Solid-state foaming of poly(ether-ether-ketone)/hydroxyapatite composites

Abstract

Poly(ether-ether-ketone) (PEEK) is a high strength and high temperature-resistant plastic with good potential for medical use. The addition of hydroxyapatite (HA) can improve the mechanical performance and biocompatibility of PEEK. However, little study has been done on solid-state foaming of PEEK or PEEK-based composites due to the semicrystalline nature and high processing temperature requirement. In this study, the solid-state foaming behavior of PEEK/HA composite is studied using supercritical ${C O}_{2}$ . A quenching process is applied to reduce the crystallinity and improve the gas saturation behavior. The mechanical properties of foamed PEEK/HA composite are characterized. The results show that quenching substantially improved the foamability of PEEK/HA composite. Foamed PEEK/HA regained crystallinity and exhibit a compressive strength comparable to human trabecular bone.

Keywords

Poly(ether-ether-ketone)hydroxyapatite solid state foaming crystallization tissue scaffolds

Introduction

In recent years, poly-ether-ether-ketone (PEEK) has become a popular candidate for bone implants. Traditionally, titanium (Ti) and titanium alloys are the main materials used for bone implant; however, these materials suffer from limitations, such as releasing harmful metal ions that could cause osteolysis and allergenic effects.^1,2 PEEK is a high temperature-resistant plastic with good chemical stability, matching mechanical strength as human bone, and X-ray transparency. However, neat PEEK prevents osteointegration with surrounding bone tissue, and can lead to fibrous encapsulation.^3,4

Several methods have been proposed to improve the bioactivity of PEEK. These include increasing surface roughness to improve the osteoblast adhesion,^5,6 adding organic/inorganic or other mineral phase such as hydroxyapatite (HA) or ${T_{i} O}_{2}$ to improve the osseointegration,^7,8 and generating porous structure to improve the in-growth of osteoblasts.^9,10 Combining bioactive materials such as HA with PEEK is considered to be an effective strategy to improve the mechanical property and biocompatibility.⁹ Moreover, creating a porous surface region allows the material to better integrate with the surrounding tissue, resulting in a higher fixation force of the implant. However, the mechanical strength reduces when making the implant porous.^9,11,12 Adding HA in PEEK could mitigate the mechanical strength loss in the porous PEEK implants.

Traditionally, porous PEEK/HA composites have been mainly fabricated using solvent casting, salt leaching, and additive manufacturing. Uddin et al. used solvent casting and salt leaching methods to create high porosity scaffold of PEEK/HA composites.¹³ Zheng et al. used a twin extruder to fabricate PEEK/HA filaments and then employed selective laser sintering to print porous scaffolds.¹¹ These methods can generate interconnect porous PEEK/HA structures; however, they either used an organic solvent to dissolve PEEK before salt particles were added or did not have enough resolution to create the pore size that is suitable for bone tissue engineering. Dissolving PEEK in organic solvent could leave residual solvent in the PEEK matrix that prevents proper cell growth.

Solid-state foaming is an organic solvent-free gas foaming process that is able to achieve high porosity structures.^14,15 During the foaming process, the pore size and porosity of the polymer foam can be controlled by adjusting the gas foaming parameters.¹⁶ However, currently, gas foaming is mainly used for amorphous plastics with a low melting temperature, such as polycaprolactone (PCL), polymethyl methacrylate (PMMA), and polyimide (PI).^16–18 There has been little research on solid-state foaming of PEEK or PEEK-based composites. PEEK is a semicrystalline polymer with high mechanical properties and a high melting temperature. Its foaming performance is different from amorphous polymers. In a previous study, Yang et al.¹⁹ examined the foaming behavior of PEEK with different crystallinity and thermal histories. It was found that foaming of high crystallinity PEEK yielded small cells of about 5–10 um, which are generally considered too small for bone tissue engineering. Li et al.²⁰ used the microcellular extrusion foaming method with ${C O}_{2}$ to fabricate porous PEEK filaments. High toughness and excellent thermal insulation properties were achieved; however, the filaments were foamed during the extrusion process, with limited control on the pore size. Liang et al.²¹ used a rapid depressurization method to foam a mixture of PEEK and glass fiber and achieved high expansion ratio with good mechanical properties. However, the method requires a rapid release of high pressure at a high temperature, putting strenuous stresses on the equipment.

In this study, we investigate the solid-state foaming behavior of PEEK/HA composites using supercritical ${C O}_{2}$ . To improve the gas concentration and foaming behaviors, a quenching process is applied to reduce the PEEK crystallinity before saturation. PEEK/HA composites of different loadings were foamed at different temperature and foaming time conditions. Mechanical properties of the foamed samples were characterized and compared with human trabecular bone. We show that the quenching process helps to achieve PEEK/HA foams of higher porosities and larger pore sizes and that adding HA can significantly improve the mechanical property of the PEEK/HA foams.

Experimental setup and procedure

Figure 1 shows a schematic of the experimental setup and procedure. PEEK pellets were obtained from Victrex (Grade 450G) and hydroxyapatite powder from Sigma-Aldrich. The pellets and powder were mixed using a Thermal Science Minilab II twin extruder at 370 $℃$ and 30 rpm for 20 mins. The weight ratio of HA ranged from 10% to 30%. The PEEK/HA composites were extruded and free cooled to room temperature. The PEEK/HA samples were then reheated to 340 $℃$ , when the polymer samples became semitransparent, and quenched in 15 $℃$ water to reduce the crystallinity. Both as-extruded and quenched samples were tested. The extrudates were sectioned into shorter samples with a size about 30 mm × 5 mm x 3 mm.

Figure 1.

Experimental setup and procedure.

The crystallinities of the PEEK and PEEK/HA samples were both assessed using differential scanning calorimetry (DSC) with a temperature range from 25 to 400 $℃$ . The crystallinity was determined as:

x_{c} = \frac{∆ H_{m} - ∆ H_{c c}}{∆ H_{m}^{0}}

(1)

where

∆ H_{m}

and

∆ H_{c c}

are the measured enthalpies of melting and cold crystallization during the DSC scan, respectively.

∆ H_{m}^{0}

is the melting enthalpy of 100% crystallinity and set to be 130 J/g according to literature.²²

For foaming, the PEEK/HA samples were first saturated at 20 MPa and 35 $℃$ . The samples were periodically retrieved from the pressure vessel and weighed to calculate the gas concentration. The saturated samples were foamed in a molten salt mixture at a temperature ranging from 300 to 360 $℃$ . After foaming, the samples were cleaned with water for characterization.

The morphology of foamed PEEK/HA samples was examined with a scanning electron microscope (SEM). The samples were also tested using DSC after foaming to study the effect of foaming on crystallinity. Compression tests were conducted to characterize the mechanical properties of both non-foamed and foamed PEEK/HA composites.

Results and discussion

Effect of quenching

The DSC results for both as-extruded and quenched samples are shown in Figure 2. Comparing as-extruded neat PEEK with 10% HA composite in Figure 2(a), the crystallinity of composite is clearly higher. The crystallinity of as-extruded neat PEEK samples was determined to be 28% and that of as-extruded PEEK/HA composite was 33%. This is probably due to the heterogeneous nucleation effect of hydroxyapatite particles during the solidification process.

Figure 2.

DSC results for (a) as-extruded samples; (b) quenched samples.

The DSC results for quenched PEEK/HA samples are shown in Figure 2(b). A significant difference of quenched samples than as-extruded samples is that a cold crystallization peak appears at around 175 $℃$ . The calculated crystallinity is 5.5%, 6.9% and 7.1% corresponding to 10%, 20%, and 30% hydroxyapatite content, respectively. The crystallinity is higher with a higher hydroxyapatite content for both as-extruded and quenched samples. More significantly, the crystallinity of as-extruded PEEK can decrease by more than 20% with quenching. This is because that PEEK is a semicrystalline polymer with long polymer chains. The material behaves as crystal in the region where polymer chains are closely packed together. As shown in Figure 3, the naturally cooled samples are opaque, and with higher hydroxyapatite loadings, the color of the sample becomes darker. Quenched samples are semi-transparent, with the neat PEEK sample having an amber color, while the PEEK/HA samples have a darker color. The quenching experimental results show that with the heat treatment and quenching process, the crystallinity of PEEK/HA composite can be reduced significantly.

Figure 3.

(a)-(d) Free cooled PEEK samples with 0%, 10%, 20%, and 30% hydroxyapatite, respectively, (e)-(h) quenched PEEK samples with 0%, 10%, 20%, and 30% hydroxyapatite, respectively.

Gas saturation results

Figure 4 shows the gas concentration of as-extruded and quenched PEEK/HA samples with time. The center line represents the average gas concentration, and the shaded area shows the range. As shown in Figure 4 the equilibrium gas concentration of as-extruded PEEK/HA composites is around 4%, and it takes about 96 hours to reach the equilibrium. All samples reached similar gas concentrations, regardless HA contents in the sample. This may be because HA particles are porous and will absorb gas as well as the PEEK matrix. This saturation behavior is the same for quenched samples, where all the samples reached similar equilibrium gas concentrations. However, the quenched PEEK/HA samples reached the equilibrium gas concentration in only 48 hours, much faster compared with as-extruded samples. The equilibrium gas concentration of the quenched samples reached 6%, which is significantly higher than that of as-extruded samples. This result shows that the reduction of crystallinity can significantly improve the gas absorption behavior of PEEK.

Figure 4.

Gas saturation. behavior of as-extruded and quenched PEEK/HA samples.

Foaming behavior of Poly(ether-ether-ketone)/hydroxyapatite composites

After full saturation, composite samples were foamed at 300, 330, and 360 $℃$ for 8 s. The SEM results of as-extruded samples are shown in Figure 5. The hydroxyapatite weight ratio in the composite was 20%. The SEM results of quenched samples are shown in Figure 6. The hydroxyapatite weight ratios of quenched samples were 10%, 20%, and 30%. Table 1 summarizes the mean pore size and cell density information for the samples tested. Figure 7 shows the pore size distributions. From Figure 5(a), it can be seen that at 300 $℃$ the as-extruded sample was barely foamed with only a few dispersed cells throughout the cross-section. The cell density was 1270 cells/ ${m m}^{3}$ , and the mean pore size was 7.5 μm. As the temperature was increased to 330 $℃$ , shown in Figure 5(b), more cells started to appear and the mean cell size slightly increased to 8 μm. The cell density increased to 4070 cells/ ${m m}^{3}$ . As the foaming temperature was increased to 360 $℃$ , which is the melting temperature of PEEK, as shown in Figure 5(c) more nucleation occurred and the cell density increased to 5490 cells/ ${m m}^{3}$ . Foamed cells were packed much closer with a mean pore size of 10.9 μm. With a higher foaming temperature, more nuclei were generated, and more gas molecules participated in the bubble growth process.

Figure 5.

Foaming result for as-extruded 20% HA samples at (a) 300 $℃$ , (b) 330 $℃$ , (c) 360 $℃$ .

Figure 6.

Foaming results of quenched samples with 10% HA (a, b, c), 20% HA (d, e, f), and 30% HA (g, h, i), each sample of the group foamed at 300 $℃$ , 330 $℃$ , and 360 $℃$ , respectively.

Table 1.

Mean pore sizes and cell densities of foamed PEEK/HA samples.

Foaming temp	300°C		330°C		360°C
Sample type	Pore size (μm)	Cell density (cells/mm³)	Pore size (μm)	Cell density (cells/mm³)	Pore size (μm)	Cell density (cells/mm³)
As-extruded 20%HA	7.5	1270	8.0	4070	10.9	5490
Quenched 10%HA	19.1	6280	28.2	9240	30.1	9670
Quenched 20%HA	22.2	6610	28.9	9730	32.9	9550
Quenched 30%HA	24.4	7660	28.3	10220	32.1	9760
Average of quenched	21.9	6850	28.5	9730	31.7	9660

Figure 7.

Pore size distributions for: (a) as-extruded 20% HA samples, (b) quenched 10% HA samples, (c) quenched 20% HA samples, and (d) quenched 30% HA samples.

With a lower crystallinity, the cell size and cell density of the quenched samples are significantly increased compared with those of as-extruded samples. The foam cells are densely packed across the entire cross section. However, when the foaming temperature is at 300 $℃$ , most of the cells are small, at around 5 μm, and only a few large cells on the level of tens of micrometers are formed, as shown in Figure 6(a), (d), and (g). This bimodal cell size distribution can also be seen in Figure 7(b)–(d). As the foaming temperature increases, the small cells tend to expand until the cells fill in the entire cross section. As the foaming temperature increased from 300 to 330 and 360 $℃$ , the average cell size increased from 21.9 to 28.5 and 31.7 μm, respectively. Increasing the HA content from 10% to 30% helps to increase the cell density at the low foaming temperature; however, the effect becomes negligible when the foaming temperature is high, as seen in Table 1. This suggests that heterogenous nucleation caused by adding HA particles only works when the foaming temperature is low. At high foaming temperatures, nucleation density is dominated by gas concentration.

In addition to foaming temperature, the effect of foaming time is studied by foaming the quenched 20% HA samples at 330 $℃$ for 2s, 4s, 8s, and 15s. The SEM images for different foaming times are shown in Figure 8. For the 2s sample, cells of 15–25 um can be observed in the outer regions of the sample, and the cells are gradually getting smaller going deeper into the sample. The center of the sample does not foam. This is because there is insufficient time to achieve uniform temperature distribution across the sample. This non-uniform temperature distribution can be used as a method to generate functionally graded porous structure. With foaming time increases, the cells expand gradually and grow toward the center of the sample; however, the cell size is nonuniform as the cells expand. When the foaming time reaches 8 seconds, the entire sample is foamed, with the average cell size increasing to 40 um. As foaming time continues to increase to 15 s, cells start to collapse, and larger cells start to show a polyhedral shape instead of a spherical one.

Figure 8.

Foaming result for 20% HA composite at 330 $℃$ for 2s, 4s, 8s, 15s.

Porosity and crystallinity of foamed Poly(ether-ether-ketone)/hydroxyapatite composites

The porosity of foamed PEEK/HA samples was estimated based on the sample volume expansion after foaming. Figure 9(a)–(c) show the results for samples foamed for 8 sec at different temperatures. The porosity of PEEK/HA composites foamed at 300 $℃$ are approximately 25%-30% for all HA loadings. When foaming at 330 $℃$ , the porosity increases to 65%–70%. When the foaming temperature is increased to 360 $℃$ , the porosity increases to 75%–80%. While the porosity also appears to increase as the HA loading increases at each foaming temperature, these increases are insignificant compared to the effect of foaming temperature. Temperature has a stronger effect than HA loading on the foaming result. Figure 9(d) shows that the porosity increases as the foaming time increases; however, there is little difference between the samples foamed at 8 s and 15 s. Samples in Figure 9(d) had 20% HA and were foamed at 330 $℃$ .

Figure 9.

Porosity of foamed samples with different foaming temperature and time.

DSC tests were conducted on foamed PEEK/HA composites to examine the crystallinity after solid state foaming. The results are shown in Figure 10. Although the samples were quenched to reduce the crystallinity before saturation and foaming, the cold crystallization peak that characterizes the amorphous PEEK disappeared after solid-state foaming. The final crystallinity of the samples after foaming recovered to 17% and 21% for 10% HA and 20% HA samples, respectively. A previous study has found that solid state foaming can induce crystallization of semicrystalline polylactic acid (PLA).¹⁸ The same phenomenon is observed for PEEK in this study.

Figure 10.

DSC results of PEEK/HA samples after foaming.

Mechanical properties of Poly(ether-ether-ketone)/hydroxyapatite composites

The mechanical properties of both non-foamed and foamed PEEK/HA samples were tested on a Instron 5980 mechanical tester. Due to the sample size limitation, a compression test was conducted on the 3 mm × 5 mm surface of prism samples. A constant strain rate of 0.5 mm/min and a preload of 10 N was used. The stress-strain curves for non-foamed PEEK/HA samples are shown in Figure 11, and Table 2 shows the compressive strength and elastic modulus results. Both the compressive strength and elastic modulus were improved with the addition of hydroxyapatite. The compressive strength increased by about 70% and the stiffness increased by 65% with the addition of 30% HA. Table 1 also compares non-foamed PEEK/HA samples with human cortical bones. It can be seen that the compression strength of non-foamed PEEK/HA can be easily adjusted to match that of cortical bones, while the stiffness is lower than that of human cortical bones.

Figure 11.

Compression test results of non-foamed samples.

Table 2.

Mechanical properties of non-foamed PEEK/HA composites.

Porosity%	Neat PEEK	10% HA	20% HA	30% HA	Cortical bone²³
Strength (MPa)	83.2	92.3	110.8	144.3	105
Stiffness (GPa)	1.01	1.32	1.42	1.67	13–15

For foamed PEEK/HA samples, the mechanical properties are greatly dependent on the porosity. The compressive test is done with 20% HA samples of different porosities after foaming, and the results are shown in Figure 12(a). Both the elastic modulus and the compressive strength reduced as the porosity increased. The compressive strength was 18.5 MPa for the 50%, 13.7 MPa for the 60%, and 7.5 MPa for the 70% porosity samples, respectively. To compare the results for different HA loadings, samples with a porosity around 70% were chosen. The compression test results for these foamed PEEK/HA samples are shown in Figure 12(b) and Table 3. It can be seen from Table 3, by adding 30% HA the compressive strength of foamed PEEK/HA samples could increase by up to 50% and the stiffness could be doubled, comparing with neat PEEK with the same porosity. The strength of foamed PEEK/HA can be adjusted to match that of human trabecular bones, which typically have a porosity of 75%–95%. The elastic modulus of the PEEK/HA sample, however, is still lower than that of human trabecular bones.

Figure 12.

Compression test result for foamed samples: (a) different porosity with 20% HA; (b) different HA content with 70% porosity.

Table 3.

Mechanical properties of foamed PEEK/HA composites.

Porosity = 70%	Foamed neat PEEK	Foamed 10%HA	Foamed 20%HA	Foamed 30%HA	Trabecular bone²⁴
Strength (MPa)	5.8	6.1	7.4	8.6	1 – 12
Stiffness (MPa)	25.1	37.4	42.7	51.1	80 – 800

The low stiffness of the foamed PEEK/HA samples could be due to two reasons. First, the stiffness of porous PEEK structure is generally in the range of 20-200 MPa,^11,21,25,26 depending on the porosity. The porous PEEK/HA structure reported in this work has a similar stiffness compared to those with the salt leaching method, which was about 50 MPa. Another possible reason is the low crystallinity in the PEEK/HA samples in this study. PEEK is a semicrystalline polymer, and its mechanical strength is dependent on crystallinity. Commercially available PEEK normally has a high crystallinity of 35%. However, the foamed PEEK/HA samples only had 21% crystallinity. Improving PEEK crystallinity will help increase the stiffness of the foamed PEEK/HA samples.

Conclusions

The solid-state foaming behaviors of PEEK and PEEK/HA composite are studied in this research. High porosity and large cell size are hard to achieve with as-extruded PEEK because of the semicrystalline nature of the polymer. Thus, a quenching process is applied to reduce the crystallinity of PEEK and to achieve better foaming results. Foaming temperature is found to be the most influential factor on the cell size and porosity of the PEEK/HA foams. Addition of hydroxyapatite has little effect on the saturation and foaming behavior when compared with neat PEEK. Heterogenous gas bubble nucleation enhanced by adding HA particles only dominates in the foaming process when the foaming temperature is low. However, by adding the HA particles, the compressive strength and elastic modulus of PEEK can be greatly improved for both foamed and non-foamed samples. After foaming, the crystallinity of PEEK recovers to the same level as before quenching. The compressive strength of the foamed PEEK/HA is similar to that of human trabecular bones, and the stiffness is slightly lower. This study developed a method to achieve high porosity PEEK/HA foams using the solid-state foaming method. It paves the way for further research on using porous PEEK for medical implant applications.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the John T. MacGuire Endowment, Walker Department of Mechanical Engineering, The University of Texas at Austin.

ORCID iD

Wei Li

References

Sun

, et al. Bioinspired modifications of PEEK implants for bone tissue engineering. Front Bioeng Biotechnol 2021; 8: 631616.

Fage

Muris

Jakobsen

, et al. Titanium: a review on exposure, release, penetration, allergy, epidemiology, and clinical reactivity. Contact Dermatitis 2016; 74(6): 323–345.

Walsh

Bertollo

Christou

, et al. Plasma-sprayed titanium coating to polyetheretherketone improves the bone-implant interface. Spine J 2015; 15(5): 1041–1049.

Evans

Torstrick

Lee

, et al. High-strength, surface-porous polyether-ether-ketone for load-bearing orthopedic implants. Acta Biomater 2015; 13: 159–167.

Han

Sharma

, et al. An in vitro study of osteoblast response on fused-filament fabrication 3D printed PEEK for dental and cranio-maxillofacial implants. J Clin Med 2019; 8(6): 771.

Liu

Wei

, et al. Nano-TiO2/PEEK bioactive composite as a bone substitute material: in vitro and in vivo studies. Int J Nanomed 2012; 7: 1215.

Wang

Yang

, et al. Strontium/adiponectin co-decoration modulates the osteogenic activity of nano-morphologic polyetheretherketone implant. Colloids Surf B Biointerfaces 2019; 176: 38–46.

Deng

Zhou

Liu

, et al. Preparation, characterization, cellular response and in vivo osseointegration of polyetheretherketone/nano-hydroxyapatite/carbon fiber ternary biocomposite. Colloids Surf B Biointerfaces 2015; 136: 64–73.

Torstrick

Lin

Potter

, et al. Porous PEEK improves the bone-implant interface compared to plasma-sprayed titanium coating on PEEK. Biomaterials 2018; 185: 106–116.

10.

Feng

Liang

, et al. Osteointegration of 3D-printed fully porous polyetheretherketone scaffolds with different pore sizes. ACS Omega 2020; 5(41): 26655–26666.

11.

Zheng

Zhao

Ouyang

, et al. Additively-manufactured PEEK/HA porous scaffolds with excellent osteogenesis for bone tissue repairing. Compos B Eng 2022; 232: 109508.

12.

Torstrick

Lin

Safranski

, et al. Effects of surface topography and chemistry on polyether-ether-ketone (PEEK) and titanium osseointegration. Spine 2020; 45(8): E417–E424.

13.

Uddin

Dhanasekaran

Asmatulu

. Mechanical properties of highly porous PEEK bionanocomposites incorporated with carbon and hydroxyapatite nanoparticles for scaffold applications. Progress in biomaterials 2019; 8(3): 211–221.

14.

Zhou

, et al. Fabrication of tissue engineering scaffolds through solid-state foaming of immiscible polymer blends. Biofabrication 2011; 3(4): 045003.

15.

Zhang

Zhou

Sun

, et al. Easy route to control density of polyimide foams and impact on mechanical and combustion behaviors. J Cell Plast 2021; 57(4): 451–470.

16.

Ding

, et al. Fabrication of rigid polyimide foams with overall enhancement of thermal and mechanical properties. J Cell Plast 2021; 57(5): 717–731.

17.

Zhou

Wang

. Fabrication of functionally graded porous polymer via supercritical CO2 foaming. Compos B Eng 2011; 42(2): 318–325.

18.

Wang

Kumar

. Development of crystallization in PLA during solid-state foaming process using sub-critical CO2. Cell Polym 2012; 31(1): 1–18.

19.

Yang

Zhang

, et al. Effects of processing parameters and thermal history on microcellular foaming behaviors of PEEK using supercritical CO2. J Appl Polym Sci 2015; 132(39): 132–132.

20.

Zhou

Jiang

, et al. Microextrusion Foaming of Polyetheretherketone Fiber: Mechanical and Thermal Insulation Properties. Adv Eng Mater 2022; 24(3): 2101110.

21.

Liang

Jiang

, et al. Preparation of Glass Fiber/Poly (Ether Ether Ketone) Composite Foam with Improved Compressive Strength and Heat Resistance. Adv Eng Mater 2022; 24(3): 2100940.

22.

Tierney

Gillespie

Jr . Crystallization kinetics behavior of PEEK based composites exposed to high heating and cooling rates. Compos Appl Sci Manuf 2004; 35(5): 547–558.

23.

Morgan

Unnikrisnan

Hussein

. Bone mechanical properties in healthy and diseased states. Annu Rev Biomed Eng 2018; 20: 119–143.

24.

Prakasam

Chirazi

Pyka

, et al. Fabrication and multiscale structural properties of interconnected porous biomaterial for tissue engineering by freeze isostatic pressure (FIP). J Funct Biomater 2018; 9(3): 51.

25.

Siddiq

Kennedy

. Porous poly-ether ether ketone (PEEK) manufactured by a novel powder route using near-spherical salt bead porogens: Characterisation and mechanical properties. Mater Sci Eng C 2015; 47: 180–188.

26.

Uddin

Dhanasekaran

Asmatulu

. Mechanical properties of highly porous PEEK bionanocomposites incorporated with carbon and hydroxyapatite nanoparticles for scaffold applications. Progress in biomaterials 2019; 8: 211–221.