poly(α-hydroxyesters) by increasing the hydrophilicity and water absorption of the polymer matrix, thus altering the scaffold degradation kinetics. For example, polymer biocomposites filled with HA particles was found to hydrolyze homogeneously due to water penetrating into interfacial regions [314].

Biocomposites of poly(α-hydroxyesters) with CaPO₄ are prepared mainly by incorporating the inorganic phase into a polymeric solution, followed by drying under vacuum. The resulting solid biocomposites might be shaped using different processing techniques. One can also prepare these biocomposites by mixing HA particles with L-lactide prior the polymerization [306] or by a combination of slip-casting technique and hot pressing [315]; however, other production techniques are known [300,302,316]. Addition of a surfactant (surface active agent) might be useful to keep the suspension homogeneity [317]. Furthermore, HA/PLA [248,249] and HA/PLGA [250] microspheres might be prepared by a microemulsion technique. More complex formulations, such as carbonated-FA/PLA [318] and PLGA/carbon nanotubes/HA [319], are also known. An interesting list of references, assigned to the different ways of preparing HA/poly(α-hydroxyesters) biodegradable composites, might be found in publications by Durucan and Brown [320,321,322]. The authors prepared CDHA/PLA and CDHA/PLGA biocomposites by solvent casting technique with a subsequent hydrolysis of α-TCP to CDHA in aqueous solutions. The presence of both polymers was found to inhibit α-TCP hydrolysis, if compared with that of single-phase α-TCP; what is more, the inhibiting effect of PLA exceeded that of PLGA [320,321,322]. The physical interactions between CaPO₄ and poly(α-hydroxyesters) might be easily seen in Figure 2 [322]. Another set of good pictures might be found in Ref. [51]. Nevertheless, it should not be forgotten that typically non-melt based routes lead to development of composites with lower mechanical performance and many times require the use of toxic solvents and intensive hand labor [125].

Open in a separate window

Figure 2

SEM micrographs of (a) α-Tricalcium phosphate (α-TCP) compact; (b) α-TCP/PLGA biocomposite (bars = 5 µm). Reprinted from Ref. [322] with permission.

The mechanical properties of poly(α-hydroxyesters) could be substantially improved by addition of CaPO₄ [323,324]. Namely, CDHA/PLLA biocomposites of very high mechanical properties were developed [119] and fixation tools (screws and plates) made of these composites were manufactured and tested. These fixation tools revealed an easy handling and shaping according to the implant site geometry, total resorbability, good ability to bond directly to the bone tissue without interposed fibrous tissue, osteoconductivity, biocompatibility and high stiffness retainable for the period necessary to achieve bone union [314,316]. The initial bending strength of ~280 MPa exceeded that of cortical bone (120–210 MPa), while the modulus was as high as 12 GPa [119]. The strength could be maintained above 200 MPa up to 25 weeks in phosphate-buffered saline solution. Such biocomposites could be obtained by precipitation from PLLA/dichloromethane solutions, in which small particles of CaPO₄ were distributed [118,325]. Porous scaffolds of PDLLA + HA [267,326,327] and PLGA + HA [328] have been manufactured as well. Upon implantation into rabbit femora, a newly formed bone was observed and biodegradation was significantly enhanced if compared to single-phase HA bioceramics. This might be due to a local release of lactic acid, which in turn dissolves HA. In other studies, PLA and PGA fibers were combined with porous HA scaffolds. Such reinforcement did not hinder bone ingrowth into the implants, which supported further development of such biocomposites as bone graft substitutes [29,300,301].

Blends (named as SEVA-C) of EVOH with starch filled with 10–30 wt % HA have been fabricated to yield biocomposites with modulus up to ~7 GPa with a 30% HA loading [329,330,331,332,333]. The incorporation of bioactive fillers such as HA into SEVA-C aimed to assure the bioactive behavior of the composite and to provide the necessary stiffness within the typical range of human cortical bone properties. These biocomposites exhibited a strong in vitro bioactivity that was supported by the polymer’s water-uptake capability [334]. However, the reinforcement of SEVA-C by HA particles was found to affect the rheological behavior of the blend. A degradation model of these biocomposites has been developed [335].

Higher homologues poly(3-hydroxybutyrate), 3-PHB, and poly(3-hydroxyvalerate) show almost no biodegradation. Nevertheless, biocomposites of these polymers with CaPO₄ showed a good biocompatibility both in vitro and in vivo [336,337,338,339,340]. Both bioactivity and mechanical properties of these biocomposites can be tailored by varying the volume percentage of CaPO₄. Similarly, biocomposites of PHBHV with both HA and amorphous carbonated apatite (almost ACP) appeared to have a promising potential for repair and replacement of damaged bones [341,342,343,344].

Along this line, PCL is used as a slowly biodegradable but well biocompatible polymer. PCL/HA and PCL/CDHA biocomposites have been already discussed as suitable materials for substitution, regeneration and repair of bone tissues [262,345,346,347,348,349,350,351,352]. For example, biocomposites were obtained by infiltration of ε-caprolactone monomer into porous apatite blocks and in situ polymerization [346]. The composites were found to be biodegradable and might be applied as cancellous or trabecular bone replacement material or for a cartilage regeneration. Both the mechanical performance and biocompatibility in osteoblast cell culture of PCL were shown to be strongly increased when HA was added [353]. Several preparation techniques of PCL/HA biocomposites are known [262,349]. For example, to make biocomposite fibers of PCL with nanodimensional HA, the desired amount of nanodimensional HA powder was dispersed in a solvent using magnetic stirrer followed by ultrasonication for 30 min. Then, PCL was dissolved in this suspension, followed by the solvent evaporation [354]. The opposite preparation order is also possible: PCL was initially dissolved in chloroform at room temperature (7%–10% w/v), then HA (~10 µm particle size) was suspended in the solution, sonicated for 60 s, followed by the solvent evaporation [355] or salt-leaching [356]. The mechanical properties obtained by this technique were about one-third that of trabecular bone. In a comparative study, PCL and biological apatite were mixed in the ratio 19:1 in an extruder [357]. At the end of the preparation, the mixture was cooled in an atmosphere of nitrogen. The authors observed that the presence of biological apatite improved the modulus while concurrently increasing the hydrophilicity of the polymeric substrate. Besides, an increase in apatite concentration was found to increase both the modulus and yield stress of the composite, which indicated to good interfacial interactions between the biological apatite and PCL. It was also observed that the presence of biological apatite stimulated osteoblasts attachment to the biomaterial and cell proliferation [357]. In another study, a PCL/HA biocomposite was prepared by blending in melt form at 120 °C until the torque reached equilibrium in the rheometer that was attached to the blender [358]. Then the sample was compression molded and cut into specimens of appropriate size for testing. It was observed that the composite containing 20 wt % HA had the highest strength [358]. However, a direct grafting of PCL on the surface of HA particles seems to be the most interesting preparation technique [345]. In another study, HA porous scaffolds were coated by a PCL/HA composite coating [359]. In this system, PCL, as a coating component, was able to improve the brittleness and low strength of the HA scaffolds, while the particles in the coating were to improve the osteoconductivity and bioactivity of the coating layer. More complex formulations, such as PDLLA/PCL/HA [360], PLLA/PCL/HA [361], FA-HA/PCL [362], magnetic PCL/Fe-doped HA [363] and supramolecular PCL/functionalized HA [364,365] biocomposites, have been prepared as well. Further details on both the PCL/HA biocomposites and the processing methodologies thereof might be found elsewhere [262,349].

An interesting phenomenon of fractal growth of FA/gelatin composite crystals (Figure 3) was achieved by diffusion of calcium- and orthophosphate + fluoride-containing solutions from the opposite sides into a tube filled with a gelatin gel [366,367,368,369,370,371,372]. The reasons of this phenomenon are not quite clear yet. Other types of CaPO₄-based composites, based on DCPD and OCP, were grown by the similar technique in an iota-carrageenan gel [373]. Up to now, nothing has yet been reported on a possible biomedical application of such unusual structural composites.

Open in a separate window

Figure 3

A biomimetically grown aggregate of fluorapatite (FA) that was crystallized in a gelatin matrix. Its shape can be explained and simulated by a fractal growth mechanism. Scale bar: 10 μm. Reprinted from Ref. [366] with permission.

To finalize this section, one should note that there are polymers possessing shape-memory properties [374] and, therefore, CaPO₄-based composites with such polymers appear to have the shape-memory properties as well [375,376,377].

4.2. Self-Setting Formulations

Inorganic self-setting CaPO₄ formulations (cements) hardening in the body were introduced in the early 1980-s. Since then, they have been broadly studied and many formulations have been proposed. These formulations set and harden due to various chemical interactions among CaPO₄ that finally lead to formation of a monolithic body consisting of either CHDA or DCPD with possible admixtures of other phases. Unfortunately, having a ceramic nature, the self-setting CaPO₄ formulations are brittle after hardening and the setting time is sometimes unsuitable for clinical procedures [442]. Therefore, to improve their properties, various attempts have been performed to transform them into biocomposites, e.g., by adding hydroxycarboxylic acids [443,444,445], gelatin [389,446,447,448,449], osteocalcin/collagen [450], alginate [451], chitin [452], chitosan [453], silk fibroin [454], silanized HPMC [455], bioactive glass [456], magnetic nanoparticles [457], etc. More to the point, various reinforcement additives of different shapes and nature are widely used to improve the mechanical properties of the self-setting formulations [458,459]. Even carbon nanotubes are used for this purpose [460,461]. Although the biomaterials community does not use this term, a substantial amount of the reinforced formulations might be defined as CaPO₄-based concretes [442]. The idea behind the concretes is simple: if a strong filler is present in the matrix, it might stop crack propagation.

Various apatite-containing formulations based on PMMA [462,463,464,465,466,467,468,469,470] and polyethylmethacrylate [471,472] have been already developed. Such biocomposites might be prepared by dispersion of apatite powder into a viscous fluid of the polymer [473] and used for drug delivery purposes [474]. When the mechanical properties of the concretes composed of PMMA matrix and HA particles of various sizes were tested, the tensile results showed that strength was independent on particle sizes. After immersion into Ringer´s solution, the tensile strength was not altered whereas the fatigue properties were significantly reduced. The biocompatibility of PMMA/HA formulations was tested in vivo and enhanced osteogenic properties of the implants compared to single-phase PMMA were observed [463,464,465,466]. It was shown that not only the mechanical properties of PMMA were improved but the osteoblast response of PMMA was also enhanced with addition of HA [464]. Thereby, by adding of CaPO₄, a non-biodegradable PMMA was made more bioactive and osteoconductive, yielding a well-processible biocomposite concrete. As a drawback, the PMMA/HA formulations possess a low flexural, compressive and tensile strength.

Biocomposites made from CaPO₄ and various types of resins appeared to possess comparable mechanical and biological properties to typical PMMA cement, leading to potential uses for implant fixation [170,172,475,476,477]. To improve the mechanical properties of the self-setting CaPO₄-based formulations and stabilize them at the implant site, various researchers have resorted to formulations that set in situ, primarily through crosslinking reactions of the polymeric matrix. For example, TTCP was reacted with PAA, forming a crosslinked CDHA/calcium polyacrylate biocomposite [478]. In aqueous solutions, TTCP hydrolyzes to CDHA [16] and the liberated calcium cations react with PAA, forming the crosslinked network [478]. Reed et al., synthesized a dicarboxy polyphosphazene that can be crosslinked by calcium cations and CDHA/polyphosphazene biocomposites with a compressive strength ~10 MPa and of ~65% porosity were prepared as a result [479]. To mimic PMMA cements, PFF/β-TCP biocomposites were prepared with addition of vinyl monomer to crosslink PPF. As a result, quick setting and degradable formulations with a low heat output and compressive strengths in the range of 1–12 MPa were prepared by varying the molecular weight of PPF, as well as the contents of the monomer, β-TCP, initiator and NaCl, as a porogen [480,481]. An acrylic formulation with Sr-containing HA as a filler [100], an injectable polydimethylsiloxane/HA formulation [482], biocomposites consisting of PLGA microspheres and self-setting CaPO₄-based formulations [483,484], as well as hybrid formulations of chitosan oligosaccharide/gelatin/CaPO₄ [485] were prepared as well.

In order to improve the mechanical properties of self-setting formulations, numerous researchers blended various polymers to them. For example, gelatin might be added to self-setting formulations, primarily to stabilize the paste in aqueous solution before it develops adequate rigidity and, secondly, to improve the compressive strength [389,446,486]. Adding rod-like fillers to the self-setting formulations also caused an improvement in the mechanical properties [486]. For example, PAA and PVA were successfully used to improve the mechanical properties of a TTCP + DCPD formulation but, unfortunately, with an inevitable and unacceptable reduction of both workability and setting time [487,488]. Similar findings were reported in the presence of sodium alginate and sodium polyacrylate [489]. Other polymers, such as polyphosphazene might be used as well [490,491,492]. Other examples of polymer/CaPO₄ self-setting formulations might be found elsewhere [493,494]. Metallic wires might be used as reinforcements as well [495].

Porous CaPO₄ scaffolds with interconnected macropores (~1 mm), micropores (~5 μm) and of high porosity (~80%) were prepared by coating PU foams with a TTCP + DCPA formulation, followed by firing at 1200 °C. In order to improve the mechanical properties of the scaffolds, the open micropores of the struts were then infiltrated by a PLGA solution to achieve an interpenetrating bioactive ceramic/biodegradable polymer composite structure. The PLGA filled struts were further coated with a 58S bioactive glass/PLGA composite coating. The obtained complex porous biocomposites could be used as tissue engineering scaffolds for low-load bearing applications [496]. A more complicated construction, in which the PLGA macroporous phase has been reinforced with a bioresorbable TTCP + DCPA formulation, followed by surface coating of the entire construct by a non-stoichiomentic CDHA layer, has been designed as well [497]. In one more study, α-TCP and poly(ethylenephosphate) sodium salt-coated PLGA microparticles were mixed with castor oil and water to form an oil-in-water emulsion. The resulting emulsion spontaneously set in a humidified atmosphere at ambient temperature. The PLGA particles incorporated within the set formulations were hydrolytically degraded, which resulted in the formation of an interconnected microporosity [498].

A porosity level of 42%–80% was introduced into self-setting CaPO₄/chitosan biocomposites by addition of the water-soluble mannitol [499]. Chitosan significantly improved the mechanical strength of the entire biocomposite [500]. A similar approach was used by other researchers who studied the effect of the addition of PLGA microparticles [501,502,503,504] (which can also be loaded with drugs or growth factors [505,506,507]) to self-setting CaPO₄ formulations. These biocomposites were implanted into cranial defects of rats and a content of ~ 30 wt. % of the microparticles was found to give the best results [501], while the addition of a growth factor to the biocomposites significantly increased bone contact at 2 weeks and enhanced new bone formation at 8 weeks [507]. The in vivo rabbit femur implant tests showed that self-setting PLGA/CaPO₄ formulations exhibited outstanding biocompatibility and bioactivity, as well as a better osteoconduction and degradability than self-setting formulations consisted solely from CaPO₄ [502].

To finalize this topic, one should mention that CaPO₄ might be added into self-setting calcium sulfate (plaster of Paris) formulations to improve their osteoconductivity [508,509]. Similarly, calcium silicates could be added to the self-setting CaPO₄ formulations [510,511,512,513].

4.3. Formulations Based on Nanodimensional CaPO₄ and Nanodimensional Biocomposites

Nanodimensional and nanophasic materials are the materials that have particles or grain sizes less than 100 nm, respectively. However, before proceeding any further, one should clearly differentiate between the nanodimensional composites and the composites containing nanodimensional particles. The former might be any type of composites but disintegrated to particles with dimensions <100 nm, while the latter consist of two or more materials, in which at least one of the materials is of a nanometer-scale. Nevertheless, both options combine the properties of both components, thus broadening their functionality beyond the pure materials.

Nanodimensional and nanophasic materials have different mechanical and optical properties if compared to the large grained materials of the same chemical composition. Namely, they possess the unique surface properties, such as an increased number of atoms, grain boundaries and defects at the surface, huge surface area and altered electronic structure, if compared to the conventional micron-sized materials. For example, nanodimensional HA (size ~67 nm) has a higher surface roughness of 17 nm if compared to 10 nm for the conventional submicron size HA (~180 nm), while the contact angles (a quantitative measure of the wetting of a solid by a liquid) are significantly lower for nanodimensional HA (6.1) if compared to the conventional HA (11.51). Additionally, the diameter of individual pores in nanodimensional HA compacts is five-times smaller (pore diameter ~6.6 Å) than that in the conventional grain-sized HA compacts (pore diameter within 19.8–31.0 Å) [514]. Besides, nanodimensional CaPO₄ promote osteoblast cells adhesion, differentiation and proliferation, osteointegration and deposition of calcium containing minerals on its surface better than microcrystalline ones; thus enhancing formation of a new bone tissue within a short period [515,516,517]. More to the point, nanodimensional HA was found to cause apoptosis of the leukemia P388 cells [518].

Natural bones and teeth are hierarchical biocomposites of biological origin based on nanodimensional compounds because they consists of nano-sized blade-like crystals of biological apatite grown in intimate contact with the organic matrix rich in bioorganic fibers and organized in complicated hierarchical structures. Given the fact that the major bioorganic phase of bones is collagen, i.e., a natural polymer (Table 1), it is obvious that biocomposites of nanodimensional CaPO₄ with biodegradable polymers should be advantageous as bone grafting materials. In such biocomposites, the inorganic phase would be responsible for the mechanical strength (hardness) and bioactivity, while the polymeric phase would provide the elasticity. In addition, the solubility of CaPO₄ depends on their crystallite size (smaller crystals have a higher solubility) and on their carbonate content (higher carbonate content increases the solubility) [519]. To the author’s best knowledge, among CaPO₄ listed in Table 3, before very recently only apatites (CDHA, HA and, perhaps, FA) have been available in nanodimensional state. However, recently, nano-sized DCPA [428,429,430] and nano-sized MCPM [520] have been synthesized and applied to prepare biocomposites with strong ionic release to combat tooth caries. Presumably, all CaPO₄ from Table 3 might be manufactured in nanodimensional and/or nanocrystalline state; however not all of them have been prepared yet [517].

A number of investigations have been conducted recently to determine the mineralization, biocompatibility and mechanical properties of the biocomposites based on various (bio)polymers and nanodimensional CaPO₄ (mainly, HA). Unfortunately, in the majority of the already published papers it often remained unclear whether “nanodimensional HA”, in fact, represented the nanodimensional stoichiometric HA or a nanodimensional non-stoichiometric CDHA; therefore, no differentiation is possible between them. These studies covered biocomposites with PGA [521], PLA [176,267,522,523,524,525,526,527,528,529,530,531,532,533] and its copolymers with PGA [534,535,536] and PCL [537], collagen [538,539,540,541,542,543,544,545,546,547,548], collagen + PLA [549,550,551,552,553,554], collagen + PVA [555], collagen + alginate [556,557], alginate [558], gelatin [559,560,561,562,563,564], PPF [565,566,567], PA [243,244,568,569,570,571,572,573,574,575], PVA [271,272,576,577,578], PVAP [276], poly(ethylene-co-acrylic) acid [579,580], chitosan [581,582,583,584,585,586,587,588,589] and its derivatives [590], konjac glucomannan + chitosan [591], chitosan + gelatin [592], xanthan [593], PHEMA + PCL [594], PCL [317,354,595,596,597], PU [598], cellulose [41,42,599,600,601,602], Ti [602,603,604], ZrO₂ [605,606,607], glasses [608,609], metals and alloys [610,611] and many other biocompatible hybrid formulations [215,256,264,270,342,612,613,614,615,616,617,618,619,620,621,622,623,624,625,626,627,628,629,630,631]. Furthermore, each from the aforementioned formulations might be covered by a layer of nanodimensional CaPO₄, as it was done by Zandi et al. [632], who coated a biocomposite of nano-sized rods HA with gelatin by nano-sized HA. Several nanodimensional biocomposites were found to be applicable as carriers for delivery of drugs and growth factors [609,633,634,635], as well as promising vectors with ultrahigh gene loading and transfection efficiency [636]. Data are available on the excellent biocompatibility of such biocomposites [545]. The dispersion state of nano-sized particles appears to be the critical parameter in controlling the mechanical properties of nanodimensional biocomposites, as nano-sized particles always tend to aggregate owing to their high surface energy [342]. A comparison was made of the mechanical properties between biocomposites of PA with the nano-sized and micron-sized HA. The results showed that the bending and tensile strengths of the biocomposite increased with increasing content of nanodimensional HA but decreased with increasing micron-sized HA content [243]. A SEM image of the mineralized collagen fibrils, demonstrating homogeneity of the nanodimensional biocomposite and the close interaction between the mineral phase and the reconstituted collagen fibrils is shown in Figure 4 [637].

Open in a separate window

Figure 4

Scanning electron microscopy image of reconstituted mineralized collagen I fibrils. An example of an organic-inorganic nanostructural composite, mimicking the extracellular matrix of bone tissue on the nanometer scale. Reprinted from Ref. [637] with permission.

Porous (porosity ~85%) biocomposites of nano-sized HA with collagen and PLA have been prepared by precipitation and freeze-drying; these biocomposites did not show a pH drop upon in vitro degradation [549,550,551]. They were implanted in the radius of rabbits and showed a high biocompatibility and partial resorption after 12 weeks. Nano-sized HA/сhitosan biocomposites with improved mechanical stability were prepared from HA/сhitosan nano-sized rods [638]. Nano-sized HA/PLLA biocomposites of high porosity (~90%) were prepared using thermally induced phase separation [639]. Besides, nanodimensional HA was used to prepare biocomposites with PAA and the nanostructure of the resulting nano-sized crystals exhibited a core-shell configuration [640,641].

Nanodimensional crystals of HA appeared to be suitable for intra-osseous implantation and offered a potential to formulate enhanced biocomposites for clinical applications [642]. Thus, the biocompatibility of chitosan in osteoblast cell culture was significantly improved by addition of nano-sized HA [643]. Similar finding was valid for nanodimensional HA/PA biocomposites [569]. Further details on nanodimensional biocomposites might be found in excellent reviews [22,644]. More to the point, a more general review on applications of nanodimensional biomaterials in orthopedics is also available [645], where the interested readers are referred.

4.4. Biocomposites with Collagen

The main constituent of the bioorganic matrix of bones is type I collagen (Table 1) with molecules about 300 nm in length. The structural and biochemical properties of collagens have been widely investigated and over 25 collagen subtypes have been identified [646,647]. This protein is conducive to crystal formation in the associated inorganic matrix. It is easily degraded and resorbed by the body and allows good attachment to cells. Collagen alone is not effective as an osteoinductive material but it becomes osteoconductive in combination with CaPO₄ [648]. Both collagen type I and HA were found to enhance osteoblast differentiation [649] but combined together, they were shown to accelerate osteogenesis. However, this tendency is not so straightforward: the data are available that implanted HA/collagen biocomposites enhanced regeneration of calvaria bone defects in young rats but postponed the regeneration of calvaria bone in aged rats [650]. Finally, addition of CaPO₄ to collagen sheets was found to give a higher stability and an increased resistance to 3D swelling compared to the collagen reference [651]. Therefore, a bone analogue based on these two constituents should possess the remarkable properties.

The unique characteristics of bones are the spatial orientation between the nanodimentional crystals of biological apatite and collagen macromolecules at the nano-scale, where the crystals (about 50 nm length) are aligned parallel to the collagen fibrils, which is believed to be the source of the mechanical strength of bones (Figure 5a) [652]. The collagen molecules and the crystals of biological apatite assembled into mineralized fibrils are approximately 6 nm in diameter and 300 nm long. Although the complete mechanisms involved in the bone building strategy are still unclear, the strengthening effect of nanodimentional crystals of biological apatite in calcified tissues might be explained by the fact that the collagen matrix is a load transfer medium and thus transfers the load to the intrinsically rigid inorganic crystals [14,15,19,20,21]. Furthermore, the crystals of biological apatite located in between tangled fibrils cross-link the fibers either through a mechanical interlocking or by forming calcium ion bridges, thus increasing deformation resistance of the collagenous fiber network [653].

Open in a separate window

Figure 5

Schematic illustrations of collagen–CaPO₄ nanocomposites: (a) structure of bone: nanodimensional crystals of biological apatite aligned along with the collagen fibrils; linking occurs between the apatite and polar groups on collagen chains in bone; (b) nanodimensional crystals of CaPO₄ physically “trapped” within the collagen matrix; (c) nanodimensional crystals of CaPO₄ deposited on and covered the collagen fibrils surface. Reprinted from Ref. [652] with permission.

When CaPO₄ are combined with collagen in a laboratory, the prepared biocomposites appear to be substantially different from natural bone tissue (Figure 5b,c) due to a lack of real interaction between the two components, i.e., the interactions that are able to modify the intrinsic characteristics of the singular components themselves. The main characteristics of the route, by which the mineralized hard tissues are formed in vivo, is that the organic matrix is laid down first and the inorganic reinforcing phase grows within this organic matrix [14,15,20,21]. Although to date, neither the elegance of the biomineral assembly mechanisms nor the intricate composite nano-sized architectures have been duplicated by non-biological methods, the best way to mimic bone is to copy the way it is formed, namely by nucleation and growth of CDHA nano-sized crystals from a supersaturated solution both onto and within the collagen fibrils. Such syntheses were denoted as ‘‘biologically inspired’’ which means they reproduce an ordered pattern and an environment very similar to natural ones [654,655,656]. The biologically inspired biocomposites of collagen and CaPO₄ (mainly, apatites) for bone substitute have a long history [648,657,658,659,660,661,662,663,664,665,666,667,668,669,670] and started from the pioneering study by Banks et al., who grew CDHA on reconstituted calfskin collagen tapes in 1977 [671], followed by initial medical applications performed by other researchers in 1982 [672,673]. Such combinations were found to be bioactive, osteoconductive, osteoinductive [648,674,675,676] and, in general, artificial grafts manufactured from this type of the biocomposites are likely to behave similarly to bones and be of more use in surgery than those prepared from any other materials. Indeed, data are available on the superiority of CaPO₄/collagen biocomposite scaffolds over the artificial polymeric and CaPO₄ bioceramic scaffolds individually [677].

It has been found that CaPO₄ may be successfully precipitated onto a collagen substrate of whatever form or source [539,671,678,679]. However, adherence of CaPO₄ crystals to collagen did depend on how much the collagen had been denatured: the more fibrillar the collagen, the greater attachment. Clarke et al., first reported the production of a biocomposite produced by precipitation of DCPD onto a collagen matrix with the aid of phosphorylated amino acids commonly associated with fracture sites [657]. Self-setting CDHA forming formulations (DCPD + TTCP) have been mixed with a collagen suspension, hydrated and allowed to set. CDHA crystals were found to nucleate on the collagen fibril network, giving a material with the mechanical properties weaker than those reported for bone. More to the point, the prepared biocomposites were without the nanostructure similar to that of bone [658,680]. The oriented growth of OCP crystals on collagen was achieved by an experimental device in which Ca²⁺ and PO₄³⁻ ions diffused into a collagen disc from the opposite directions [680,681,682]. Unfortunately, these experiments were designed to simulate the mechanism of in vivo precipitation of biological apatite only; due to this reason, the mechanical properties of the biocomposites were not tested [683].

Conventionally, collagen/CaPO₄ biocomposites can be prepared by blending or mixing of collagen and CaPO₄, as well as by biomimetic methods [539,550,634,654,655,656,657,658,668,678,683,684,685,686,687,688,689,690,691,692,693,694]. Tampieri et al., [656] produced and compared artificial bone like tissue apatite/collagen biocomposites prepared by using two different methodologies: (1) dispersion of apatite in a collagen aqueous suspension and then freeze dried and (2) direct nucleation of an apatitic phase on assembling collagen fibrils. Biocomposites obtained using first way were similar to uncalcified natural collagen. The crystallite sizes were not uniform and were often aggregated and randomly distributed into the matrix (Figure 5b), proving that there was no real interaction between apatite and collagen fibers. However, the second method allowed the direct nucleation of nano-sized crystals of apatite on self-assembled collagen fibers. In this case, the two components (CDHA and collagen) exhibited strong interactions (Figure 5c), highlighted by several analysis techniques, which showed a complete analogy of the composite with calcified natural tissue [656]. Other production techniques are also possible. For example, using a polymer-induced liquid-precursor process, collagen/apatite biocomposites mimicking the nanostructure of bones, wherein nano-sized crystals of apatite were embedded within the collagen fibrils, were prepared [693]. More complicated formulations, such as a magnetite enriched HA/collagen [695] and HA/collagen/PVA [696] biocomposites, have been developed as well. Recent investigations revealed that CaPO₄/collagen formulations might be printed by means of 3D printers [697].

In addition, collagen might be incorporated into various self-setting CaPO₄-based formulations [658,680,698,699,700,701,702]. Typically, a type I collagen sponge is presoaked in PO₄³⁻-containing a highly basic aqueous solution and then is immersed into a Ca²⁺-containing solution to allow mineral deposition. Also, collagen I fibers might be dissolved in acetic acid and then this solution is added to phosphoric acid, followed by a neutralization synthesis (performed at 25 °C and solution pH within 9–10) between an aqueous suspension of Ca(OH)₂ and the H₃PO₄/collagen solution [654,655]. To ensure the quality of the final product, it is necessary to control the Ca/P ionic ratio in the reaction solution. One way to do this is to dissolve a commercial CaPO₄ in an acid; another one is to add Ca²⁺ and PO₄³⁻ ions in a certain ratio to the solution and after that induce the reaction [703]. Biomimetically, one can achieve an oriented growth of CDHA crystals onto dissolved collagen fibrils in aqueous solutions via a self-organization mechanism [688]. Besides, CDHA crystallization from aqueous solutions might be performed in the presence of a previously dispersed collagen [539]. More to the point, collagen might be first dispersed in an acidic solution, followed by addition of calcium and orthophosphate ions and then co-precipitation of collagen and CDHA might be induced by either increasing the solution pH or adding mixing agents. Although it resulted in biocomposites with poor mechanical properties, pressing of the apatite/collagen mixtures at 40 °C under 200 MPa for several days is also known [704]. Attempts have been performed for a computer simulation of apatite/collagen composite formation process [705]. It is interesting to note, that such biocomposites were found to possess some piezoelectric properties [706].

As the majority of the collagen/HA biocomposites are conventionally processed by anchoring micron-sized HA particles into collagen matrix, it makes quite difficult to obtain a uniform and homogeneous composite graft. Besides, such biocomposites have inadequate mechanical properties; over and above, the proper pore sizes have not been achieved either. Further, microcrystalline HA, which is in contrast to nanocrystalline bone apatite, might take a longer time to be remodeled into a new bone tissue upon the implantation. In addition, some of the biocomposites exhibited very poor mechanical properties, probably due to a lack of strong interfacial bonding between the constituents. Furthermore, in all blended composites, the crystallite sizes of CaPO₄ were not uniform and the crystals were often aggregated and randomly distributed within a fibrous matrix of collagen. Therefore, no structural similarity to natural bone was obtained and only a compositional similarity to that of natural bone was achieved. The aforementioned data clearly demonstrate that the chemical composition similar to bone is insufficient for manufacturing the proper grafts; both the mechanical properties and mimetic of the bone nanostructure are necessary to function as bone in recipient sites. There is a chance for improving osteointegration by reducing the grain size of HA crystals by activating of ultrafine apatite growth into the matrix. This may lead to enhance the mechanical properties and osteointegration with improved biological and biochemical affinity to the host bone. Besides, the porosity was found to have a positive influence on the ingrowth of the surrounding tissues into the pores of collagen/HA biocomposites [707,708].

Bovine collagen might be mixed with CaPO₄ and such biocomposites are marketed commercially as bone-graft substitutes those further can be combined with bone marrow aspirated from the iliac crest of the site of the fracture. Bioimplant^®, Bio-Oss Collagen^®, Boneject^®, Collagraft^®, CollapAn^®, Healos^®, Integra Mozaik^®, and LitAr^® are several examples of the commercially available CaPO₄/collagen grafts for the clinical use [17,22]. Application of these materials was compared with autografts for the management of acute fractures of long bones with defects, which had been stabilized by internal or external fixation [709,710]. These biocomposites are osteogenic, osteoinductive and osteoconductive; however, they lack the structural strength and require a harvest of the patient’s bone marrow. OCP/collagen biocomposites have been investigated [711] and clinically tested [712].

Collagen sponges with an open porosity (30–100 μm) were prepared by a freeze-drying technique and then their surface was coated by a 10-μm layer of biomimetic apatite precipitated from simulated body fluid [713]. The researchers found a good in vitro performance with fibroblast cell culture. Other preparation techniques are also possible [714]. Collagen/HA microspheres or gel beads have been prepared in the intention of making injectable bone fillers [715,716,717]. Liao et al. succeeded in mimicking the bone structure by blending carbonateapatite with collagen [718]. A similar material (mineralized collagen) was implanted into femur of rats and excellent clinical results were observed after 12 weeks [719]. Collagen/HA biocomposites were prepared and their mechanical performance was increased by cross-linking the collagen fibers with glutaraldehyde [540,541,542]. These biocomposites were tested in rabbits and showed a good biological performance, osteoconductivity and biodegradation. A similar approach was selected to prepare HA/collagen microspheres (diameter ~5 μm) by a water-oil emulsion technique in which the surface was also cross-linked by glutaraldehyde [716]. That material showed a good in vitro performance with osteoblast cell culture. A porous bone graft substitute was formed from a nano-sized HA/collagen biocomposite combined with PLA by a freeze-drying method; the resulting material was found to mimic natural bones at several hierarchical levels [550]. Subsequent in vitro experiments confirmed a good adhesion, proliferation and migration of osteoblasts into this composite [549]. A further increase in biocompatibility might be achieved by addition of various dopants. For example, to enhance bone substitution, Si-substituted HA/collagen composites have been developed with silicon located preferentially in the collagen phase [541]. Porous (porosity level ~95% with interconnected pores of 50–100 μm) biocomposites of collagen (crosslinked with glutaraldehyde) and β-TCP have been prepared by a freeze-drying technique, followed by sublimation of the solvent; the biocomposites showed a good biocompatibility upon implantation in the rabbit jaw [720].

Biocomposites of CaPO₄ with collagen were found to be useful for delivery of drugs, growth factors and other important biomolecules [557,660,702,721,722,723]. Namely, an HA/collagen – alginate (20 µL) with the rhBMP-2 (100 µg/mL, 15 µL) showed bone formation throughout the implant 5 weeks after implantation without obvious deformation of the material [557]. Gotterbarm et al., developed a two-layered collagen/β-TCP implant augmented with chondral inductive growth factors for repair of osteochondral defects in the trochlear groove of minipigs. This approach might be a new promising option for the treatment of deep osteochondral defects in joint surgery [722].

To conclude this part, one should note that biocomposites of apatites with collagen are a very hot topic for research and up to now, just a few papers have been devoted to biocomposites of other CaPO₄ with collagen [685,686,687,722,724,725,726]. These biomaterials mimic natural bones to some extent, while their subsequent biological evaluation suggests that they are readily incorporated into the bone metabolism in a way similar to bone remodeling, instead of acting as permanent implant [550,673]. However, the performance of these biocomposites depends on the source of collagen from which it was processed. Several attempts have been made to simulate the collagen-HA interfacial behavior in real bone by means of crosslinking agents such as glutaraldehyde [540,542,543,678,716,720] with the purpose to improve the mechanical properties of these biocomposites. Unfortunately, a further progress in this direction is restricted by a high cost, difficulty to control cross-infection, a poor definition of commercial sources of collagens, as well as by a lack of an appropriate technology to fabricate bone-resembling microstructures. Further details on CaPO₄/collagen biocomposites might be found elsewhere [652,663,727].

4.5. Formulations with Other Bioorganic Compounds

The biggest practical problems with collagen type I are its cost and the poor definition of commercial sources of this material, which makes it difficult to follow up on well controlled processing. Therefore, collagen type I can be replaced by other compounds. One should notice, that, besides collagen, both human and mammalian bodies contain dozens types of various bioorganic compounds, proteins and biological macromolecules. Therefore, the substantial amount of them potentially might be used to prepare biocomposites with CaPO₄. For example, a biologically strong adhesion (to prevent invasion of bacteria) between teeth and the surrounding epithelial tissues is attributed to a cell-adhesive protein, laminin [728]. In order to mimic the nature, a laminin/apatite biocomposite layer was successfully created on the surface of both titanium [729] and EVOH [730,731] using the biomimetic approach. A more complicated laminin/DNA/apatite biocomposite layer was found to be an efficient gene transfer system [732]. Further details on this subject are available in a topical review [733].

CaPO₄/gelatin biocomposites are widely investigated as potential bone replacement biomaterials [253,383,384,385,386,387,388,389,390,409,446,447,448,559,560,561,562,563,564,734,735,736,737,738,739]. For example, gelatin foams were successfully mechanically reinforced by HA and then crosslinked by a carbodiimide derivative [253]. Such foams were shown to be a good carrier for antibiotic tetracycline [735]. Several biocomposites of CaPO₄ with alginates have been prepared [387,556,557,560,655,740,741,742,743,744]. For example, porous HA/alginate composites based on hydrogels were prepared both biomimetically [655] and by using a freeze-drying technique [740]. More complicated formulations have been developed as well [745].

Various biocomposites of CaPO₄ with chitosan [234,403,425,436,453,462,499,581,582,583,584,585,586,587,588,589,590,591,618,638,643,746,747,748,749,750,751,752,753,754,755,756,757] and chitin [392,452,758,759,760,761,762] are also very popular. For example, a solution-based method was developed to combine HA powders with chitin, in which the ceramic particles were uniformly dispersed [758,759]. Unfortunately, it was difficult to obtain the uniform dispersions. The mechanical properties of the final biocomposites were not very good; due to a poor adhesion between the filler and the matrix both the tensile strength and modulus were found to decrease with increase of the HA amount. Microscopic examination revealed that HA particles were intervened between the polymer chains, weakening their interactions and decreasing the entire strength [758,759]. Other manufacturing techniques might be found in the aforementioned references, for example, a HA/chitosan biocomposite was produced by a hydrothermal process from natural CaCO₃/chitosan biocomposite of crab shells [751]. Similarly, HA/chitosan biocomposite was produced by a hydrothermal process from previously prepared DCPD/chitosan biocomposite [752]. Biocomposites of natural HA with chitosan were found to possess both a good hard tissue biocompatibility and an excellent osteoconductivity, which is suitable for artificial bone implants and frame materials of tissue engineering [748]. Data are available that addition of CaPO₄ into chitosan improved cell attachment, provided a higher cell proliferation and well-spread morphology when compared to chitosan alone [584,754]. More complex formulations, such as silk fibers reinforced HA/chitosan [763] and HA/collagen/chitosan [764] biocomposites, have been studied as well. Besides biomedical applications, biocomposites of nano-sized HA with chitin/chitosan might be used for removal of Fe(III) [765] and fluorides [766,767] from aqueous solutions.

Biocomposites of CDHA with water-soluble proteins, such as BSA, might be prepared by a precipitation method [493,768,769,770,771]. In such biocomposites, BSA is not strongly fixed to solid CDHA, which is useful for a sustained release. However, this is not the case if a water/oil/water interfacial reaction route has been used [247]. To extend this subject, inclusion of DNA into CDHA/BSA biocomposites was claimed [247,772,773,774]. Furthermore, biocomposites of an unspecified CaPO₄ with DNA [775,776], as well as those of nano-sized crystals of biomimetic apatite with C₆₀ fullerene and Au-DNA nano-sized particles [777] were prepared as well.

Akashi and co-workers developed a procedure to prepare CaPO₄-based biocomposites by soaking hydrogels in supersaturated by Ca²⁺ and PO₄³⁻ ions solutions in order to precipitate CDHA in the hydrogels (up to 70% by weight of CDHA could be added to these biocomposites) [778]. This procedure was applied to chitosan; the 3D shape of the resulting biocomposite was controlled by the shape of the starting chitosan hydrogel [779]. Another research group developed biocomposites based on in situ CaPO₄ mineralization of self-assembled supramolecular hydrogels [780]. Other experimental approaches are also possible [781].

Various biocomposites of CDHA with glutamic and aspartic amino acids, as well as poly-glutamic and poly-aspartic amino acids have been prepared and investigated [277,278,782,783,784,785]. These (poly)amino acids were quantitatively incorporated into CDHA crystals, provoking a reduction of the coherent length of the crystalline domains and decreasing the crystal sizes. The relative amounts of the (poly)amino acid content in the solid phase, determined through HPLC analysis, increased with their concentration in solution up to a maximum of about 7.8 wt % for CDHA/aspartic acid and 4.3 wt % for CDHA/glutamic acid biocomposites. The small crystal dimensions, which implied a great surface area, and the presence of (poly)amino acids were suggested to be relevant for possible application of these biocomposites for hard tissues replacement [277,278,782,783,784,785]. A schematic description of a biocomposite formation from amino acids and ACP is shown in Figure 6 [786].

Open in a separate window

Figure 6

A proposed mechanism for the formation of amorphous calcium phosphates (ACP)/amino acid biocomposites in aqueous solutions. Reprinted from Ref. [786] with permission.

Furthermore, BCP (HA + β-TCP)/agarose macroporous scaffolds with controlled and complete interconnection, high porosity, thoroughly open pores and tailored pore size were prepared for tissue engineering application [787,788]. Agarose, a biodegradable polymer, was selected as the organic matrix because it was a biocompatible hydrogel, which acted as gelling agent leading to strong gels and fast room temperature polymerization. Porous scaffolds with the designed architecture were manufactured by combining a low temperature shaping method with stereo-lithography and two drying techniques. The biocompatibility of this BCP/agarose system was tested with mouse L929 fibroblast and human SAOS-2 osteoblast during different colonization times [789].

Fibrin sealants are non-cytotoxic, fully resorbable, biological matrices that simulate the last stages of a natural coagulation cascade, forming a structured fibrin clot similar to a physiological clot [790]. Biocomposites of CaPO₄ with fibrin sealants might develop the clinical applications of bone substitutes. The 3D mesh of fibrin sealant interpenetrates the macro- and micro-porous structure of CaPO₄ ceramics. The physical, chemical and biological properties of CaPO₄ bioceramics and the fibrin glue might be cumulated in biocomposites, suitable for preparation of advanced bone grafts [791,792,793,794,795,796,797].

Furthermore, there are biocomposites of CaPO₄ with bisphosphonates [798], silk fibroin [245,619,620,799,800,801,802,803,804], cellulose [805], chitosan + silk fibroin [806], chitosan derivatives [807], fibronectin [808], chondroitin sulfate [235,675,809], casein phosphopeptides [810], okra hydrocolloids [811], keratin hydrogel [812], amyloid fibrils [813], agarose [814] and vitamins [815]. Photopolymerisable formulations have been developed as well [816]. Besides, the reader’s attention is pointed out to an interesting approach to crystallize CDHA inside poly(allylamine)/poly(styrene sulfonate) polyelectrolyte capsules resulting in empty biocomposite spheres of micron size. Depending on the amount of precipitated CDHA, the thickness of the shell of biocomposite spheres can be varied between 25 and 150 nm. These biocomposite capsules might find application as medical agents for bone repairing and catalytic microreactors [817].

4.6. Injectable Bone Substitutes (IBS)

With the development of minimally invasive surgical methods, for example percutaneous surgery, directly injectable biomaterials are needed. The challenge is to place a biomaterial at the site of surgery by the least possible invasive method. In this regard, IBS appear to be a convenient alternative to solid bone-filling materials. They represent ready-to-use suspensions of CaPO₄ microspheres [818,819], nano-sized rods [820] or powder(s) in a liquid carrier phase. However, addition of other phases, such as calcium sulfate [821], is possible. IBS look like opaque viscous pastes with the rheological properties, sufficient to inject them into bone defects by means of surgical syringes and needles. Besides, IBS could be easily produced in a sterile stage. Their stable composition and mechanical properties are suitable for reproducibility of the biological response [822,823]. All types of IBS are divided into two major groups: self-setting formulations (see Section 4.2.) and those, which do not set. The latter ones are described here.

IBS requires suitable rheological properties to ensure bonding and cohesion of the mineral phase in situ with good cell permeability. Usually, the necessary level of viscosity is created by addition of suitable water-soluble polymers [92,251,564,824,825,826,827,828,829,830,831,832]. However, IBS based on hydrogels are known as well [833,834,835]. Therefore, the majority of CaPO₄-containing IBS formulations might be considered as a subgroup of CaPO₄/polymer biocomposites. For example, an IBS was described that involved a silanized hydroxyethylcellulose carrier with BCP (HA + β-TCP). The suspension is liquid at pH within 10–12, but gels quickly at pH < 9 [826]. Injectable composites can be formed with β-TCP to improve mechanical integrity [480]. Similarly, a polydioxanone-co-glycolide-based biocomposite reinforced with HA or β-TCP can be used as an injectable or moldable putty [827]. During the crosslinking reaction following injection, carbon dioxide is released allowing the formation of interconnected pores. Furthermore, HA/poly(L-lactide-co-ε-caprolactone) biocomposite microparticles were fabricated as an injectable scaffold via the Pickering emulsion route in the absence of any molecular surfactants. A stable injectable oil-in-water emulsion was obtained using water dispersed HA nano-sized crystals as the particulate emulsifier and a dichloromethane solution of poly(L-lactide-co-ε-caprolactone) as an oil phase [251]. CaPO₄-containing IBS based on other (bio)organic hydrogels, such as gelatin [564], CMC [830], oligo(poly(ethylene glycol) fumarate) [828,829], chitosan + collagen [831] have been developed as well. In addition, photocrosslinkable formulations are known [832].

Daculsi et al. developed viscous IBS biocomposites based on BCP (60% HA + 40% β-TCP) and 2% aqueous solution of HPMC that was said to be perfectly biocompatible, resorbable and easily fitted bone defects (due to an initial plasticity) [825,836,837,838,839,840,841,842,843]. The best ratio BCP/HPMC aqueous solution was found to be at ~65/35 w/w. To extend this subject further, IBS might be loaded by cells [844,845], radiopaque elements [846] or microparticles [847], as well as functionalized by nucleic acids [848]. Self-hardening formulations, based on Si-HPMC hydrogel, are known as well [844]. The list of the commercially available CaPO₄-based IBS formulations is presented in Table 5 [848].

The advanced characteristics of IBS come from their good rheological properties and biocompatibility and the ease of tissue regeneration. Although the fabrication of IBS biocomposites in most cases improved the mechanical properties of the system and provided the material with resistance to fluids penetration, these achievements were limited by the amount of polymer that can be added to the paste. For instance, Mickiewicz et al. reported that after a critical concentration (that depended on the type and molecular weight of the polymer, but was always around 10%), the polymer started forming a thick coating on the crystal clusters, preventing them from interlocking, originating plastic flow and, as a consequence, decreasing mechanical properties [493]. More to the point, Fujishiro et al. reported a decrease in mechanical properties with higher amounts of gel, which was attributed to formation of pores due to leaching of gelatin in solution [486]. Therefore, it seems that mechanical properties, although improved by the addition of polymers, are still a limitation for the application of CaPO₄-based IBS formulations in load-bearing sites [125]. Further details on IBS are available elsewhere [822,823].

4.7. Biocomposites with Inorganic Compounds, Carbon and Metals

To overcome the problem of poor mechanical properties of CaPO₄ bioceramics, suitable biocomposites of CaPO₄ reinforced by various inorganic materials, glasses and metals have been developed as well. Such biocomposites are mainly prepared by the common ceramic processing techniques such as thermal treatment after kneading [849,850,851], powder slurry coating [852] and metal-sol mixing [853]. For example, HA was combined with Bioglass^® (Novabone Products, Alachua, FL) [854,855] and with other glasses [856] to form glass-ceramics biocomposites. Other reinforcement materials for CaPO₄ could be differentiated by either shape of the fillers, such as particles [857,858], platelets [859,860], whiskers [520,861,862,863], fibers [864,865,866,867,868], or their chemical composition. Namely, zirconia and PSZ [849,850,851,852,861,869,870,871,872,873,874,875,876,877,878,879,880,881], alumina [857,860,882,883,884,885,886,887], titania [888,889,890,891,892,893,894,895,896] and their mixtures [897], other oxides and their mixtures [863,898,899,900,901,902,903,904,905], silica and/or glasses [906,907,908,909,910,911,912,913,914,915], wollastonite [916,917,918,919,920], mullite [921,922,923,924], various metals and alloys [495,610,611,866,925,926,927,928,929,930,931,932,933,934,935,936,937,938,939,940,941,942,943], calcium sulfate [944,945,946,947,948,949], calcium carbonate [950,951], silicon carbide [623,862], barium titanate [952], zeolite [953], boron nitride [954], zirconium nitride [955], carbon [956] and several other materials [957,958,959] were added to CaPO₄ to improve reliability. In addition, CaPO₄-based biocomposites possessing piezoelectric properties have been proposed [960]. Interestingly, Fe₃O₄/HA composites possess photocatalytic properties [902,903]. More complicated formulations, such as HA/aluminum oxide/carbon nanotubes [961], have been developed as well. Unfortunately, significant amounts of the reinforcing phases are needed to achieve the desired properties and, as these materials are either bioinert, significantly less bioactive than CaPO₄ or not bioresorbable, the ability of the biocomposites to form a stable interface with bone is poorer if compared with CaPO₄ bioceramics alone. Due to the presence of bioinert compounds, such formulations occasionally are called bioinert/bioactive composites [908]. The ideal reinforcement material would impart mechanical integrity to a biocomposite at low loadings, without diminishing its bioactivity.

There are several types of HA/glass biocomposites. The first one is also called bioactive glass-ceramics. A dense and homogeneous biocomposite was obtained after a heat treatment of the parent glass, which comprised ~38 wt % oxy-FAP (Ca₁₀(PO₄)₆(O,F)₂) and ~34 wt % β-wollastonite (CaO·SiO₂) crystals, 50–100 nm in size in a MgO-CaO-SiO₂ glassy matrix [916,917,918,919,920]. A-W glass-ceramics is an assembly of small apatite particles effectively reinforced by wollastonite. The bending strength, fracture toughness and Young’s modulus of A-W glass-ceramics are the highest among bioactive glass and glass-ceramics, enabling it to be used in some major compression load bearing applications, such as vertebral prostheses and iliac crest replacement. It combines a high bioactivity with the suitable mechanical properties [962]. β-TCP/wollastonite biocomposites are also known [963,964,965]. More complicated formulations have been developed as well. For example, A-W/HDPE composite (AWPEX) biomaterials have been designed to match the mechanical strength of human cortical bone and to provide favorable bioactivity, with potential use in many orthopedic applications [966,967,968,969]. Similar can be said on A-W/silk fibroin biocomposites [970]. Other examples comprise wollastonite-reinforced HA/Ca polycarboxylate [971], glass-reinforced HA/polyacrylate [972], as well as collagen- [973] and gelatin- [974] calcium phosphate silicate/wollastonite biocomposites.

HA/glass biocomposites can be prepared by simple sintering of appropriate HA/glass powder mixtures [975,976,977,978]. If sintering is carried out below 1000 °C, HA does not react with the bioactive glass [976,977] or this reaction is limited [978]. Besides, reaction between HA and glasses depends on the glass composition. In another approach, small quantities of bioactive glass have been added to HA bioceramics in order to improve densification and/or mechanical properties [23]. In addition, biocomposites might be sintered from HA and silica [908]. In general, bioactive glass-ceramics maintain a high strength for a longer time than HA bioceramics under both the in vitro and in vivo conditions [910].

Due to a huge difference in shapes, it is a challenge to prepare homogeneous mixtures of CaPO₄ and carbon nanotubes: “one can imagine something similar to achieving a homogeneous mixture of peas and spaghetti” [979] (p. 7). Nevertheless, different strategies might be employed to prepare CaPO₄/carbon nanotube biocomposites. For example, apatites might be chemically synthesized by using carboxyl functionalized carbon nanotubes as a matrix [238,239,240,241,980,981,982,983,984]. Physico-chemical characterization of these biocomposites showed that nucleation of CDHA initiates through the carboxyl group [238]. Hot pressing [985], compaction [986], plasma spraying [987], laser surface alloying [988,989,990], spark plasma sintering [991], shear mixing [992], sol-gel [993] and precipitation [994] techniques might be applied as well. Due to carbon oxidation at elevated temperature, sintering of CaPO₄/carbon nanotube biocomposites must be performed in a deoxidizing atmosphere [995]. The research on CaPO₄ (up to now, only apatite)/carbon nanotube biocomposites is in its early stages, with the first papers published in 2004 [980]. Due to this reason, the mechanical property data for such biocomposites have been reported only in few papers; however, these results are encouraging. For example, Chen et al., performed nano-indentation tests on biocomposite coatings to give hardness and Young’s modulus values [990]. They found that the higher the loading of the nanotubes, the better the mechanical properties. Namely, at 20 wt % loading, hardness was increased by ~43% and Young’s modulus by ~21% over a single-phase HA coating [990]. Scratching test results indicated that as alloyed HA biocomposite coatings exhibited improved wear resistance and lower friction coefficient with increasing the amount of carbon nanotubes in the precursor material powders [989]. Additionally, measurements of the elastic modulus and hardness of the biocomposite coatings indicated that the mechanical properties were also affected by the amount of carbon nanotubes [988]. Another research group performed compression tests on bulk HA/carbon nanotubes biocomposites and found an increase in strength over single-phase HA [980]. However, the highest compressive strength they achieved for any material was only 102 MPa, which is similar to that of cortical bone but much lower than the typical values for dense HA [979]. More complex formulations, such as poly-L-lysine/HA/carbon nanotube hybrid biocomposites, have been also developed [996]. Furthermore, CaPO₄/carbon nanotube biocomposites might be immobilized by hemoglobin [997]. Unfortunately, carbon nanotubes are very stable substances; they are neither bioresorbable nor biodegradable. Therefore, during in vivo bioresorption, the nanotubes will get into the human body from the biocomposite matrix and might cause uncertain health problems. Certainly, this problem must be solved. To conclude the carbon subject, one should mention on application of carbon fibers of microscopic dimensions [998,999,1000], nanodimensional diamonds [1001], graphene [1002,1003,1004,1005,1006,1007], fullerenes [777] and its derivatives [1008] to reinforce CaPO₄ bioceramics.

As seen from the amount of the references, CaPO₄/zirconia biocomposites appear to be a hot topic of the research. The main disadvantage of HA reinforced by PSZ is degradation of zirconia in wet environments [861,870,871,874]. Transformation of the tetragonal ZrO₂ to the monoclinic phase on the surface results in formation of microcracks and consequently lowers the strength of the implant [1009,1010].

Various biocomposites of CaPO₄ with metals and alloys have been fabricated as well [495,610,611,866,925,926,927,928,929,930,931,932,933,934,935,936,937,938,939,940,941,942,943]. For example, an HA-based biocomposite reinforced with 20 vol.% of Ti particles was fabricated by hot pressing [927]. Besides, biocomposites of CaPO₄ with metals might be prepared by powder metallurgy processing [928,930,941,942] and mechanical compaction followed by sintering [611,936,939]. Furthermore, silver nanoparticles/HA biocomposites could be synthesized through γ-irradiation reduction of silver ions incorporated into HA particles [610]. At high temperatures, the presence of Ti metal phase was found to promote dehydration and decomposition of HA into β-TCP and TTCP [927,928,939,1011] and/or partial formation of β-TCP and calcium titanates instead of HA [603,930,942,1011]. Comparing with pure HA bioceramics manufactured under the same conditions, the HA/Ti biocomposites possessed a higher fracture toughness, bending strength, work of fracture, which is better suitable for the biomedical applications [936]. However, the mechanical properties appeared to be not high enough to use HA/Ti biocomposites in load-bearing applications. Luckily, the histological evaluations revealed that HA/Ti biocomposites could be partially integrated with newborn bone tissues after 3 weeks and fully osteointegrated at 12 weeks in vivo [927]. Similar findings had been earlier made for HA bioceramics reinforced by addition of silver particulates (5–30 vol.%) and subsequent sintering of the HA/Ag powder compacts [925,926]. Besides, addition of silver imparts an antimicrobial activity [932]. Other studies on CaPO₄/Ti biocomposites are available elsewhere [929,931,942]. In all the aforementioned studies, CaPO₄ was used as a matrix; however, the opposite situation is also possible. There is a study, in which composites made of Mg3Zn0.8Zr alloy as a matrix and β-TCP particles as reinforcements were prepared for investigating the effect of β-TCP spherical particles sized in 100 nm on the microstructure of that alloy [1012].

To conclude this part, biocomposites consisting of CaPO₄ only should be briefly described. First, all types of biphasic, triphasic and multiphasic CaPO₄ formulations should be mentioned [412]. For example, in 1980-s, BCP was called as “TCP ceramics complexed with HA” [1013]. Even nowadays BCP is occasionally called as a “nanocomposite” [1014]. Furthermore, fluoridated HA (described by a chemical formula Ca₁₀(PO₄)₆(OH)_2−xF_x, where 0 < x < 2) might be mentioned as composites [1015]; however, an applicability of the “composite” term for such systems is doubtful. One should better consider 70% HA-powder + 30% HA-whisker biocomposites, which were fabricated by pressureless sintering, hot pressing and hot isostatic pressing. These biocomposites were found to exhibit an improved toughness, attaining the lower fracture-toughness limit of bone without a decrease of bioactivity and biocompatibility [1016,1017]. Besides, a dual HA biocomposite that combined two HA materials with different porosities: HA with 75% porosity, for bone ingrowth and HA with 0% porosity, for load bearing was manufactured. This dual HA biocomposite appeared to be suitable for use as an implant material for spinal interbody fusion as a substitute for iliac bone grafts, which could eliminate the disadvantages associated with autograft harvesting [1018]. A biodegradable biocomposite porous scaffold comprising a β-TCP matrix and nano-sized fibers of HA was developed and studied for load-bearing bone tissue engineering. The nano-sized fibers of HA were prepared by a biomimetic precipitation method, the inclusion of which significantly enhanced the mechanical property of the scaffold, attaining a compressive strength of 9.87 MPa, comparable to the high-end value (2–10 MPa) of cancellous bone [1019]. Furthermore, HA/β-TCP biocomposites with different β-TCP content (10, 20, and 30 wt %) were prepared by wet mixing of HA and β-TCP powders, compaction of the powder mixtures and sintering [1020]. Finally, it is interesting to mention on a successful reinforcement of carbonateapatite porous blocks by newly prepared carbonateapatite crystals (i.e., by the same compound; thus, a biocomposite of two different carbonateapatites was obtained) [1021]. First, a calcium salt was introduced to micropores of carbonateapatite blocks. Then, the calcium salt was carbonated to form calcite inside the micropores of the carbonateapatite blocks by exposing the blocks to carbon dioxide at the second step. On the third step, the blocks were immersed in a Na₂HPO₄ aqueous solution. In this process, calcite inside the micropores of the carbonateapatite blocks was transformed to carbonateapatite and the newly formed crystals of carbonateapatite entangled on those of the existing carbonateapatite blocks. Due to bonding between the newly formed carbonateapatite crystals and the existing ones in the carbonateapatite blocks, a mechanical strength of the blocks became ~1.5 times higher when compared to that before the treatment [1021].

4.8. Functionally Graded Formulations

Although, in most cases, the homogeneous distribution of filler(s) inside a matrix is required [347], there are composites, in which this is not the case. For example, functionally graded materials (FGM) might be characterized by the intentional variations in composition and/or structure gradually over volume, resulting in corresponding changes in the properties of the composite. The main feature of such materials is the almost continuously graded composition that results in two different properties at the two ends of the structure. Such composites can be designed for specific function and applications. Various approaches based on the bulk (particulate processing), pre-form processing, layer processing and melt processing are used to fabricate the functionally graded materials.

Bone is a biologically formed composite with variable density ranging from very dense and stiff (cortical bone) to a soft and foamed structure (trabecular bone). Normally the outer part of long bones consists of cortical bone with the density decreasing towards the core, where the trabecular bone is found. The trabecular bone is porous and the porosity is filled with osseous medulla [14,15]. This brief description clearly indicates that bones are natural functionally graded composites.

The concept of FGM has been increasingly used for biomaterial design and currently it remains to be an important area of the research. For example, many studies have been performed to fabricate porosity-graded CaPO₄ bioceramics in attempts to mimic the porous structure of bones [1022,1023,1024,1025]. This is a structural approach to fabricate FGM. Besides, there is a compositional approach. For example, functionally graded composite skull implants consisting of polylactides, carbonateapatite and CaCO₃ have been prepared [312,313]. In addition, powder metallurgy methods have been used to fabricate HA/Ti functionally graded biocomposite dental implants offering the biocompatible HA on the tissue side and titanium on the outer side for mechanical strength [1026,1027,1028]. The graded structure in the longitudinal direction contains more Ti in the upper section and more HA in the lower section. Actually, in the upper section the occlusal force is directly applied and Ti offers the required mechanical performance; in the lower part, which is implanted inside the bone, the HA confers the bioactive and osteoconductive properties to the material [1026]. Since the optimum conditions of sintering for Ti and HA are very different, HA/Ti functionally graded biocomposites are difficult to fabricate and the sintering conditions for their mixtures are obliged to compromise. The expected properties of this implant are shown in Figure 7 [1027]. Such biocomposites might be both symmetrical [1029] and asymmetrical [1030].

Open in a separate window

Figure 7

Expected properties of functionally graded biocomposite dental implant. For comparison, the upper drawing shows a functionally graded implant and the lower one shows a conventional uniform implant. The properties are shown in the middle. The implant with the composition changed from a biocompatible metal (Ti) at one end (left in the figure), increasing the concentration of bioceramics (hydroxyapatite (HA)) toward 100% HA at the other end (right in the figure), could control both mechanical properties and biocompatibility without an abrupt change due to the formation of discrete boundary. This functionally gradient material (FGM) biocomposite was designed to provide more titanium for the upper part where occlusal force is directly applied and more HA for the lower part, which is implanted inside the jawbone. Reprinted from Ref. [1027] with permission.

A series of functionally graded CaPO₄ coatings incorporated with various percentages of silver were deposited on titanium substrates using ion beam-assisted deposition. The analysis of the coating’s cross-section revealed a decreased crystallinity as well as a distribution of nano-sized (10–50 nm) silver particles from the coating/substrate interface to top surface [1031]. Similarly, compositionally graded HA/Ti biocomposite coatings were prepared by rf-plasma spraying [1032]. In addition, both laser cladding [1033] and combinatorial matrix-assisted pulsed laser evaporation [1034] techniques were used to put down gradient deposits of carbon nanotubes/HA and Sr-substituted HA and zoledronate-modified HA, respectively. A functionally graded HA/PMMA biocomposite was developed based on sedimentary HA distributions in a PMMA viscous fluid, using a centrifuge to avoid stress convergence on the interface. The stress-strain curves of this biocomposite showed a sufficient strength for biomedical applications along with the relaxation of brittleness and fragility [473]. A compositionally graded collagen/nanodimensional HA biocomposite scaffold might be prepared by an in situ diffusion method. Chemical and microstructural analysis revealed a gradient of the Ca to P ratio across the width of the scaffold template, resulting in the formation of a Ca-rich side and a Ca-depleted side of scaffold. The Ca-rich side featured low porosity and agglomerates of the nanodimensional HA crystallites, while the Ca-depleted side featured higher porosity and nanodimensional HA crystallites integrated with collagen fibrils to form a porous network structure [1035]. A three-layered graded biocomposite membrane, with one face of 8% nanodimensional carbonateapatite/collagen/PLGA porous membrane, the opposite face of pure PLGA non-porous membrane and the middle layer of 4% nanodimensional carbonateapatite/collagen/PLGA as the transition was prepared through the layer-by-layer casting method [552]. A similar PLGA/nano-HA/lauric acid graded biocomposite was prepared as well [1036]. Functionally graded non-woven meshes of PCL incorporated by nano-sized particles of β-TCP were prepared using a hybrid twin-screw extrusion/electrospinning process [1037]. A functionally graded HA/silk fibroin biocomposite was prepared by pulse electric current sintering [1038]. HA/glass FGM layers were coated on titanium alloy (Ti6Al4V) substrates. The design of these layers and the use of the glass were for achieving a strong bonding between the FGM layered coatings and the substrates [1039,1040].

Functionally graded β-TCP/FA biocomposites combine the biostability of FA with the bioresorbable properties of β-TCP. An interesting multilayered (each layer of 1 mm thick) structure consisting of β-TCP/FA biocomposites with different molar ratios has been prepared, giving rise to formation of a FGM (Figure 8). After implantation, the preferential dissolution of β-TCP phase would result in functionally gradient porosity for bone ingrowth [1041]. In addition, compositionally graded HA-alumina-zirconia biocomposites were prepared by the same multilayered technique [1042]. Functionally graded fluoridated HA with a gradient of fluoride [1043] and carbonated HA with a gradient of carbonate [1044] were synthesized as well. HA/zirconia graded biocomposites were fabricated to enhance the mechanical properties of HA while retaining its bone bonding property [879]. TiO₂ and HA were found to be a good combination for FGM providing both a gradient of bioactivity and a good mechanical strength [1045]. Besides, graded HA/CaCO₃ biocomposite structures for bone ingrowth were developed [1046]. Functionally graded composite skull implants consisting of polylactides, carbonateapatite and CaCO₃ are known as well [312,313]. Thus, the research in this field is quite promising but currently the mechanical properties of the available biocomposites do not match the similar properties of bones [1047].

Open in a separate window

Figure 8

A schematic diagram showing the arrangement of the FA/β-TCP biocomposite layers. (a) A non-symmetric functionally gradient material (FGM); (b) symmetric FGM. Reprinted from Ref. [1041] with permission.