www.elsevier.com/locate/jnoncrysol

Journal of Non-Crystalline Solids 353 (2007) 2367–2373

Biodegradation behavior of chitosan/calcium phosphate composites

Shinn-Jyh Ding *

Institute of Oral Materials Science, Chung-Shan Medical University, Taichung 402, Taiwan, ROC

Received 30 January 2007; received in revised form 14 March 2007

Abstract

A variety of biomimetic materials with structural and mechanical equivalence to bone have been developed to repair bone defects.Chitosan/calcium phosphate composites composed of bioactive calcium phosphate and flexible chitosan were made by a simple mix-ing-and-heating method. Mechanical properties, morphology, phase composition, and weight change after immersion in Hanks’ solutionwere evaluated. Experimental results showed that the formation of pores/cracks on immersed sample surface obviously depended on thecalcium phosphate content and immersion time. The immersion time imposed in this study did have a statistically significant effect onmechanical properties. When immersed for 90 days in Hanks’ solution, the strength of immersed composites containing 10 wt/v% cal-cium phosphate with the initial strength of 27 MPa was about 2 MPa, having a reduction of 92%. Based on the above results, theorganic–inorganic hybrid composites with high initial strength might be an acceptable material candidate for bone tissue repair.� 2007 Elsevier B.V. All rights reserved.

PACS: 87.68.+z; 81.05.Qk; 68.35.Gy

Keywords: Biomaterials; Chemical durability; Strength

1. Introduction

Natural bone is an inorganic/organic composite mainlyconsisting of collagen matrix and nano-structuredhydroxyapatite [1,2]. When the hard tissue is damaged orfails, a number of bioactive implant materials have beenused to repair bone defects [3–8]. Chitosan is an abundant,naturally occurring polysaccharide obtained by deacetyla-tion of natural chitin [9]. Its biocompatibility, biodegrad-ability, and nontoxicity make chitosan to become anatural choice as drug-delivery carrier [10,11], cartilage/skin tissue engineering scaffolds [9,12,13], and regenerativemembrane [14]. For example, Desai and Park reported thatchitosan-tripolyphosphate microspheres of controlled sizeshad a high encapsulation efficiency using acetaminophen asa model drug substance [11]. However, the mechanicalproperties of chitosan cannot satisfy bone tissue repairrequirements. Therefore, some studies have attempted to

0022-3093/$ - see front matter � 2007 Elsevier B.V. All rights reserved.

doi:10.1016/j.jnoncrysol.2007.04.020

* Tel.: +886 4 24718668x55529; fax: +886 4 24759065.E-mail address: sjding@csmu.edu.tw

modify the properties of chitosan to overcome these draw-backs [9,10,15].

Due to their chemical compositions being similar to theinorganic component of bone, calcium phosphates (CaP),particularly hydroxyapatite (HA), have been used asimplant materials for bone repair and regeneration. None-theless, the brittle and rigid nature of calcium phosphateseverely limits its usefulness in biomedical applications[16–18]. It must be emphasized at this point that the suc-cessful design of a bone substitute material requires anappreciation of the structure of bone. Thus, the use of ahybrid composite that makes up of chitosan and calciumphosphate resembles the morphology and properties ofnatural bone. This may be one-way to solve the problemof CaP’s brittleness, besides possessing good biocompati-bility, high bioactivity and great bone-bonding properties[17–26]. In a study by Matsuda et al., the mechanicalstrength of the chitosan/calcium phosphate composite fiberwith core-shell structure increased with an increased con-centration of chitosan solution [26]. Additionally, Zhanget al., stated that the addition of b-tricalcium phosphate

2368 S.-J. Ding / Journal of Non-Crystalline Solids 353 (2007) 2367–2373

to mechanically weak chitosan greatly improved thecompressive modulus and yield strength of the compositescaffolds [17]. In an earlier study by the present author[27], chitosan/CaP composites were made by a simple mix-ing-and-heating method, with the aim of improving themechanical properties of chitosan. The major advantagesof this approach used were its simplicity, low cost, and fab-rication by casting method. The results revealed that aCaP/chitosan ratio of 5% by weight to volume in the com-posite achieved the significantly highest bending strength of45.7 MPa. It seems that chitosan/calcium phosphate com-posites are another potential candidate as implant materialand deserve further research, such as immersion study.

Biodegradation of materials can be characterized bychanges in the mechanical and physicochemical properties(disintegration and dissolution) of the material afterimplantation or after immersion in physiological solution.Little information on the variations in the mechanicalproperties of chitosan/calcium phosphate composites,when immersed in simulated body fluid, has been reported.Hence, there is a need to use immersion method to test thebond durability of the organic–inorganic hybrids for eval-uation of feasibility in clinical applications. Following theprevious study [27], the degradation behavior in Hanks’solution of the different chitosan/calcium phosphate com-posites investigated was focused by monitoring changesin morphology, phase composition, three-point bendingstrength and modulus as well as weight loss.

2. Experimental

2.1. Preparation of the composites

Chitosan (85% deacetylated) was obtained from Sigma(St. Louis, MO, USA) and used without further purifica-tion. Glutaraldehyde (50% aq. solution) was likewise pur-chased from Sigma. A commercial Merck product calledTricalcium Phosphate (Catalog number 2143, Darmstadt,Germany) was used as reinforcement, but in fact it wasan apatitic phase with a small amount of tricalcium phos-phate [27]. All other chemicals used were of reagent gradeand used as obtained. A series of chitosan/calcium phos-phate (designated hereafter as CTS/CaP) composites weremade, as described elsewhere in detail [27]. Briefly, anappropriate amount of CaP was added to the 2% chitosansolution and the mixture was stirred for 30 min using ahybrid defoaming mixer (Thinky, Tokyo, Japan, ARE-250) to achieve CaP/CTS ratios of 2.5%, 5%, 10%, and20% by weight to volume. After which, this reaction mix-ture was stirred vigorously for 5 min after adding 5% aque-ous glutaraldehyde (GA) at a 40:1 volume ratio of chitosanto GA. Sequentially, the mixture was cast into a polyethyl-ene die of 12-mm diameter. To avoid phase separationbetween CaP and chitosan solution, the mixture was rap-idly transferred into a refrigerator at 4 �C for 24 h toundergo the cross-linking reaction. Following which, dryheat treatment was done in an oven at 60 �C for 3 days

to completely remove the solvent and to reduce excessshrinkage.

2.2. Preparation of Hanks’ solution

Hanks’ balanced salt solution, an extracellular solutionwith an ionic composition similar to that of human bloodplasma, was used as the supporting solution for the bend-ing strength test and in vitro fatigue test. The simulatedsolution consisted of 8.00 g NaCl, 0.35 g NaHCO3, 0.40 gKCl, 0.06 g KH2PO4, 0.10 g MgCl2 Æ 6H2O, 0.14 g CaCl2,0.06 g Na2HPO4 Æ 2H2O, 0.06 g MgSO4 Æ 7H2O, 1.00 g glu-cose in 1000 ml distilled H2O. This solution had an initialpH of 7.4 and was recommended by Pourbaix [28] for test-ing the degradation behavior of implant materials. Thespecimens were immersed in 15 mL of the solution forthe predetermined periods of time at 37 ± 0.5 �C. Afterimmersion, some specimens were removed from the vialsand placed in a container with fresh Hanks’ solution toevaluate the mechanical properties, in addition to somedried in an oven for analysis of phase and morphology.

2.3. Analysis of phase and morphology

The surface of composites was coated with gold andobserved under a Hitachi S-4200 field emission scanningelectron microscope (FESEM, Tokyo, Japan). Phase anal-ysis was performed using Shimadzu XD-D1 X-ray diffrac-tometry (XRD, Kyoto, Japan) with Ni-filtered CuKaradiation operated at 40 kV and 40 mA at a scanning speedof 2�/min.

2.4. Measurement of weight loss

The extent of the in vitro degradation was also moni-tored through sample weight change before and afterimmersion. Prior to weighing with a 4-digital balance(AE 240S, Mettler-Toledo AG, Greifensee, Switzerland),the immersed specimens were dried at 120 �C for 3 h inan oven. At least seven samples were tested for each mea-surement. Weight measurement was taken until constantweight was reached. The percentage of weight loss wasdetermined using the following equation: %W =(W0 �Wt) · 100/W0, where W0 is the initial weight of thespecimen and Wt the weight of the dried specimen atimmersion time t.

2.5. Three-point bending test

Three-point bending test was conducted on an EZ-Testmachine (Shimadzu, Kyoto, Japan) at a loading rate of0.5 mm/min. Span length was 10 mm. With the bendingof samples, ultimate bending strength (rb) and Young’sbending modulus (Eb) were calculated as follows [22]:

rb ¼8F maxL

pd3and Eb ¼

4L3

3pd4

DFDl

;

Fig. 1. XRD patterns of chitosan/calcium phosphate composites of 5 (a),10 (b), and 20 wt/v% CaP (c) before and after immersion in Hanks’solution.

S.-J. Ding / Journal of Non-Crystalline Solids 353 (2007) 2367–2373 2369

where Fmax is the maximum load (N), L the support span(mm), d the diameter of sample (mm), DF/Dl the slope ofthe initial linear elastic portion of the load-deflection curve(N/mm).

2.6. Statistical analysis

One-way analysis of variance (ANOVA) was used toevaluate the significant differences between the means inbending strength. In the event of a significant differencebetween test groups, it necessitated testing all the possibledifferences via multiple comparisons that were character-ized by considering any significant differences between allpossible pairs of groups. Scheffe’s multiple comparison testwas used to determine the significance of the standard devi-ations in the strength of each sample for different experi-mental conditions. In all cases, the results wereconsidered statistically different at p < 0.05.

3. Results and discussion

3.1. Phase composition

The synergistic combination of organic and inorganiccompounds in hybrid ceramic-polymer biomaterials notonly results in unique properties compared with those ofthe individual constituents, but also resembles the mor-phology and properties of natural bone, making them veryuseful in hard tissue repair and replacements. Fig. 1 showsthe XRD measurement results of the CTS/CaP compositesbefore and after immersion in Hanks’ solution. As previ-ously stated [27], diffraction intensities of the heat-treatedcomposites resulted mainly from crystalline CaP, althoughchitosan did contribute to making the diffraction peaks ofthe apatitic phase of CaP particles more diffuse. This couldbe explained by the preparation method used in the presentstudy whereby calcium phosphate particles were incorpo-rated within the chitosan matrix.

The XRD patterns of immersed samples for a series oftimes are also shown in Fig. 1. After immersion in Hanks’solution, The CTS/CaP composites of 2.5 wt/v% CaPentirely disintegrated within 7 day, but maintaining up to30 days for 5 wt/v% CaP sample. This was because the dis-solution rate of chitosan was much higher than calciumphosphate [29]. As for higher content of CaP such as 10and 20 wt/v%, there were no obvious changes observed inthese XRD patterns, even after immersion up to 90 days.The failure to see significant changes in the XRD patternsdid not necessarily imply that nothing had happened at thevery surface during immersion. The difficulty for XRD todetect small changes on immersed surfaces was largelyresolved by utilizing the FESEM technique.

3.2. Morphology

The morphologies of the four different as-made CTS/CaP composites have been characterized as described in

Ref. [27]. In brief, the morphology of the compositesbecame heterogeneous with CaP addition, and this ten-dency became more obvious with increasing CaP content.

2370 S.-J. Ding / Journal of Non-Crystalline Solids 353 (2007) 2367–2373

A 5 wt/v% ratio of CaP avoided powder agglomerationwith a homogenous distribution of CaP within the chitosanmatrix, thereby achieving maximum mechanical strength.On the contrary, at 20 wt/v% ratio, the composite exhibitedsevere agglomeration and a loose structure and had thelowest initial strength between the four composites.

Concerning immersion effect, FESEM microphoto-graphs confirmed the degradation of all composites inHanks’ solution (Fig. 2). When the composites contained5 wt/v% CaP particles, the 15-day immersed sample surfaceapparently appeared pores/cracks, as shown in Fig. 2a. Asa sign of degradation, the porosity increased with increas-ing immersion time in Hanks’ solution, likely the 30-dayimmersed sample. Nevertheless, the more CaP content inthe composites, the greater resistance to disintegration insimulated physiological solution. It was not until immer-sion for 90 days that the 20 wt/v% CaP-containing com-posites appeared a lot of pores. The hybrid compositeseemed to be dissociated possibly due to the degradationof the chitosan. The development of the porous microstruc-ture, as expected, could lead to the decreased bendingstrength of all immersed samples (to be discussed below).

Fig. 2. Surface FESEM micrographs of chitosan/calcium phosphate composite

However, Fernandez et al. [30] stated that bone tissue cellscould use the increasing porosity to improve the osteointe-gration of the implant and to accelerate its complete trans-formation into real bone tissue.

3.3. Weight loss

To further study the immersion-induced degradationprocess, a series of weight change measurements were per-formed for all immersed samples. Fig. 3 shows that thesamples containing lower CaP content such as 5 and10 wt/v% continue to dissolve after immersion in Hanks’solution, consistent with SEM observation results. Thecomposites containing 5 wt/v% CaP sharply varied inweight loss of approximately 30% after 7-days of immer-sion and were entirely dissolved in Hanks’ solution in30 days. After 90 days of degradation, the composite mate-rial made up of 10 wt/v% CaP experienced a weight loss of45%. On the contrary, the 20 wt/v% CaP sample keptweight loss to be remained 5% over 90 days of immersion.As discussed earlier, this faster dissolution of the compos-ites was due to organic chitosan matrix. These weight loss

s of 5 (a), 10 (b), and 20 wt/v% CaP (c) after immersion in Hanks’ solution.

Fig. 3. Weight loss of chitosan/calcium phosphate composites afterimmersion in Hanks’ solution.

S.-J. Ding / Journal of Non-Crystalline Solids 353 (2007) 2367–2373 2371

data further confirmed the lower dissolution rate of thecomposite containing higher CaP amount, when immersedin Hank’s solution.

3.4. Mechanical properties

Besides possessing bioactivity, a desirable material forbiodegradable implants should provide enough initialmechanical strength to support the strength of the healingtissue during an early stage. As heat-treated chitosan can-not be formed into the sound shape of a rod, its strengthwas not measured. Table 1 shows the relationship betweenthree-point bending strength of CTS/CaP composites andCaP content prior to and after immersion in Hanks’ solu-tion. One-way ANOVA analysis of the bending strengthdata prior to immersion showed that the variation instrength between samples was significant (p < 0.05). Com-posites containing 2.5 wt/v% CaP showed bending strengthof 30.8 MPa on average. Addition of CaP up to 5 wt/v%achieved significantly increased bending strength of up to45.7 MPa (p < 0.05) compared to the other three. However,higher CaP content at 20 wt/v% adversely affected themechanical properties. This was probably due to the loose

Table 1Mean bending strength and standard derivation (SD) of the four different chitsolution

CaP content (%) Mean bending strength ± SD (MPa)

As-made 7-day 15

2.5 30.8 ± 1.2 – –5 45.7 ± 3.6 23.3 ± 1.9 1710 27.0 ± 2.1 7.2 ± 1.1 620 10.6 ±1.7 5.5 ± 0.5 5

Number of samples is at least 10 in each subgroup.

connections between excess CaP particles, as evidenced bysevere agglomeration and a loose structure revealed byFESEM. Therefore, it could be expected that excess CaPmight disrupt the chitosan network, resulting in weakerinterfacial bonding between chitosan matrix and CaP con-tent [27].

The variations in the bending strength with immersiontime revealed that, after immersion for 7 days, the differenttypes of composite samples largely declined the strength(Table 1). For 2.5 wt/v% CaP-containing composites, noneof the immersed specimens was tested due to its severe deg-radation even in the 7-day specimen. As for the other threecomposites, the strength was significantly reduced from theinitial strength of 45.3, 27.0 and 10.6 MPa down to 23.3,7.2, 5.5 MPa for 5, 10 and 20 wt/v% CaP additives, respec-tively, after 7-day immersion (p < 0.05). In contrast to thelower CaP content, the composite with higher CaP amountseemed to have a more stable bonding. The 90-day strengthof the composite containing 10 and 20 wt/v% CaP declineddown to about 2 MPa, instead of complete declination.This deterioration in the strength seems unavoidable forbiodegradable composites immersed in simulated physio-logical solution and has also been observed in other studies[4,31,32]. Ignatius et al. found that composite pins made ofpolylactide containing 30 w/w% b-tricalcium phosphateprepared by injection molding process lost half of theirstrength after 16 weeks [31]. Zhang and Xu reported thatthe calcium phosphate cement with suture fiber and chito-san had a bending strength of 40.5 MPa after 1 day immer-sion; its strength was 4.2 MPa at 119 days [32]. Theimmersion-induced degradation in mechanical strengthwas due to the dissolution of the chitosan and calciumphosphate resulting in a porous structure. In addition,the penetration of water/ions resulting from the solutioncould reduce the adhesion between interfaces. Althoughhaving a quite low strength, the present CTS/CaP compos-ite might provide initial supporting the in vivo loadingwhen implanted.

To further understand the degradation behavior, Fig. 4shows the degree of the strength decrease as a function ofimmersion time. The bending strength of immersed com-posites mixed with 10 wt/v% CaP having a higher initialstrength of 27.0 MPa largely decreased by about 92% afterimmersion for 90 days (p < 0.05). Contrary to the findings,20 w/v% CaP sample with 10.6 MPa initial strength had a

osan/calcium phosphate composites before and after immersion in Hanks’

-day 30-day 60-day 90-day

– – –.0 ± 1.4 15.5 ± 1.9 – –.0 ± 0.4 5.4 ± 1.3 4.3 ± 1.0 2.2 ± 0.2.3 ± 0.7 4.5 ± 0.8 3.6 ± 0.4 1.8 ± 0.1

Fig. 4. Degree of strength decrease of immersed chitosan/CaP samples asa function of immersion time.

Table 2Mean Young’s modulus and standard derivation (SD) of the four different chitosan/calcium phosphate composites before and after immersion in Hanks’solution

CaP content (%) Mean Young’s modulus ± SD (MPa)

As-made 7-day 15-day 30-day 60-day 90-day

2.5 77.3 ± 5.1 – – – – –5 60.8 ± 5.4 45.3 ± 1.3 30.0 ± 1.9 20.8 ± 0.6 – –10 25.0 ± 5.7 23.2 ± 1.7 22.0 ± 1.8 21.0 ± 1.2 17.1 ± 0.9 14.9 ± 1.420 10.2 ± 1.6 8.5 ± 1.3 7.7 ± 0.7 5.9 ± 0.5 6.3 ± 0.8 5.9 ± 1.1

Number of samples is at least 10 in each subgroup.

2372 S.-J. Ding / Journal of Non-Crystalline Solids 353 (2007) 2367–2373

strength reduction of 83% (p < 0.05). The chitosan/calciumphosphate composites containing 10 wt/v% might be anoptimal material in terms of initial strength and degrada-tion behavior. Its initial strength was comparable to thetensile strength of about 3.5 MPa for cancellous bone [33].

The variations of the elastic modulus prior to and afterimmersion time for hybrid composites are listed in Table 2.The amount of CaP added influenced the elastic modulusof the composites in the range of 10–80 MPa; however,the trend presented was not similar to the changes inthree-point bending strength. A significant difference(p < 0.05) in Young’s modulus among the composites wasfound. Although the detailed mechanism of this ‘monoto-nous’ change was not fully understood, it could be attrib-uted to the chitosan content. Chitosan possesses theadhesive and flexible properties and it binds the CaP parti-cles together. Upon immersion in solution, there was also apronounced decrease in the elastic modulus for the presentcomposites, similar to the bending strength. With increas-ing immersion time, the modulus decreased and reacheda value of 6–15 MPa.

Based on the above results, the important factor affect-ing the degradation of CTS/CaP composites appeared tobe the amount of calcium phosphate in the chitosan matrix.

Hence, the different ratio of calcium phosphate to chitosanprepared by the simple mixing-and-heating method led todifferent initial mechanical properties and biodegradationdegree tailoring the requirements of a specific medicalapplication such as bone tissue defect. Further studies arecertainly needed in order to evaluate the cytotoxic effectof glutaraldehyde as a crosslinker in the present chitosan/calcium phosphate.

4. Conclusions

The biodegradable composites based upon chitosan andcalcium phosphate have been prepared using a simple mix-ing-and-heating method. The detrimental effects of the sim-ulated physiological environment on mechanical propertiesof the hybrid composites resulted in the significantlydecrease in strength and modulus. The chitosan/calciumphosphate composites containing 10 wt/v% might an opti-mal material in terms of initial strength and degradationbehavior. Although susceptibility to solution attack, thistype of chitosan/calcium phosphate composites with highinitial strength might be acceptable for use in bone tissuerepair.

Acknowledgement

The authors acknowledge with appreciation the supportof this research by the National Science Council of theRepublic of China under the contract, No. NSC 94-2320-B-040-013.

References

[1] K. Piekarski, Int. J. Eng. Sci. 11 (1973) 557.[2] C. Durucan, P.W. Brown, Adv. Eng. Mater. 3 (2001) 227.[3] M.C. Chang, T. Ikoma, M. Kikuchi, J. Tanaka, J. Mater. Sci. Lett. 20

(2001) 1199.[4] S.J. Ding, C.W. Wang, D.C. Chen, H.C. Chang, Ceram. Int. 31

(2005) 691.[5] L. Lu, B.L. Currier, M.J. Yaszemsk, Cur. Opin. Orthop. 11 (2000)

383.[6] D. Baksh, J.E. Davies, S. Kim, J. Mater. Sci. Mater. Med. 9 (1998)

743.[7] M. Kikuchi, Y. Koyama, K. Takakuda, H. Miyairi, N. Shirahama, J.

Biomed. Mater. Res. 62 (2002) 256.[8] C.M. Agrawal, R.B. Ray, J. Biomed. Mater. Res. 55 (2001) 141.[9] J.K. Francis Suh, H.W.T. Matthew, Biomaterials 21 (2000) 2589.

[10] E.B. Denkbas�, M. Odabas�i, J. Appl. Polym. Sci. 76 (2000) 1637.

S.-J. Ding / Journal of Non-Crystalline Solids 353 (2007) 2367–2373 2373

[11] K.G.H. Desai, H.J. Park, Drug Dev. Res. 64 (2005) 114.[12] P.J. Vande Vord, H.W.T. Matthew, S.P. DeSilva, L. Mayton, B. Wu,

P.H. Wooley, J. Biomed. Mater. Res. 59 (2002) 585.[13] L. Ma, C. Gao, Z. Mao, J. Zhou, J. Shen, X. Hu, C. Han,

Biomaterials 24 (2003) 4833.[14] A.N.L. Rocha, T.N.C. Dantas, J.L.C. Fonseca, M.R. Pereira, J.

Appl. Polym. Sci. 84 (2002) 44.[15] L.Y. Lim, E. Khor, C.E. Ling, J. Biomed. Mater. Res. (Appl.

Biomater.) 48 (1999) 111.[16] T. Kitsugi, T. Yamamuro, T. Nakamura, Y. Kakutani, T. Hayashi, S.

Ito, T. Kokubo, M. Takagi, T. Shibuya, J. Biomed. Mater. Res. 21(1987) 467.

[17] Y. Zhang, M. Zhang, J. Biomed. Mater. Res. 55 (2001) 304.[18] R. Murugan, S. Ramakrishna, Comp. Sci. Technol. 65 (2005) 2385.[19] T. Kawakami, M. Anotoh, H. Hasegawa, T. Yamagishi, M. Ito, S.

Eda, Biomaterials 13 (1992) 759.[20] A. Yoshida, T. Miyazaki, E. Ishida, M. Ashizuka, Mater. Trans. 45

(2004) 994.[21] Y.J. Yin, F. Ye, J. Cui, F. Zhang, X. Li, K.D. Yao, J. Biomed. Mater.

Res. A 67 (2003) 844.

[22] Q. Hu, B. Li, M. Wang, J. Shen, Biomaterials 25 (2004) 779.[23] M. Ito, Y. Hidaka, M. Nakajima, H. Yagasaki, A.H. Kafrawy, J.

Biomed. Mater. Res. 45 (1999) 204.[24] I. Yamaguchi, K. Tokuchi, H. Fukuzaki, Y. Koyama, K. Takakuda,

H. Monma, J. Tanaka, J. Biomed. Mater. Res. 55 (2001) 20.[25] W. Tachaboonyakiat, T. Serizawa, M. Akashi, J. Biomater. Sci.

Polym. Edn. 13 (2002) 1021.[26] A. Matsuda, T. Ikoma, H. Kobayashi, J. Tanaka, Mater. Sci. Eng. C

24 (2004) 723.[27] S.J. Ding, Dent. Mater. J. 25 (2006) 706.[28] M. Pourbaix, Biomaterials 5 (1984) 122.[29] Y. Zhang, M. Zhang, J. Biomed. Mater. Res. 61 (2002) 1.[30] E. Fernandez, M.D. Vlad, M.M. Gel, J. Lopez, R. Torres, J.V.

Cauich, M. Bohner, Biomaterials 26 (2005) 3395.[31] A.A. Ignatius, P. Augat, L.E. Claes, J. Biomater. Sci. Polym. Edn. 12

(2001) 185.[32] Y. Zhang, H.H.K. Xu, J. Biomed. Mater. Res. A 75 (2005) 832.[33] C.J. Damien, J.R. Parsons, J. Appl. Biomater. 2 (1991) 187.