Biodegradable polymer/bioceramic composite scaffolds can overcome the limitations of conventional ceramic bone substitutes, such as brittleness and difficulty in shaping. To better mimic the mineral components and microstructure of natural bone, a novel nano-hydroxyapatite (nHAp)–chitosan composite scaffold including gelatin and polymer (poly(lactic acid)) with high porosity was developed using a sol-gel method and subsequently lyophilized for efficient bone tissue engineering. The nanocrystalline structure of hydroxyapatite was observed using X-ray diffraction analysis and the composite showed crystallinity due to the presence of nHAp. The pore diameter of the composite containing 5% nHAp was found to be 125 μm, while the composites with 10%, 15%, and 20% nHAp revealed a smaller pore size in the range of 15–28 μm. The highest compressive strength of 5.5 MPa was observed for the 10% nHAp-containing scaffold, whereas thermogravimetric analysis showed 90%–94% degradation at a temperature of 600°C, which demonstrated its excellent thermal stability. Antibacterial and cytotoxicity test results revealed that the composite is resistant toward microbial attack and has low sensitivity in cytotoxicity. The compressive strength data suggests that the composite does not have enough strength as that of human compact bone; however, the highly porous structure as observed in scanning electron microscopy makes it possible for use as an excellent substrate in the spongy bone of humans.
In the tissue engineering approach, the temporary three-dimensional (3D) scaffold serves an important role in the control of the functions of osteoblasts as well as a central role in the guidance of new bone formation into desired shapes . The most widely used materials for bone replacement are autografts, allografts, and artificial bone substitutes . Although autografting is a popular procedure for reconstructive surgery, it has several disadvantages, such as the shortage of donor supply, the persistence of pain, nerve damage, fractures, and cosmetic disability at the donor site. On the other hand, there are no donor site problems for allografting, although it has some clinical risks including disease transmission and immunological reaction. The scaffolds for bone regeneration mainly act as the substrate for the attachment and proliferation of osteogenic cells and, in this context, 3D biodegradable materials with a porous structure  have gained great attention as bone grafts not only because of their composition and structural similarity with natural bone but also because of their unique functional properties such as larger surface area and superior mechanical strength compared with their single phase constituents. Many of the natural biodegradable polymers such as gelatin, fibrin, chitosan, and starch exhibit good biocompatibility and some osteoconductive properties , . Synthetic biodegradable polymers offer a tremendous versatility and are widely used as scaffolds for tissue engineering. Degradation at an appropriate rate without undesirable by-products can be modulated by varying the ratios of the components , , , , , . In particular, synthetic bioresorbable polymers including poly(e-caprolactone), a slowly degrading aliphatic polyester, and poly(a-hydroxy acids), have been studied extensively. The latter group includes poly(lactic acids) (PLA) and poly(glycolic acids) that have been approved for other medical applications , . Porous structures can be produced by using several processing techniques such as solvent casting, phase inversion, fiber bonding, emulsion freeze-drying, and gas forming , . To overcome the drawbacks of the different biomaterials used as substrate for bone tissue engineering, an increasing amount of attention has been given to composite structures. This approach allows the reproduction of the entire bone structure that is more precise from a biomimetic perspective. For instance, polymer matrices combined with bioceramics (hydroxyapatite) or bioactive glasses allow mimicking the organic and mineral phases found in bone. Mineral components shaped as powders or granules can be associated with polymers and formulated as macroporous scaffolds by various means such as thermally induced phase separation, solvent casting/particulate leaching, solid freeform fabrication, or coating. Polymer microfibers have been used to mimic collagen fibers whereas the biomineralization process of bone has been reproduced by mixing with ceramic particles ,  or coating with calcium phosphates. Besides benefiting from both the biodegradability of polymers and the bioactivity of ceramics and glasses, the interest in composites can be found in the improvement of the resulting mechanical properties or coating with calcium phosphates. For instance, ceramics with low elastic modulus are substantially improved when associated with polymers. Similarly, for hydrogels, composites that combine the toughness of the polymer phase with the compressive strength of an inorganic phase have been produced , . However, despite promising results, the mechanical properties of current composites are still less than those of natural bone tissue. In this article, biodegradable and biocompatible scaffolds were fabricated from porous PLA, gelatin (GEL), chitosan (CHT), and nano-hydroxyapatite (nHAp) using the sol-gel technique for bone tissue engineering applications. The complete characterization of the scaffold will be discussed in the following sections.
2 Materials and methods
Nitric acid (65%), potassium dihydrogen phosphate, ammonia solutions, sodium hydroxide, hydrochloric acid (37%), and glacial acetic acid (99.5%) were purchased from Merck, Germany. Type B gelatin granules were purchased from Merck, India. PLA, chloroform, dimethyl sulfoxide (DMSO), and thioacetic acid were obtained from Sigma Aldrich, Germany.
2.2 Synthesis of nHAp from eggshells
Eggshells were crushed slightly for easy drying and placed in an oven at 100°C for 2 h. After that, the dried eggshells were ground in a grinder. The finely crushed powder was then calcined in air using an electric furnace at 900°C (heating rate of 10°C/min) for 1 h. At this temperature, the eggshells transform into calcium oxide (CaO) by freeing carbon dioxide (CO2). The CaO thus obtained from the eggshells was then dissolved in concentrated HNO3 and diluted with distilled water such that 1 m of Ca(NO3)2 solution was obtained and the solution was filtered. A total of 0.6 m of monobasic potassium dihydrogen phosphate (KH2PO4) solution was slowly added to the filtrate. The pH of the solution was kept at 10 by adding ammonia solution. The solution was then rigorously stirred (1000 rpm) using a hot plate magnetic stirrer for 1 h and kept for ageing overnight at room temperature. A bilayer solution of clear ammonia at the top and a white precipitate (nHAp) at the bottom of the beaker was obtained. The overall reactions were
The excess ammonia solution at the top was poured out using a decantation method and the precipitate was washed with distilled water to remove NH4+ and NO3− ions. The neutral white solution was then centrifuged for 20 min and the precipitates were collected. Then, the precipitate was dried at 65°C for 24 h in an oven, and finally, the dried white powder of nHAp was obtained.
2.3 Extraction of chitosan from prawn shell
Chitosan was extracted from waste prawn shell using the modified method of Rahman et al. . The extraction of chitosan from waste prawn shell involves three major steps: (a) deproteination, (b) demineralization, and (c) deacetylation. The washed and crushed waste shell was treated with 4% (w/w) sodium hydroxide (NaOH) for 3 h for deproteination. The residue was washed until neutral and dried. The dried shell was treated with 3N HCl and stirred for 3 h. This step is called demineralization. Then, the prawn shell was washed and dried and pure chitin was obtained. The chitin was deacetylated by heating at 80°C–100°C with 50% NaOH (w/w) for 4 h. Then, the mixture was washed properly to remove the NaOH from the product and finally allowed to dry. The degree of deacetylation of chitosan ranges from 56% to 99% with an average of 80% (details of the calculation of degree of deacetylation are given in Supporting Information, Supplementary Figure S1). Moreover, the molecular weight of chitosan (viscosity average molecular weight, Mv) was 155,245.5 Da (details are given in the Supporting Information, Supplementary Figure S2).
2.4 Preparation of scaffolds
Biodegradable biomimetic composite scaffold C1–C5 were prepared by varying the amounts of gelatin and nHAp in chitosan solution as shown in Table 1. In a typical procedure, accurately weighed gelatin granules were dissolved in deionized water with constant stirring and heated at 50°C. On the other hand, chitosan was dissolved in 1% acetic acid solution at 40°C. In the next step, 18.5% PLA solution was prepared by dissolving 3.7 g of PLA in 50 ml of chloroform under constant stirring at room temperature. The required amount of nHAp was incorporated into the PLA solution as shown in Table 1. A total of 5 ml of chitosan solution was then mixed with 15 ml of gelatin solution under constant stirring at 50°C and the solution was then mixed with different PLA/nHAp solutions. Then, the final nHAp/GEL/CHT/PLA solution was poured into aluminum foil as cube shapes (4 cm×4 cm×4 cm) and frozen to –4°C for 24 h in a normal freezer, and then the scaffolds were placed in the freeze dryer. After removal of solvents, the final scaffolds were dried in atmospheric temperature for 24 h and cut into definite shapes as required for mechanical testing and then preserved in desiccators.
|Types of composites||Amount (g)||Percentage based on composite||Total weight (g)|
|nHAp||Gelatin||CHT||PLA||nHAp (%)||Gelatin (%)||CHT (%)||PLA (%)|
Fourier transform infrared (FT-IR) spectra of the prepared samples were recorded with FT-IR 8400S (Shimadzu, Japan) spectrophotometer in the range of 4000–400 cm−1. The surface morphology of the sample was measured with a JEOL JSM-6490 LA (Jeol, Japan) scanning electron microscope (SEM) at an accelerating voltage of 10 kV. All the samples were coated (to make them conductive) with platinum using a JEOL JFC-1600 Auto Fine Coater. The pore length and width were measured with a stereomicroscope (Euromax, Germany) equipped with an optical micrometer. Pore size was calculated by using the expression d=√(l.h), where l and h are the average length and width of the pores, respectively. The compressive properties of the composite scaffolds were determined with a universal testing machine (H50KS-0404, Hounsfield Series S, UK) with a 500N load cell at a cross-head speed of 0.5 mm/min until failure. The samples were cut into cubes of approximately 9 mm in length, 8 mm in width, and 10 mm in height in accordance with the compression mechanical test guidelines set in the American Society for Testing and Materials (ASTM F 451-95). The compressive strength (δc) was calculated using the following formula: δc=F/A, where F is the maximum stress and A is the cross-sectional area of the cylinder. The values were expressed as the means±standard error (n=5). Images of the composite used for the compressive strength test are shown in Figure 1.
X-ray diffraction (XRD) analyses were carried out with a powder diffractometer A D8 Philips Advance X-ray diffractometer using Cu radiation of wavelength, λ=1.54 and a graphite monochromator with a current of 40 mA and a voltage of 40 mV was used to evaluate nHAp powder, chitosan, gelatin, PLA, and the composite scaffolds. The diffraction intensity was in the range of 5°–40° of 2θ (Bragg angle) and the scanning speed was 2°/min. Thermogravimetry (TG)/differential thermal analysis (DTA) of nHAp, chitosan, gelatin, PLA, and composite scaffolds were done using a thermogravimetric analyzer (Seiko Extar TG/DTA 6300, Seiko, Japan). Thermogravimetric analysis (TGA) of each sample was performed using a Perkin-Elmer setup (TAQ-500) and a heating rate of 20°C/min under a nitrogen atmosphere.
2.6 Antimicrobial test
The antimicrobial activity of the composites was done using the Kirby–Bauer method . The Kirby–Bauer test, also known as the disk diffusion method, is the most widely used antibiotic susceptibility test to determine what antibiotics should be used when treating an infection. This method relies on the inhibition of bacterial growth measured under standard conditions. For this test, a culture medium, specifically the Mueller–Hinton agar, is uniformly and aseptically inoculated with the test organism. Mueller–Hinton agar medium is the only susceptibility test medium that has been validated by the National Committee for Clinical Laboratory Standards. Mueller–Hinton tryptic soy agar media was used for disk diffusion susceptibility testing. Two bacterial strains including selected Gram-positive bacteria (Staphylococcus aureus) and Gram-negative bacteria (Escherichia coli-0157) were selected to assess susceptibility patterns. The antibacterial test was done in the food laboratory of the Centre for Advanced Research in Science, University of Dhaka. A brief description of the process is as follows:
Sterile tryptic soya agar (TSA) plate was prepared according to the direction for the growth of bacterial strain. After autoclaving, Mueller–Hinton TSA media was allowed to cool to 50°C. The selected bacterial strain was injected into the prepared broth solutions using an inoculating loop and kept in an incubator at 37°C for 18 h. Each sample (0.1 g) was dissolved into DMSO to perform the test. The cultured bacterial strains were swabbed using a cotton bud into the TSA plate and the sample and the standard disc (nalidixic acid) were injected within 15 min after swabbing. Then, the Petri dishes containing the sample and bacterial strain were incubated at 37°C for 18 h and the inhibition zone that appeared for both sample and standard were measured with a scale and recorded in mm.
2.7 Cytotoxicity test
The cytotoxic effect was examined at the Centre for Advanced Research in Science. Biological Biosafety Cabinet (NU-400E, Nuaire, USA), CO2 incubator (Nuaire, USA), a trinocular microscope with a camera (Olympus, Japan), and hemocytometer were used for the analysis. Sample stock solutions were prepared by dissolving composite materials in 10% DMSO using a serial dilution technique. Then, the samples were placed in an autoclave to prevent any bacterial growth. The next day, HeLa, a human cervical carcinoma cell line, was maintained in Dulbecco’s modified Eagles’ medium containing 1% penicillin–streptomycin (1:1) and 0.2% gentamycin and 10% fetal bovine serum. Cells (4×104/400 μl) were seeded onto 24-well plates and incubated at 37°C+5% CO2. The next day, 100 μl of the sample was added to each well. Cytotoxicity was examined under an inverted light microscope after 24 and 48 h of incubation. Duplicate wells were used for each sample.
3 Results and discussion
3.1 FT-IR analysis
The FT-IR spectra of pure HAp, chitosan, gelatin, and PLA is given in Supplementary Figure S3, and Figure 2 shows the FT-IR spectra of prepared composites. From Figure 2, it can be stated that a peak in the region of 3500–3000 cm−1 of the broad absorption band was seen for every composite (3378, 3427, 3428, 3442, and 3443 cm−1 for composites C1–C5, respectively), which is related to the –OH stretching vibrations. It was also observed from the figure that all composites showed peaks at 1763 cm−1 for >C=O groups. The absorption bands between 1700–1000 cm−1 and 875–870 cm−1 originated from the –CH2 and –CH3 stretching and bending vibrations, respectively. An absorption band at 1037 cm−1 was indicated for the PO43−group. The FT-IR data suggested that there may be a noncovalent interaction or that mechanical blending occurred in the nHAp-GEL-CHT-PLA composite scaffolds.
3.2 Surface morphology
In Figure 3A, C-1 having 80% gelatin, 1.5% chitosan, and 18.5% PLA yielded three phases and there are also cavities at the interface among the phases. The cavities showed poor interfacial adhesion due to the hydrophobic character of PLA. From the figure, it is obvious that the surface is rough and uneven. This feature dominated the morphology of pure gelatin. Also, there are crystalline flat lamellae with leaf-like shapes that dominated the morphology of pure chitosan. From the micrographs of Figure 3B, it is clear that composite C-2 displayed heterogeneous 3D porous structure with interconnecting polyangular pores, whereas nHAp particles were homogeneously distributed throughout the scaffold (Supplementary Figure S4). Meanwhile, the pore diameter of composite containing 5% nHAp was found to be 125 μm, whereas it is observed from Figure 3C to Figure 3E that the sample with 10%, 15%, and 20%, respectively, nHAp in gelatin/chitosan/PLA matrix revealed a smaller pore size in the range of 15–28 μm. Because one osteoblast occupies an area of approximately 700 μm2  and the pore size suitable for cell infiltration and ingrowth of host bone tissue was reported to be in the range of 100–350 μm; ,  hence, the pore size of 125 μm (diameter of a spherical pore) of composite C-2 is the best compatible for osteoconduction. The SEM image of hydroxyapatite at ×25,000 magnification is shown in Figure 3F. From the SEM pictures, it is clearly observed that the precipitated crystallites of nHAp were of spherulite morphology with the narrow size distribution of about a few microns in diameter. The spherulite is composed of tiny nanosize platelets of loosely aggregated stabilized structure, which was confirmed by TEM analysis (Figure 4). From the TEM micrograph illustrated in Figure 4, it can be seen that synthesized HAp is in nanoscale with the crystal structure; however, some HAp particles are rod-like. The average diameter of prepared HAp has been found to be ~15 nm.
3.3 XRD analysis
The XRD patterns of nHAp, chitosan, gelatin, and PLA film are shown in Supplementary Figure S5 in the Supporting Information. The peak analysis of Supplementary Figure S5 shows that the average particle size of the synthesized HAp crystal is 14.9 nm at 2θ angle of 25.8° and it was done by using the Debye–Scherrer formula represented by the following equation: 
The XRD pattern for the chitosan exhibits peaks at 2θ=10.88° and 19.56°. The peaks are prominent and sharp and showed the crystalline nature of chitosan. Gelatin exhibits a small and broad peak at 2θ=14.69° with one less prominent peak and can be considered as semicrystalline in nature. Molten PLA film exhibits a small and broad peak at 2θ=16.6° with one less prominent peak at 32.1° and it can also be considered as semicrystalline in nature. Supplementary Figure S5 showed the maximum peak at 16.6°, which is due to the planes (200) and (110) of the orthorhombic structure . The XRD patterns for composites are plotted in Figure 5. Composite C-1 showed peaks at 10.8°, 20.5°, and 32.1°. C-2 showed sharp peaks at 25.5° and 32.5° and a broad peak at 16.6°. Composite C-3 produced sharp peaks at 25.8° and 32.5° with some less prominent peaks. C-4 and C-5 showed the same peaks described above and the peaks become more prominent. The higher crystallinity of C-2–C-5 compared with C-1 was due to the fact that nHAp has incorporated more crystallinity into GEL-CHT-PLA blends.
3.4 Compressive strength
The mechanical properties of the composites in terms of compressive strength were determined from the stress–strain curve by applying the load until the scaffold was cracked and the results were plotted in Figure 6. Among all the composites, the highest compressive strength was found to be 5.49 MPa for the 10% nHAp-containing composite. It exerts the highest compressive strength due to its compact and less porous structure, which indicates good dispersion of nHAp into GEL-CHT-PLA blend. Whereas composite containing 0% nHAp showed higher compressive strength, i.e. 4.76 MPa compared with C-2 (2.62 MPa), C-4 (3.34 MPa), and C-5 (3.62 MPa) because it has the higher percentage of gelatin and chitosan and gelatin is uniformly bonded with PLA. The average ultimate compressive strengths of human spongy bone and compact bone were found to be 25 and 152.2 MPa, respectively . The composites showed very low compressive strength compared with that of the natural human bone. Because the composites (C2–C5) are highly porous and spongy, they have the potential to be used as the spongy bone in humans.
3.5 Thermogravimetric analysis
The TGA curves for thermal degradation of hydroxyapatite, chitosan, gelatin, and PLA film are shown in Figure 7A. The weight loss curve for nHAp has two decomposition stages: approximately 3% weight loss at around 0°C–200°C is attributed to the loss of residual water and thermal decomposition of nHAp is 212°C–250°C, which causes weight loss of approximately 8%. A total degradation of 12% was observed at the maximum temperature of 600°C. The TGA curves for all composites are shown in Figure 7B, which also showed the initial loss due to moisture or bound water. C-1 yielded the weight loss of approximately 65% at 309°C–336°C and 90% at 600°C. C-2 showed 44% decomposition at 321°C–354°C and 60% at 600°C, and C5 showed 14% at 313°C–358°C and 32% at 600°C. The results clearly indicated that thermal stability increases with increasing nHAp in the composites.
3.6 Differential thermal analysis
The DTA curve of nHAp, GEL, CHT, and PLA are given in Supplementary Figure S6 in the Supporting Information. The DTA curve of the composites is shown in Figure 8. C-1 in Figure 8 yielded endothermic peaks at approximately 69°C, 164°C, 277°C, and 332°C and maximum degradation occurred at 332°C. C-2 showed endothermic peaks at approximately 77°C, 167°C, and 233°C and maximum degradation 359°C. Then, the maximum degradation for C3 to C5 decreases compared with C1, as observed in Figure 8, which might be due to the reduced uniformity of nHAp in the composite. Composite C-5 showed endothermic peaks at approximately 79.8°C, 166°C, 306°C, and 355.7°C and maximum degradation occurred at 355.7°C.
3.7 Antimicrobial test
The measurement of the antimicrobial activity of five composites (C1–C5) was done in two bacterial strains including selected Gram-positive bacteria (S. aureus) and Gram-negative bacteria (E. coli-0157). Zone inhibition was measured (including the diameter of the disk) with a ruler under the surface of the plate without opening the lid. The zones of growth inhibition were compared with standard drug-loaded disks. From Table 2, it could be observed that all the composites showed exactly the same inhibition zone with excellent antibacterial properties.
|Gram-positive bacteria||S. aureus||10||10||10||10||10|
|Gram-negative bacteria||E. coli-0157||8||8||8||8||8|
3.8 Cytotoxicity test
In vitro cytotoxicity tests of different composite scaffolds were performed using HeLa cells containing 1% penicillin–streptomycin (1:1) and 0.2% gentamycin and 10% fetal bovine serum. Cytotoxicity was examined under an inverted light microscope after 24 and 48 h of incubation and the effect was observed by comparing with the cytotoxic effect of control (10% DMSO solution) as shown in Supplementary Figure S7 (in the Supporting Information) and the analytical data is shown in Table 3.
|Type||Survival rate (%)||Morphology|
|Control (10% DMSO)||100||Normal|
From the above table, it is observed that the composite scaffolds are moderately biologically active. Here, the survival rate of control was considered as 100%, but in actuality, DMSO itself is cytotoxic. That is why the survival rate of composites decreases slightly; however, among the composites, having nHAp C-2 showed the lowest cytotoxic effect.
Regular and highly interconnected macroporous (150 μm) PLA/CHT/GEL/nHAp scaffold was prepared using the sol-gel technique. The composite scaffold showed excellent thermal strength, which is due to the presence of nHAp. The influences of nHAp content in the composites were studied and the results confirmed that 5% nHAp in the composite showed the best mechanical and thermal properties, and the least cytotoxic effect compared with other composites. Although the composites did not yield enough compressive strength similar to that of compact human bone, its highly porous nature makes it possible for use in the spongy bone of human body for cell attachment and migration in bone tissue engineering.
The authors highly acknowledge the International Foundation for Sciences for the research grant (Funder Id: 10.13039/501100001724, grantee F/4903-2) to carry out the research project.
 Porter A, Patel N, Skepper J, Best S, Bonfield W. Biomaterials 2003, 24, 4609–4620.10.1016/S0142-9612(03)00355-7Search in Google Scholar
 Shin H, Temenoff JS, Mikos AG. Biomacromolecules 2003, 4, 552–560.10.1021/bm020121mSearch in Google Scholar
 Murugan R, Ramakrishna S. In Molecular Building Blocks for Nanotechnology, Springer: New York, USA, 2007.Search in Google Scholar
 Lee CH, Singla A, Lee Y. Int. J. Pharm. 2001, 221, 1–22.10.1016/S0378-5173(01)00691-3Search in Google Scholar
 Bensaıd W, Triffitt J, Blanchat C, Oudina K, Sedel L, Petite H. Biomaterials 2003, 24, 2497–2502.10.1016/S0142-9612(02)00618-XSearch in Google Scholar
 Salgado AJ, Coutinho OP, Reis RL. Macromol. Biosci. 2004, 4, 743–765.10.1002/mabi.200400026Search in Google Scholar PubMed
 Wu S, Liu X, Yeung KW, Liu C, Yang X. Mater. Sci. Eng. R: Reports 2014, 80, 1–36.10.1016/j.mser.2014.04.001Search in Google Scholar
 Quinlan E, López-Noriega A, Thompson E, Kelly HM, Cryan SA, O’Brien FJ. J. Control. Release 2015, 198, 71–79.10.1016/j.jconrel.2014.11.021Search in Google Scholar PubMed
 García-Gareta E, Coathup MJ, Blunn GW. Bone 2015, 81, 112–121.10.1016/j.bone.2015.07.007Search in Google Scholar PubMed
 Schieker M, Seitz H, Drosse I, Seitz S, Mutschler W. Eur. J. Trauma 2006, 32, 114–124.10.1007/s00068-006-6047-8Search in Google Scholar
 Ignatius AA, Augat P, Hollstein E, Schorlemmer S, Peraus M, Pokinskyj P, Claes L. J. Biomed. Mater. Res. B Appl. Biomater. 2005, 75, 128–136.10.1002/jbm.b.30274Search in Google Scholar PubMed
 Wu C, Chang J. Biomed. Mater. 2013, 8, 032001.10.1088/1748-6041/8/3/032001Search in Google Scholar
 Sachlos E, Reis N, Ainsley C, Derby B, Czernuszka J. Biomaterials 2003, 24, 1487–1497.10.1016/S0142-9612(02)00528-8Search in Google Scholar
 Yang F, Both SK, Yang X, Walboomers XF, Jansen JA. Acta Biomater. 2009, 5, 3295–3304.10.1016/j.actbio.2009.05.023Search in Google Scholar
 Erisken C, Kalyon DM, Wang H. Biomaterials 2008, 29, 4065–4073.10.1016/j.biomaterials.2008.06.022Search in Google Scholar
 Niu X, Feng Q, Wang M, Guo X, Zheng Q. J. Control. Release 2009, 134, 111–117.10.1016/j.jconrel.2008.11.020Search in Google Scholar
 Wang J, Valmikinathan CM, Liu W, Laurencin CT, Yu X. J. Biomed. Mater. Res. A 2010, 93, 753.Search in Google Scholar
 Rashid TU, Rahman MM, Kabir S, Shamsuddin SM, Khan MA. Polym. Int. 2012, 61, 1302–1308.10.1002/pi.4207Search in Google Scholar
 Farzana K, Shah SNH, Jabeen F. J. Res. (Sci.) 2004, 15, 145–151.Search in Google Scholar
 Schwartz I, Robinson BP, Hollinger JO, Szachowicz EH, Brekke J. Otolaryngol. Head Neck Surg. 1995, 112, 707–713.10.1016/S0194-5998(95)70180-XSearch in Google Scholar
 Hulbert S, Hench L, Forbers D, Bowman L. Ceramics Int. 1982, 8, 131–140.10.1016/0272-8842(82)90003-7Search in Google Scholar
 Jin HH, Lee CH, Lee WK, Lee JK, Park HC, Yoon SY. Mater. Lett. 2008, 62, 1630–1633.10.1016/j.matlet.2007.09.043Search in Google Scholar
 Cullity BD, Stock SR. Elements of X-Ray Diffraction, Prentice Hall: Upper Saddle River, 2001.Search in Google Scholar
 Hellmich C, Ulm FJ. J. Eng. Mech. 2002, 128, 898–908.10.1061/(ASCE)0733-9399(2002)128:8(898)Search in Google Scholar
 Keller TS. J. Biomech. 1994, 27, 1159–1168.10.1016/0021-9290(94)90056-6Search in Google Scholar
The online version of this article offers supplementary material (https://doi.org/10.1515/polyeng- 2018-0103).
©2019 Walter de Gruyter GmbH, Berlin/Boston