J. Am. Ceram. Soc., 88 [12] 33533360 (2005) DOI: 10.1111/j.1551-2916.2005.00623.

x r 2005 The American Ceramic Society

Synthesis of HA-Seeded TTCP (Ca4(PO4)2O) Powders at 12301C from Ca(CH3COO)2 . H2O and NH 4H2PO 4
Sahil Jalota, A. Cune yt Ta s, w and Sarit B. Bhaduri
School of Materials Science and Engineering, Clemson University, Clemson, South Carolina 29634

is,

display extensive twinning and suspected to have a high-temperature orthorhombic polymorph. 5 The positions of the P atoms and the Ca and oxide ions lie close to those required by an orthorhombic space group, Pmcn. This probably explains the appreci able twi nning e xhibited by C a4(PO4)2O.4 How can TTCP transform into apatite? Ca 4(PO 4)2Ocontains a crystallographic layer similar to an apatitic layer present in both Ca 10(PO 4)6(OH) 2 and Ca 8H2(PO 4)6 5H 2O (octacalcium phosphate, OCP) . As a consequence of this layer, an epitaxial relationship between TTCP and HA, as well as between TTCP and OCP, may be present. 4 Epitaxy is often thought of as being governed chiefly by the metric fits of crystallographic networks based on the unit cell translations. 4 The combination of good metric and chemical fits between HA and OCP makes the epitaxy between these two compounds possible. 6 Undetected epitaxy between HA and TTCP would increase the Ca/P molar ratio of an apparent apatite above 1.667, and on the microscopic scale the material would resemble a s olid solution. HA therefore, an u nusual substance in t hat it h as two closely r elated salts with which it may enter into epitaxial relationships. One salt, Ca 8H2(PO 4)6 5H 2O, is more acidic, and the

other, Ca 4(PO 4)2O, is more basic than the bone mineral, HA. 4 TTCP is the most soluble compound among all the calcium phosphates. 1 Producti on of C a4(PO4)2O in an a queous environment is extremely difficult, if not impossible, because of the oxygen ion in its formula . In aqueous solutions hydroxyl (OH) ions inc orporate t hemselves into t he formed precipitates. A carbonate- and/or hydroxyl ion-containing apatitic product will form readily if one attempts the synthesis of calcium in a basic aqueous solution with a Ca/P molar ratio of 2. For this reason, synthesis of TTCP, which is nowadays used as one of the main components of self-setting orthopedic and dental cements, has only been limited to solid-state reactive firing practices. 79 TTCP was typically syn thesized by the s olid-state reaction of calcium carbonate (CaCO 3) and dicalcium phosphate anhydrous (CaHPO 4) powders, mixed with one another at a Ca/ P molar ratio of 2. Powders prepared by Brown and Chow, 7 Chow and Takagi, 8 and Matsuya et al. 9 were usually heated at 1450115001C for 612 h, and then rapidly quenched to room temperature. Slow cooling of TTCP, instead of quenching, high temperatures resulted in the formation of undesired secondary phases, such as HA, CaO, CaCO 3, and b-Ca3(PO4)2. Gr eish et a l.10 very recently synthesized TTCP powders by mixing CaCO 3 and Ca(H 2PO 4)2 H2O powders in n-heptane, followed by heating and quenching from 13101C. As of now, it been impossible to produce TTCP powders without quenching from a temperature in excess of 1300 1C. Rapid cooling is essential in order to prevent the total decomposition of TTCP HA and CaO/Ca(OH)
2

phosphates

from

has

into

.11 For use as a constituent of a cement,

sintered bodies of TTCP need grinding to a particle size less than 45 mm.9 The use of h igher synthesis temperatures, which further promotes sintering and grain growth, places a significant economic burden on the manufacturer in the form of costly grinding oper ations. Such grinding practices may also be regarded as sourc es of f urther co ntamination. Sargin et al. 12 studied, for the first time, the influence of different starting materials on the synthesis of TTCP powders . In

3354

Journal of the American Ceramic SocietyJalota et al.

Vol. 88, No. 12

this important study, Sargin et al.12 showed that by using CaCO3 and NH 4H2PO4 as r eactants, it w as possible to decrease the temperature of synthesis of single-phase TTCP down to about 1 3501C. An interesting feature reported in Sargin et al. 12 was that, for the first time, separate sources of Ca and P were used. In Brown and Chow,7 Chow and Takagi,8 Matsuya et al.,9 Greish et al.,10 and Bian et al. 13 calcium needed to form TTCP came from both CaCO3 and CaHPO4 or from CaCO3 and Ca(H 2PO 4)2 H2O. During heating, CaCO3 decomposes by evolving CO 2 at around 900 1C, and CaHPO 4 first transforms into Ca2P2O7.14 Therefore, in Chows solid-state approach, high temperatures (49001C) w ere deemed necessary to even start the reactions between the starting materials with one another. 79,13 The discovery that the basic TTCP reacts with acidic dicalcium phosphate anhydrous, CaHPO 4, to form pure HA led to the development of a self-setting calcium phosphate cement (CPC) by Brown and Chow. 7,8,15 TTCP-containing self-setting CPCs are now commercially available (Norian SRS , Skeletal Repair System, Synthes Corp., Oberdorf, S witzerland; Bi opex , Mitsubishi Materials Corp., Tokyo, Japan)16 and have a proven track record in clinical/surgical use. In t he present study we i nvestigated a n ew c hemical s ynthesis route to prepare HA-seeded TTCP powders. One major aim was to decrease the synthesis temperature even below that reported by Sargin et al. 12, i.e., 1350 1C. Formation of TTCP, from the starting chemicals, was monitored by using X-ray diffraction, thermogravimetric/differential thermal analysis (TG/DTA), scanning electron microscopy, and Fourier-transform infrared spectroscopy. The powder synthesis route we present here simply consisted of physical mixing of the stoichiometric amounts of calcium acetate monohydrate (Ca(CH3COO)2 H2O) and ammonium dihydrogen phosphate (NH4H2PO4) powders, followed by heating at 12301C in air in a box furnace for 12 h and quenching. This process enabled us to obtain powders at the lowest temperature ever reported for the formation of TTCP phase. Any significant decrease in the synthesis temperature of such self-setting orthopedic cement constituents is extremely important, because well-sintered TTCP samples produced by soaking bet ween 14001 and 15001C nee d to be ground int o a powder prior to their use as a biomaterial . Grinding of TTCP powders would damage their stoichiometry and consequently alter their in vivo performance. 17 Biomimetic (i.e., with the help of an aqueous solution mimicking the ion concentrations of human blood plasma at 371C and pH 7.4) formation of carbonated apatitic calcium phosphates on these powders was tested and confirmed by imm ersing these i nto a s ynthetic body fl uid ( SBF) from 36 to 96 h. 18

II. (1)

Experimental Procedur e

Powder Synthesis

HA-seeded TTCP powders were synthesized in small batches. Fir st, 0. 77 g of N H4H2PO4 (499%, Fisher Sci entific, F airlawn, NJ) was ground into a very fine powder by using an agate mortar and pestle . 2.45 g of Ca(CH 3COO) 2 H2O (499%, Fisher) was then added to it. Two powders were dry mixed in the mortar for about 45 min w ithout a pplying too much force. The pow der mixture w as then h eated at 3001C in air for 30 min, followed by re-mixing in a mortar for 15 min . Powders were then heated at 8001C in air for 1 h, f ollowed by a h omogenization mixing in an agate mortar . These powders had a gray/bluish tint, and they were heated (with a heating rate of 21C/min) at 12301C in air for 6 h. At the end of 6 h of heating at 1230 1C in a box furnace, t he red-hot alumina boat containing the powders was taken out and immediately placed into a desiccator . Obtained powders were only lightly ground and homogenized in an agate mortar for few minutes and re-fired at 1230 1C for another 6 h, followed by simi lar quenching to r oom t emperature. Powders obtained were easy to grind in the agate mortar . This firing scheme was only suitable for the synthesis of HA-seeded TTCP powders . We were careful not to over grind the resultant TTCP powders, be-

December 2 005

Synthesis of HA-Seeded TTCP (Ca 4(PO 4)2O) Powders

3355

Fig. 1. (a) X-ray diffraction (XRD) spectrum of pure Ca(CH 3COO) 2 H2O powders . (b) Thermogravimetric and differential thermal analyses (TG/ DTA) traces for pure Ca(CH 3COO) 2 H2O p owders (in air). (c) XRD sp ectrum of p ure N H4H2PO4 powders. (d) TG/ DTA t races for pure N H4H2PO 4 powders (in air).

traces of water or other adsorbed gases that may be present on the powder surfa ces.

III.

Results and Discussion

XRD and TG/DTA data for the pure starting chemicals (Ca-acetate monohydrate and ammonium dihydrogen phosphate) of powder synthesis are given in Fig. 1. The FTIR traces of the starting chemicals, i.e., pure NH 4H2PO 4 and Ca(CH 3 COO) 2 H2O, are given in Fig. 2. The FTIR trace for the unheated powder mixture, which has the stoichiometry of TTCP, is also given in Fig. 2 for comparison purposes . NH 4H2PO4 me lts at around 1901C and forms an acidic phosphate liquid.19 NH 4H2PO 4 was chosen here over (NH 4)2HPO 4 mainly because of (1) its lower ammonium content and (2) its definite melting point. This phosphate liquid immediately reacts with Ca(CH 3 COO) 2 H2O, which prior to reaching the melting point of NH 4H2PO4 (1901C) would have already lost i ts wat er of hydration and transformed into calcium acetate anhydrous . The following reac tions took place until the temperature reached 2001C: CaCH 3 CO O2 H2 Os ! CaCH 3CO O2 s H2 Og NH4 H2 PO 4 s ! NH 4 H2 PO 4 l Equation (1) envisages a weight loss of 10.22% . Calcium acetate anhydrous started its decomposition at around 400 1C releasing volatiles, acetone in an air atmosphere, leaving behind

3356

Journal of the American Ceramic SocietyJalota et al.

Vol. 88, No. 12

for ammonium dihydrogen phosphate, on the other hand, can be given by the following equation: 2 NH 4 H2 PO 4 s ! 2 NH 3 g 3 H2 Og P2 O5 s (4)

continued

Equation (4) indicated a total weight loss of 38.3%, and it explained well the weight loss observed in the pure NH 4H2PO 4 samples (Fig. 1(d)) until the 680 16901C TGA s houlder. With a further increase in temperature, the P2O5 already formed starts evaporating, bringing the total weight loss in pure NH 4H2PO 4 samples to around 96%. The main purpose in heating the powder mixtures first to 3001C followed by grinding was the homogenization of the reaction products of molten ammonium phosphate and the solid phase of partially decomposed calcium acetate. Upon heating to 800 1C, volatilization of NH 3 from NH 4H2PO 4, followed by its condensation to highly reactive P2O5 was completed, whereas acetone, carbon dioxide, and water vapor were the main decomposition products of Ca-acetate . At this point, the samples were ground into a fine po wder, followed by heating to the final temperature of 1230 1C. Figure 3(a) showed a typical TG/DTA trace for the Ca(CH 3 COO) 2 H2O and NH 4H2PO 4 mixtures (i.e., TTCP mix) up

to 12001C. To s ummarize t he thermal events seen in t his T G/DTA chart, first, NH 4H2PO 4 melted at around 190 1C (endothermic peak) engulfing the calcium a cetate particles; second, the abrupt decomposition of Ca-acetate accompanied by the drastic crystallizati on of C aCO3 (indicated by t he sharp e xothermic p eak in Fig. 3(a)) t ook p lace at about 4001C; and t hird the tion of CaCO 3 was observed in the vicinity of 750 1C. Crystallization of CaCO 3 from the decomposing calcium acetate particles, which are embedded in a molten phosphate bath, occurred through an exothermic reaction. X-ray diffraction es given in Fig. 3(b) d epict the stepwise phase e volution in mixtures, with the help of isothermal heatings (2 h at the peak temperatures, with 51C/min heating and cooling rates) at individual temperatures indicated on this chart . Table II lists the phases assigned to the individual peaks shown in Fig. 3(b). TTCP w as not detected even in samples quenched from FTIR data of the TTCP p owder mixtur es were a lso u seful in determining the nature of phase evolution sequence in these samples. Figure 3(c) shows the IR spectra of samples isothermally heated (2 h at peak temperatures, 51C/min heating and cooling rates) at the temperatures indicated (RT to 1000 1C) in the chart . Experimentally determined IR bands are given in Table III together with the corresponding phases. An amorphous intermediate (consisting of long chains of condensed phosphate of the form ( PO3)n) is expected to form as a result of the reactions between an acidic phosphate liquid and calcium acetate, which subsequently converts itself into crystalline b-Ca(PO 3)2 (calcium metaphosphate) with an increase in temperature to around 25014001C.21 XRD and FTIR results both confirmed this, and after an isothermal heating at 350 1C, calcium metaphosphate, calcium hydrogen phos-

decarboxyla-

tracthese

10001C.

partially

Table II. Phases Identified (by XRD) in the Isothermal Heatings of Ca-Acetate Monohydrate 1NH 4H2PO4 Mixed Powders
Phase name Phase formula ICDD -PDFw

Dicalcium h ydrogen phosphate Calcium pyrophosphate

CaHPO4 b-Ca2P2O7

9-0080 9-0346

Calcium carbonate (calcite) Calcium metaphosphate Tricalcium phosphate Calc ium hydroxyapatite Calcium hydroxide Calcium oxide
ICDDPDF, Internation File.

CaCO3 Ca(PO3)2 b-Ca 3(PO4)2 Ca 10(PO 4)6(OH) 2 Ca(OH) 2 CaO

5-0586 50-0584 9-0169 9-0432 44-1481 37-1497

al Centre for D iffra ction Da taPowder Diffraction

December 2 005

Synthesis of HA-Seeded TTCP (Ca 4(PO 4)2O) Powders

3357

samples

can not be dete cted by X RD. Below 3501C, calcium acetate ( i.e., COO bands) and ammonium phosphate (i.e., orthophosphate and acid phosphate groups) IR bands and XRD peaks were visible. Therefore, a step of 3001C calcination (followed by grin ding) was j ust necessary to hom ogenize t he powder prior to the crystallization of any calcium phosphates . Calcium metaphosphate disappeared upon heating the powder mixtures to 4301C, and the samples consisted of a biphasic mixture of CaCO 3 and Ca 2P2O7 at this temperature . However, upon h eating to 7001C, the presence of CaCO 3 significantly diminished with the simultaneous appearance of b-tricalcium phosphate, Ca3(PO4)2 (TCP), and HA. Because of the overall stoichiometry of the initial powder mixtures (Ca/P molar ratio 5 2), det ection of HA or TCP in the powders should automatically call for the prese nce of C a(OH)2 or CaO in t hose. Major pe aks of C aO and Ca(OH) 2 are at 37.36 1 and 34.17 1 2y, respectively . Most HA peaks overlap with those of TTCP; however, the significant (i.e., 002 reflection) of HA was encountered at 25.88 1 2y. The presence of Ca(OH)2 was unmistakably detected by the IR

peak spec-

TCP,

tra, with its characteristic h ydrox yl str etching b and observed at 3640 cm 1 in the samples heated at 700 1C and higher . Samples heated at 10001C contained the phases of carbonated HA, Ca(OH) 2, and Ca 2P2O7. Complete elimination of TCP and Ca2P2O7 and the forma tion of TTCP were observed at 412001C. The characte ristic XRD d iagr am of T TCP p owders obtained after the second step of 12301C firing/quenching is given in Fig. 4(a). These powders contained approximately 56 wt% (me dian: 5. 43, estimated standard deviation: 0.26) of HA

phase, together with Ca(OH) 2. The amounts of HA in these powders

Fig. 5. Morphology of hydroxyapatite-seeded sized at 1230 1C.

TTCP powders synthe-

3358

Journal of the American Ceramic SocietyJalota et al.

Vol. 88, No. 12

were determined, on three batches of preparation, by quantitatively comparing the peak intensities of TTCP and HA phases . The FTIR pattern of the HA-seeded TTCP powders is given in Fig. 4(b). This experimental pattern precisely matched those then previously reported for tetracalcium phosphate by Sargin et al.12 than and Posset et al.,22 respectively. Vibrational wavenumbers w ere assigned by the above-mentioned researchers . In the data of Fig. 4(b), threefold degenerate stretching (1105989 cm 1) mode surface (n3), symmetric str etching ( 962941 cm 1) mode ( n1), and t hreefold d egenerate deformational (620571 c m 1) mode (n4) vibr athe tions were all visible.12,22 The inset in Fig. 4(b) displays the characteristic OH stretching vibration (3640 cm 1), originating from Ca(OH) 2.12 On the other hand, the weak band at 872 cm 1 belongs to the carbonate groups, and these are considered to be associated with the carbonated HA seeds present in interthe sample s. The addition of HA seeds to TTCP powders has previously been tested by Hamanishi et al.; 23 however, they had intentionally added about 40 wt% precipitated HA into a cement mixture of TTCPCaHPO 4 2H2O. The level of HA seeding in the scatterin g, present study is obviously much lower than this . in Particle morphology of the TTCP powders in this study is given in Figs . 5(a) and (b). Powders displayed a unique vermicular microstructure (less than 5 mm in size), which was indicsuitable ative of the vapor- and liquid-phase reactions that took place . The remnants of the amorphous-looking chunks of Ca-doped conden sed p hosphates (i.e., P 2O7, (PO3)n, and P2O5 in comparison with PO 4)2426 were observed in the form of larger chunks of about 40 mm. The as-quenched powder samples depicted in

Figs. 5(a) and (b) had a BET surface area of 3274 m2/g. Conventional TTCP powders prepared by 1500 1C heating (for 6 h) of CaCO 3 and CaHPO 4 are first ground in ball mills to reduce their particle sizes to around 1320 mm ( coarse TTCP), and further gro und f or 24 h to r educe t heir particle sizes to l ess 5 mm (fine TTCP). 27 SBF solutions are able28 to induce apatitic calcium phosphate formation on metals, ceramics, or polymers (with proper treatments) soaked in them. It is also known that an amorphous calcium phosphate (ACP) precursor is always present during precipitation of apatitic calcium phosphates from highly supersaturated solutions. 29 Posner and Betts 30 proposed that the process of ACP formation in solution involved the formation of Ca 9(PO 4)6 clusters, which then aggregated randomly to produce the larger spherical particles or globules, with the cluster space fi lled w ith water. Such clusters, we believe, are the transient solution precursors to the formation of carbonated globules with the nominal stoichiometry of Ca10 x(HPO4)x (PO4)6 x(OH) 2 x, where x might be converging to 1. 31 Onuma and Ito 32 have demon strated, by u sing dynamic l ight the presence of calcium p hosphate clusters from 0.7 to 1.0 nm size even in clear simulated body fluids . They reported the presence of calcium phos phate clu sters in S BF ev en when the re was no visible precipitation. Therefore, one only needs some surface to be immersed into an SBF solution, which would probably trigger the hexagonal packing 32 of those nanoclusters into globules or spherulites of apatitic calcium phosphates. SEM photomicrographs given in Figs . 6(a)(d) exhibited the rapid formation of those apatitic calcium phosphate globules after

Fig. 6. (a) Thirty-six hours in synthetic body fluid (SBF) at 371C (twinning; towards lower le (c, d) Ninety-six hours in SBF at 371C.

ft-hand c orner). (b) Seventy-two hours in SBF at 371C.

December 2 005

Synthesis of HA-Seeded TTCP (Ca 4(PO 4)2O) Powders

3359

Fig. 9.

Change in surface pH of HA-seeded TTCP powders in SBF.

This was ascribed to the transformation of HA-seeded TTCP powders into carbonated apatitic calcium phosphates formed under biomimetic condi tions pr ovided in vitro by the SBF solutions. On the other hand, for comparison purposes, pure, stoichiometric, crystalline HA has a pH value of about 9.5. After 96 h of soaking in SBF, recorded pH value was 9.5670.04 at RT. It is reasonable to assume that this pH value would have continued to drop toward the physiological pH if we had continued the soaking beyond 4 days.34,35 The pH diagram of TTCP powders given in Fig. 9 may very well serve the practitioner to precisely t ailor t he amount of an acidic calcium phate 2H2O; MCPM, Ca(H 2PO 4)2 H2O; and MCPA, Ca(H2PO4)2) that needs to be blended with these powders in designing a TTCP-based cement formulation having a neutral pH at the very start of the cement-forming process. Being the calcium phosp hate compound of highest sol ubility at the physiological pH, the hydrolysis of TTCP should be regarded as a process of initial dissolution followed by the precipitation of a less soluble phase . In an SBF environment (i.e., 371C and pH 7.4), t he precipitating phase can o nly be nanos ize, carbonated HA, i.e., the bone mineral . The presence of HA seeds in TTCP simply accelerates this in vitro biomineralization process. Nevertheless, it is yet to be s een if t hese powders can made into useful self-setting CaP cements. (such as DCPA, CaHPO4; DCPD, CaHPO 4

phos-

be

IV.

Conclusions

syn-

(1) Five to six weight percent HA-seeded TTCP (Ca4 (PO 4)2O) powders were synthesized at 1230 1C, by using the starting chemicals of calcium acetate monohydrate and ammonium dihydrogen phosphate, for the first time, followed by quenching from the synthesis temperature. (2) This is t he lowest temperature e ver r eported for the thesis of TTCP powders, and such low synthesis temperatures can r eadily be a chieved in i nexpensive el ectrical r esistance-heated furnaces. (3) HA-s eeded T TCP p owders underwent ap atitic transformation when they were brought into contact with SBF of pH 7.4 at 371C, within the first 3696 h. The HA seeds can perhaps be seen as having a catalytic function for the TTCP transformation.

solutions

Acknowledgments

This work was a part of the MSc thesis of Mr . Sah il Jalota.

References
W. E. Brown and E. F. Epstein, Crystallography of Tetracalcium Phosphate, J. Res. Natl. Bur. Stand., 69A, 54751 (1965).

3360
schlack e, Stahl Eisen. , 7, 2459 (1887) .

Journal of the American Ceramic SocietyJalota et al.

Vol. 88, No. 12

D. R. Lideed. Handbook of Chemistry and Physics , 72nd edition, pp. 439. CRC Press, Boston, 1992. J. Adanez, L. F. de Diego, and F. Garcia-Labiano, Calcination of Calcium Acetate and Calcium Magnesium Acetate: Effect of the Reacting Atmosphere, Fuel, 78, 58392 (1999). to U. Posset, E. Loecklin, R. Thull, and W. Kiefer, Vibrational Spectroscopic Study of Tetracalcium P hosphate in Pure Polycrystalline Form and as Constituent of a Self-Setting Bone Cement, J. Biomed. Mater. Res., 40, 6405 (1998). C. Hama nishi, K. Kitamoto, K. Ohura, S . Tanaka, and Y . Doi, Self-Setting, Bioactive, and Biodegradable TTCP-DCPD Apatite Cement, J. Biomed. Mater. Res., 32, 3839 (1996). B. T. Lynch, P. W. Brown, and J. R. Hellmann, Synthesis of Castable Sodium Zirconium Monoliths Employing Reactions between Zirconyl Nitrate Hydrate and Condensed Phosphates, J. Mater. Sci., 34, 180913 (1999). M. Plydme, J. Plydme, and K. Utsal, Phase Transformations in Mixtures

Tetracalcium Diphosphate Monoxide: Crystal Structure and Relationships Ca 5(PO 4)3(OH) and K3Na(SO 4)2, Acta Crysta llogr., B29, 204656 (1973).

Methods for Preparing and Using Them; U.S . Patent No . 5,525,148, June 11, 1996. Akamine, Effect of Powder Grinding on Hydroxyapatite Formation in a Polymeric Calcium Phosphate Cement Prepared from Tetracalcium Phosphate and Poly(Methyl Vinyl Ether Maleic-Acid), Biomaterials, 20, 6917 (1999). Y. E. Greish, J. D. Bender, S. Lakshmi, P. W. Brown, H. R. Allcock, and C. T. Laurencin, Low Temperature Formation of Hydrox yapatite-Poly(Alkyloxybenzoate)Phosphazene Composites for Biomedical Applications, Biomaterials, 26, 19 (2005). H. Monma, M. Goto, H. Nakajima, and H. Hashimoto, Preparation of Y. Sargin, M. Kizilyalli, C. Telli, and H. Guler, A New Method for the SolidState Synthesis of Tetracalcium Phosphate, A Dental Cement: X-Ray Powder Diffraction and IR Studies, J. Eur. Ceram. Soc., 17, 96370 (1997). Phosphate Cements with Different Amounts of Tetracalcium Phosphate and Dicalcium Phosphate Anhydrous, J. Biomed. Mater. Res., 46, 50410 (1999).

T. Kokubo, Apatite Formation on Surfaces of Ceramics, Metals and Polymers in Body Environment, Acta Mate r., 46, 251927 (1998).

A. S. Posner and F . Betts, Sy nthetic Amo rphous Ca lcium P hosphate a nd its

Ionic Strength and Carbonate on the CaP Coating Formation from SBF 5 Solution, Biomaterials, 23, 192130 (2002). K. Onuma and A. Ito, Cluster Growth Model for Hydroxyapatite, Chem. Cement; pp . 35179 in Cement Resear ch Pro gress, Edited by P. W. Brown. The American Ceramic Society, Westerville, OH, 1986. C. D. Friedman, P. D. Cos tantino, S . Takagi, and L. C. Chow, BoneSource (TM) Hy droxyapatite Cement: A N ovel Biomaterial f or Craniofacial S keletal Tissue Engineering and Reconstruction, J. Biomed. Mater. Res., 43, 42832 (1998). U. Gbureck, J. E. Barralet, M. Hoffmann, and R. Thull, Mechanical Activation of Tetracalcium Phosphate, J. Am. Ceram. Soc., 87, 3113 (2004).

phate in H3PO 4 and KH 2PO 4, J. Mater. Sci., 31, 32639 (1996). scopic Study on Setting Mechanism of Tetracalcium Phosphate/Dicalcium Phosphate Anhydrous-Based Calcium Phosphate Cement, J. Biomed. Mater. Res., 64A, 66471 (2003) . &