montmorillonite nanocomposite prepared by powder sintering

montmorillonite nanocomposite prepared by powder sintering

Results in Physics 15 (2019) 102540 Contents lists available at ScienceDirect Results in Physics journal homepage: www.elsevier.com/locate/rinp Phy...

10MB Sizes 0 Downloads 8 Views

Results in Physics 15 (2019) 102540

Contents lists available at ScienceDirect

Results in Physics journal homepage: www.elsevier.com/locate/rinp

Physicochemical properties of hydroxyapatite/montmorillonite nanocomposite prepared by powder sintering

T

Rosnah Nawanga, Mohd Zobir Husseina, , Khamirul Amin Matoria,b, Che Azurahanim Che Abdullaha,b, Mansor Hashima ⁎

a b

Materials Synthesis and Characterisation Laboratory (MSCL), Institute of Advanced Technology, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia Faculty of Science, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia

ARTICLE INFO

ABSTRACT

Keywords: Mechanical properties Porosity Hydroxyapatite Montmorillonite

This study investigated the effect of the addition of montmorillonite into a hydroxyapatite nanocomposite for biomedical application. Hydroxyapatite (HA)/montmorillonite (MMT) nanocomposite was prepared using a powder sintering technique at 800 °C for 2 h. The specific surface area and pore volume were found to decrease linearly with the addition of MMT. The addition of 10–20% MMT led to a more homogeneous pore size distribution, which resulted in an increase in flexural strength by 18.9–17.1% and an increase in compressive strength by 107.9–63.1%. However, further addition of 30–60% MMT led to a less homogeneous pore size distribution, resulting in a decrease in flexural and compressive strength. The homogeneity of the pore size distribution was found to offer significant control over the strength of the nanocomposites. The addition of MMT resulted in the presence of an anhydrite phase, and this phase is useful to enhance the bioactivity of the nanocomposite. This study shows that the addition of MMT to HA for the formation of HA/MMT nanocomposite has a beneficial effect and has the potential to be used as a biomaterial, especially for non-load bearing sites in bone substitutions.

Introduction Hydroxyapatite (HA) has the chemical formula, Ca10(PO4)6(OH)2. HA is the same type of mineral that is found in bone and teeth. Having a structure similar to the mineral in bone, HA is very biocompatible with the human body. HA is osteoconductive, stable towards bioresorption, bioactive, and has no adverse effect on the human body [1]. Owing to these special characteristics, HA has attracted many scientists to conduct their research on the use of HA as a bone substitute [2–4]. HA has been applied successfully for several years in dentistry and orthopaedics due to its biocompatibility, and it forms a strong bond with the surrounding tissue when implanted [5,6] in the human body. However, HA has low mechanical properties (fracture toughness < 1 MPa m1/2) and therefore, the application of HA as a bone defect filler has been limited to non-load bearing sites [7]. Montmorillonite (MMT) is a member of the smectite clay group, with the chemical formula, Al2Si4O10(OH)2·nH2O. MMT is composed of two tetrahedral sheets of silica that sandwich a central octahedral sheet of alumina. The MMT layer has a thickness of 1 nm, and the layers are stacked on each other and loosely bound by van der Waals forces [8]. MMT is biocompatible with the human body and is available in ⁎

abundance at a relatively low cost. MMT finds a wide range of application in our lives such as in civil engineering [9], as a catalyst [10], in the paper industry [11], in the oil-drilling, paint and textile industries [12], as a cleaning and detergent agent [10], in sugar and oil refining [9], in foodstuffs and beverages [10], in cosmetics [13] and in pharmaceuticals [12]. MMT is also used as a reinforcing agent to increase the strength of a polymer/HA composite for a bone replacement application. The addition of MMT even in a low amount has been reported to enhance the mechanical properties of composite scaffolds [14]. Katti et al. [15] studied the effect of MMT addition on the properties of a chitosan/HA/ MMT composite. The composite was prepared in aqueous media followed by a drying process in an oven to obtain a thin sheet. They found that the addition of MMT had increased the elastic modulus and the hardness of the composite. In different studies, other researchers reported an improvement in tensile strength with the incorporation of MMT into chitosan/MMT/HA-ZrO2 [16] and chitosan/MMT/HA-ZnO composites [17]. A study by Olad and Farshi [18] found that the addition of MMT into a chitosan/gelatin/HA composite had increased the tensile and compressive strength as well as the in vitro bioactivity of the composite. Balakrishnan et al. [19] prepared a high-density

Corresponding author. E-mail address: [email protected] (M.Z. Hussein).

https://doi.org/10.1016/j.rinp.2019.102540 Received 30 January 2019; Received in revised form 20 July 2019; Accepted 26 July 2019 Available online 31 July 2019 2211-3797/ © 2019 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/BY-NC-ND/4.0/).

Results in Physics 15 (2019) 102540

R. Nawang, et al.

Fig. 2. XRD patterns of sintered samples: (a) HA, (b) MMT, (c) HA:MMT (90:10), (d) HA:MMT (80:20), (e) HA:MMT (70:30), (f) HA:MMT (60:40), (g) HA:MMT (50:50), (h) HA:MMT (40:60). The ratio is indicated in the bracket. HA (■), MMT (♦), Muscovite (●), Quartz (♢), Whitlockite (+), Anhydrite (○).

Pores are introduced through the fabrication of ceramics. During the compaction of starting powder material into the desired shape, pores or void spaces will exist between the powder particles. When the ceramics are subjected to heat treatment, most of the pores will be eliminated due to the expulsion of trapped gases and moisture and the formation of oxides. However, the elimination process is frequently incomplete, and some pores will still remain. During the early stage of sintering, bonds will grow between the contacting particles and will reduce the specific surface area. The bonding will substitute surface area with lower energy grain boundary area. As the sintering process continues, the surface area will decrease, and the density will increase. Grain boundaries will form in the contacts. As the sintering process goes on, the grain size will increase with densification, and the surface area will continue to decrease. In the final stage of sintering, grain growth will remove the grain boundary area and as densification continues, the mechanical strength will improve [26]. Understanding the role of pore and porosity, the way of controlling them and how they affect the strength of ceramics is vital, which could help to produce ceramics with good strength. Many previous studies have been conducted to study the pore and porosity and their relationship with strength. The particle size of the starting powder has a great effect on the porosity, where decreasing the particle size will decrease the porosity. Furthermore, by tuning the pore size distribution from bimodal to monomodal, we could increase the strength [27]. A study on the effect of powder compaction and sintering temperature on the porosity of hydroxyapatite found that it was possible to produce ceramics with various porosities by applying various pressures during the compaction process of the powder. The higher the pressure applied, the less pore volume will be obtained. For a given pressure applied, pore volume and pore size distribution are dependent largely on the particle size distribution of the starting powder. For the sintered samples, it was found that as the sintering temperature increased, the distribution of the pores became more uniform, narrower in distribution and larger in size [28]. In this work, we prepared HA/MMT nanocomposite by a powder sintering technique to study the effect of MMT incorporation on the physicochemical properties of the nanocomposite. To our knowledge, preparation of the HA/MMT composite by a powder sintering technique has not yet been widely reported. This present paper served as our preliminary study and we only focus on the sintering temperature of 800˚C. From the literature search, we found that there was a study

Fig. 1. (a) Particle size distribution of starting materials; HA and MMT powders. (b): XRD patterns of (i) unsintered HA; (ii) sintered HA; (iii) unsintered MMT; (iv) sintered MMT. HA (■), MMT (♦), Alunogen (△), Muscovite (●), Quartz (♢), Whitlockite (+).

polyethylene (HDPE)/HA/MMT composite by the melt-compounding method and found that the tensile and flexural strength improved with the incorporation of MMT. Another study by Abdul Haq et al. [20] also reported the increase of flexural strength in the PCL/HA/MMT composite with the addition of MMT. From the literature, many studies have reported on the effect of MMT addition on the properties of polymer/HA. There are very limited reports on the effect of MMT addition on HA alone [21]. Therefore, in this paper, we report the effect of MMT addition on the properties of HA. The nanocomposite was prepared by the powder sintering technique at various HA to MMT ratios. The effect of MMT addition on the physicochemical properties of the nanocomposite was studied. Since the mechanical properties of ceramics are strongly affected by the porosity, the effect of MMT addition on the porosity, specific surface area, pore volume and pore size distribution and their relationship to the mechanical properties was also studied. Mechanical properties of ceramics are very dependent on their porosity, pore size, pore size distribution, total pore volume and specific surface area. Ceramics with high porosity [22], large pore size and inhomogeneous pore size distribution [23], large specific surface area and high total pore volume [24] will have low strength. Pores reduce the strength of ceramics in two ways: pores reduce the cross-sectional area across which a load is applied, and they act as stress concentration [25]. Therefore, reducing the porosity, pore size, specific surface area and total pore volume will help to improve the strength of ceramics. 2

Results in Physics 15 (2019) 102540

R. Nawang, et al.

Fig. 3. FTIR spectra of (a) unsintered MMT, (b) sintered MMT, (c) unsintered HA, (d) sintered HA.

reported on a compressive strength value of about 0.15 MPa for the hydroxyapatite prepared from bones which was sintered at 1100 °C for about 2.5 h [29]. Another study reported a hardness value of 17 Hv for hydroxyapatite prepared from fish bone which was sintered at 1000 °C for 2 h [30]. Another study by Thangamani et al. [31] reported a flexural strength value of about 15 MPa for the hydroxyapatite sintered at 1100 °C for 3 h. A study by Irza Sukmana et al. [32] reported a hardness value of 14.8 Hv for hydroxyapatite sintered at 1000 °C for 2 h. The values reported by these previous studies were comparable with our HA flexural strength (11.1 MPa) and hardness (37.2 Hv) value, that can already be obtained with a lower sintering temperature of 800 °C for 2 h soaking time. In fact, our compressive strength is much higher (176 MPa) compared to the value obtained by one of the researchers stated earlier [29].

Microstructural analysis of the specimen was performed with a field emission scanning electron microscope (FESEM), FEI Nova Nanosem 230 (Netherlands). The relative density of the samples was determined using the Eq. (1) below:

Relative density =

a/ s

× 100

(1)

whereas, the porosity of the samples was calculated using Eq. (2), below [33]:

(%) = (1

a/ s)

× 100

(2)

where ф is the percentage porosity, ρa is the apparent density which is calculated from the outer dimension of the samples, and ρs is the skeletal density of the sample that was obtained using a helium gas pycnometer (Metromeritics, AccuPyc 1330, USA). Flexural strength was measured using the three-point bending test. The testing was carried out on rectangular samples using a universal testing machine (Instron 5566, USA) with 10 kN load and a crosshead speed of 0.5 mm/min. A minimum of three replicates of the samples was used for this test. The flexural strength of the samples was calculated using Eq. (3).

Materials and methods Hydroxyapatite was purchased from Acros Organic (Germany), and montmorillonite KSF clay was obtained from Aldrich (USA). The particle size of the starting powders was measured using a dynamic light scattering (DLS) technique on a Malvern Zetasizer Nano-S (UK). The nanocomposites were prepared by mixing HA powders with MMT in the ratios of HA:MMT of 90:10, 80:20, 70:30, 60:40, 50:50 and 40:60 (wt %) on a rolling mill at 50 rpm for 26 h. The nanocomposite powders were then pressed into a rectangular shape of (45 × 3 × 4) mm and cylindrical shape of (6 × 10) mm size samples. The samples were then sintered at 800 °C for 2 h in an electrical furnace (Elite, UK). X-Ray Diffraction (XRD) analysis of the powder samples was carried out on a Philips Expert Pro PW3040 (Netherlands) instrument with the 2θ range of 4–55°. Fourier transform infrared spectroscopy (FTIR) analysis was performed on a Thermo Nicolet Nexus (UK) using the Attenuated Total Reflectance (ATR) technique with the resolution of 4 cm−1 in the range of 400–4000 cm−1. Thermogravimetric analysis (TGA) of the samples was performed on a Mettler Toledo TGA/DSC 1HT Stare System (Switzerland) from 30 to 1300 °C with the scanning rate of 10 °C/min. The Brunauer-Emmett-Teller (BET) specific surface area analysis for the premixed powders and ground sintered samples was done by adsorption-desorption of nitrogen gas at liquid nitrogen temperature (77 K) using a BELSORP-mini (Japan). Before the analysis, the samples were degassed at 120 °C in a vacuum environment.

f

= 3FL /2bd 2

(3)

where σf is the flexural strength, F is the fracture load, L is the length of the support span, b and d are the width and the thickness of the sample, respectively. Compression testing was carried out on cylindrical-shaped samples using a universal testing machine (Instron 5566, USA) with 10 kN load and a crosshead speed of 0.5 mm/min. A minimum of three replicates of samples was used for this test. The compressive strength of the samples was calculated using Eq. (4). cs

= F /A

(4)

where F is the fracture load, and A is the cross-sectional area. Fracture toughness of the sample was measured using the indentation method. Microhardness and fracture toughness tests were performed on a Wilson Wolpert Micro-Vickers Hardness Tester 401MVD (USA), using a 200 gf load with a dwell time of 10 s. Five indentations were made for each sample. The crack length was measured using an optical microscope after indentation. Fracture toughness (KIC) was 3

Results in Physics 15 (2019) 102540

R. Nawang, et al.

Table 1 Assignment of FTIR bands for HA, MMT and their nanocomposites obtained by sintering at 800 °C for 2 h. Bands in MMT (cm−1) 3621 3001 2511 1627 993 912 795 747 693 671 592 511 416 408

Bands in HA (cm−1)

Assignment O-H stretching O-H stretching of water in alunogen O-H stretching of water in alunogen H-O-H bending Si-O stretching in Si-O-Si Al-Al-OH bending Si-O stretching of quartz Si-O stretching of quartz Al-Al-OH bending Si-O deformation υ4 SO4 in alunogen Al-O-Si bending out of plane OH vibration of muscovite Al-O-Al vibrational mode of muscovite

Assignment −

3571 3392 1654 1415 1089 1019 963 628 600 559 476

υ(OH ) H-O-H H-O-H CO32− υ3(PO43−) υ3(PO43−) υ1(PO43−) OH υ2(PO43−) υ2(PO43−) PO43−

Bands in the Nanocomposites (cm−1)

Assignment

1030 676 594 547

υ1(SO4) υ4(SO4) υ4(SO4) Whitlockite

Fig. 4. FTIR spectra of (a) unsintered HA and sintered samples, (b) HA, (c) HA:MMT (90:10), (d) HA:MMT (80:20), (e) HA:MMT (70:30), (f) HA:MMT (60:40), (g) HA:MMT (50:50), and (h) HA:MMT (40:60). The ratio is indicated in the brackets.

calculated using the equation given by Niihara as given below [34]:

KIC = 0.067(E / Hv )0.4Hv

using FESEM.

(5)

0.5 (c / a) 1.5

Results and discussion

1/2

where KIC is the fracture toughness (MPa. m ), Hv is the Vicker hardness (MPa), E is the young modulus (MPa), c is the average length of the cracks obtained in the tips of the Vickers marks (micron) and a is half average length of the diagonal of the Vickers marks (microns). A preliminary study on bioactivity behaviour of the samples was performed by soaking the cylindrical-shaped samples in phosphate buffer saline (PBS) solution for 14 days in a water bath (Memmert, Germany) at 37 °C. Then the samples were rinsed carefully with deionized water and dried in an oven (Memmert, Germany) at 37 °C until constant weight. The changes of samples morphology were analysed

X-Ray diffraction The average particle size of HA and MMT starting powder is about 11.3 nm and 11.1 nm, respectively as given in Fig. 1a. Fig. 1b shows the X-ray diffraction (XRD) patterns for unsintered and sintered HA and MMT. The unsintered HA diffractogram shows only the HA phase existed in the sample. On sintering at 800 °C, the HA phase had increased in crystallinity. However, very small peaks at 2θ = 31.1 and 2θ = 34.5 that were assigned to the whitlockite phase had been observed, 4

Results in Physics 15 (2019) 102540

R. Nawang, et al.

Fig. 5. TGA/DTG thermograms of unsintered samples (a) HA, (b) MMT, (c) HA:MMT (90:10), (d) HA:MMT (80:20), (e) HA:MMT (70:30), (f) HA:MMT (60:40), (g) HA:MMT (50:50), (h) HA:MMT (40:60). The ratio is indicated in the brackets. 5

Results in Physics 15 (2019) 102540

R. Nawang, et al.

Fig. 6. TGA/DTG thermograms of sintered samples (a) HA, (b) MMT, (c) HA:MMT (90:10), (d) HA:MMT(80:20). The ratio is indicated in the brackets.

indicating that a small amount of HA had transformed to whitlockite during the heating. The XRD pattern of the unsintered MMT demonstrated the presence of MMT, muscovite, quartz and alunogen (aluminium sulphate) phases in the sample. On sintering of the MMT at 800 °C, the diffractogram revealed that the alunogen phase had disappeared while the other phases still remained. The missing of the alunogen phase indicated that it was decomposed when the sample was sintered at 800 °C or totally transformed into other phases. Fig. 2 depicts the XRD patterns of the nanocomposites prepared at various HA to MMT ratios sintered at 800 °C. The addition of MMT contributed to the presence of new peaks that were assigned to the existence of whitlockite, a type of calcium phosphate and anhydrite (CaSO4) phase in the resulting nanocomposite. Whitlockite formation is possible due to the presence of trace amounts of magnesium and iron in the starting materials, HA or MMT, or both. However, anhydrite formation is also possible due to the reaction of Ca from HA and SO42− from alunogen during the heating process. The figure shows that with the 10% MMT addition, the small whitlockite peak starts to appear more clearly at 2θ = 31.1 and 2θ = 34.8. A small anhydrite peak starts to appear at 2θ = 25.5, whereas the intensity peak for HA at 2θ = 34.1 and at 2θ = 31.8 starts to decrease. With the addition of 40% MMT, the whitlockite peaks at 2θ = 31.2 and 2θ = 34.5 become more prominent. A more intense peak for anhydrite at 2θ = 25.5 can be observed clearly, whereas the peak for HA at 2θ = 34.1 has disappeared, and the peak for HA at 2θ = 31.8 has decreased significantly. The increase in MMT content leads to an increase in the intensity peak for whitlockite and anhydrite phases, leading to an increase in the amount of whitlockite and anhydrite content in the nanocomposite. The existence of anhydrite can produce a beneficial effect to enhance the bioactivity of the nanocomposites. Anhydrite has been widely used as bone defect filler [35],

and it was first used at the end of the 19th century. Anhydrite has the ability to fully resorb when placed in the body and can facilitate the formation of new bone in the diseased tissues of human tuberculosis sufferers [36]. FTIR Fig. 3 shows the FTIR spectra of unsintered and sintered HA and MMT, and Table 1 lists the assignment of the bands. Fig. 4 shows the FTIR spectra of unsintered HA, sintered HA and the nanocomposites. For the FTIR spectrum of the MMT (Fig. 3a), the band at 3621 cm−1 is due to the eOH functional group stretching vibration [37,38]. A band at 1627 cm−1 is due to the H-O-H bending of the absorbed water [8]. Another band at 993 cm−1 is assigned to Si-O stretching of the Si-O-Si group in clay minerals [8], the band at 671 cm−1 is assigned to Si-O deformation [39], bands at 912 cm−1 [37,40] and 693 cm−1 [40] are assigned to Al-Al-OH bending vibrations, while another band at 511 cm−1 is assigned to Si-O-Al bending [37]. A broad band in the range of 3500–2500 cm−1 that was concentrated at 3001 cm−1 and a band at 2511 cm−1 reflected the OH stretching mode of the absorbed water in aluminium sulphate (alunogen), while a band at 592 cm−1 was attributed to υ4 SO4 in aluminium sulphate (alunogen) [41]. Bands at 795 cm−1 [42] and 747 cm−1 [43] were assigned to the Si-O stretching of quartz. A band at 416 cm−1 was assigned to out of plane OH vibrations of muscovite [44], and a band at 408 cm−1 was assigned to AlO-Al vibrational mode of muscovite [45]. The FTIR spectrum of sintered MMT (Fig. 3b) shows that a band at 3621 cm−1 had disappeared, which indicated the loss of the eOH group in the MMT structure due to dehydroxylation process during the heating. Bands at 3001 cm−1 and 2529 cm−1 also disappeared due to

6

Results in Physics 15 (2019) 102540

R. Nawang, et al.

the loss of absorbed water in MMT. A band at 993 cm−1 has been shifted to 1028 cm−1, while a band at 795 cm−1 was shifted to 791 cm−1, and another band at 511 cm−1 was shifted to 539 cm−1 while a band at 416 cm−1 disappeared. These are as a result of dehydroxylation of muscovite. The FTIR spectrum of HA (Fig. 3c) shows bands at 3392 cm−1 and 1654 cm−1 due to the absorbed water in HA [47]. A band at 1415 cm−1 is attributed to the CO32− group [48]. Other bands at 3571 cm−1 [47] and 628 cm−1 [39,41] are due to apatitic OH− in the HA matrix. Bands at 1089 cm−1 and 1019 cm−1 are assigned to υ3 (PO43−) [49], while a band at 963 cm−1 is attributed to υ1(PO43−) [39–41]. Bands at 600 cm−1 and 559 cm−1 are due to υ2(PO43−) [39,41], and 476 cm−1 is assigned to PO43− groups in HA [39,40]. The FTIR spectrum of sintered HA (Fig. 3d) shows the disappearance of bands at 3392 cm−1 and 1654 cm−1, indicating the loss of absorbed water as a result of the heating process. The sintered HA bands (Fig. 4b) at 561 cm−1 and 599 cm−1 are less intense compared to the band for the unsintered HA (Fig. 4a). This difference may be attributed to the presence of the whitlockite phase in the sintered HA [50]. With the 10% addition of MMT (Fig. 4c), the intensity ratio of the 561 cm−1 band to the 599 cm−1 band is reduced, compared to intensity ratio of these peaks in HA (Fig. 4a) and sintered HA (Fig. 4b). A band at 1023 cm−1 which is attributed to υ1(SO4) from anhydrite [51] has appeared. This band may overlap with the band at 1019 cm−1 from υ3(PO43−) of HA. With a 30% addition of MMT (Fig. 4e), the intensity of the peak at 561 cm−1 is almost the same as the intensity of the peak at 599 cm−1, demonstrating the presence of a higher amount of the whitlockite phase [50]. A band at 678 cm−1 starts to appear at this MMT content, indicating the presence of υ4(SO4) from the anhydrite [51]. With the addition of 40% MMT (Fig. 4f), a band at 548 cm−1 assigned to whitlockite [50] was appeared, indicating that a higher amount of this phase existed in the sample. A band at 1030 cm−1 becomes broad, whereas a band at 963 cm−1 that was attributed to υ1(PO43−) and a band at 629 cm−1 that was assigned to apatitic OH− have disappeared. With the 50 and 60%, the addition of MMT (Fig. 4g and h), more intense bands at 1030, 676, 605 and 547 cm−1 were observed, indicating that more whitlockite and anhydrite phases are present in the sample. Therefore, the FTIR results indicate that the increase in the amount of MMT had contributed to the increase of whitlockite and anhydrite phases in the sample, which is in good agreement with the XRD results, as discussed earlier.

Table 2 TGA/DTG analysis of MMT, HA and their nanocomposites. Temperature Range (°C)

Tmax (°C)

% Weight Loss

Assignment

17.85

343–425

83 120 384

515–678

611

3.12

683–800

750

4.99

805–934

862

2.43

1132–1305

1230

1.32

Loss of absorbed water Loss of interlayer water Dehydration of absorbed water in alunogen Further dehydration of alunogen, dehydroxylation of MMT Decomposition of alunogen, dehydroxylation of MMT Dehydroxylation of MMT, Dehydroxylation of muscovite Solid phase structural decomposition & crystallization of mullite, spinel, cristobalite, cordierite and sillimanite

HA 31–240 240–352 385–566 706–966

71 293 455 890

2.63 0.52 0.42 0.68

966–1169 1169–1304

1075 1233

0.47 0.34

MMT 36–289

0.89

Loss of absorbed water Loss of water lattice – Decomposition of HA into whitlockite phase – –

The residue of Sintered and Nonsintered Samples of HA/MMT Nanocomposite Residue (%) HA:MMT

Unsintered

Sintered

100:0 90:10 80:20 70:30 60:40 50:50 40:60 0:100

94.79 92.05 91.02 87.54 87.24 82.73 80.25 67.22

96.70 95.60 95.07 – – – – 92.84

TGA Fig. 5 depicts the TGA/DTG thermograms of unsintered HA, MMT and the nanocomposites prepared at various HA to MMT ratios. Fig. 6 illustrates TGA/DTG thermograms of sintered samples, whereas Table 2 shows the analysis of TGA/DTG for the samples. Several steps of mass loss occurred in MMT, as listed in Table 2. The first step occurred at 36–289 °C with the loss of approximately 17.85%, which was due to the loss of absorbed water at 83 °C and the loss of interlayer water in MMT at 120 °C. At 343–425 °C, a minor mass loss occurred that was attributed to the dehydration of alunogen. Further loss was recorded at 515–678 °C, presumably due to further dehydration of alunogen and dehydroxylation of MMT. Another mass loss step was observed at 683–800 °C as a result of dehydroxylation of MMT and the decomposition of alunogen. These findings show a good agreement with the XRD results, demonstrating the loss of the alunogen peak when the nanocomposite was heated at 800 ˚C. A mass loss at 805–934 °C is due to dehydroxylation of MMT and muscovite. At 1132–1305 °C, the mass loss is due to solid phase structural decomposition and crystallization of mullite, spinel cristobalite, cordierite and sillimanite phases [46]. The final residue for MMT was 67.22% with a total loss of approximately 30.60%. HA is more stable to thermal degradation compared to MMT

Fig. 7. The effect of MMT content on the final residue of the unsintered samples obtained from TGA/DTG analysis under N2 gas with the scanning rate of 10 °C/ min from 30 to 1300 °C.

the loss of absorbed water in alunogen, while the disappearance of a band at 592 cm−1 indicated the decomposition of alunogen, which will be decomposed to other phases when it is heated at 800−850 °C [46]. These results are also in agreement with the XRD results, which demonstrated the loss of the alunogen peak when the nanocomposite was heated to 800 °C. A band at 1627 cm−1 also disappeared as a result of 7

Results in Physics 15 (2019) 102540

R. Nawang, et al.

Fig. 8. Nitrogen adsorption–desorption isotherms of unsintered samples: (a) HA, (b) MMT, (c) HA:MMT (90:10), (d) HA:MMT (80:20), (e) HA:MMT (70:30), (f) HA:MMT (60:40), (g) HA:MMT (50:50), (h) HA:MMT (40:60). The ratio is indicated in the brackets. 8

Results in Physics 15 (2019) 102540

R. Nawang, et al.

(caption on next page) 9

Results in Physics 15 (2019) 102540

R. Nawang, et al.

Fig. 9. Nitrogen adsorption-desorption isotherms of sintered samples: (a) HA, (b) MMT, (c) HA:MMT (90:10), (d) HA:MMT (80:20), (e) HA:MMT (70:30), (f) HA:MMT (60:40), (g) HA:MMT (50:50), (h) HA:MMT (40:60). The ratio is indicated in the bracket. Table 3 BET specific surface area, total pore volume, average pore diameter, relative density and porosity of the nanocomposite prepared at various HA to MMT ratios. Sample

HA HA:MMT HA:MMT HA:MMT HA:MMT HA:MMT HA:MMT MMT

(90:10) (80:20) (70:30) (60:40) (50:50) (40:60)

BET Specific Surface Area (m2g−1)

Total Pore Volume (cm3g−1)

Average Pore Diameter (nm)

Relative Density (%) (Eq. (1))

Porosity (%) (Eq. (2))

Unsintered

Sintered

Unsintered

Sintered

Unsintered

Sintered

Sintered

Sintered

60.9 56.9 54.8 43.5 41.4 34.0 40.9 5.4

12.1 10.4 10.0 9.7 8.4 6.5 5.9 14.4

0.53 0.52 0.44 0.39 0.34 0.22 0.25 0.04

0.24 0.24 0.21 0.17 0.15 0.14 0.09 0.11

35 36 32 36 33 26 25 32

79 94 84 71 72 84 60 31

31.46 48.76 48.78 53.73 50.76 47.81 50.93 –

68.54 51.24 51.22 46.19 49.24 52.19 49.07 –

increased, which means that the total weight loss increased with the increase of the MMT content (Table 2). Fig. 7 shows that the residue decreased linearly with the increase of MMT content. For the sintered samples, the final residue was found to be higher than the unsintered samples, which means that their total weight loss was lower than in the unsintered samples (Table 2). Specific surface area and porosity Figs. 8 and 9 show the nitrogen adsorption-desorption isotherms of unheated and heated samples, respectively. All the samples are dominated by the Type III isotherm, which is normally associated with nonporous or macroporous materials. The Type III isotherm indicates the characteristics of weak interaction between the adsorbate-adsorbent, the formation of multilayers and no monolayer formation. As demonstrated in Figs. 8 and 9, all the samples show a very low uptake until P/Po > 0.8, which reveals the poor interaction between adsorbate-adsorbent. When P/Po exceeds 0.8, the amount of adsorbed gas is increasing exponentially, implying the much stronger interactions of adsorbate-adsorbate. The unheated samples (Fig. 8) show more total adsorbed gas compared to the heated samples (Fig. 9). Among all the samples, HA and HA:MMT (90:10) show the highest amount of total adsorbed gas due to their high total pore volume (Table 3). However, MMT shows the lowest amount of total adsorbed gas due to its low total pore volume. For the nanocomposites, as the MMT content was increased, the amount of adsorbed gas was decreased, due to the decrease in the total pore volume. HA exhibited the Type III isotherm, which explains the formation of multilayers at the higher relative pressure. However, the formation of a monolayer at the initial stage cannot be ruled out (Fig. 8a). When HA was heated at 800 ˚C, the N2 adsorption capacity of HA was reduced due to the reduction in total pore volume, and the Type III isotherm remained. Similarly, the heated MMT sample maintained the Type III isotherm with N2 adsorption capacity increased slightly, as a result of the increased in total pore volume. The heated nanocomposites also show a reduction in the amount of the adsorbed gas due to the decrease in the total pore volume. Table 3 and Fig. 10 show BET specific surface area (SSA), BET total pore volume, BET average pore diameter and porosity of the samples. The unsintered samples had higher SSA and total pore volume compared to the sintered samples. However, their pore size diameter is lower compared to the sintered samples. HA has 60.9 m2g−1, 0.53 cm3g−1 and 35 nm of SSA, total pore volume and average pore diameter, respectively, compared to 5.4 m2g−1, 0.04 cm3g−1 and 36 nm,

Fig. 10. The effect of MMT addition on the BET specific surface area, BET total pore volume and BET average pore diameter of samples.

because it shows only minor mass loss with the total amount of approximately 5.06% (Fig. 5a). The mass loss recorded at 31–240 °C can be attributed to the loss of absorbed water. A minor loss at 240–352 °C results from the loss of water in the lattice while another mass loss occurs at 706–966 °C, possibly due to the decomposition of HA and the presence of some whitlockite phase in the sample. This finding is in agreement with the XRD and FTIR results, as discussed earlier. The final residue for HA was found to be high, 94.79%. As expected, for the nanocomposites, the final residue decreased as the MMT content

10

Results in Physics 15 (2019) 102540

R. Nawang, et al.

Fig. 11. Pore size distribution of unsintered samples: (a) HA, (b) MMT, (c) HA:MMT (90:10), (d) HA:MMT (80:20), (e) HA:MMT (70:30), (f) HA:MMT (60:40), (g) HA:MMT (50:50), (h) HA:MMT (40:60). The ratio is indicated in the brackets. 11

Results in Physics 15 (2019) 102540

R. Nawang, et al.

(caption on next page) 12

Results in Physics 15 (2019) 102540

R. Nawang, et al.

Fig. 12. Pore size distribution of sintered samples: (a) HA, (b) MMT, (c) HA:MMT (90:10), (d) HA:MMT (80:20), (e) HA:MMT (70:30), (f) HA:MMT (60:40), (g) HA:MMT (50:50), (h) HA:MMT (40:60). The ratio is indicated in the brackets.

respectively, for MMT, indicating that HA is more porous than MMT. The addition of MMT into HA has resulted in the reduction in SSA and pore volume of the resulting nanocomposites. Fig. 10a and b shows for SSA, the total pore volume generally decreased linearly with the increased of the MMT content. Such an effect is presumably due to the lower pore volume and SSA value of MMT. When MMT is added into HA, the low pore volume and SSA value of MMT will reduce the pore volume and SSA value of the HA/MMT nanocomposite, leading to a lower pore volume and SSA value of the resulting material. The higher amount of MMT in the sample will contribute to the lowering of the resulting pore volume and SSA value of the nanocomposites. Sintering at 800 °C drastically reduced the SSA of HA to 11 m2g−1. The total pore volume was reduced to 0.24 cm3g−1, but the average pore diameter was increased to 84 nm. The reduction of the specific area was due to the densification and grain growth during the sintering process [26] as seen from the FESEM study. The increase in pore size might be due to the indirect consequences of interparticle neck growth, which led to the surface rounding of pores [52]. For MMT, sintering caused the value of SSA and total pore volume to increase significantly, but the average pore diameter decreased only slightly. During the heating process, the interlayer water molecules in the MMT structure were removed, resulting in the collapse of the interlayer spacing in MMT (Fig. 1b-iii and 1b-iv). At the same time, the decomposition process of alunogen was taking placed, which produced SO2 [46] and contributed to the increase in SSA of the sintered MMT. For the nanocomposites, sintering also reduced the SSA value due to the decrease in total pore volume, but sintering increased the pore diameter due to the interparticle neck growth. The porosity (Table 3) of the samples also decreased with the addition of MMT. When MMT was added into HA, the MMT particles filled the pore spaces between the HA particles, and this would reduce the porosity and the accessible surface area. At the same time, the pore volume and pore size of the resulting nanocomposite would also be reduced. The higher amount of MMT in the sample will contribute to the lowering of the resulting SSA, pore volume, pore size and porosity value of the nanocomposites. The relative density (Table 3) of the samples increased with the increase in MMT content. Sintered HA had a relative density of about 31.46%. With the addition of MMT, the relative density had increased about 47.81–53.73%. The increase in relative density could be contributed by the decrease in SSA and pore volume of the nanocomposites due to the improvement in particle packing resulted from the addition of MMT into HA. Fig. 11 depicts pore size distribution (PSD) of the Barrett-JoynerHalenda (BJH) adsorption branch for unsintered samples. All the samples exhibited multimodal pore size distribution with the existence of mesopores and macropores. For HA, the PSD was concentrated at < 40 nm. There were multiple peaks in the range of 0–20 nm, the main peak in the range of 20–35 nm and two minor peaks in the range of 35–75 and 75–100 nm. MMT shows a broad multimodal peak with all the peaks showing almost the same intensity for all the pore size ranges. With the addition of 10% MMT, the peak in the range of 75 to 100 nm disappeared, and the intensity for the peaks in the range of 0–20 nm had increased. The packing of raw powder particles can affect the pore size distribution of the nanocomposites. When two different particle sizes are mixed together, the smaller particle size will fill the bigger unoccupied spaces among the bigger particle. If the amount of smaller particle is adequate, the number of large particles can reach the condition where they can touch each other and result in optimal packing. If

all the spaces are filled with the smaller particles, further addition of the smaller particles will force the larger particle apart, and this will lead to the existence of bigger pore sizes and the packing density cannot be further enhanced [35]. In this case, when 10% MMT is added into HA, the particles would first fill the bigger pore spaces between HA particles, resulting in more of the smaller pore sizes and a decreased amount of the bigger pore sizes, which is indicated by the disappearance of the peak in the range of 75–100 nm and the increase in the intensity of the peak in the range of 0–20 nm. With 20% MMT addition, the intensity of the peaks at the range of 0–40 nm had been reduced. Two minor peaks in the range of 40–60 and 60–85 nm were appeared, which might indicate that the packing density had decreased. Further addition of 30–60% MMT resulted in almost all the peaks in all the pore size range sharing the same intensity, which shows the inhomogeneous pore size distribution that might result from the decrease in the packing density. Fig. 12 demonstrates pore size distributions of the sintered samples. Sintered samples also show multimodal pore size distribution with the existence of mesopores and macropores. However, compared to unsintered samples, the sintered samples have more uniform peaks. Furthermore, the peak for sintered samples has moved slightly to the right, showing that the pore diameter of the sintered samples becomes bigger compared to the pore diameter of the unsintered samples, due to the intra-neck growth that led to the surface rounding of pores, as discussed earlier. For sintered HA, the distribution was more concentrated at > 20 nm with two main peaks at 20–40 nm and 40–75 nm. There were also two minor peaks in the range of < 20 nm and one minor peak in the range of 75–100 nm. For sintered MMT, the distribution was more concentrated at < 20 nm with higher intensity compared to MMT. For the nanocomposite prepared at HA:MMT of 90:10, two main peaks are observed in the range of 20–40 nm and 50–75 nm and two minor peaks in the range of < 20 nm. The nanocomposite prepared at HA:MMT of 80:20 also shows a similar pattern with HA:MMT prepared at 90:10 but with reduced intensity. The rest of the four nanocomposites (HA:MMT 70:30, HA:MMT 60:40, HA:MMT 50:50 and HA:MMT 40:60) show almost the same pattern with each other and had a broad distribution with all the peaks showing almost the same intensity for all the pore size ranges. When 10% MMT was added into HA, the pores with the size range of 75–100 nm were eliminated (Fig. 12a and c), leaving two dominant peaks at 25 nm and 55 nm, with higher intensity, as could be seen in Fig. 13c, indicating more homogeneous PSD, compared to the sintered HA. With the addition of 20% MMT, the intensity of the peak at the pore size range of 45–75 nm becomes less intense with broader distribution, which implies the less homogeneous PSD, compared to the composite prepared with 10% MMT. With further addition of 30–60% MMT, almost all the peaks for all the pore size range have the same intensity, which indicates the inhomogeneous nature of PSD. The packing of raw powder particles can affect the pore size distribution of the sintered samples. Homogeneous packing of raw powder particles will lead to homogeneous PSD that, in turn, will result in homogeneous PSD in the sintered samples [53], as shown in Figs. 11 and 12. As mentioned earlier, all the sintered samples demonstrated a multimodal pore size distribution. However, compared to the others, the nanocomposite prepared at 10% MMT demonstrated more homogeneous distribution, followed by the nanocomposite prepared at 20% MMT, and then the sintered HA, parallel with their mechanical trend that will be discussed in the following subsection.

13

Results in Physics 15 (2019) 102540

R. Nawang, et al.

Fig. 13. The effect of MMT content on the specific surface area (SSA), total pore volume, average pore diameter, relative density, porosity, flexural and compressive strength of the samples.

Fig. 14. The effect of specific surface area on the flexural (a) and compressive (b) strength of the HA/MMT nanocomposite prepared at various ratios of HA to MMT. Table 4 Hardness and fracture toughness of HA and the nanocomposites. Sample HA HA:MMT HA:MMT HA:MMT HA:MMT HA:MMT HA:MMT

(90:10) (80:20) (70:30) (60:40) (50:50) (40:60)

Hardness (gf/mm2)

Fracture toughness (MPa. m1/2)

37.2 ± 6.1 24.63 ± 0.59 21.23 ± 2.79 13.8 ± 3.33 13.98 ± 0.81 18.5 ± 0.45 –

0.08 ± 0.01 0.09 ± 0.01 – – – – –

Mechanical properties Fig. 13 shows the effect of MMT content on the flexural and compressive strength of the nanocomposites and their relationship to the SSA, pore volume, pore size and porosity. Generally, the addition of MMT has decreased the SSA, pore volume, pore size and porosity, and increased the relative density. As shown in Fig. 13, the addition of 10 and 20% MMT had increased the flexural strength by 18.9 and 17.1%, respectively. For the compressive strength, the addition of 10 and 20% MMT had improved the strength by 107.9 and 63.1%, respectively. The increase in strength is resulted from the decrease in porosity. Pores give a negative effect on strength by two ways [54]. They act as stress raiser, meaning that the stress concentration is high at these pores. If this stress 14

Results in Physics 15 (2019) 102540

R. Nawang, et al.

produce a higher flexural and compressive strength of the nanocomposite, with higher SSA, if the nanocomposite has a homogeneous PSD. Table 4 shows the hardness and fracture toughness for the sintered HA and the nanocomposites. Fig. 15 depicts the hardness of the sintered HA and the nanocomposites. The addition of MMT has decreased the hardness of the nanocomposites. MMT has a hardness value of 1–2 Mohs scale [57], which is lower than HA, therefore it was thought that the addition of MMT has lowered the hardness value of nanocomposites. No significant changes occurred with the addition of 10% MMT on the fracture toughness of the nanocomposites. The addition of 20–50% MMT caused the difficulties in observing the crack length. Therefore the crack length for the nanocomposites could not be measured and hence the fracture toughness of the nanocomposites could also not be measured. Surface morphology

Fig. 15. The effect of MMT addition on the hardness of the nanocomposites.

Fig. 16 shows the effect of MMT addition on the microstructure of HA/MMT nanocomposites sintered at 800 °C for 2 h. The microstructure of the nanocomposite at 10% MMT content (Fig. 16b) shows the most homogeneous distributions of the pores and matrices, which leads to a more homogeneous pore size distribution (Fig. 12c), hence resulting in the highest flexural and compressive strength (Fig. 13e and f). Similar properties were also observed for the nanocomposite prepared at 20% MMT (Fig. 16c). However, the nanocomposites with MMT content of 30–60% (Fig. 16d–g) show that the microstructure was inhomogeneous, which led to inhomogeneous pore size distribution and hence resulted in lower flexural and compressive strength (Fig. 13e and f). Therefore, the addition of MMT to the HA/MMT nanocomposite plays an important role in increasing the strength if the addition resulted in better homogenous pores.

together with the applied stress surpasses a certain limit, a crack will develop. The crack will propagate spontaneously and the material will fail. Pores also reduce the cross-sectional area at which the load is applied. In these nanocomposites, the decrease in porosity helped reduce the negative effect of pores and therefore improved their strength. Furthermore, the addition of 10–20% MMT gave relatively more homogeneous PSD (refer to Fig. 12c and d), and hence improved the strength. It is believed that homogeneous pore size distribution is useful to improve the strength of a porous ceramics [27]. With the addition of 30–60% MMT, the strength of the nanocomposites was not improved even though they had lower porosity compared to HA. This might be contributed by their relatively less homogeneous PSD as can be seen in Fig. 12e–h. It was reported that the broad multimodal pore size distribution had resulted in lower strength of ceramics [55]. Another factor that might contribute to the decrease in the strength of the nanocomposites is the chemical composition of the nanocomposite itself. As the XRD results show (Fig. 2), the addition of 30–60% MMT has led to a more prominent existence of whitlockite and anhydrite phases. The presence of these two phases in the nanocomposites was thought to give a less homogeneous PSD, therefore reducing the flexural and compressive strength of the nanocomposites. In addition, anhydrite phase itself has a low mechanical property [56], and this low mechanical property might affect the strength of the nanocomposites. Fig. 14 shows the relationship between flexural and compressive strength with SSA. For the flexural strength, the value increased drastically when SSA was increased. The flexural strength reached the maximum value when the SSA value was approximately at 9.5–10 m2g−1. Then, the strength was reduced when SSA was further increased. For compressive strength, the value was gradually increased when SSA was increased, until it reached the SSA value of approximately 9.5 m2g−1, where the compressive strength increased drastically and reached the maximum value when the SSA was approximately at 10 m2g−1. When SSA was further increased, the compressive strength was reduced. The optimum value of SSA (10 m2g−1) that gives the maximum flexural and compressive strength is correlated with the 10% MMT addition, which results in more homogeneous PSD. However, the second and third higher value is correlated to 20% MMT addition and the sintered HA, which shows the second and third homogeneous PSD, respectively. The lower SSA values (5.8–9 m2g−1) that give the lower strength are correlated with the addition of 60–30% MMT, giving the inhomogeneous PSD. From this finding, we can deduce that we might

Bioactivity Fig. 17 shows the morphology of the samples before and after they were immersed in PBS solution. There were no changes can be clearly observed on the surface of HA (Fig. 17b) and the nanocomposite prepared at 90:10 (Fig. 17d) after 14 days of immersion in PBS. However, changes were observed on the surface of the nanocomposite prepared at 80:20 (Fig. 17f), whereby an apatite layer formation was found to cover on the surface of the nanocomposite. The formation of an apatite layer on the nanocomposite prepared at 80:20 could be due to the existence of anhydrite phase in the nanocomposite which enhanced the bioactivity of the sample. Compared to hydroxyapatite, anhydrite can dissolve easily in water [58] releasing Ca2+ into the PBS solution. The released Ca2+ will react with phosphate ion in the PBS solution to form apatite compound and then precipitate on the surface of the sample to form an apatite layer. The feasibility of anhydrite to dissolve in water and release the Ca2+ ion will increase the capability of apatite forming in the nanocomposite 80:20 and resulting in an increase in the bioactivity. Conclusions This study shows that the hydroxyapatite/MMT nanocomposite can be prepared using a simple powder sintering technique. The addition of only 10–20% (w/w) of MMT clay to HA for the formation of the HA/ MMT nanocomposite had a beneficial effect in which the flexural and compressive strength could be enhanced by approximately 17–108%, due to the improvement of the homogeneity of the pore size distribution together with the formation of oxide in the resulting

15

Results in Physics 15 (2019) 102540

R. Nawang, et al.

(caption on next page) 16

Results in Physics 15 (2019) 102540

R. Nawang, et al.

Fig. 16. Microstructure of sintered (a) HA, (b) HA:MMT (90:10), (c) HA:MMT (80:20), (d) HA:MMT (70:30), (e) HA:MMT (60:40), (f) HA:MMT (50:50) and (g) HA:MMT (40:60). Insert shows FESEM micrograph of the sample at lower magnification, 50,000×.

Fig. 17. Surface morphology of the samples before and after immersion in PBS for 14 days. (a) HA before immersion, (b) HA after immersion, (c) HA:MMT 90:10 before immersion, (d) HA:MMT 90:10 after immersion, (e) HA:MMT 80:20 before immersion (f) HA:MMT 80:20 after immersion, showing the surface of the nanocomposite was covered by an apatite layer. Inset showing the surface of the nanocomposite covered by apatite layer, at a magnification of 2500×.

17

Results in Physics 15 (2019) 102540

R. Nawang, et al.

nanocomposite. However, further addition of MMT of 30–60% resulted in the strength reduction, presumably due to the inhomogeneity of the pore size distribution of the resulting nanocomposite, showing that the strength of HA can only be enhanced by the addition of a small amount of MMT (10–20%). The addition of MMT clay also had a positive effect in which it helped the formation of the anhydrite phase that could enhance the bioactivity of the nanocomposite. In this present paper, we only focus on sintering temperature of 800 °C. For our future work, we will focus on different sintering temperatures and compare their resulting physicochemical properties.

[15] Katti KS, Katti DR, Dash R. Synthesis and characterization of a novel chitosan/ montmorillonite/hydroxyapatite nanocomposite for bone tissue engineering. Biomed Mater 2008;3(3):034122. [16] Bhowmick A, Jana P, Pramanik N, Mitra T, Banerjee SL, Gnanamani A, et al. Multifunctional zirconium oxide doped chitosan based hybrid nanocomposites as bone tissue engineering materials. Carbohydr Polym 2016;151. [17] Bhowmick A, Banerjee SL, Pramanik N, Jana P, Mitra T, Gnanamani A, et al. Organically modified clay supported chitosan/hydroxyapatite-zinc oxide nanocomposites with enhanced mechanical and biological properties for the application in bone tissue engineering. Int J Biol Macromol 2018;106. [18] Olad A, Farshi Azhar F. The synergetic effect of bioactive ceramic and nanoclay on the properties of chitosan–gelatin/nanohydroxyapatite–montmorillonite scaffold for bone tissue engineering. Ceram Int 2014;40(7):10061–72. [19] Balakrishnan H, Husin MR, Wahit MU, Rafiq M, Kadir A. Preparation and characterization of organically modified montmorillonite-filled high density polyethylene/hydroxyapatite nanocomposites for biomedical applications. Polym Plast Technol Eng 2014;53:790–800. [20] Haq RHA, Wahab MS, Wahid MU. Fused Deposition Modeling of PCL/HA/MMT Biocompatible Polymer Nano-Composites. Singapore: Springer; 2016. p. 23–32. [21] Hesaraki S, Zamanian A, Hafezi M. Montmorillonite-added calcium phosphate bioceramic foams. Key Eng. Mater. I 2008;361–363:111–4. [22] Wagh AS, Singh JP, Poeppel RB. Dependence of ceramic fracture properties on porosity. J Mater Sci 1993;28(13):3589–93. [23] Yan W, Chen J, Li N, Han B, Wei Y. Lightweight cordierite-mullite refractories with low coefficients of thermal conductivity and high mechanical properties. Bull Mater Sci 2015;38(2):409–15. [24] Seuba J, Deville S, Guizard C, Stevenson AJ. Mechanical properties and failure behavior of unidirectional porous ceramics. Sci Reports, Nat Publ Gr 2016;6: (24326). [25] William D, Callister J, Rethwisch DG. Materials science and engineering. Eighth ed. John Wiley & Sons, Inc; 2010. [26] German RM. Sintering trajectories: description on how density, surface area, and grain size change. J Miner Met Mater Soc 2016;68(3):878–84. [27] Yan W, Li N, Tong J, Liu G, Xu J. Effect of particle size on the pore characterization and strength of porous cordierite-mullite ceramics prepared by a pore-forming insitu technique. Sci Sinter 2013;45(2):165–72. [28] Rootare HM, Craig RG. Characterization of the compaction and sintering of hydroxyapatite powders by mercury porosimetry. Powder Technol 1974;9(4):199–211. [29] Heidari F, Razavi M, Ghaedi M, Forooghi M, Tahriri M, Tayebi L. Investigation of mechanical properties of natural hydroxyapatite samples prepared by cold isostatic pressing method. J Alloy Compd 2017:1150–6. [30] Hariani PL, Said M, Salni. Effect of Sintering on the Mechanical Properties of Hydroxyapatite from Fish Bone (Pangasius Hypophthalmus). IOP Conference Series: Materials Science and Engineering. 2019. [31] Thangamani N, Chinnakali K, Gnanam FD. The effect of powder processing on densification, microstructure and mechanical properties of hydroxyapatite. Ceram Int 2002;28(4):355–62. [32] Sukmana I, Hendriyanto A, Savetlana S, Tarkono. The Effect of sintering parameter on the properties of hydroxyapatite from local limestone for bone implant application. AIP Conf. Proc. 1977;2018. [33] Engstrand Unosson J, Persson C, Engqvist H An. Evaluation of methods to determine the porosity of calcium phosphate cements. J Biomed Mater Res Part B Appl Biomater 2015;103(1):62–71. [34] Rocha-Rangel E. Fracture Toughness Determinations by Means of Indentation Fracture. InTech; 2011. p. 360. [35] Li K. 1.12 – Ceramic hollow fiber membranes and their applications. Compr Membr Sci Eng 2010;3:253–73. [36] Hughes EAB, Grover LM. Characterisation of a novel poly (Ether Ether Ketone)/ calcium sulphate composite for bone augmentation. Biomater Res 2017;21(1):7. [37] Chen Q, Zhu R, Ma L, Zhou Q, Zhu J, He H. Influence of interlayer species on the thermal characteristics of montmorillonite. Appl Clay Sci 2017;135:129–35. [38] Altinişik A, Seki Y, Yurdakoç K. Preparation and characterization of chitosan/KSF biocomposite film. Polym Compos 2009;30(8):1035–42. [39] Sharma S, Sarasan G. Influence of acid activation on natural calcium montmorillonite clay. IOSR J Appl Chem 2017;10(6):71–7. [40] Maina EW, Wanyika HJ, Gacanja AN. Instrumental characterization of montmorillonite Clay by FT-IR and XRD from J.K.U.A.T Farm, in the Republic of Kenya. Chem Mater Res 2015;7(10):43–9. [41] Kloprogge JT, Ruan H, Frost RL. Near-infrared spectroscopic study of basic aluminum sulfate and nitrate. J Mater Sci 2001;36(3):603–7. [42] Adikary SU, Ashokcline M, Nirojan K. Characterization of montmorillonite clay from naturally occurring clay deposits in murunkan area. Int Res Conf KDU 2015:163–9. [43] Thambavani DS, Kavitha B. Mineralogical Characterization of River Bed Soil from Tamilnadu by FT-IR, XRD and SEM/EDAX. Int J Adv Res 2014;2(2):656–9. [44] Sayin M, Reichenbach VON. Infrared spectra of muscovite as affected by chemical composition. Heating and particle size. Clay Miner 1978;13:241. [45] Singha M, Singh L. Vibrational spectroscopic study of muscovite and biotite layered phyllosilicates. Indian J Pure Appl Phys 2016;54(2):116–22. [46] Földvári M. Handbook of the thermogravimetric system of minerals and its use in geological practice. Budapest: Geological Institute of Hungary; 2011. p. 213. [47] Ramesh S, Aw KL, Tolouei R, Amiriyan M, Tan CY, Hamdi M, et al. Sintering properties of hydroxyapatite powders prepared using different methods. Ceram Int 2013;39(1):111–9. [48] Popa C, Albu M, Costescu A, Luculescu C, Trusca R, Antohe S. Structural characterization and optical properties of hydroxyapatite/collagen matrix. Rom Reports Phys 2016;68(3):1149–58. [49] Peon E, Fuentes G, Delgado JA, Morejon L, Almirall A, García R. Preparation and characterization of porous blocks of synthetic hydroxyapatite. Lat Am Appl Res

Acknowledgements We are very grateful to the Ministry of Education, Malaysia for financial support through Fundamental Research Grant Scheme [FRGS/ 1/2018/STG07/UPM/01/2; Grant Vot No. 5540162] and Universiti Putra Malaysia for supporting us by providing Dana Tautan Penyelidikan, Vot No. 9201202 and UPM/ITMA for RN study leave. This paper is also dedicated to the memory of our wonderful former colleague, Assoc. Prof. Dr Mansor Hashim, who passed away during the course of this study. Author contributions Rosnah Nawang, Mohd Zobir Hussein, Khamirul Amin Matori, Che Azurahanim Che Abdullah and Mansor Hashim conceived and designed the experiments; Rosnah Nawang performed the experiments; Rosnah Nawang and Mohd Zobir Hussein analysed the data; Rosnah Nawang, Mohd Zobir Hussein, Khamirul Amin Matori and Che Azurahanim Che Abdullah contributed reagents/materials/analysis tools; Rosnah Nawang wrote the paper. Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. References [1] Orlovskii VP, Komlev VS, Barinov SM. Hydroxyapatite and hydroxyapatite-based ceramics. Inorg Mater 2002;38(10):973–84. [2] Vuola J, Göransson H, Böhling T, Asko-Seljavaara S. Bone marrow induced osteogenesis in hydroxyapatite and calcium carbonate implants. Biomaterials 1996;17(18):1761–6. [3] Yamamoto T, Onga T, Marui T, Mizuno K. Use of hydroxyapatite to fill cavities after excision of benign bone tumours. Clinical results. J Bone Joint Surg Br 2000;82(8):1117–20. [4] Matsumine A, Myoui A, Kusuzaki K, Araki N, Seto M, Yoshikawa H, Uchida A. Calcium hydroxyapatite ceramic implants in bone tumour surgery. A long-term follow-up study. J Bone Jt Surg 2004;86(5):719–25. [5] Ohgushi H, Dohi Y, Tamai S, Tabata S. Osteogenic differentiation of marrow stromal stem cells in porous hydroxyapatite ceramics. J Biomed Mater Res 1993;27(11):1401–7. [6] Okumura M, Ohgushi H, Dohi Y, Katuda T, Tamai S, Koerten HK, et al. Osteoblastic phenotype expression on the surface of hydroxyapatite ceramics. J Biomed Mater Res 1997;37(1):122–9. [7] Tan CY, Ramesh S, Tolouei R, Sopyan I, Teng WD. Synthesis of high fracture toughness of hydroxyapatite bioceramics. Adv Mater Res 2011;264–265:1849–55. [8] Bertuoli PT, Piazza D, Scienza LC, Zattera AJ. Preparation and characterization of montmorillonite modified with 3-aminopropyltriethoxysilane. Appl Clay Sci 2014;87(January):46–51. [9] Hartwell JM. The diverse uses of montmorillonite. Clay Miner 1965;6(3):111–8. [10] Tyagi B, Chudasama CD, Jasra RV. Determination of structural modification in acid activated montmorillonite clay by FT-IR spectroscopy. Spectrochim Acta – Part A Mol Biomol Spectrosc 2006;64(2):273–8. [11] Seo YS, Nah C. Effects of montmorillonite on properties of methyl cellulose carvacrol. J Korea TAPPI 2006;38(1):34–44. [12] Obaje SO, Omada JI, Dambatta UA. Clays and their industrial applications: synoptic review. Int J Sci Technol 2013;3(5):264–70. [13] Pusch R. Flow and ductility of smectite clay for skin treatment. J Cosmet Dermatological Sci Appl 2014:67–72. [14] Kar S, Kaur T, Thirugnanam A. Microwave-assisted synthesis of porous chitosanmodified montmorillonite–hydroxyapatite composite scaffolds. Int J Biol Macromol 2016;82:628–36.

18

Results in Physics 15 (2019) 102540

R. Nawang, et al. 2004;34(4):225–8. [50] Piga G, Gonçalves D, Thompson TJU, Brunetti A, Malgosa A, Enzo S. Understanding the crystallinity indices behavior of burned bones and teeth by ATR-IR and XRD in the presence of bioapatite mixed with other phosphate and carbonate phases. Int J Spectrosc 2016;2016(March):1–9. [51] Bishhop Janice L, Lane MD, Darby Dyar M, King Sara J, Brown AJ, Swayze GA. Spectral properties of ca-sulfates- gypsum, bassanite, and anhydrite. Am Mineral 2014;99:2105–15. [52] Akash A, Mayo MJ. Pore growth during initial-stage sintering. J Am Ceram Soc 1999;82(11):2948–52. [53] Rahaman Mohamed N. Ceramic Processing. Second Edition CRC Press; 2017. [54] K., B. A.; Sharma, C. P. Mechanical Behaviour and Testing of Materials; PHI Learning Pvt. Ltd: Delhi, 2011.

[55] Dittmann J, Willenbacher N. Micro structural investigations and mechanical properties of macro porous ceramic materials from capillary suspensions. J Am Ceram Soc 2014;97(12):3787–92. [56] Chern Lin J-H, Hung S-H, Chen W-L, Chen C-K, Lee J-W, Lin R-M, et al. Properties of TTCP/DCPA/CSH cement immersed in Hanks’ solution. J Med Biol Eng 2012;32(3):201–4. [57] Uddin F. Montmorillonite: an introduction to properties and utilization. Current Topics in the Utilization of Clay in Industrial and Medical Applications References. IntechOpen; 2018. p. 1–23. [58] Ferguson J, Diefenbeck M, Mcnally M. Ceramic biocomposites as biodegradable antibiotic carriers in the treatment of bone infections. J Bone Jt Infect 2017;2(1):38–51.

19