The present invention relates to group 2 metal phosphates, and particularly, although not exclusively, to methods for preparing calcium phosphates, such as hydroxyapatite. The invention extends to the phosphate compounds per se that can be produced by these methods, and to a variety of biomedical applications of such materials, for example as bioceramic compounds, which can be used in whole or as part of materials for bone or teeth replacement procedures.

Calcium phosphates (CaP) are well-known for their use as bone graft substitutes, coatings on metallic implants, reinforcements in biomedical composites and in bone and as components in dental cements. Synthetic hydroxyapatite (HA), [Cai ₀ (P0 ₄ )6(OH) ₂ ] is similar to biological apatite, the main mineral constituent of teeth and bone because of its composition, biocompatibility, bioactivity and low solubility in wet media. Synthetic HA and other calcium phosphates have been employed as scaffold materials to encourage new bone formation for osteoinductive coatings on metal implants, as the hard segment in biocomposites, as components in calcium phosphate bone cements and as bulk bone fillers. The dissolution and reactions of calcium phosphate bioceramics (in vivo and in vitro) is found to be largely dependent on the composition, crystallinity, and pH of the surrounding solution. Multiple techniques have been used for the preparation of HA and calcium phosphates including solid state synthesis, spray pyrolysis, sol gel techniques, chemical vapour deposition (CVD), solvothermal processes, and wet precipitation methods.

The majority of the reported literature for the synthesis of nano-sized HA and calcium phosphates are multi-step, inconsistent, energy intensive or time-consuming processes and often involve very careful control over reaction pH during the mixing of reagents. For example, batch hydrothermal syntheses of HA as reported in the literature are normally conducted in the temperature range 6o - 150 °C for up to 24 hours to yield crystalline HA rods that are usually agglomerated. It is very difficult in batch reactions to control the pH as it fluctuates. Batch co-precipitation reactions for the synthesis of HA can also be carried out at room temperature. In this case, if the pH is typically less than 10 during the synthesis, non-stoichiometric HA can be produced. Furthermore, for such a reaction and near room temperature, failure to permit enough maturation of the reagents (sometimes as long as a day), can give a phase-separated product upon additional heat-treatment (due to a non- stoichiometric hydroxyapatite being formed). This can significantly affect biological properties in vivo and in vitro. Reports of continuous production of HA are carried out at temperatures in excess of the range of 200-300°C, which is very energy intensive. Furthermore, at such high

temperatures, although nucleation occurs, there is substantial growth or agglomeration of smaller nuclei to form substantially larger particles, typically >ioo nm in size (i.e. the length of needles). Additionally, there is a disadvantage of having to use high temperature continuous reactors, in that they are conducted in metal pipes due to the high reaction temperatures and therefore if such a process were used to make bioceramics, they would contain substantial levels of leached metals from the steel (e.g. Fe, Cr, etc). This would mean that the bioceramics may not be acceptable for clinical use according to the latest

international standards and guidelines.

There is therefore a need to develop a simple, low cost, clean, synthesis technique, which could work under mild conditions and allow the synthesis of high purity stoichiometric HA and other bioceramic materials in a considerably short time period with a fine and controllable particle size (range from 20-150 nm) and controlled surface area (typically range from 95-300 rr^g ^"1 ), depending on reaction conditions and containing trace contaminating elements which are well below acceptable levels as stipulated by international recognised standards for clinical use of bioceramics. Thus, in a first aspect of the invention, there is provided a method of preparing a group 2 metal phosphate, or a doped, substituted or surface-functionalised compound thereof, the method comprising contacting a first precursor comprising a group 2 metal and a second precursor comprising a phosphate source, to thereby form a group 2 metal phosphate, or a doped, substituted or surface-functionalised compound thereof, characterised in that the method is carried out under continuous flow conditions at a temperature in the range of 20- ioo°C.

In a second aspect of the invention, there is provided an apparatus for continuously producing a group 2 metal phosphate, or a doped, substituted or surface-functionalised compound thereof, the apparatus comprising a reactor in which a group 2 metal phosphate, or a doped, substituted or surface-functionalised phosphate compound thereof is formed, a first feed means for continuously feeding a first precursor comprising a group 2 metal to the reactor, a second feed means for continuously feeding a second precursor comprising a phosphate source to the reactor, characterised in that the reactor is a continuous flow reactor. As described in the Examples, the inventors have demonstrated that the continuous hydrothermal flow synthesis (CHFS) method of the first aspect and the apparatus of the second aspect can be used to produce a wide range of group 2 metal phosphate compounds (e.g. calcium phosphates, such as stoichiometric hydroxyapatite or mixtures of two or more calcium phosphates, including doped, substituted and surface-functionalised group 2 metal phosphate compounds). Advantageously, the inventors have demonstrated the rapid, single- step synthesis of stoichiometric hydroxyapatite (HA) nanorods using a novel two pump continuous plastic flow synthesis (CPFS) method in the range of 6o-8o°C in 5 minutes (i.e. the residence time) under conditions of pH 10-11 from an aqueous solution of calcium nitrate tetrahydrate and diammonium hydrogen phosphate. To date, there have been no reports of direct and rapid syntheses of phase pure stoichiometric HA by using a single-step continuous flow synthesis at reasonably mild temperature conditions, i.e. between 20 and ioo°C, and at atmospheric pressure. Advantageously, the method of the first aspect makes it possible to rapidly make stoichiometric HA in a matter of seconds or minutes and under certain conditions can achieve nucleation but, importantly, retards the growth step substantially, such that the HA is stoichiometric HA, but having substantively smaller particle size than that which is made using more conventional synthesis methods. The properties, nature and crystallinity of the precipitated HA formed by the method depended on the reaction variables which include, overall concentration, temperature, pH, time and variation in Ca:P ratio. Indeed, other calcium phosphate phases (e.g. brushite, monetite, β-TCF, CD HA and biphasic HA / -TCP) may be prepared by changing the Ca/P ratio and pH of the precursors (some of them also require an additional heat-treatment step). Among these phases, brushite has raised great attention as a highly resorbable bioceramic, but it is rather difficult to prepare in a batch reaction because its synthesis is highly pH- and time-sensitive.

The inventors have observed that the method of the first aspect works very well at higher concentrations of precursors. However, the particle size increases, with a subsequent lowering of surface area. Therefore, at higher concentrations of precursors, the method can still be used for the rapid and continuous manufacture of stoichiometric hydroxyapatite or other bioceramics (e.g. calcium phosphates and their derivatives). Under such conditions, the particles are typically around one hundred nanometres or more in length and surface areas are about 100 rr^g- ¹ . The properties of the precipitated material depended on the temperature, concentration, pH, time and other precipitation conditions such as variation in Ca:P ratio and residence time. By tuning the reaction parameters in the method of the invention, products of any required particle size (ranging from 20-150 nm) and surface area (ranging from 95-300 m ² g ^_1 ) can be produced. The method of the invention has the potential to be used for the production of synthetic HA with trace element purity levels and thermal stability within the tolerances stipulated by ISO standards [e.g. ISO13779-3] for such bioceramics. Advantageously, the use of continuous flow has resolved the issues of pH control and long maturation times (at room temperature) that are normally required. Furthermore, the process overcomes the problems of batch to batch variations in synthesis and overcomes difficulties presented for scale-up of batch processes. As such, the method and apparatus of the invention provides excellent control over pH conditions (with no need to monitor/control pH during synthesis as in a batch process) to synthesize high purity HA and other ceramic materials in a considerably short time period with very fine particle size. A variety of ion substituted calcium phosphates (Mg, Sr, Zn, Fe, Ag, Mn, SiCV ^" , C0 ₃ ² ), nanocomposite materials (Fe ₃ 0 ₄ -HA, T1O2-HA) and surface modified organopolymer-nanoparticle dental composites have also been developed successfully by using the method.

For synthesis, the first precursor may be selected from a range of group 2 metal-containing compounds, for example hydrated or dehydrated group 2 metal salts, such as nitrate, hydroxide, chloride, carbonate, or oxide. Preferably, the first precursor comprises a group 2 metal selected from the group consisting of: beryllium, magnesium, calcium, strontium and barium. It is most preferred that the first precursor comprises calcium. Suitably calcium- containing compounds which may be used as the first precursor may include calcium nitrate, calcium hydroxide or calcium chloride. Preferably, the first precursor comprises calcium nitrate, more preferably calcium nitrate tetrahydrate. It is preferred that the first precursor comprises an aqueous solution. The second precursor may be selected from a range of phosphate-containing compounds, for example ammonium hydrogen phosphate, orthophosphoric acid, or sodium or potassium dihydrogen phosphate. Preferably, the second precursor comprises ammonium hydrogen phosphate, more preferably diammonium hydrogen phosphate. It is preferred that the second precursor comprises an aqueous solution.

The method preferably comprises feeding a calcium nitrate tetrahydrate solution via the first feed means and diammonium hydrogen phosphate solution via the second feed means to the reactor, such as tee-piece mixer or similar, where they react. The apparatus preferably comprises a T-junction comprising three arms, wherein two of the arms correspond to the first and second feed means where the two precursors initially react, and the third arm corresponds to the reactor, where the two precursors react further together whilst being heated and the reaction reaches completion therein. Preferably, each feed means comprises a pump. Preferably, one pump is configured to pump the first precursor to the first arm of the T-junction, and the other pump is configured to pump the second precursor to the second arm of the T-junction. Preferably, the reactor comprises a tube in communication with the third arm along which the mixed and reacting precursors continuously pass as a result of the first and second pumps, preferably while being heated, and product material is collected at the end thereof. For a small-scale laboratory-based process, the tube is preferably at least lorn long and can be made of a fluoropolymer which would give good resistance to strong pH's and be easy to clean. For given flow rates for the two pumps of 10 mL per minute (total flow= 20 mL/min ¹ ), this gives an approximate residence time of five minutes as a typical set of process parameters. Preferably, the tube is heated externally via a heat exchanger disposed downstream of the T-junction mixing point. For example, the tube may be located inside an oil bath for some kind of heat exchanger downstream of the mixing point, allowing the reaction to go to completion. The product is preferably collected from the end of the tube at ambient pressure. The tube diameters, type of heat exchanger and flow rates and types of pumps would differ considerably depending on what is required to achieve similar performance and lab scale.

Some reaction or nucleation may occur in the T-junction when the two feeds meet and the pH will be well-defined and controlled at this point, allowing for phase pure materials to be formed collected when the mixture has passed through the heat exchanger, but in a relatively short amount of time. The pH of the prior art batch processes is usually susceptible to pH fluctuation; this is because it usually requires the drop-wise addition of a basic precursor into another reagent resulting in a fluctuating pH. When the reaction is carried out normally at room temperature for example is in a batch process it ideally needs to be kept above a pH of 10 for example in order to ultimately deliver stoichiometric hydroxyapatite. Therefore pH cannot be readily controlled in the same way as the method of the invention.

The residence time in the reactor for producing the compounds described herein depends on the flow rates, pipe diameters (i.e. construction of the process), the pressure, reaction temperature. Preferably, the method is conducted at atmospheric pressure.

Preferably, the method is carried out under continuous flow conditions at a temperature of 22-ioo°C, more preferably 30-ioo°C, even more preferably 50-ioo°C, even more preferably 55-ioo°C, still more preferably 6o-ioo°C, and most preferably 70-ioo°C. Preferably, the method is carried out under continuous flow conditions at a temperature of 20-95°C, more preferably 20-90°C, even more preferably 20-85°C, and more preferably 20-8o°C, and most preferably 20-70°C. Preferably, the method is carried out under continuous flow conditions at a temperature of 22-ioo°C, more preferably 30-90°C, more preferably 50-90°C, even more preferably 55-85°C, more preferably 6o-8o°C, and most preferably 70-8o°C. Any of the lower and upper temperatures in the above ranges maybe combined with each other. Preferably, the residence time for the method is between 10 seconds and 20 minutes, more preferably between 30 seconds and 10 minutes, still more preferably between 1 minute and 5 minutes, more preferably between 3 and 5 minutes, and most preferably about 4 minutes. Any of the lower and upper residence times in the above ranges may be combined with each other. Furthermore, it will be appreciated that any of temperature ranges given herein may be combined with any of the residence times.

As described in the Examples, a range of different calcium phosphate materials can be produced with the method and apparatus of the invention. Thus, preferably the group 2 metal phosphate comprises calcium phosphate. In one preferred embodiment, the group 2 metal phosphate comprises hydroxyapatite, i.e. Ca _io (P0 ₄ ) ₆ (OH) ₂ Preferably, the Ca:P ratio is as close to 1.67 as possible. Preferably, hydroxyapatite is produced within a pH range of 9.5- 12, which is described in Example 1. Preferably, the temperature of the reaction is maintained at 6o-8o°C. Preferably, the residence time for the reaction is 1-5 minutes. This gives an apatite-like phase (also referred to as prepared hydroxyapatite) which can be heat-treated at iooo°C (or even as high as 1200°C) to obtain highly crystalline phase pure synthetic hydroxyapatite, with no secondary phases being obtained.

In another preferred embodiment, the prepared group 2 metal phosphate comprises calcium- deficient hydroxyapatite, i.e. Ca ₁₀ _ _a (HPO ) _b (PO ) _6< (ΟΗ) ₂ _ - Preferably, the Ca:P ratio is in the range of 1.5 to 1.67. Preferably, prepared calcium-deficient hydroxyapatite is produced within a pH range of 6.5-9.5, which is described in Example 2. Preferably, the reaction temperature is maintained at 6o-8o°C. Preferably, the residence time for the reaction is 1-5 minutes.

In another preferred embodiment, the group 2 metal phosphate comprises β-tricalcium phosphate, i.e. -Ca ₃ (P0 ₄ ) ₂ _ Preferably, the Ca:P ratio in the reagents is about 1.5. Preferably, a prepared calcium-deficient hydroxyapatite is initially produced at a pH of about 8, which is described in Example 2. Preferably, the temperature of the reaction is maintained at 60- 8o°C. Preferably, the residence time for the reaction is 1-5 minutes. After collection and cleaning of the resulting slurry and a further heat -treatment at iooo°C for lh in air (of the dry prepared powder), β-tricalcium phosphate is formed. If the heat-treatment is carried out above 1200°C, then oc-tricalcium phosphate is formed instead. In yet another preferred embodiment, the group 2 metal phosphate comprises dicalcium phosphate dihydrate (brushite), i.e. CaHP0 ₄ .2H ₂ 0 (can directly be made in the process).

Preferably, the Ca:P ratio in the reagents is about 0.8. Preferably, brushite is produced within a pH range of 2.0-6.0, which is described in Example 3. Preferably, the temperature of the reaction is maintained at 20-30°C, more preferably at 22-30°C. Preferably, the residence time for the reaction is 1-3 minutes.

In a further preferred embodiment, the group 2 metal phosphate comprises anhydrous dicalcium phosphate (monetite), i.e. CaHP0 ₄ Preferably, the Ca:P ratio for the reagents is about 0.8. Preferably, brushite is initially produced within a pH range of 2.0-6.0, which is described in Example 3. Preferably, the temperature of the reaction is maintained at about 22°C. Preferably, the residence time for the reaction is about 3 minutes (in this embodiment, time is important in order to obtain phase pure product). After collection of the slurry, cleaning and isolation of the dry powder, a further heat-treatment is preferably conducted at about 300°C for about 1 h, to form phase pure monetite. Monetite phase is typically formed in the wide temperature range of iio-300°C (from Brushite by heating in air at a heating rate of 10 °C/ min for 1 hr).

In a still further preferred embodiment, the group 2 metal phosphate comprises calcium pyrophosphate i.e. Ca ₂ P ₂ 0 _y- Preferably, the Ca:P ratio of the reagents is about 0.8. Preferably, brushite is initially produced within a pH range of 2.0-6.0, which is described in Example 3. Preferably, the temperature of the reaction is maintained at about 22°C. Preferably, the residence time for the reaction is about 3 minutes. After collection of the slurry, cleaning and isolation of the dry powder, a further heat-treatment is preferably conducted at about 500°C for 1 h in air, to form calcium pyrophosphate. Calcium pyrophosphate phase typically forms after heat treatment in the temperature range of 440-700°C using heating rate of 10 °C/ min for 1 hr. Alternatively, calcium pyrophosphate can be made from heat-treatment of monetite (which in itself is made from Brushite) at about 500°C for 1 h. As described in Example 4, the inventors have demonstrated that hydroxyapatite can be doped with a range of group 2 metal ions to produce a doped group 2 metal phosphate. In one embodiment, a doped group 2 metal phosphate comprises strontium doped in a group 2 metal phosphate, preferably strontium doped in a calcium phosphate, most preferably strontium doped in calcium hydroxyapatite. In another embodiment, a doped group 2 metal phosphate comprises barium doped in a group 2 metal phosphate, preferably barium doped in a calcium phosphate, most preferably barium doped in calcium hydroxyapatite. Preferably, the temperature of the synthesis reaction is maintained at 6o-8o°C. Preferably, the residence time for the reaction is 1-5 minutes. Advantageously, doping of the strontium or barium in calcium hydroxyapatite, results in materials that are phase pure and can be used in inorganic bone filler and related applications. Doping of these ions into calcium hydroxyapatite can also be used to alter the crystallinity, bioactivity, resorbability and antibacterial or X-ray opacity properties of the calcium phosphate.

As described in Examples 5 and 6, the inventors have demonstrated that hydroxyapatite can be substituted with a range of different anions or cations to produce a substituted group 2 metal phosphate. In relation to the apparatus of the second aspect, the dopant cation or substituted anions are added on purpose to alter the biological or other properties of the HA.

Thus, in one embodiment, a preferred substituted group 2 metal phosphate comprises an anion substituted group 2 metal phosphate. Preferred anions which may be used include carbonate ions or silicate ions. Accordingly, preferably the substituted compound comprises carbonate substituted group 2 metal phosphate, more preferably carbonate substituted calcium phosphate, most preferably carbonate substituted hydroxyapatite. In another embodiment, preferably the substituted compound comprises silicate substituted group 2 metal phosphate, more preferably silicate substituted calcium phosphate, most preferably silicate substituted hydroxyapatite. Preferably, the temperature of the reaction is maintained at 6o-8o°C. Preferably, the residence time for the reaction is 1-5 minutes. Advantageously, as described in Example 5, these materials are more resorbable and thus can allow faster bone growth rates in embodiments where they are used as biomedical bone fillers or coatings and related bioceramics applications. Preferred cations which may be used include magnesium, strontium, zinc, silver or manganese ions. Accordingly, preferably the substituted compound comprises magnesium, strontium, silver, zinc or manganese substituted group 2 metal phosphate, more preferably magnesium, strontium, silver, zinc or manganese substituted calcium phosphate, most preferably magnesium, strontium, silver, zinc or manganese substituted hydroxyapatite. Preferably, the temperature of the reaction is maintained at 6o-8o°C. Preferably, the residence time for the reaction is 1-5 minutes. Advantageously, as described in Example 6, doping of such ions can alter the resulting solubility, bioactivity, resorbability, bone growth rates, antibacterial properties of the bioceramic material. As described in Example 7, the inventors have demonstrated that hydroxyapatite can be surface-modified with a range of different organic surface coordinated compounds. Indeed, such surface-modified (as prepared) hydroxyapatites may be obtained following a similar procedure as for pure nano-HA (mentioned above), except the calcium containing precursor additionally contained the appropriate amount of surface agents e.g. polyvinyl alcohol, carboxylic acid etc. Hence, in one preferred embodiment, the surface-functionalised group 2 metal phosphate is preferably polyvinyl alcohol surface-functionalised group 2 metal phosphate, more preferably polyvinyl alcohol surface-functionalised calcium phosphate, most preferably polyvinyl alcohol surface-functionalised hydroxyapatite. This compound is termed PVA-HA (polyvinyl alcohol).

In another preferred embodiment, the surface-functionalised group 2 metal phosphate is preferably adipic acid surface-functionalised group 2 metal phosphate, more preferably adipic acid surface-functionalised calcium phosphate, most preferably adipic acid surface- functionalised hydroxyapatite. This compound is termed AA-HA (adipic acid).

In yet another preferred embodiment, the surface-functionalised group 2 metal phosphate is preferably citric acid surface-functionalised group 2 metal phosphate, more preferably citric acid surface-functionalised calcium phosphate, most preferably citric acid surface- functionalised hydroxyapatite. This compound is termed CA-HA, (citric acid).

In still another preferred embodiment, the surface-functionalised group 2 metal phosphate is preferably vinylphosphonic acid surface-functionalised group 2 metal phosphate, more preferably vinylphosphonic acid surface-functionalised calcium phosphate, most preferably vinylphosphonic acid surface-functionalised hydroxyapatite. This compound is termed VPA- HA (vinylphosphonic acid).

In another preferred embodiment, the surface-functionalised group 2 metal phosphate is preferably methacrylic acid surface-functionalised group 2 metal phosphate, more preferably methacrylic acid surface-functionalised calcium phosphate, most preferably methacrylic acid surface-functionalised hydroxyapatite. This compound is termed MA-HA (methacrylic acid).

Preferably, the temperature of the reaction for each of the surface-functionalised compounds is maintained at 6o-8o°C. Preferably, the residence time for the reaction is 1-5 minutes. As can be seen from above, further advantages of the method of the invention are that a bioceramic material can be produced which is surface functionalised in less than a minute at about 8o°C and only 5 minutes at 6o°C. Thus, the method involves a single-step, faster synthetic route by reducing the reaction completion time from hours to only minutes. It allows the use of low cost precursors to produce biocompatible bioceramic materials with low levels of impurities. If the surface functionalised bioceramics are heat-treated, they normally form non-stoichiometric calcium phosphates depending on the conditions (because some of the organic acid groups are incorporated and replace some of the phosphate ions that would have typically been in the as-prepared material).

Surprisingly, the materials obtained using the method of the invention can possess a substantially superior high temperature stability (e.g. hydroxyapatite can be made which is stable up to 1200°C because it is stoichiometric with a calcium to phosphorus ratio of 1.67) with remarkably high surface area (up to 263.9 rn ² /g) and small particle size (typically 2onm) compared to that reported in the literature. As such, these high surface area materials (i.e. nanoparticles) are novel per se, and have a great range of applications for use in replacement of living hard tissues such as bone and teeth, as bone graft substitutes, injectables, coatings on metallic implants, as fillers or additives into commercial products such as toothpastes, materials for the controlled release of drugs or other controlled release therapies, reinforcements in biomedical composites and in bone and dental cements.

Hence, in a third aspect, there is provided a group 2 metal phosphate, or a doped, substituted or surface-functionalised compound thereof obtained or obtainable by the method of the first aspect.

In a fourth aspect, there is provided a group 2 metal phosphate, or a doped, substituted or surface-functionalised compound thereof, of the second aspect, for use in therapy or as a medicament.

In a fifth aspect, there is provided use of a group 2 metal phosphate, or a doped, substituted or surface-functionalised compound thereof, of the second aspect, as a bone substitute material, optionally as a bone graft substitute material.

In a sixth aspect, there is provided use of a group 2 metal phosphate, or a doped, substituted or surface-functionalised compound thereof, of the second aspect, as a dental substitute material. Preferably, the compound may be used in bone and dental cements or any biomaterials material which is used to replace hard tissues in the body. In a seventh aspect, there is provided use of a group 2 metal phosphate, or a doped, substituted or surface-functionalised compound thereof, of the second aspect, as a bioceramic material. In an eighth aspect, there is provided use of a group 2 metal phosphate, or a doped, substituted or surface-functionalised compound thereof, of the second aspect, as an injectable formulation for bone repair, such as spinal fusion.

In a ninth aspect, there is provided use of a group 2 metal phosphate, or a doped, substituted or surface-functionalised compound thereof, of the second aspect, as a bioactive phase component in biomedical composites.

In a tenth aspect, there is provided use of a group 2 metal phosphate, or a doped, substituted or surface-functionalised compound thereof, of the second aspect, as a source for bioceramic coatings on metallic or polymeric implants. One can use a range of methods such as electrolytic or spray methods typically.

In a eleventh aspect, there is provided use of a group 2 metal phosphate, or a doped, substituted or surface-functionalised compound thereof, of the second aspect, as an additive for a toothpaste or other formulation.

In a twelfth aspect, there is provided use of a group 2 metal phosphate, or a doped, substituted or surface-functionalised compound thereof, of the second aspect, as a component of a drug delivery system for the controlled release of a therapeutic agent.

All of the features described herein (including any accompanying claims, abstract and drawings), and/or all of the steps of any method or process so disclosed, may be combined with any of the above aspects in any combination, except combinations where at least some of such features and/or steps are mutually exclusive.

For a better understanding of the invention, and to show how embodiments of the same may be carried into effect, reference will now be made, by way of example, to the accompanying diagrammatic drawings, in which: - Figure 1 shows transmission electron microscope images of hydroxyapatite nano-rods of sample HA60 (a) at x50,ooo times magnification (bar = 200 nm), (b) at x 10,000 magnification (bar = 50 nm), (c) and (d) at x 10,000 magnification ( bar = 20 nm); Figure 2 shows transmission electron microscope images of hydroxyapatite nano-rods of samples HA70, (a) at x50,ooo times magnification ( bar = 500 nm), and for (b) at x 10,000 magnification (bar = 100 nm);

Figure 3 shows transmission electron microscope images of hydroxyapatite nano-rods of samples HA80 (c) at x50,ooo times magnification ( bar = 500 nm), and (d) x 10,000 magnification (bar = 100 nm);

Figure 4[a] shows the effect of temperature on BET surface area and TEM particle size analysis; [b] shows the effect of concentration on BET surface area and TEM particle size analysis.

Figure 5 shows powder X-ray diffraction patterns of phase pure as-prepared hydroxyapatite made at temperature (a), 60 °C (sample HA60), (b), 70 °C (sample HA70), and (c), 80 °C (sample HA80);

Figure 6 shows heat-treatment of phase pure HA60 at different temperatures;

Figure 7 shows XPS survey spectrum of phase pure HA60 sample;

Figure 8 [a-c] shows XPS Spectra of Ca 2p,0 is and P 2p recorded from the phase pure HA60;

Figure 9 shows FTIR spectra of phase pure hydroxyapatite synthesized at (a), 60 °C (HA60) (b), 70 °C (HA70) and (c), 80 °C (HA80), respectively;

Figure 10 shows Raman spectra of phase pure hydroxyapatite synthesized at (a), 60 °C (HA60) (b), 70 °C (HA70) and (c), 80 °C (HA80), respectively;

Figure 11 shows MG63 viability on HA discs. Cell viability was measured at days 1, 4 and 7 via Alamar blue assay for MG63s cultured on HA 60, HA 70, HA 80 and cHA (Mean + SD, n = 6). * p <0.05;

Figure 12 shows MG63 morphology on HA discs. Cells were visualised at day 7 for cell nucleus (DAPI in blue) and cytoskeleton (phalloidin-FITC in green) for MG63S cultured on HA 60, HA 70, HA 80 and commercial HA. Scale bar is 100 μπι;

Figure 13 shows powder X-ray diffraction patterns of nonstoichiometric calcium

phosphates made at temperature 30 °C, 40 °C and 50 °C;

Figure 14 shows transmission electron microscope images of hydroxyapatite nano-rods (a) magnification 200k (bar = 100 nm) and for (b) CDHA and (c) Whitlockite, magnification 50k (bar = 500 nm);

Figure 15 shows powder X-ray diffraction patterns of phase pure HA, CDHA, HA+TCP and Whitlockite, respectively;

Figure 16 shows FTIR spectra of different phases of calcium phosphates;

Figure 17 shows human osteoblast cell proliferation after 1, 4 and 7 days of culture;

Figure 18 shows transmission electron microscope images of brushite (a) magnification 50k (bar = 500 nm) and (b) magnification 200k (bar = 100 nm); Figure 19 shows heat-treatment of brushite at 300 ⁰ , 500 ⁰ and iooo°C;

Figure 20 shows FTIR spectra of brushite nanoparticles obtained at different temperatures; Figure 21 shows human osteoblast cell proliferation after 1, 4 and 7 days of culture;

Figure 22 shows X-ray diffraction pattern of as-prepared Ba-HA;

Figure 23 shows radiographical examination of pure strontium and Ba-HA;

Figure 24 shows transmission electron microscope images of CHA nano-rods (a) magnification 50,000 (bar = 500 nm), and for (b) magnification 10,000k (bar = 100 nm); Figure 25 shows transmission electron microscope images of Si-HA nano-rods (a) magnification 50,000 (bar = 500 nm), and for (b) magnification 10,000 (bar = 100 nm); Figure 26 shows BET surface area of carbonate and silicate substituted samples as a function of substitution level;

Figure 27 shows powder X-ray diffraction patterns of heat-treated (iooo°C,ihr) C0 ₃ ^2~ - substituted hydroxyapatite powders made using CPFS at 70°C;

Figure 28 shows XRD patterns of heat-treated (1000 °C, 1 hr in air) Si-substituted hydroxyapatite powders made using CPFS at 70°C;

Figure 29 shows FTIR spectral data of C0 ₃ ^2~ substituted hydroxyapatite powders made using CPFS at 70°C;

Figure 30 shows FTIR spectral data of Si-substituted substituted hydroxyapatite powders made using CPFS at 70°C;

Figure 31 shows Human osteoblast cell proliferation after 1, 4 and 7 days of culture;

Figure 32 shows powder X-ray diffraction patterns of heat-treated (1000 °C, ihr in air) Mg-substituted calcium phosphates. The wt% values quoted are named values according to the amount of magnesium in the precursor feed;

Figure 33 shows transmission electron microscope images of sample (a) lMg-CaP, bar = 100 nm, (b) 6Mg-CaP, bar = 100 nm, (c) loMg-CaP, bar = 100 nm;

Figure 34 shows BET surface area of as precipitated HA, Biphasic Mg-HA and Mg- whitlockite calcium phosphates. The data points from left to right corresponded to samples oMg-CaP, lMg-CaP, 2Mg-CaP, 3Mg-CaP, 4Mg-CaP, 5Mg-CaP, 6Mg-CaP, 8Mg-CaP, loMg- CaP;

Figure 35[a-d] shows XPS Spectra of Ca 2p,0 is and P 2p and Mg is recorded from the synthesized CaP sample 8Mg-CaP;

Figure 36 shows human osteoblast cell proliferation after 1, 4 and 7 days of culture;

Figure 37 shows powder X-ray diffraction patterns of heat-treated (1000 °C, ihr in air) Sr- substituted calcium phosphates. The wt% values quoted are named values according to the amount of magnesium in the precursor feed; Figure 38 shows transmission electron microscope images of sample (a), (c) and (e), bar = 200 nm, and (b), (d) and (f) with bar size 20 nm for samples, 5 Sr-HA, ioSr-HA, and 2oSr- HA, respectively;

Figure 39 shows human osteoblast cell proliferation after 1, 4 and 7 days of culture;

Figure 40 shows XPS survey spectrum of Strontium substituted hydroxyapatite (loSrHA); Figure 41 [a-c] shows XPS Spectra of Ca 2p,0 is and P 2p recorded from loSrHA;

Figure 42 shows radiographical examination of 5, 10, 50 and 100% strontium- substituted HA;

Figure 43 shows transmission electron microscope images of surface modified

hydroxyapatite nano-rods (a) methacrylic acid modified hydroxyapatite with bar size = 200 nm, (c) vinylphosphonic acid modified hydroxyapatite and (e) adipic acid modified hydroxyapatite with bar size = 500 nm, and for (b), (d) and (f) bar = 20 nm;

Figure 44 shows BET surface area analysis of pure HA, surface modified PVA-HA

(polyvinyl alcohol), AA-HA (adipic acid), CA-HA, (citric acid), VPA-HA (vinylphosphonic acid) and MA-HA (methacrylic acid), respectively, made at 70°C using CPFS system;

Figure 45 shows powder X-ray diffraction patterns of surface modified hydroxyapatite with (a) = Polyvinylalcohol modified hydroxyapatite (b) = adipic acid modified hydroxyapatite (c)= citric acid modified hydroxyapatite (d) = vinylphosphonic acid modified hydroxyapatite (e) = methacrylic acid modified hydroxyapatite, respectively;

Figure 46 shows FTIR Spectra of surface modified hydroxyapatite with (a) =

Polyvinylalcohol modified hydroxyapatite (b) = adipic acid modified hydroxyapatite (c)= citric acid modified hydroxyapatite (d) = vinyphosphonic acid modified hydroxyapatite (e) = methacrylic acid modified hydroxyapatite, respectively;

Figure 47 shows XPS spectra of Ca 2p, P 2p, O is and C is;

Figure 48 shows lH NMR spectrum of surface modified vinylphosphonic acid

hydroxyapatite with HB = 5.73 ppm, HC =6.i6ppm; and

Figure 49 shows BET surface area measurements of ungrafted and polymer grafted HA at 60, 70 and 80 °C. Examples

Using continuous plastic flow synthesis (CPFS), Example 1 describes the synthesis of synthetic apatite like nanoparticles (<ioo nm) in any dimension using a two pump continuous (plastic) flow reactor at reaction temperatures in the range 6o-8o°C and total reaction time of ca. 5 minutes from mixing of an aqueous solution of calcium nitrate tetrahydrate with diammonium hydrogen phosphate (at pH 10) The prepared material was heat-treated 1200°C in air for lh to give phase pure hydroxyapatite (HA). The lattice constants of hydroxyapatite (HA) were similar to reference JCPDS Pattern no. 09-432. No decomposition of HA as-prepared into other phases was observed after heating at 1200°C in air for lh. Thus, the product was obtained as a phase pure material with a stoichiometric Ca:P molar ratio of 1.67, without the need for a long initial ageing step. The influence of initial Ca/P molar ratio, pH and precipitation temperature on the phase evolution and crystallinity of the as-prepared hydroxyapatite nanopowder were systematically investigated and optimised. An in-vitro study was used to evaluate biocompatibility and osteoblast cell proliferation / attachment on the surface of the as-prepared hydroxyapatite nanoparticles as pressed disks. The samples were further characterized by techniques such as transmission electron microscopy (TEM), BET surface area analysis, X-ray powder diffraction, FTIR, Raman and X-ray photoelectron spectroscopy (XPS). Particles synthesized at 6o°C in 5 minutes residence time possessed remarkably high surface area value of 264m ² g ^~1 . The proposed synthesis strategy provides a facile and economical pathway to obtain nano-sized as-prepared hydroxyapatite with high purity, very small size and ultra-low level of impurities. The use of a continuous process also offers good potential for scale up manufacture in the future.

Other Materials made using the CPFS Process

1. Other phase pure calcium phosphates were also obtained by changing the Ca/P ratio and pH of the precursor solutions and in some cases using an additional heat treatments ( -TCP, calcium-deficient hydroxyapatite (CD HA), brushite, monetite, calcium pyrophosphates) - Examples 2 and 3;

2. Sr-HA and Ba-HA Example 4;

3. A variety of anion and cation-substituted calcium phosphates (Mg, Sr, Ba, Zn, Ag, Mn Si, CO ₃ ² ) - Examples 5 and 6;

4. Surface-modified organopolymer nano dental composites have also been developed successfully - Example 7.

Example 1 - phase-pure stoichiometric hydroxyapatite using apatite made at 60. 70 and 8o°C

Materials and methods

Calcium nitrate tetrahydrate [Ca(N0 ₃ ) ₂ .4H ₂ 0, 99%] and diammonium hydrogen phosphate [(NH ₄ ) ₂ HP0 ₄ , 98 %] were purchased from Sigma-Aldrich Chemical Company (Dorset, UK). Ammonium hydroxide solution [NH ₄ OH, 28 vol%] was supplied by VWR International (UK). 10 mega-ohms deionised water was used throughout all experiments. Experimental procedure

The nano-hydroxyapatite (HA) was-prepared as follows. The reaction was carried out at 6o°C using a simple continuous (plastic) flow synthesis system (CPFS). 0.3 M diammonium hydrogen phosphate solution and 0.5 M calcium nitrate solutions were pumped into the CPFS (Ca:P molar ratio: 1.67) using pump 1 (Pi) and pump 2 (P2), respectively. The pH of both the solutions prior to the reaction was kept at pH 10. 5.0 mL and 15.0 mL of ammonium hydroxide were added to calcium nitrate (500 mL) and diammonium hydrogen phosphate solutions (500 mL), respectively. Both reagent solutions were pumped at 20 mL min ¹ , to meet at a 1/ 16 in. Polyflon™ T-piece through a 1/8 in. Polyflon™ Straight union reducer (D6-D1/8", PFA). This initial mixture were connected to 8 m long 1/16 in. Polyflon™ PTFE tubing (ID4.omm x OD6.omm) surrounded by an oil bath at the desired temperature. For flow rates of calcium nitrate and diammonium hydrogen phosphate for Pi and P2, respectively, this gives a total residence time of 5 minutes for the reaction. The product suspension was collected in a beaker at the exit point of the reactor. The aqueous suspension obtained was centrifuged at 4500 rpm for 10 minutes. The clear liquid supernatant was removed and wet solid residue was redispersed in DI water using a vortex mixer (VWR model VM-300) for 5 minutes followed by three further centrifugation and washing cycles. The wet residue obtained was freeze- dried using a VirTis™ Genesis Pilot Lyophilizer 35XL (SP Scientific, UK) at 0.3 Pa for 24 hours. The same experiment was carried out by tailoring reaction temperature ideally at 70 and 8o°C, by keeping all other parameters the same. The fine calcium apatite powders obtained from all three reactions were thus carried out at three different reaction temperatures (60, 70 and 8o°C) and are hereafter referred as HA60, HA70 and HA80, respectively. The powders were compacted into cylindrical discs of 13 mm diameter and 2 mm thickness by using a hydraulic press at 10 MPa. These as-prepared pressed samples were used for cell viability studies after sterilization. The powders were also separately heat-treated to show that the as-prepared materials were able to form phase pure synthetic hydroxyapatite that was stable at 1200°C in air (this suggested that it was stoichiometric hydroxyapatite as it didn't decompose). Material and surface characterisation

Chemical analysis

Chemical analysis of all samples was carried out using Thermo Scientific K- Alpha X-ray photoelectron spectrometer with a two chamber vacuum system (loadlock and analysis chamber). The XPS used a monochromated Al K- alpha source (E=1486.6 eV) with maximum power of 72 W. X-rays were microfocused at source to give a spot size on the sample in the range of 30 - 400 microns. The monochromator was comprised of a single toroidal quartz crystal set in a Rowland circle with a radius 250mm. The surface sensitivity (typically in the range 40 - 100 A) makes this technique ideal for measurements of elemental ratio as oxidation states. The vacuum analysis chamber pressure was at ~ 3x io ^~8 Torr. The spectrum collected included one at an energy of 150 eV for survey scans and one at 50 eV for high resolution regions. The detector was a 128 channel position sensitive detector. The spectral intensity of the Ag 3d 5/2 peak from a clean metal sample is >2.5 Mcps at a FWHM of 1.0 eV. The XPS spectra were processed using Casa™ software. The binding energy scale was calibrated by a C 1 s peak at 285.0 eV.

Transmission Electron Microscopy

Transmission electron microscopy (TEM) images were collected using a JEOL JEM-1200EX II Electron Microscope. Digital images were taken with a side mounted AMT 2K high sensitivity digital camera (Debens, UK). A small amount of sample (less than 10 mg) was dispersed in neat methanol and then gently ultrasonicated for 2 minutes to yield a very dilute suspension. A few drops of the resulting suspension were then deposited on a carbon-coated copper grid (procured from Agar Scientific, UK) which was used as the TEM specimen. The grid was dried prior to use in the double tilt holder of the TEM. Image J software (version 5.0) was used for estimating particle sizes.

Powder X-ray Diffraction

Bruker AXS D4 Endeavour™ XRD diffractometer was used for XRD collection of all samples. The data was collected in the 2Θ range from 5 to 8o° with a scanning step of 0.05 ⁰ and a count time of 2 sec-step ¹ using Cu-Κα radiation (λ = 1.5406 A). DIFFRACplus Eva™ software was used for the phase analysis of the data by spectral matching with standard patterns.

FTIR Spectroscopy

The functional groups present on the surface of HA were interrogated using Fourier

Transform Infrared spectroscopy (FTIR) using a Nicolet 6700 FTIR (ThermoScientific, UK). Photoacoustic accessory cell was used as detector for the direct analysis of the samples as powders using a helium gas. The FTIR spectra were collected in the range 400-4000 cm ¹ at resolution of 4 cm ¹ averaging 256 scans.

Raman Spectroscopy

A Confocal Raman DXR Spectrometer (SP Thermo-Scientific, UK) was used. Each powders sample was deposited onto a 316L stainless steel block using a spatula. The block was wiped clean first using distilled water then acetone prior to sample analysis. The data was collected using 780 nm laser, ten times magnification lens with the scan time of 90 seconds for each sample.

BET Surface Area Analysis

BET surface area (N ₂ adsorption) measurements of all samples were made using an ASAP 2420 micromeritics BET machine. Prior to use, sample tubes were washed with methanol and then dried overnight in an oven at 100 °C. The powder samples were accurately weighed and degassed for 12 hours at 180 °C prior to BET analyse. After degassing, they were weighed again and then analysed.

Culture of human osteoblast cells

MG63 human osteocarcoma-osteoblast cells were expanded in complete culture medium consisting of Dulbecco's Modified Eagle's Medium (DMEM) (Biosera, UK) supplemented with 10 % fetal calf serum (FCS), 2 mM L-glutamine and 100 mgmL ¹ penicillin and streptomycin (Sigma-Aldrich, Dorset, UK). Cells were kept incubated at 37°C with a 95 % Oxygen and 5 % C0 ₂ humidified atmosphere and media changes were performed every 2-3 days. HA disc-shaped pellets (13 mm diameter) were sterilised in 70 % ethanol for 1 hour and then washed well with PBS. HA discs were then either soaked in PBS or FCS for 30 min prior to seeding with 25,000 cells per disc and MG63S were used between passages 60-65. The response of MG63S to the HA discs was evaluated by cell viability assay and cell

morphological appearance.

Cell viability

Cell viability of MG63S on HA discs was assessed at days 1, 4 and 7 of culture using

Resazurin (7-hydroxy-3H-phenoxazin-3-one-io-oxide) assay. The Resazurin assay incorporates an oxidation-reduction (REDOX) indicator that fluoresces in response to metabolic activity from growing cells. Briefly, culture medium was removed from cells and a known amount of fresh medium containing 0.1 mM Resazurin sodium salt (Sigma-Aldrich, Dorset, UK) was added and cells were incubated at 37°C for 4 h. Negative controls or HA discs with no cells showed a blue coloured solution (oxidised) while samples with cells showed a purple-pink coloured solution (reduced). Reduction of Resazurin was detected using fluorescence in opaque 96-well plates on a FLx8oo microplate fluorescence reader (BioTek, UK) using wavelengths of 540 nm excitation and 635 nm emission.

Cell morphology

Cellular morphology of MG63S on HA discs was visualised at day 7 using fluorescence microscopy. Briefly, cells were fixed in 3.7 % formaldehyde and permeabilized with 0.5 % Triton-Xioo before staining with DAPI (4',6-diamidino-2-phenylindole dihydrochloride) (^g-ml/ ¹ ) and phalloidin-TRITC (phalloidin-tetramethylrhodamine B isothiocyanate) (^g/ml) (Sigma-Aldrich, Dorset, UK) for cell nucleus and actin-cytoskeleton respectively. Images were captured using an Image Express™ fluorescent microscope (Axon Instruments, UK) using the built in x20 objective and the preset DAPI (blue wavelength) and Rhodamine (red wavelength) filters. Individual images of nucleus and cytoskeleton were compiled together using Image J (Image Processing and Analysis in Java) (National Institute of Health, Bethesda, Maryland, U.S.).

Results and discussion

Transmission electron microscope images of the samples as shown in Fig. l confirmed that small crystallites had been obtained. As-prepared apatite (the precursor to phase pure HA) synthesized at 6o°C in the CPFS, the particles had a rod like morphology and average length along the longest axis of each particle was ca. 25 ± 5 nm and 5 ± 1 nm along the smaller axis (200 particles sampled). On the other hand, particles obtained at 70°C and 8o°C possessed particle size of ~ 80 ± 15 nm and ~ 95 ± 15 nm along the longest axis ca. 12 ± 5 nm and 15 ± 5 nm along the smaller axis, respectively.

It was observed that tailoring reaction temperature (6o-8o°C) and concentration while keeping the reaction time same as 5 minutes in all reactions, greatly influence the particle size and specific surface area as shown in Fig. 4[a-b]. BET surface area measurements of as- prepared HA60 possessed a remarkably high surface area of 263.9 rn ² g ^_1 . This is the highest surface area ever reported in literature for as-prepared HA powdered needles to our knowledge. Whilst HA70 and HA80 samples made using the same residence time of 5 minutes, revealed a surface area of 195.48 and 113 m ² g ^_1 , respectively. On the other hand, samples HA60, HA70 and HA80 heat-treated at 1200°C for one hour in air, revealed surface areas of 9.5 ± 0.1, 5.9 ± 0.1 and 4.9 ± 0.1 m ² g ^_1 , respectively. The powder X-ray diffraction data of as-prepared hydroxyapatites displayed broad peak typical of an apatite like structure (Fig. 5). Upon heat-treatment (i200°C for 1 hour), the X-ray diffraction peaks became considerably sharper and well resolved and gave a good match to the phase pure

hydroxyapatite reference pattern JCPDS [09-432] as shown in Fig. 6.

Other studies carried out by lowering temperature (below 55 °C and time (below 5 minutes) have shown that apatite obtained below 55 °C after heat-treatment at 1000 °C for 1 hr was not phase pure synthetic hydroxyapatite as shown in Figure 13. This observation showed that the residence time is important in order to achieve a stoichiometric hydroxyapatite and that CPFS can confer high-temperature stability of the apatite powders if the correct conditions are used. By tailoring temperature, time and pH, specific apatite products can be obtained with controlled particle size (typically in the range 25 - 95 nm) within 5 minutes at the temperature range of 60 to 80 °C. A chemical analysis of the prepared HA sample HA60 was analysed by using XPS analysis as shown in Fig. 7. The peaks at 134 eV corresponded to P 2p spectra of as-prepared

hydroxyapatite. While the binding energy values for O is and Ca 2p were measured as 532 and 347 eV, respectively. The Ca/P ratio in the analysed sample was closed to stoichiometric ratio of 1.67. The Ca 2p spectrum could be resolved into two peaks for Ca 2p ₃ / ₂ and 2 i/ ₂ (two spin-orbit pairs) at 347.4 and 351.3 eV, respectively, which are related to as-prepared hydroxyapatite

In Fig. 8b, the 2P peak can also be deconvoluted into two peaks with a spin orbit splitting for and p levels with binding energy 134.2 and 133.4 eV, respectively. Fig. 8c, depicts the core level spectrum of O is and the peaks at 530.4 and 531.8 eV are attributed to the phosphate group, and contribution of hydroxyl group in as-prepared hydroxyapatite crystal, respectively.

FTIR and Raman spectroscopy were used to analyze the samples and aid identification of different calcium phosphates. FTIR data of as-prepared apatite samples (HA60, HA70 and HA80) revealed peak at 3570 cm ¹ corresponding to stretching vibrations of the hydroxyl group associated with as-prepared HA. The intensity of peak at 1639 cm ¹ corresponding to the bending mode for lattice water suggested that loosely bonded OH groups became more incorporated into the lattice with an increase in temperature. Peaks at 1453 and 1414 cm ¹ were assigned to the stretching modes, respectively, of some adsorbed carbonate ions on the surface of as-prepared HA. Peaks at 1100 and 1031cm ¹ corresponded to the (P-O) asymmetric stretching mode of phosphate, whilst the peaks at, 602 and 564 cm ¹ revealed the presence of O-P-0 bending modes as shown in fig. 9. Raman spectroscopy was conducted on the as-prepared HA in order to supplement crystallographic data and detect substitutions in the apatite lattice. The peak at 965cm ¹ corresponded to a symmetric stretching mode of the P-0 bond in phosphate. Peaks at 610, 593 and 583 cm ¹ are likely to correspond to the bending mode of the O-P-0 linkage in phosphate. Peaks at 1078, 1049 and 1030 cm ¹ corresponded to asymmetric stretching modes of the P-0 bonds in phosphate as shown in Fig.10. An increase in intensity and sharpness of these peaks with temperature, indicated an increase in crystallinity with increasing temperature. Human osteoblast cells (MG63) were cultured on HA samples synthesised using a continuous plastic flow synthesis and cell viability was studied over the course of 7 days (Figure 11). At day 1, cell viability was similar on all HA samples indicating that initial cell attachment was not enhanced by any surface. At day 4, cell viability increased on both HA 60 and HA 70, while it remained constant on HA 80, when compared with day 1. A similar trend was observed at day 7; however, all HA samples had an increased cell viability compared with day 1 and 4. At both days 4 and 7, HA 60 had significantly higher cell viability then both HA 70 and HA 80, while HA 70 had significantly higher cell viability than HA 80. MG63 viability on commercial HA (cHA) was compared with HA samples prepared using novel CPFS method and was initially higher (day 1) than all prepared HA samples. Cell viability on cHA remained higher than HA 70 and HA 80 at all time points but at days 4 and 7 it was equal to and less than HA 60 respectively. These results indicate that newly developed CPFS technique for the production of HA is biocompatible and supports osteoblast proliferation, but that cell viability at day 7 was dependent on preparation temperature (60 > 70 > 80).

MG63 morphology was observed at day 7 (Figure. 12) on each HA type. Cells were evenly distributed on cHA and showed a typical square-like osteoblast morphology. MG63S were also evenly spread across the surface of HA 60 and showed an osteoblastic morphology but the cell cytoskeleton was also more elongated than on cHA and the cells appeared to form cell-cell bridges. On HA 70, cells were sparser and generally less-well spread than those on cHA and HA 60, with the majority showing a rounded morphology. MG63S on HA 80 were also generally quite rounded and formed clusters. These results indicate that even cell spreading and an ostoeblastic morphology are favoured on HA 60 compared with HA 70 and HA 80 and that the compare well with cHA.

The combined cell observations indicate that high surface area nanosized HA rods may have the potential to be used for biomedical applications where bone regeneration /replacement is required. Future work should address the ability of CPFS reactor produced HA to support osteogenic differentiation.

Conclusions

In summary, continuous plastic flow synthesis (CPFS) technique was used as a rapid, facile and economical route to obtain high surface area, nano-sized bioceramics with high purity and small size. In many cases, the as-prepared materials can be heat-treated to give phase pure bioceramics such as highly crystalline hydroxyapatite. Example 2 - Phase pure calcium phosphates other than stoichiometric Ca-HA. Calcium- deficient hydroxyapatite CCD HA), ίβ-tri calcium phosphate. (β-TCP) from heat-treatment of CDHA)

Experimental

0.3 M diammonium hydrogen phosphate solution and 0.5 M calcium nitrate solutions were used (Ca : P molar ratio: 1.67). The pH of both the solutions prior to the reaction was kept above pH 10. 5.0 ml and 15.0 ml of ammonium hydroxide were added to calcium nitrate (500 ml) and diammonium hydrogen phosphate solutions (500 ml), respectively. The continuous plastic flow synthesis was carried out as previously described. Other calcium phosphate phases (β-TCP, CDHA) were also obtained by changing the Ca/P ratio (i.5~ 1.67) by keeping other conditions same. β-TCP is obtained by the heat-treatment of as-prepared CDHA at iioo°C. The influence of initial Ca/P molar ratio, pH and precipitation temperature on the phase evolution and crystallinity of the nanopowder were systematically investigated and optimised. Calcium deficient hydroxyapatite, (CDHA) has gained importance as components in injectable bone cements; regulating particle properties is often used to control cement setting behaviour. In contrast, beta-tricalcium phosphate, β-TCP,

[Ca3(P04)2], another bioceramic, has a high dissolution rate compared to HA in vivo. Thus, the use of biphasic calcium phosphate ceramics composed of HA and β-TCP mixtures for bone grafts can assist rapid bone formation around the implant site compared to HA alone.

Transmission electron microscope images of the samples as shown in Fig. 14 confirmed the synthesis of small crystallites. The average length along the longest axis of each particle was ca. 80 ± 15 nm (200 particles sampled) for CPFS phase pure as-prepared hydroxyapatite with the particles having rod-like morphology. On the other hand, as- prepared apatites which form CDHA (after heat treatement) had different morphology and possessed particle size of ca. 95 ± 15 nm. TEM images of β-TCF (heat-treated product) possessed spherical morphology with the average particle size of ~ 45 ± 15 nm (100 particles sampled). BET surface area measurements (N2 adsorption) of as precipitated HA and CDHA samples yielded sufrace areas of ca 195 and 80 m ² g ¹ , respectively, while heat-treated samples at iooo°C for one hour revealed surface area of 4.9 ± 0.1 and 5.9 ± 0.1 m ² g ¹ , respectively. The results of the present study are in agreement with previous reports that indicate that the apatite made at ambient conditions possess a surface area of 90 m ² g ¹ after oven-drying and 113 m ² g ¹ after freeze-drying. The XRD results (Figure 15) confirmed that the products obtained after heat-treatements at various conditions (HA, β-TCP, CD HA and biphasic HA / β-TCP, whitlockite), were phase pure. The lattice constants of heat-treated hydroxyapatite were similar to reference JCPDS Pattern no. 09-432. No decomposition of HA into other phases was observed even after heating at 1200°C in air for lh. This observation showed the superior high-temperature stability of such "biomimetic' HA powders made under the right conditions where the as-prepared material is stoichiometric. The XRD plot obtained for the sample -TCP (from heat -treatment of CD HA at iooo°C) showed good agreement and excellent match with JCPDS pattern 09-0426 Synthetic TCP [Ca ₃ (P0 ₄ ) ₂ ].

The weak absorption peak at 880 cm ¹ was assigned to the P-O-H vibration in the HPO ₄ ^2" group which exists in an apatite which was heat-treated to give non- stoichiometric HA. Peak at 1031 cm ¹ corresponds to the P-0 asymmetric stretching mode of phosphate, whilst the peaks at, 534 and 466 cm ¹ correspond to O-P-0 bending modes. An increase in intensity and sharpness of these peaks with temperature, indicates an increase in crystallinity.

FTIR was used to analyze the samples and aid identification of different calcium phosphates. FTIR data revealed peaks at 3420 and 632 cm ¹ (corresponding to stretching vibrations of the hydroxyl group in as-prepared HA. The intensity of a peak at 1637 cm ¹ , corresponding to the bending mode for lattice water, was observed to decrease with increasing synthesis temperature. Peaks at 1453 and 1414 cm ¹ correspond to the stretching modes, respectively, of some substituted carbonate in the as-prepared HA. Human osteoblast cell proliferation study was conducted on all 4 samples. The osteoblast cells cultured on all ceramic samples showed continuous proliferation. Cells were seen to attach, spread and grow on all types of samples as shown in fig. 17. A significant increase in the growth of osteoblast cells with culturing time was observed on all samples, and significantly higher level of viability was observed for heat-treated β-TCP product. This could be due to the faster dissolution rate of β-TCP sample as compared to as-prepared HA.

In summary, Continuous plastic flow synthesis (CPFS) technique provides a rapid (as low as 5 minutes), facile and economical pathway to obtain nano-sized HA and other calcium phosphate bioceramics with high purity, suitable size and ultra low level of impurities. The results of the in vitro studies showed that the all 4 synthesized samples are non-cytotoxic and biocompatible. Hence the current work deals with the preparation of synthetic calcium phosphates with optimum properties closer to those of living hard tissues like bone and teeth, aiming at better and more effective biomedical ceramics for use as powders or as nanocomposites in future efforts. Importantly, by using a flow system the work is readily scalable and thus has promise for scale-up.

Example 3 - Dicalcium phosphate dihydrate fbrushite, CaHPO _d .2H ₂ o), Dicalcium phosphate anhydrous fmonetite. CaHPO^) and Calcium pyrophosphate (Ca ₂ P ₂ C from heat-treatment of brushite) Experimental

The phase pure brushite was-prepared at room temperature using a simple continuous (plastic) flow synthesis (CPFS) system. 0.6 M diammonium hydrogen phosphate solution and 0.5 M calcium nitrate solutions were pumped in the CPFS (Ca : P molar ratio: 0.8) using pump 1 and pump 2, respectively. The continuous plastic flow synthesis was carried out as previously described.

Introduction

Phase pure, stable brushite (DCPD: dicalcium phosphate dihydrate: CaHP0 ₄ .2H ₂ 0) nanoparticles (<ioonm) were rapidly synthesised using a novel two pump continuous plastic flow synthesis (CPFS) at room temperature (22°C) in 3 minutes (residence time) at the conditions of pH 5.5 from aqueous solution of calcium nitrate tetrahydrate and diammonium hydrogen phosphate. The product was obtained as a phase pure material with a Ca:P molar ratio for the reagents of 0.8, and without the need for an initial prolonged ageing step. The most common approach for synthesizing brushite is by adding water soluble calcium (e.g., CaCl ₂ .2H ₂ 0, Ca(N0 ₃ ) ₂ .4H ₂ 0, or Ca(CH ₃ C00) ₂ .H ₂ 0) and phosphate (e.g., NH ₄ H ₂ P0 ₄ , (NH ₄ ) ₂ HP0 ₄ , Na ₂ HP0 ₄ , NaH ₂ P0 ₄ , KH ₂ P0 ₄ or K ₂ HP0 ₄ ) salts upon adjusting the Ca/P molar ratio to 1. The resultant product is usually washed with distilled H ₂ 0, and air-dried. On the other hand, pure DCPD powders can also be synthesized by reacting a suspension of Ca(0H) ₂ with stoichiometric amounts of H ₃ P0 ₄ by keeping the solution pH in the acidic range. It may also be prepared by mixing two phosphate powders in the presence of water. The starting powders are β-TCP (Ca ₃ (P0 ₄ ) ₂ ) and monocalcium phosphate monohydrate [Ca(H ₂ P0 ₄ ) ₂ Ή ₂ 0]. A small amount of sodium pyrophosphate (Na ₂ H ₂ P ₂ 0 ₇ ) is added to the starting powders as a setting regulator. Mixing of the powders is done in a sulfuric acid solution. Brushite is usually transformed into monetite by losing its crystal water upon heating at or above 110 °C. Relying on the experimental solubility values of different calcium phosphate phases recently reported by Tang et al., it is noted that brushite has 3.4 times greater dissolution rate as compare to TCP at a pH value of 5.5.

The main purpose of this study was to develop robust chemical synthesis procedure for the synthesis of high purity brushite and its thermal transformation to monetite and calcium pyrophosphates, respectively, by heat-treatment of the as-prepared materials at different temperatures. The precipitation of brushite does not take place at the start of the reaction.

Ferreira et al. reported different stages involved in the formation of brushite. The first step is the instantaneous precipitation of HA (Eq. 1) and as a result decrease in pH followed by the formation of brushite nucleate. At this stage two species, brushite and HA, coexist in the solution. Recent reports have indicated that the crystallization rate of brushite is much higher than that of HA at room temperature.

Thus, the existing newly formed HAP is transformed into brushite, accounting for the increase of pH as shown in Eq. 2.

Ca ₅ OH(P0 ₄ ) ₃ + 2H ₃ PO ₄ + 9H ₂ 0 * 5CaHP0 ₄ + 2H ₂ 0 (1)

Ca(0H) ₂ + H ₃ PO ₄ CaHPC-4 2H ₂ 0 (2)

Results and Discussion

Transmission electron microscope images of the samples as shown in Fig. 18 confirmed the synthesis of small crystallites. The average length along the longest axis of each particle was ~ 35 ± 15 nm (200 particles sampled) for CPFS phase pure as-prepared HA, with the particles having rod-like morphology. XRD analysis indicated that the as-prepared products obtained at a lower temperature were brushite particles. The XRD spectra shown in Fig. 19 indicate that the maximum peak intensities are at 2Θ values of around 11.6 ⁰ , 20.9 ⁰ , 23.6° and 29.3 attributed to the Miller indices of (020), (121), (040) and (112) of brushite. The spectra confirmed that the products obtained were phase pure and were mainly composed of brushite. The intensity of the diffraction peaks indicated that the particles were fairly well- crystallized.

Brushite is transformed into monetite by losing its crystal water upon heating above 110 °C as described in Eq. 3. The presence of calcium pyrophosphate in a heat-treated sample indicates the presence of HP0 ₄ ^2_ in the as-prepared material. Calcium pyrophosphate is formed by the loss of one H ₂ 0 molecule from two HP0 ₄ ^2~ groups of brushite under high- temperature conditions. The transition of monetite into pyrophosphate typically occurs at temperatures above 400 °C as described in Eq. 4.

CaHP0 ₄ -2H ₂ 0→ CaHP0 ₄ + 2H ₂ 0 (3) 2CaHP04→ Ca2P207 + H2O (4)

The results of the present study are in agreement with previous reports that indicate that the brushite transformed into monetite at around i8o°C and the monetite to calcium

pyrophosphate transition was completed above 440°C. Thus, alpha calcium pyrophosphate phase was observed at 500°C and 700°C. On the other hand, beta-calcium pyrophosphate phase obtained after heat-treatment in the range 850 - iooo°C. Brushite formation was also confirmed by FTIR spectral analysis. Fig. 20 shows the FTIR spectra of products obtained at different temperatures. The spectra exhibit easily distinguishable bands attributed to P0 ₄ 3- for the as-prepared form of DCPD. Bands at around 525 crrr ¹ and 575 cm-i were attributed to the v4 bending vibrations of the P-O-P mode. Bands at 873 cm-i and 1225 cm-i assigned to the HPO ₄ ^2" group of as-prepared brushite were observed in all samples. A band at 985 cm ^-1 originated from the P-O(H) vi symmetric stretching vibration of P0 ₄ ^3~ . In addition, the peak at around 1060 cm ^-1 was attributed to the νβ vibration of the P0 ₄ ^3~ group. A broad band between 3400-3600 crrr ¹ was observed, due to the stretching (vs) mode of H-bonded OH- or water. It was clear that the latter was formed by the loss of one H ₂ 0 molecule from two brushite HP0 ₂ ^4_ groups under high-temperature conditions. The FTIR spectra thus show that the as-prepared nanoparticles obtained were brushite, whereas calcium pyrophosphate was obtained after high-temperature heat-treatment of this as-prepared brushite.

The osteoblast cells cultured on as-prepared brushite samples (pressed disks) showed continuous proliferation. Cells were seen to attach, spread and grow on all types of samples as shown in Fig. 21. A significant increase in the growth of osteoblast cells with culturing time was observed. Different phases were observed in the synthesis of brushite. The first phase was poorly crystalline HA followed by the appearance of few crystals of brushite. Both phases coexist in the solution for first two minutes and then finally poorly crystalline HA transformed to phase pure stable brushite crystals. The reaction pH and temperature play an important role in the formation of single phase as-prepared brushite. Because the crystallization rate of brushite at room temperature and pH 6 is 1000 times greater than the poorly crystalline HA, which explains why the final as-prepared product of reaction was single phase brushite. Table 1 - A summary of calcium phosphates and the pH ranges under which the as-prepared materials were made fin some cases a heat-treatment was required to access the named phase)

Example 4 - Phase pure Sr-HA and Ba-HA

Experimental

The as-prepared pure strontium and Ba-hydroxyapatite (containing no calcium) were prepared at 6o°C using a simple continuous (plastic) flow synthesis system (CPFS) . In the synthesis process, 0.5M basic solutions of strontium nitrate with (Sr/P = 1.67) or barium hydroxide solutions with (Ba : P molar ratio: 1.67) and 0.3 M diammonium hydrogen phosphate, respectively, were pumped to meet at a T-piece .The pH of both the solutions prior to the reaction was ideally kept above pH 12. This initial mixture was connected to 8 m long pipe which was coiled inside an oil bath which gave an effective 5 minute residence time from the tee to the exit of the pipe.

Introduction

Strontium and barium ions are divalent and have the ability to be incorporated into hydroxyapallte crystals. Among the many bivalent cations that can be substituted for calcium in hydroxyapatite, strontium is of interest because of its beneficial effect on bone formation and prevention of bone resorption. Strontium and barium hydroxyapatite (with no calcium) have gained interest in recent, years because of their high density and radiopaque properties. The doping of Sr in the hydroxyapatite structure for implants has been shown to increase bone mineral density (BMD), bone mineral content (BMC), bone volume and

microarchitecture, reduced bone fracture risk and improved the mechanical performance of whole bone without affecting the estimated material properties of the bone tissue. Kurata et al. prepared Sr- and Ba-HA (no calcium) at various pH and temperatures, and indicated that Ba-HA was formed above pH 13 at room temperature. However, sintering and cytotoxicity studies of Ba-HA were not reported. Ba-HA has greater density compared to calcium hydroxyapatite, and this property is useful for a X-ray opaque bone filler if Ba-HA is nontoxic.

Radiopaque dental materials are beneficial in making diagnosis and monitoring existing restoration .However, there is still a lack of dental material that is tooth coloured, minimally invasive and has satisfactory radiopacity. In a recent study on dental restorative materials containing Sr and Ba-HA (no calcium), it was been found that Sr-HA appeared to be a better radiopaque filler when it is mixed as a composite with a polymeric matrix as shown in figure 23·

Example 5 - Anion-substitution - Carbonate C0^ ^2~ ). Silicate (SiO^ ^4" )

Experimental

Carbonate (C0 ₃ ² )-substituted calcium phosphate:

A series of carbonate substitution reactions were carried out for this study. The continuous polymer flow synthesis was carried out as previously described. Urea (carbonate ion source) was added to diammonium hydrogen phosphate (adjusted pH 10). This solution was reacted with calcium nitrate solution (adjusted pH 11). A Ca:[P0 ₄ 3+C0 ₃ ² ] molar ratio of 1.67 was maintained in all precursor solutions. The pH of both the solutions was adjusted using neat ammonium hydroxide solution. It was found that carbonate replaces phosphate in HA lattice, this is known as to B-type carbonate substitution. Samples are labelled as 2CHA, 4CHA, 6CHA and 8CHA. The initial numbers in the sample IDs show the nominal carbonate content (wt%).

Silicate (Si0 ₄ ² )- substituted calcium phosphate:

a similar procedure was adopted for the synthesis of silicon substituted calcium phosphates using CPFS system. In total, nine reactions were carried out using silicon acetate as a silicate ion source. The samples are labelled as iSi-HA, 2S1-HA, 3S1-HA, 4S1-HA, 5S1-HA, 6S1-HA, 7S1-HA, and 8S1-HA, respectively. The numbers in the sample IDs represent the nominal silicon content (wt%) according to how much silicate precursor was used in the precursors.

Introduction

Biological and physiochemical properties of HA can be enhanced by the substitution with ions, some of which are also usually present in natural bone apatite. Most natural apatites are non-stoichiometric because of the presence of minor constituents such as metal cations (Mg ²⁺ , Mn ²⁺ , Ag ⁺ Zn ²⁺ , Na ⁺ , Sr ²⁺ ) or anions (HP0 ₄ ^2" or C0 ₃ ² -). HA is capable of

accommodating substitute ions within its lattice. Trace ions substituted into apatite-like phases can have a dramatic effect crystal structure parameters, crystallinity, dissolution kinetics, bioactivity and other physical properties of as-prepared bioceramics.

Silicon-substituted HA (incorporating silicate anions) has been synthesised using wet- precipitation and batch hydrothermal techniques. Gibson et al. produced phase pure Si-HA by an aqueous precipitation of a calcium containing solution and a phosphate containing solution at high pH by using Si acetate as the source of silicate ions. There have also been many reports on the development of silicon-substituted HA coatings on metallic substrates for enhanced osseointegration. Silicon enters the HA lattice in the form of silicate ions which substitute phosphate ions. Silicon levels up to 4 wt% in HA have been identified using a batch hydrothermal process.

In this research, carbonate and silicate substituted hydroxyapatite powder was synthesized via a continuous plastic flow synthesis reactor at 70°C in 5 minutes (residence time) at the conditions of pH 10-11. An in-vitro study evaluated the biocompatibility and osteoblast cell proliferation / attachment on the surface of these nanoparticles as pressed disk. The obtained powders were physically characterized using transmission electron microscopy, BET surface area analysis, X-ray powder diffraction analysis, and FTIR. Dynamic light scattering was used to evaluate the size of particles which were made at different

concentrations. Results and Discussion

Transmission electron microscopy was used to analyse the particle size and morphology. TEM images of 6 wt% C0 ₃ ^2~ sample as shown in Fig. 24(a) and (b) confirmed the synthesis of small crystallites. The average length along the longest axis of each particle was ~ 70 ± 15 nm (200 particles sampled), with the particles having rod-like morphology. TEM images were also collected for Si substituted HA to investigate the particle morphology and size with increasing silicon content. Fig. 25(a) and (b) for sample 6S1-HA reveal distinct nanorods of size ~ 110 ± 15 nm (200 particles sampled), along the longest axis. It was observed that the presence of silicon in HA lattice slightly increased the particle size as compared to pure HA as shown in Fig. 25(a) and (b). Fig. 26 shows the trends in BET surface areas for the carbonate substituted calcium phosphate samples (determined using XPS). There was a little change in the BET surface area by increasing urea content. All carbonate substituted samples had the surface areas in the range 111 - 136 rr^g- ¹ . One of the possible reasons for lower surface area with small particle size was the increase in particle agglomeration with small size as observed in TEM (Fig 24(b)). In contrast, sample 6S1-HA showed a noticeable increase in surface area in the range 113 - 163 m ² g ^_1 which might be due to a slight difference in size or agglomeration as compared to the carbonate [see Fig. 24 and 25(a) and (b)].

Particle size distribution was also calculated for selected samples, pure HA, 2CHA, 4CHA, 8CHA, 2S1HA, 4S1HA and 8S1HA, respectively. DLS measurements of pure hydroxyapatite sample synthesized at 70 °C in 5 minutes residence time reveals average hydrodynamic radius of ca. 154 and polydispersity value of 0.201. For samples 2CHA, 4CHA, 8CHA, 2S1HA, 4S1HA and 8S1HA, respectively, DLS measurements yielded average hydrodynamic radii of ca. 146, 142, 149, 156, 164, and 174 nm. Whilst PDI values of 0.268, 0.264, 0.272, 0.207, 0.239 and 0.231, respectively, were recorded. It was observed that the trends rather than the absolute values of the DLS measurements are in good agreement with TEM determined distributions.

Powder X-ray diffraction data was obtained for all samples to investigate how carbonate and silicate substitution influenced phase composition and phase purity. The XRD pattern suggests an apatite-like structure (sample 2wt% C0 ₃ -HA in Fig. 27) and showed a good match to phase pure HA [compared to JCPDS pattern 09-432]. All XRD patterns for silicate- substituted samples in Fig. 28 have a good match to hydroxyapatite JCPDS pattern 09-432 (and are therefore phase pure).

FTIR spectroscopy was carried out on all as-prepared carbonate- and silicate-substituted samples in order to aid observations made using XRD. The FTIR analysis (Fig. 29) detected strong peaks at the wavenumbers of the B-type CHA (870, 1430 and 1450 cm ¹ )- The typical peaks of the A-type CHA (880, 1450 and 1540 cm ⁴ ) were not evident. The FTIR spectrum for as-prepared sample SiHA in Fig. 30 revealed peaks similar to those observed for as-prepared carbonate-substituted samples. However, the weak band in the range 1565-1380 cm ¹ (corresponding to asymmetric stretching of the C-0 band of C0 ₃ ^2~ group in both the as- prepared A- and B-type carbonate substitutions in HA) was understandably much lower in intensity as compared to the similar band seen in the FTIR spectrum of the carbonate substitution samples in Fig 30. The weak peak centred at 872 cm ¹ was due to the bending mode of the O-C-0 linkage in a small amount of carbonate which is present in the as- prepared material. This was also lower in intensity as compared to a similar peak observed at 876 cm ¹ in Fig. 29 (due to higher amount of carbonate ions present in the carbonate- substituted samples). A notable difference from the FTIR spectra in Fig. 29 was that the OH - ¹ stretching peak at about 3571 cm ¹ decreased in intensity with increasing silicon substitution (Fig. 30). This supported the hypothesis that silicon was being substituted in the HA lattice in the form of silicate, Si0 ₄ ^4_ , partially substituting phosphate, P0 ₄ ³ . The most notable effect of silicon substitution on the FTIR spectra of hydroxyapatite is the changes in the P0 ₄ bands between the ranges 800 - 1100 crrr ¹ and 500-700 cm- ¹ .

Human osteoblast cell proliferation study was conducted on selected carbonate (2CHA, 4CHA, 6CHA) and silicate (2S1HA, 4S1HA, 8S1HA) samples. The pure as-prepared HA sample made via CPFS at 70 °C (reaction time = 5 min) was used as a control. The osteoblast cells cultured on all as-prepared ceramic samples showed continuous proliferation. Cells were seen to attach, spread and grow on all types of as-prepared samples as shown in Fig. 31. Cell spreading was less pronounced in as-prepared carbonate samples compared to as- prepared Silicate substituted samples. A significant increase in the growth of osteoblast cells with culturing time was observed on all samples, and significantly higher level of viability was observed for as-prepared samples 4S1HA and 6S1HA, respectively. At day 7, cells were seen adhering and spreading on top of each other for all samples, forming a multilayer of cells without any specific orientation. Results revealed better bone cell cytoskeletal organisation and greater growth activity for osteoblast cells cultured on as-prepared SiHA than on as-prepared CHA. This could be due to the faster dissolution or higher surface area of the SiHA samples compared to CHA.

A CPFS reactor was used to successfully synthesise ion substituted calcium phosphates from calcium nitrate tetrahydrate [(Ca(N0 ₃ ) ₂ .4H ₂ 0), and diammonium hydrogen phosphate (NH ₄ ) ₂ HP0 ₄ )] precursor solutions at (near) ambient conditions in a rapid single step.

Careful control of the quantities of reactants used, resulted in phase pure nano-sized carbonate and silicon substituted hydroxyapatite that retained the apatite structure after heat-treatment at 1000 °C for 1 h. Cell toxicity analysis on the as-prepared materials confirmed excellent biocompatiblity of these materials. With its unique features, the obtained product is a promising material with biological properties and has potential to be used in biomedical applications where small size and fine particle size distribution control may be beneficial e.g. injectables for spinal fusion or as a filler in a biocomposite.

Example 6 - Cation-substitution of HA with Magnesium (Ms ). Strontium (Sr )

Experimental

All samples were made using the continuous plastic flow synthesis as previously described. In the synthesis process, basic solutions of calcium nitrate with different wt% of metal ions ((Ca + Mg)/P = 1.67 - magnesium nitrate hexahydrate [Mg(N0 ₃ ) ₂ .6H ₂ 0, 97%] or (Ca + Sr)/P = 1.67 - strontium nitrate hexahydrate [Sr(N0 ₃ ) ₂ .6H ₂ 0, 97%]) and diammonium hydrogen phosphate were used, respectively.

Introduction

Nano-sized magnesium substituted calcium phosphate bioceramics (less than 100 nm) were prepared by using a continuous plastic flow synthesis (CPFS) system at 70 °C in 5 minutes (residence time) at a pH of ca. 10. Initially, phase pure hydroxyapatite and magnesium substituted hydroxyapatite were prepared with a BET surface area of 160 m ² g ^_1 and 139 m ² g ^_1 , respectively using the CPFS system. Biphasic mixtures of as- prepared Mg-HA and phase pure Mg-whitlockite were also obtained upon increasing the amount of magnesium in the reagents. This was accompanied by a BET surface area to 75 rr^g- ¹ in case of pure Mg-whitlockite. The as-prepared apatites appeared to be less crystalline and more agglomerated with an increase of magnesium content. Continuous plastic flow synthesis has been used as a simple, low cost and efficient route to produce a range of as- prepared strontium and barium-substituted calcium phosphate bioceramics. Such bioceramics are candidates for injectable bone replacement applications and provide better mechanical strength to calcium phosphate implants. In this research, strontium and barium substituted hydroxyapatite powder was synthesized via a continuous plastic flow synthesis reactor at 70°C in 5 minutes (residence time) at the conditions of pH 10-11.

An in-vitro study evaluated the biocompatibility and osteoblast cell proliferation / attachment on the surface of these as-prepared nanoparticles as a pressed disk. The as- prepared powders were physically characterized using transmission electron microscopy, BET surface area analysis, X-ray powder diffraction analysis, FTIR, Raman spectroscopy and X-ray photoelectron spectroscopy (XPS). Radiographs of the as-prepared sample discs were taken adjacent to an Al wedge for radiopacity investigation.

Most biological apatites are non-stoichiometric because of the presence of minor

constituents which include cations e.g. Mg ²⁺ , Mn ²⁺ , Ag ⁺ , Zn ²⁺ , Na ⁺ , Sr ²⁺ or anions e.g.

HP0 ₄ ^2" or C0 ₃ ^2" . Synthetic HA is capable of accepting substitute ions within its lattice. Trace ions substituted in bioceramics can effect lattice parameters as well as crystallinity, dissolution kinetics and other physical properties.

Being in group 2, magnesium (Mg ²⁺ ) is a divalent cationic substitute for calcium (Ca ²⁺ ) in the HA lattice. Such substitution often reduces crystallinity, increases solubility, and lowers the temperature at which conversion of as-prepared hydroxyapatite into β-TCP can occur. As the amount of substituted magnesium affects thermal stability of the apatite or other phases, this has implications for sintering behavior. For example, the β-TCP to a-TCP phase transformation normally occurs at ca. ii8o°C for the pure calcium-based compounds.

However, Mg ²⁺ substitution for Ca ²⁺ can increase the transformation temperature to ca. 1500 °C. This enables enhanced sintering of β-TCP at elevated temperatures without deleterious formation of a-TCP (the latter is a less bioactive polymorph). Magnesium substitution levels of up to 1.6 wt% have been reported using wet precipitation reactions. Surprisingly, as much as ca. 28.4 wt% substitution of Mg ²⁺ for Ca ²⁺ in HA has also been claimed in the literature using mechano-chemical synthesis routes. However, an excess of Mg ²⁺ can also be detrimental as it is known to reduce bioactivity in certain biomaterials.

Mg ²⁺ as one of the key substitutes for Ca ²⁺ in natural apatites is expected to have reasonable biocompatibility and biological properties. Human bone, enamel and dentine comprise 0.72, 0.44, and 1.23 wt% of Mg ²⁺ , respectively, and this may play an important role in the initial formation of tooth apatites and have a significant effect on their physiochemical properties. The complete substitution of Mg ²⁺ for Ca ²⁺ in HA has been shown to inhibit the formation of an extracellular matrix and has deleterious effect on bone cells.

As well as Mg-HA, other phases such as Mg-whitlockite (or Mg a-TCP), can also be made directly from precipitation reactions at relatively low temperatures (below 95°C) and acidic or neutral pH [15-16]. For Example, Mg ²⁺ ions can stabilise the formation of

Ca ₂ . ₇ Mgo.3(P0 ₄ )2.H ₂ 0 [17]. The main aim of this present study is to produce ion substituted bioceramics with unique physical or compositional attributes (ion substitution level, particle size, crystallinity and phase composition) that may possess novel properties. Among the bivalent cations that can replace calcium in CaHA, strontium has gained interest for its possible biological role. Strontium is present in the mineral phase of the bone, and connective tissue has 320-400 mg of strontium. Clinically strontium has been observed to exert several in vivo effects on bone especially at regions of high metabolic turn-over, and its beneficial effect in the treatment of osteoporosis is well known. In vitro, strontium increases the number of osteoblast cells followed by the reduction osteoclast cells, whereas strontium administration decreases bone resorption and stimulates bone formation. In addition to its antiresorptive activity, strontium has been found to have anti-osteoporosis and

antiosteopenic activity in animal models. Strontium compounds have been demonstrated to have beneficial effects in osteoporosis by increasing mechanical performance of bone in animal models.

Strontium can replace calcium in the HA structure in the complete range of compositions. The solid solutions, which have been obtained by hydrothermal methods or by treatment at high temperatures, display a linear variation in the lattice parameters with composition, whereas different data are reported on the preferential replacement site of Sr for Ca in Ca- HA.

The present study was undertaken to evaluate the effect of strontium dopant concentration in calcium hydroxyapatite on the radiopacity and its application as filler component to develop a radiopaque resin infiltrant for dental applications. As radiopaque dental materials are beneficial in making diagnosis and monitoring existing restoration.

Results and discussion - Mg-substituted

A customized synthesis of magnesium substituted calcium phosphates was made using CPFS system which has capacity to prepare nanobioceramic materials with controllable particle size by tailoring the reaction parameters like temperature, pH and flow rates of precursors. Powder X-ray diffraction data were collected for all samples to investigate how Mg substitution affected phase composition and phase purity. The PXRD data for samples up to 3wt% Mg-CaP, shown in Fig. 32 gave a good match to the line pattern for crystalline HA

[ICDD pattern 09-0432, Ca ₅ (P0 ₄ ) ₃ (OH)]. Samples 6Mg-CaP and 7Mg-CaP were identified as biphasic mixtures, with a good match to the aforementioned patterns for HA and whitlockite, respectively. Increasing the magnesium content further (8Mg-CaP and loMg-CaP) resulted in significantly broader PXRD peaks assigned as phase pure Mg-whitlockite. This was expected as magnesium is known to inhibit the crystallisation and growth of calcium phosphates in solution and to stabilise whitlockite. For selected samples, Transmission electron microscope images were collected to investigate the particle morphology with increasing magnesium content or with changes in phase. TEM images of samples oMg-CaP are shown in Fig. 33(a) reveals distinct nanorods of size ~ 85 ± 15 nm (100 particles sampled), along the longest axis and ~ 15 ± 5 nm (100 particles sampled), along the smaller axis, respectively. TEM images of sample lMg-CaP [Fig. 33(b) ] reveal small rod shaped agglomerates with the average particle size of ~ 70 ± 16 nm (100 particles sampled). Whilst sample 6Mg-CaP [Fig. 33(c)] reveals semi-spherical shaped morphology, suggesting a deviation from the rod like morphology. On the other hand, TEM images of sample loMg-CaP possessed spherical morphology in large agglomerations with the average particle size of ~ 35 ± 15 nm (100 particles sampled). The TEM images reveal that rounded Mg-whitlockite (identified from PXRD) particles are possibly hollow as shown in Fig. 33(d). Fig. 34 shows the trends in BET surface areas for the magnesium substituted calcium phosphate samples. For phase pure HA, the BET surface area was 160 m ² g ¹ . Samples from 4Mg-CaP and 7Mg-CaP had lower surface area of 91 and 75 m ² g ^_1 , respectively and phase pure loMg-whitlockite had lower surface area than pure HA. One of the possible reason of lower surface area is the particle agglomeration observed in TEM (Fig. 33 (c-d). On the other hand, samples 8Mg-CaP and loMg-CaP (phase pure Mg-whitlockite) revealed a noticeable increase in surface area (102 rr^g- ¹ ). The afformentioned results can be supported by XRD and TEM as shown in Fig. 32 and Fig. 33 (d).

A chemical analysis of magnesium substituted sample (8Mg-CaP) were performed by using XPS analysis as shown in Fig. 35. The general scan and the C is, P 2p, Ca 2p, O is, and Mg 2p core level spectra of Mg-HA were taken. The Ca 2p spectrum could be resolved into two peaks for Ca 2p ₃ / ₂ and 2 i/ ₂ at 347.4 and 351.3 eV, respectively, which are related to hydroxyapatite. In Fig. 35b, the 2P peak can also be deconvoluted into two peaks for p^ and p levels with binding energy 134.2 and 133.4 eV, respectively. Fig. 35C, depicts the core level spectrum of O is and the peaks at 530.4 and 531.8 ev are attributed to the phosphate group, and adsorbed water in hydroxyapatite crystal, respectively. The Mg is core level spectrum could be resolved into two peaks with the binding energy 1304 and 1307 eV, respectively.

The magnesium substitution levels in powder samples as suggested by XPS analysis were 0.5 wt% for the small rods in lMg-CaP, 3.5 wt% for the agglomerates in 4Mg-CaP and 8.3 wt% for loMg-CaP, respectively.

Human osteoblast cell proliferation study has been conducted on some selected as-prepared and heat-treated samples oMg-CaP, 4Mg-CaP and 8Mg-CaP, respectively (Fig. 36). The osteoblasts cultivated on all ceramic samples showed the continuous proliferation. But significantly higher level of viability was noticed for as synthesized samples as compare to heat-treated samples at 700°C. Among these samples, 8Mg-CaP AP (Mg-whitlockite as shown by PXRD) stimulated the maximum cell proliferation. The cell proliferation results on day 7 showed that there was a continuous growth of cells on all experimental samples. Cells formed bridges across the undulations and spread over them which showed that material was non-toxic for cells. In summary, continuous (plastic) flow synthesis (CPFS) technique provides a rapid pathway to synthesize pure HA and a series of Mg-substituted calcium phosphates from calcium nitrate tetrahydrate (Ca(N0 ₃ ) ₂ .4H ₂ 0), and diammonium hydrogen phosphate (NH ₄ ) ₂ HP0 ₄ ) solutions as starting material at (near) ambient conditions. Mg- substituted HA nanorods were obtained at low Mg-substitution levels. However at higher Mg-substitution levels, biphasic mixture and phase pure Mg-whitlockite were obtained, respectively. Cell toxicity analysis confirmed the high biocompatiblity of this material. With its unique features, the obtained nanopowder is a promising material that has potential to be used in biomedical applications where bone regeneration / replacement or controlled resorbibility of bone grafts is a vital requirment. Results and discussion - Sr-Substituted

The X-ray diffraction patterns of the solid products synthesized with different Sr molar ratios are shown in Figure 37. All the patterns indicated that they are constituted of hydroxyapatite as a unique crystalline phase. The patterns of the samples containing both Ca and Sr generally exhibit broader diffraction peaks, in agreement with a reduced degree of crystallinity of the mixed strontium doped calcium hydroxyapatite. The broadening is more evident for the samples with smaller Sr contents, suggesting a greater difficulty for Ca-HA to host the larger strontium ion than for Sr-HA to host the smaller calcium ion.

Human osteoblast cell proliferation study was carried out on selected samples sSrHA, loSr- HA and isSrHA, respectively (Figure 39). Pure HA sample made on CPFS system was used as a control. There were no substantial differences in the cell morphology of all three samples. The cells were round and almost flattened against the surface that showed the colony formation and attributed to continuous proliferation. The deconvoluted spectra for O, Ca and P for as-prepared strontium substituted

hydroxyapatite (loSrHA) made at 70 °C are presented in Figure 40. The peaks at 134 eV corresponded to P 2p of the phosphate groups in hydroxyapatite. The Ca 2p spectrum could be resolved into two peaks for Ca 2p ₃ / ₂ and 2 i/ ₂ (two spin-orbit pairs) at 347.4 and 351.3 eV, respectively, which are related to hydroxyapatite. In Fig. 41b, the 2P peak can also be deconvoluted into two peaks with a spin orbit splitting for p^ and p ₃ / ₂ levels with binding energy 134.2 and 133.4 eV, respectively. Fig. 41c, depicts the core level spectrum of O is and the peaks at 530.4 and 531.8 eV are attributed to the phosphate group, and adsorbed water in hydroxyapatite crystal, respectively. While the binding energy values for Sr 3d spectrum were measured as 133.2 and 134.8 eV, respectively.

Strontium substituted calcium hydroxyapatite and pure strontium hydroxyapatite

(containing no calcium) have great potential to use for dental nanocomposites because of radiopacity properties. It was observed that 100% substitution of strontium for Calcium in the hydroxyapatite structure has highest level of radiopacity as compared to 5, 20 and 50% substitution and could be used as a suitable filler component to develop a novel radiopaque resin infiltrant for dental applications (Figure 42). In summary, continuous plastic flow synthesis (CPFS) technique was used as a rapid, facile and economical route to obtain high surface area, nano-sized Sr with high purity and better opacity in x-ray images. The results of present study have shown that the addition of Sr in dental composites is a useful method for increasing the radiopacity of dental restorative. The findings reported here should be helpful in future clinical investigations.

Table 2 - ion substituted calcium phosphates showing which ions they can replace upon doping

i Caibonalc fCO ' ) Phosphate (PO "' ^" )

j

SilicaU Si() ¹ ) i iosphaLc (|>() ^' )

1 Magnesium (M ) Calcium (Ca ^"

1 Slionlium (Sr ^" ) Calcium (Ca ^"

I liai'ium ( l'>a- ) Calcium (Ca ^"

6 Silwr (Ag-- ) Calcium (Ca ^"

7 Χΐικ·(ΧιΓ ) Calcium (Ca ^"

8 3

Tron (Tv . Ι·\ ) Calcium (Ca ^"

9 Manganese (Μι ) Calcium (Ca~ Example 7 - Surface Modifed-HA using Polyvinyl alcohol, Adipic acid. Citric acid,

Vinylphosphonic acid, Methacrylic acid

Experimental

The surface modified hydroxyapatites were obtained following a similar procedure as for pure nano-HA (mentioned above), except the calcium containing precursor additionally contained the appropriate amount of (0.05 M) functionalised carboxylic acid or organic phosphoric acid. Resulting nanopowders were termed PVA-HA (polyvinyl alcohol), AA-HA (adipic acid), CA-HA, (citric acid), VPA-HA (vinylphosphonic acid) and MA-HA (methacrylic acid).

Introduction

Hydroxyapatite (HA) is of interest as a substitute for defective bones and teeth and also has a wide range of applications in industry and medicine. Dentine a natural composite is based on an organic matrix of mainly collagen with small quantity of citrate and an inorganic mineral phase (filler) consisting of nanosized (lo-ioonm) hydroxyapatite crystals. Synthetic HA resembles some of the properties of natural teeth in hardness and offers numerous promising advantages (intrinsic radio-opaque response, enhanced polishability, improved wear performance) in restorative dentistry. The most abundant functional groups on HA surfaces are P-OH groups. Thus, many organic or inorganic substances have been introduced on HA surfaces using P-OH groups to control the physicochemical properties of colloidal HA particles. Increasing the number of surface P-OH groups has been of interest since P-OH groups can act as anionic surface charges and offer a favourable environment for protein adsorption. Furthermore, increased numbers of P-OH groups lead to increased electrophoretic mobility and good dispersibility in an aqueous phase (via the charge stabilization mechanism).

There is considerable interest in the use of bioceramics in composites as bone replacements. The reinforcement of dental resin with finally dispersed as-prepared hydroxyapatite nanocrystals seems, in principle, a favourable restorative material for human tooth tissues as it mimics the tooth. A key factor in the failure of composites occurs at the interface between the HA particles and polymer matrix. This is usually due to a weak mechanical bond between the two phases. The bonding mechanism is often simply a mechanical interlock, formed during cooling after manufacture. Chemical coupling of the HA to a polymer matrix provides a means of improving interfacial bonding between the particles and polymer in a composite. This can be achieved by creating reactive bonding sites on the surface of the inorganic filler particle, which can be used to form bonds between the inorganic filler and polymer. In this report, a continuous (plastic) flow synthesis (CPFS) reactor was exploited for the rapid production of phase pure and surface modified HA particles for potential

biomedical/dental applications. The main objective of this present study was to modify the surface of hydroxyapatite by using various reactive organic acid and then to use these modified HA's as a reinforcing filler component to develop radiopaque resin infiltrants for potential dental restorative materials.

Results and Discussion

The nanoparticles obtained after surface grafting possessed small particle size of ca. 25 ± 5 nm (200 particles sampled) along the longest axis and of ca. 7 ± 2 nm along the smaller axis as shown in Figure 43. BET surface area measurements of as precipitated HA sample typically has BET sufrace areas of 180 m ² g ^_1 while surface modified PVA-HA (polyvinyl alcohol), AA-HA (adipic acid), CA-HA, (citric acid), VPA-HA (vinylphosphonic acid) and MA-HA (methacrylic acid synthesized at the same conditions as pure HA possessed a surface area of 143, 208, 201, 225 and 231 m ² g ^_1 , respectively (Fig. 44).

The X-ray diffraction data of the as-prepared surface modified hydroxyapatites displayed broad peak of an apatite-like structure (Figure 45) and gave a good match to hydroxyapatite reference pattern JCPDS [-09-432].

The FTIR spectra of all five as-prepared surface modified HAs are shown in Figure 46. All spectra revealed peaks assigned to phosphate stretching and bending vibrations. The peaks at 1093 cm ¹ and 1023 cm ¹ correspond to asymmetric (P-O) stretching due to phosphate groups whilst peaks at 602 and 560 cm ¹ correspond to the symmetric (P-O). The weak peak at 470 cm ¹ was assigned to the phosphate bending mode. FTIR data revealed a peak at 3573 cm ¹ corresponding to the (O-H) stretching vibrations in HA. For methacrylic acid and vinylphosphonic acid surface-modified samples which contain a C=C bond, a peak for C=C stretching was observed at -1640 cm ¹ . In the FTIR data there were additional peaks due to symmetric and asymmetric (C-O) stretching in the carboxylate groups, centred at ca. 1433, 1460, and 1531 cm ¹ .

A chemical analysis of as-prepared surface grafted VPA- HA sample was done by using XPS analysis as shown in Fig. 47. It was observed that in surface modified HA, Ca/P molar ratio decreased from 1.67 to 1.37 by the addition of organic modifiers. The deconvoluted spectra for O, Ca and P are presented in Figure 48. The peaks at 134 eV corresponded to P 2p of the phosphate groups in hydroxyapatite. While the binding energy values for O is, Ca 2p and Cis were measured as 533, 347 and 285 eV, respectively.

The organic surface modified molecules that were used for the hydroxyapatites were investigated by NMR spectroscopy in solution, which required complete dissolution of in a deuterated solvent. The coupling interaction between different proton environments (doublet, triplet) in vinylphosphonic acid were evaluated as 2J(HB,HA) = 21Hz, 3J(HB, P) =44.8Hz, 3J(HA, P) =22.5 Hz, 3<J(HA, HC) =19.1 Hz, 3<J(HB, HC) =12.7 Hz, 2J(HC,P) =19.1 Hz. The presence of different proton environments associated with the carbon atoms in VPA has proved that surface modification of VPA on HA has occurred as shown in Fig. 48.

The synthesis of surface modified HA for a range of surface modifiers was achieved in ca. 5 minutes a continuous reaction at 70 °C. Surface modification with different carboxylic containing organic agents resulted in the production of nano-sized HA crystals which could be readily dispersed and possessed remarkably high surface areas. These fine, highly dispersed nanoparticles contain reactive groups on the surface; therefore they have great range of potential dental / biomedical applications.

BET Surface area analysis of pure HA fungrafted) and surface grafted HAs at 60, 70 and 8o°C

BET surface area measurements of as-prepared HA sample typically has BET surface areas of 264, 195 and 113 m ² g ^_1 . While surface modified VPA-HA (vinylphosphonic acid) and MA- HA (methacrylic acid synthesized at the same conditions as pure HA possessed a surface area of 254, 247, 210 m ² g ^_1 and 265, 244 and 221 m ² g ^_1 , respectively at reaction temperature of 60, 70 and 8o°C. This fact is attributed to the growth restriction of HA nanoparticle in the presence of surface modified agents and smooth increase in surface area at various selected

temperatures as shown in Fig. 49.