Accepted Manuscript

Reinforcement of calcium phosphate cement with multi-walled carbonnanotubes and bovine serum albumin for injectable bone substituteapplications

Kean-Khoon Chew, Kah-Ling Low, Sharif Hussein Sharif Zein, DavidS. McPhail, Lutz-Christian Gerhardt, Judith A. Roether, Aldo R.Boccaccini

PII: S1751-6161(10)00161-XDOI: 10.1016/j.jmbbm.2010.10.013Reference: JMBBM 231

To appear in: Journal of the Mechanical Behavior ofBiomedical Materials

Received date: 11 August 2010Revised date: 28 October 2010Accepted date: 31 October 2010

Please cite this article as: Chew, K.-K., Low, K.-L., Sharif Zein, S.H., McPhail, D.S.,Gerhardt, L.-C., Roether, J.A., Boccaccini, A.R., Reinforcement of calcium phosphate cementwith multi-walled carbon nanotubes and bovine serum albumin for injectable bone substituteapplications. Journal of the Mechanical Behavior of Biomedical Materials (2010),doi:10.1016/j.jmbbm.2010.10.013

This is a PDF file of an unedited manuscript that has been accepted for publication. As aservice to our customers we are providing this early version of the manuscript. The manuscriptwill undergo copyediting, typesetting, and review of the resulting proof before it is published inits final form. Please note that during the production process errors may be discovered whichcould affect the content, and all legal disclaimers that apply to the journal pertain.

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

Research Highlight:

CPC composites comprise β-TCP and DCPA incorporated with BSA and MWCNTs.

The concomitant admixture of BSA and MWCNTs improved mechanical properties of

CPC.

MWCNTs-OH containing composite studied exhibited highest compressive strength.

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

1

Reinforcement of Calcium Phosphate Cement with Multi-walled Carbon 1

Nanotubes and Bovine Serum Albumin for Injectable Bone Substitute 2

Applications 3

4

5

Kean-Khoon Chew 1, Kah-Ling Low

1, Sharif Hussein Sharif Zein

1*, 6

David S. McPhail 2, Lutz-Christian Gerhardt

2 #, Judith A. Roether

2 ^, Aldo R. Boccaccini

2, 3 7

8

1 School of Chemical Engineering, Engineering Campus, Universiti Sains Malaysia, Seri 9

Ampangan 14300 Nibong Tebal, Seberang Perai Selatan, Pulau Pinang, Malaysia 10

2 Department of Materials, Imperial College London, Prince Consort Road, London SW7 2BP, 11

UK 12

3 Institute of Biomaterials, Department of Materials Science and Engineering, University of 13

Erlangen-Nuremberg, 91958 Erlangen, Germany 14

15

16

* Corresponding author: 17

Sharif Hussein Sharif Zein 18

School of Chemical Engineering, Engineering Campus, Universiti Sains Malaysia, Seri 19

Ampangan 14300 Nibong Tebal, Seberang Perai Selatan, Pulau Pinang, Malaysia; 20

Tel.: +604 599 6442; Fax: +604 594 1013 21

Email: chhussein@eng.usm.my 22

# Present address: Technische Universiteit Eindhoven, Biomedical Engineering, PO Box 513, 23

5600 MB Eindhoven, The Netherlands 24

^ Present address: Department of Materials Science and Engineering, University of Erlangen-25

Nuremberg, 91058 Erlangen, Germany. 26

27

28

29

*Revised ManuscriptClick here to view linked References

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

2

Abstract 1

This paper presents the development of novel alternative injectable calcium phosphate cement 2

(CPC) composites for orthopaedic applications. The new CPC composites comprise of β-tri-3

calcium phosphate (β-TCP) and di-calcium phosphate anhydrous (DCPA) mixed with bovine 4

serum albumin (BSA) and incorporated with multi-walled carbon nanotubes (MWCNTs) or 5

functionalized MWCNTs (MWCNTs-OH and MWCNTs-COOH). Scanning electron 6

microscopy, compressive strength tests, injectability tests, Fourier transform infrared 7

spectroscopy and X-ray diffraction were used to evaluate final products properties. Compressive 8

strength tests and SEM observations demonstrated particularly that the concomitant admixture of 9

BSA and MWCNT improved the mechanical properties resulting in stronger CPC composites. 10

The presence of MWCNTs and BSA influenced the morphology of the hydroxyapatite (HA) 11

crystals in the CPC matrix. BSA was found to act as a promoter of HA growth when bounded to 12

the surface of CPC grains. MWCNT-OH containing composites exhibited the highest 13

compressive strengths (16.3MPa), being in the range of values for trabecular bone (2-12 MPa). 14

15

Keywords: Calcium phosphate cement (CPC); Carbon nanotubes (CNTs); Bovine serum 16

albumin (BSA); Compressive strength; Injectability 17

18

19

20

21

22

23

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

3

1. Introduction 1

The improved life expectancy in the developed countries has led to a significant rise in the 2

number of musculoskeletal disorders, such as osteoporosis and osteoarthritis (Bohner, 2000). 3

Minimally invasive surgical techniques have been shown to have important clinical potential for 4

stabilising such disorders (Mermelstein et al., 1998). Bio-resorbable calcium phosphate cements 5

(CPCs) represent an interesting alternative to traditional bone graft materials. Moreover, CPC is 6

a highly desirable material for orthopaedic applications due to its mouldability, in situ self-7

hardening ability, excellent osteoconductivity, adjustable resorbability rate and bone replacement 8

capability (Costantino et al., 1992; Friedman et al., 1998; Wang et al., 2007; Chow, 2009). Chow 9

(2009) has reviewed the current knowledge on CPCs in terms of their setting chemistry, 10

mechanical properties and in vivo tissue response, and has attributed a key role to the 11

‘injectability’ capability of the formulations for substantially improved next generation calcium 12

phosphate cements. The state of knowledge and recent developments in carbon nanotube 13

composites and coatings for hard and soft tissue engineering applications has been highlighted 14

by Boccaccini and Gerhardt (2010). 15

The development of self-setting CPCs has extended the application of calcium 16

phosphates to injectable bone substitutes that can be shaped and moulded to fit irregular defects, 17

and exhibit osteo-integrative properties comparable to or better than bulk calcium phosphates 18

(Brown and Chow, 1985). A number of CPC formulations are currently available commercially. 19

However, due to their limited compressive strength, CPCs are restricted primarily to non-stress-20

bearing applications. These include maxillofacial surgery, or the repair of cranial defects and 21

dental fillings (Friedman et al., 1998; Schmitz et al., 1999; Bohner, 2000). Several strategies are 22

being investigated to develop stronger CPC materials; of those the development of CPC-based 23

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

4

composites represents one particular and attractive approach (Fernandez et al., 1998; Watson et 1

al., 1999). In this context, a variety of reinforcing elements ranging from particulate bio-ceramic 2

inclusions to polymer fibres (Dos Santos et al., 2000) and carbon nanotubes (CNTs) (Wang et al., 3

2007) have been considered. 4

The modified interfacial bonding between bio-mineralised (i.e. pre-treated in simulated 5

body fluid (SBF)) CNTs and CPCs accounted for the significantly improved mechanical 6

properties in CPC/CNT composites (Wang et al., 2007). The use of polyamide fibres also 7

allowed a moderate increase in the compression strength in the CPC composites where the 8

observed reinforcement mechanism was related to the joining of the fibre to the matrix, and the 9

appearance of cracks in radial direction to the insertion cavity of the fibres (Dos Santos et al., 10

2000). Moreover, increasing mechanical properties with decreasing liquid-to-powder (L/P) ratio 11

is normally associated with a lower overall porosity of CPCs (Fernandez et al., 1998; Watson et 12

al., 1999). 13

Since the detailed investigations on carbon nanotubes (CNTs) published by Iijima (1991), 14

an increasing interest in the applications of CNTs has focused on their use as reinforcement in 15

different matrix materials because of their excellent mechanical properties (Treacy et al., 1996; 16

Wong et al., 1997; Gojny et al., 2003). CNTs are also being investigated for biomedical 17

applications (e.g. neural implants and tissue scaffolds) that utilise their high tensile strength, 18

electrical conductivity and chemical stability (Mattson et al., 2000; Correa-Duart et al., 2004; 19

Gheith et al., 2005; Boccaccini and Gerhardt, 2010). One particular application of interest is the 20

use of composite materials containing CNTs for bone tissue engineering leading to novel 21

matrices which can sustain bone cell growth and new bone tissue formation (MacDonald et al., 22

2005; Marrs et al., 2006; Meng et al., 2006; Shi et al., 2006; Leonor et al., 2009). CNTs have 23

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

5

also been suggested as reinforcement for inorganic bioactive coatings, e.g. hydroxyapatite 1

(Casagrande et al., 2008; Kaya and Boccaccini, 2008), or as a coating on 3D scaffolds 2

(Boccaccini et al., 2007). Another strategy to prepare improved calcium phosphate ceramics is 3

their incorporation with proteins, such as bovine serum albumin (BSA) (Peters Theodore, 1996; 4

Zieba et al., 1996; Combes et al., 1999; Burke et al., 2000; Leonor et al., 2009). For example, the 5

addition of low concentrations of BSA has been shown to enhance calcium phosphate crystal 6

growth (being favourable for bone tissue mineralisation), whereas higher concentrations 7

inhibited calcium phosphate crystallisation (Combes et al., 1999; Leonor et al., 2009). 8

In the present paper, we report on novel CPC composites incorporated with three 9

different types of multi-walled CNTs (MWCNTs) and BSA. The specific aims were to 10

investigate the influence/role of (a) filler content (MWCNT weight percentage), (b) surface 11

functionalization of MWCNT, and (c) BSA admixture on HA formation and the resultant 12

mechanical properties of CPC matrices. The new composites are intended for applications as 13

injectable bone substitutes. 14

15

2. Materials and methods 16

2.1 Composition of the cement and preparation of materials 17

Pristine MWCNTs, hydroxylated MWCNTs (MWCNTs-OH) and carboxylated MWCNTs 18

(MWCNTs-COOH) with a diameter of 30-50 nm and length of ≈ 30 μm were provided by the 19

Chinese Academy of Science. The purification processes were carried out according to the 20

experimental protocol presented in our previous work (Hong et al., 2006; Hassan et al., 2007). β-21

tri-calcium phosphate, β-Ca3(PO4)2, (β-TCP) and di-calcium phosphate anhydrous, CaHPO4, 22

(DCPA) were supplied by Sigma-Aldrich. The mean diameter of β-TCP and DCPA particles 23

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

6

were measured using CILAS 1180 laser particle size analyzer and found to be 17.30 µm and 1

11.50 µm, respectively. Equimolar fractions of β-TCP and DCPA were mixed with de-ionised 2

water, 0.25-1.0 wt% of MWCNTs and 15 wt% of BSA (supplied by Fluka) to produce different 3

CPC/MWCNTs/BSA composites to be investigated. The final solution volume was determined 4

by the amount required to produce a workable paste, i.e., a viscous cement with an L/P ratio of 5

0.27 ml/g (see Section 3.6) was prepared. The paste was blended using a mechanical overhead 6

stirrer at 30-50 rotations per minute until a homogeneous paste was obtained (approximately 1 7

hour) and then firmly packed by manual spatulation into a cylindrical stainless steel mould with 8

a diameter of 25 mm. The packed stainless steel mould was wrapped with a water-soaked wipe 9

to prevent the sample from drying out and was then stored in a Gyro-Rocker Incubator (Model: 10

S170) at 37 °C and 97 % humidity for 24 hours. All experiments were carried out under 11

controlled conditions at temperatures of 24-26 °C and at a relative humidity of 50-60 %. 12

13

2.2 Mechanical Testing 14

The compressive strength of the cylindrical specimens (nominal dimensions: Ø 25 mm, height: 15

10 mm) was determined using an Instron 3367 universal testing machine operating with a 16

crosshead speed of 0.1 mm/min. Before mechanical testing, the sample ends were filed flush to 17

ensure that the test specimens had plane-parallel surfaces. 18

19

2.3 Structural Characterisation 20

2.3.1 Scanning Electron Microscopy (SEM) 21

In order to investigate the microstructure of the as-prepared hardened cements both on the 22

surface and on the inner part of the composite, SEM examinations were performed using a 5 kV 23

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

7

accelerating voltage in a Leo Supra 35VP-24-58 microscope. All the samples were inspected at 1

various magnifications directly in the pellet form; except for the pure CPC specimen which was 2

not strong enough to be directly inspected in its pellet form. Thus, this sample was tested in its 3

crushed form which was spread evenly on top of a double-sided carbon tape attached to an 4

aluminium sample stub. 5

6

2.3.2 Fourier Transform Infrared (FTIR) spectroscopy 7

A Fourier Transform Infrared (FTIR) spectrometer (Perkin-Elmer FTIR 2000) was employed to 8

characterise the presence of specific surface functional groups in the composites. The FTIR 9

spectra were recorded in the wave number interval of 400 to 4000 cm-1

using the transmission 10

mode. Since possible incomplete mixing of the starting materials could have led to 11

inhomogeneous HA formation and to increase measurement accuracy/sensitivity, ten different 12

locations of CPC sample were scanned and averaged for further analysis. The resolution of the 13

spectrometer was 4 cm-1

. Before analysis, calibration of the spectrometer was performed by 14

using polystyrene as control sample. Then, the test sample was mixed with potassium bromide 15

using a weight-ratio of approximately 1:10. The mixture was ground to a fine, homogeneous 16

powder, which was then poured into a mould. The powder was densified and compacted using a 17

hydraulic press applying a pressure of ≈ 600 MPa to form thin pellets (thickness ≈ 100 μm). The 18

thin and transparent pellet was then placed in the sample holder for analysis. 19

20

2.3.3 X-ray Diffraction (XRD) analysis 21

XRD was used to determine the crystalline structure of the cement. The analysis was recorded on 22

a Siemens D5000 diffractometer using a diffraction angle of 2θ in the range 10-70° at a sweep 23

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

8

rate of 0.04°/sec. The qualitative analysis of different characteristic patterns of the materials 1

investigated was achieved by comparing peaks of the XRD spectrum with the standard 2

diffraction patterns of specific compounds based on the International Centre for Diffraction Data 3

(ICDD). 4

5

2.4 Injectability Tests 6

Injectability was qualitatively assessed and evaluated by extruding the paste through a disposable 7

syringe. A 10 ml syringe with a diameter of 16 mm and needle with an inner diameter of 2 mm 8

was filled with CPC paste, being then extruded from the syringe manually within a few seconds 9

at relatively constant speed. The injectability test was carried out in two parts. The objective of 10

the first part was to examine the L/P ratio required to produce a workable and injectable CPC 11

paste. Whilst the second part investigated the injectability, which was determined by considering 12

the percentage mass of the CPC paste extruded from the syringe divided by the original mass of 13

the paste inside the syringe (Eq. 1) (Burguera et al., 2008): 14

15

(1) 16

17

3. Results and Discussion 18

3.1 Fabrication of CPCs 19

CPC, the biomaterial for bone repair considered in this study, was created by mixing β-TCP and 20

DCPA with de-ionised water. After mixing, CPC becomes a paste-like material, which can 21

suitably be shaped and injected according to the contours of a bone defect or to fill miniscule 22

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

9

pores and cracks. The material is capable of self-setting at 37 °C (normal body temperature, 1

setting time ≈ 30 mins) and form calcium deficient hydroxyapatite (CDHA) as an end product. 2

3

3.2 Compressive strength 4

In order to identify the minimum amount of pristine MWCNTs required for the substantial 5

improvement of the mechanical properties of CPC/MWCNTs/BSA composite cement, we 6

investigated their compressive strength as a function of filler content prior to the reinforcement 7

effects of MWCNTs surface functionalization. Figure 1 shows the results of compressive 8

strength tests on CPC/BSA composites incorporated with 0.25-1 wt % un-functionalized 9

MWCNTs. Compared to the neat CPC (see Figure 2), the addition of 0.25, 0.50, 0.75, and 1 wt% 10

MWCNT considerably improved the mechanical properties of the CPC/BSA matrix, resulting in 11

compressive strength values ranging from 6.3 – 12.5 MPa (Figure 1). For a similar CPC matrix 12

and CNT loading, Wang et al. (2007) found higher values for the CPC matrix (25 MPa) and for 13

0.5 wt% containing bio-mineralised CNT/CPC composites (55 MPa). These results can be 14

explained by the pre-treatment of the CNTs with simulated body fluid, as well as by different 15

compressive strength test parameters (0.1 vs. 0.5 mm/min.), nanotube dimensions (20 nm -30 nm 16

vs. 60 nm -100 nm), and composite fabrication methods (manual spatulation vs. hydraulic 17

compaction with 700 kPa). In our study, CPC/BSA composites containing 0.5 wt. % MWCNT 18

were further studied in terms of CNT surface functionalization and their effect on the mechanical 19

properties. 20

The results of the compressive strength tests for the neat CPC, CPC/MWCNTs, CPC/BSA, 21

CPC/MWCNTs/BSA, CPC/MWCNTs-OH/BSA and CPC/MWCNTs-COOH/BSA composites 22

are shown in Figure 2. The addition of MWCNTs-OH gave the highest value of compressive 23

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

10

strength (16 ± 4 MPa) in the composite, followed by un-functionalized MWCNTs (12 ± 3 MPa) 1

and MWCNTs-COOH (9 ± 4 MPa), as further discussed below. For human bone, compressive 2

strengths in the range 2-12 MPa (trabecular bone) and 100-230 MPa (cortical bone) have been 3

reported (Kokubo et al., 2003). The present results indicate that a further improvement (including 4

a detailed investigation of the effect of MWCNT size and concentration) is necessary to increase 5

the mechanical performance towards the level of cortical bone, required to fill bone defects at 6

high load-bearing anatomical sites. 7

This study showed that the concomitant admixture of BSA and MWCNTs to the pure 8

CPC considerably increased the compressive strength of the CPC. In particular, the addition of 9

BSA has contributed to a further increase (by a factor of more than 10) of the compressive 10

strength of CPC/MWCNTs composites. This further improvement in the mechanical properties 11

can be plausibly explained by considering that appropriate amounts of BSA are capable of 12

promoting CPC crystal growth (Boccaccini et al., 2007) (see Figure 3 and 4). 13

In this research, physical blending was applied as a first approach to create a 14

homogeneous dispersion of MWCNTs in the cement matrix, ensuring uniform properties 15

throughout the CPC composite. However, this aim poses certain challenges due to the high 16

aspect ratio (length to diameter) of MWCNTs making them hard to mix. In particular, non-17

functionalized MWCNTs are known to have a tendency to agglomerate and form bundles (Cho 18

et al., 2009). In addition, MWCNTs are insoluble in water and organic solvents. Nevertheless, 19

hydroxyl group (-OH) functionalization has been confirmed to enable MWCNTs to disperse 20

more easily in water and organic solvents, and at the same time the surface functionalization 21

improves the interfacial bonding with the matrix (White et al., 2007; Kaya and Boccaccini, 2008). 22

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

11

This effect had probably led to the highest compressive strength in the MWCNTs-OH 1

composites investigated here. 2

In addition, since MWCNTs are chemically inert and exhibit a highly hydrophobic 3

surface nature, the un-functionalized MWCNTs are expected to have low chemical wettability 4

for their dispersion in CPC matrices (Wang et al., 2007). It is possible that lack of bonding at the 5

interface and different HA crystal morphologies (as discussed in Section 3.3) have probably 6

resulted in lower compressive strength values for CPC/MWCNT/BSA and CPC/MWCNTs-7

COOH/BSA cements in comparison to hydroxylated CPC/MWCNTs-OH/BSA composites. The 8

reasons for this phenomenon require more detailed investigation on the micromechanical 9

mechanisms involved as well as larger samples sizes for reliable statistical analyses. 10

From the point of view of the composites theory, it is well known that transfer of stress 11

and stiffness of the filler to the matrix depends on the quality of the interfacial bonding between 12

the two phases (MWCNT filler and CPC matrix in the present case) (White et al., 2007), which 13

in our study is influenced by the wettability and interfacial area between MWCNTs and CPC. 14

Thus, strong MWCNT/CPC interfacial bonding is an essential factor and significant condition 15

for improving the mechanical properties of CPCs enabling load transfer across the MWCNTs-16

CPC interface (Zhao and Gao, 2004; Wang et al., 2007). However, it should be pointed out that 17

for cement composites, stronger interfaces can also lead to increased brittleness (White et al., 18

2007). In this context, relatively weak interfaces might be advantageous in composites with 19

brittle matrices as toughening mechanisms can be activated thereby increasing fracture toughness 20

(Matthews and Rawlings, 2006). 21

22

23

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

12

3.3 Scanning Electron Microscopical Characterization 1

Different morphologies were observed in the HA crystal structures of CPC, 2

CPC/MWCNTs/BSA, CPC/MWCNTs-OH/BSA, and CPC/MWCNTs-COOH/BSA cements, 3

respectively (Figure 3 and 4). The SEM image of CPC/MWCNTs-OH/BSA composite presented 4

in Figure 3 (a) shows that HA crystals grew on MWCNTs-OH, whereas only a few HA crystals 5

grew on CPC/MWCNTs-COOH/BSA composites, as illustrated by arrows in Figure 3 (b). 6

Similar morphologies of HA crystals of Figure 3 (a-b) and Figure 4 (a-d) were obtained by Ratiel 7

et al. (2004), Carey et al. (2005) and Xu et al. (2006, 2008). However, it is still unknown and 8

unclear how the microstructure of cement interfaces and the HA morphology affect the 9

mechanical properties of CPC-based composites. Despite the different HA morphologies (Figure 10

3 and 4), relatively similar compressive strength values were found for all the 11

CPC/MWCNT/BSA composites tested. This observation indicates that the addition of BSA 12

(Figure 2) caused the improved mechanical strength of the composites and masked the 13

enhancement of the composites normally associated with surface functionalization of CNTs. This 14

finding also suggests that in CPC composites, BSA addition has a much higher enhancement 15

effect than CNT surface functionalization. In order to exploit the unique mechanical properties of 16

CNTs in composites, the CNTs are usually functionalized to improve and enhance their 17

reactivity, solubility, wettability and interfacial bonding (Niyogi et al., 2002; White et al., 2007; 18

Kaya and Boccaccini, 2008; Cho et al., 2009). Interestingly, in this study the addition of 19

carboxylated carbon nanotubes (MWCNTs-COOH) was not particularly effective at improving 20

the compressive strength of CPC/BSA composites. Zhao et al. (2005) showed that -COOH 21

groups present on the surface of SWCNTs were relatively inefficient at nucleating and growing 22

HA, probably due to impaired capability of attracting both Ca2+

and PO43-

ions for nucleating and 23

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

13

growing HA crystals. This effect can explain the slightly lower compressive strength for COOH-1

terminated MWCNTs. On the other hand, the effective attraction of both Ca2+

and PO43-

ions by 2

specific functional groups on MWCNTs is expected to enhance mechanical properties 3

(compressive strength), as observed for hydroxylated MWCNTs. The above explanation and in 4

particular the HA crystal morphology (Figure 3 and 4) of CPC/BSA/MWCNT composites can be 5

used to plausibly interpret/explain the differences in compressive strength of cements containing 6

hydroxylated and carboxylated MWCNTs, as further discussed below. 7

From Figure 4 (a), one can observe that the HA crystals grown in CPC were shorter, 8

wider, flatter and appear to be more plate-like than needle-shaped and less entangled. It is 9

possible that shorter, flatter plate-like HA crystals reduce compressive strength, as observed for 10

the pure CPC compared to the different CPC composites. For example, the CPC/MWCNTs/BSA 11

composite showed clusters of HA crystals, orientated in the same direction (Figure 4 (b)). The 12

HA crystals formed were thinner and longer than in CPC and existed in a medium size needle-13

like form. Longer crystals with higher aspect ratios would increase the mechanical properties 14

(Matthews and Rawlings, 2006). However, the interfacial bonding between un-functionalized 15

MWCNTs and the CPC matrix might be lower, as compared to composites with MWCNTs-OH, 16

as previously discussed in relation to the compressive strength results (Section 3.2 and Figure 2). 17

Figure 4 (c) shows the microstructure of the CPC/MWCNTs-COOH/BSA composite composed 18

of different structures of HA crystals. There were some regions with longer needle-like HA 19

crystals, illustrated by the solid line circle, as well as some regions with shorter plate-like HA 20

crystals demonstrated by the dashed line circle in Figure 4 (c). These different crystal formations 21

give rise to an inhomogeneous morphology, possibly inducing porosity, which can explain the 22

lower compressive strength of the carboxylated MWCNT/BSA calcium phosphate composites. 23

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

14

On the other hand, the SEM image in Figure 4 (d) shows that in the CPC/MWCNTs-OH/BSA 1

composite, the HA crystals are homogeneously distributed in all directions. The scale of the 2

microstructure is finer than in the previous case forming a composite of well-interconnected HA 3

crystals with a high surface/contact area, and the HA crystals appear to be finer and well-packed. 4

We hypothesised that this particular microstructure has led to an increased compressive strength 5

of these composites (Figure 2) and might be also favourable/beneficial to withstand shear 6

stresses. 7

As already mentioned, BSA has the capability of enhancing calcium phosphate crystal 8

growth (Combes et al., 1999). At low concentrations (< 10 g/l), BSA has been hypothesised to 9

stabilise nuclei and promote growth of octa-calcium phosphate crystals, while at higher 10

concentrations, crystal growth seems to be impeded by high BSA coverage. Although the net 11

charge on BSA at neutral pH is −17 mV, the protein contains both positively and negatively 12

charged residues (Peters Theodore, 1996). The arrangement of these charges on the protein, as 13

well as the complementarities between the charged groups on the protein and the growing apatite 14

surfaces, may influence crystal growth behaviour and also lead to more cohesive cements for 15

higher BSA contents (Zieba et al., 1996; Burke et al., 2000). 16

17

3.4 Fourier Transform Infrared Analysis 18

Figure 5 (a-c) shows the FTIR spectra of the CPC/MWCNTs/BSA, CPC/MWCNTs-OH/BSA 19

and CPC/MWCNTs-COOH/BSA composites, respectively. The spectra show absorption bands 20

at 3297-3302 cm-1

which correspond to the strong characteristic peak of the stretching mode of 21

the hydroxyl group (-OH) (Mahabole et al., 2005; Janusz et al., 2008). The characteristic bending 22

mode of intercalated H2O can be observed at 1655-1656 cm-1

(Mahabole et al., 2005). A very 23

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

15

strong, broad phosphate band derived from the P-O asymmetric stretching mode ( 3) of the PO43-

1

group was identified in the region 943-1128 cm-1

, indicating a deviation of phosphate ions from 2

their ideal tetrahedral structure (Mahabole et al., 2005; Janusz et al., 2008). 3

The absorption bands appearing at about 400 to 600 cm-1

can be attributed to the triple 4

( 4) 409, 417, 548, 551, 554, 586, 587 and 604 cm-1

and double ( 2)-degenerated fundamental 5

bending mode of the PO43-

functional group (Mahabole et al., 2005; Janusz et al., 2008). The 6

bands observed at 1543 cm-1

( 3 mode), 1547 cm-1

( 3 mode), and 943 cm-1

( 2 mode) were 7

assigned to the CO32-

group (Komath et al., 2000). 8

Furthermore, the characteristic bands at 943 cm-1

indicate the presence of HPO42-

in the 9

crystal lattice (Tsuchiya et al., 2006). As a result, all the bands discussed above and also their 10

positions in the FTIR spectra confirm the formation of apatite in composites fabricated with both 11

un-functionalized MWCNTs and functionalized MWCNTs. 12

The FTIR results further confirm the observations made on SEM images discussed above. 13

By comparing the results of the different FTIR spectra, it was found that the absorption bands of 14

CPC/MWCNTs/BSA (Figure 5 (a)) and bands of CPC/MWCNTs-OH/BSA (Figure 5 (b)) give 15

rise to sharper peaks at wave numbers of 550, 587, 1065, 1128 and 1655 cm-1

, compared to 16

CPC/MWCNTs-COOH/BSA (Figure 5 (c)). 17

For the CPC/MWCNTs/BSA and CPC/MWCNTs-OH/BSA samples, an extra band was 18

detected in the region of 780 cm-1

, being not observed in the CPC/MWCNTs-COOH/BSA 19

composite. The origin of this extra band is unclear, but might arise from the different extents of 20

apatite formation on CPC/MWCNTs/BSA and CPC/MWCNTs-OH/BSA composites compared 21

to CPC/MWCNTs-COOH/BSA composites. 22

23

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

16

3.5 X-Ray Diffraction analysis 1

The XRD patterns of the CPC/MWCNTs/BSA, CPC/MWCNTs-OH/BSA and CPC/MWCNTs-2

COOH/BSA composites are shown in Figure 6 (a-c). Diffraction peaks corresponding to HA 3

crystalline phase were detected at 2 angles of 26, 29, 32, 40 and 53°. It is therefore evident that 4

it is possible to obtain self-setting injectable HA by mixing β-TCP and DCPA with de-ionized 5

water. The sharp and narrow diffraction peaks observed in the regions of relevance to HA 6

suggest that the HA formed is crystalline, which can be correlated with the crystal morphology 7

observed by SEM (Ratier et al., 2004; Carey et al., 2005; Xu et al., 2006; Xu et al., 2008) (Figure 8

4 (a-d)). However, from Figure 6, the XRD analysis also revealed two extra phases of the 9

starting materials corresponding to β-TCP and DCPA, indicating that the reaction to form HA is 10

not complete, which might be associated with factors such as the liquid to powder ratio, the 11

hydration time and the hydration environment. As a whole, the XRD, SEM and FTIR results 12

showed that the investigated CPC composites developed a crystalline HA phase, which is in its 13

chemical and crystallographic composition similar to the mineral phase of bone (Suchanek and 14

Yoshimura, 1998). 15

16

3.6 Injectability Test 17

The injectability test was performed only with the CPC containing hydroxylated carbon nanotube 18

and BSA (CPC/MWCNT-OH/BSA), because this composite showed the most promising result 19

in terms of compressive strength. The desired physical condition of workable CPC/MWCNTs-20

OH/BSA composite paste was found at an L/P ratio of 0.27 ml/g, resulting in an injectability of 21

97%, i.e., 97% of the CPC paste could be extruded. It is important to note here that the maximum 22

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

17

percentage of cement paste extruded can never achieve 100 %, due to small amounts of residual 1

cement paste inside the syringe. 2

It is clear that the injectability of cement pastes can be influenced by varying the L/P ratio. 3

The injectability of cement pastes with an L/P ratio < 0.25 ml/g was not tested because the 4

specimen was not workable (too viscous). The injectability of cement pastes with an L/P ratio > 5

0.28 ml/g was not tested because the resulting cement paste was too liquid. For example, Bohner 6

and Baroud (2005) suggested that a well-injectable cement paste should have the capacity to stay 7

homogeneous during injection, independently of the injection force. They suggested that this 8

approach can be achieved by increasing the cement L/P ratio. As a result, the ability of cement 9

paste to harden in an aqueous condition will be reduced because the viscosity of the cement paste 10

is reduced at the same time. This reduced stability will cause a total degradation of the cement 11

paste. 12

Summarising, an L/P ratio of CPC/MWCNTs-OH/BSA composite paste of 0.27 ml/g 13

yielded mechanically strong and injectable CPCs with an injectability of 97 %. This material is 14

thus suitable for bone repair applications as an injectable bone substitute. 15

16

4. Conclusion 17

The present work demonstrated the possibility of developing high compressive strength CPCs by 18

reinforcement with MWCNTs and BSA for the use as injectable bone substitute. Drawing on the 19

results from the compressive strength tests, the CPC/MWCNTs-OH/BSA composite exhibited 20

substantially improved compressive strength (≈ 16 MPa) compared to pure cement (≈ 1 MPa). 21

Of all MWCNTs studied, functionalized MWCNTs-OH were found to be the most effective to 22

increase the compressive strength of CPC. It was suggested that hydroxyl functional groups on 23

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

18

the surface of MWCNTs improved the reactivity and wettability of MWCNTs leading to strong 1

interfacial bonding. In addition, the effective attraction of both Ca2+

and PO43-

by the functional 2

groups of MWCNTs-OH is expected to promote the nucleation and growth of HA crystals. The 3

XRD, SEM and FTIR analyses confirmed the formation of crystalline HA during the synthesis of 4

CPC. SEM observations demonstrated that the addition of MWCNTs modifies the morphology 5

of HA crystallites. 6

The HA crystals in CPC/MWCNTs-OH/BSA composites were fine, homogeneously 7

grown and distributed in all directions, forming a well-packed composite microstructure, which 8

resulted in the highest compressive strength. A simple test of injectability on the 9

CPC/MWCNTs-OH/BSA composite demonstrated the high percentage of extrusion (97%) being 10

thus easily workable and clinically applicable as an in-situ hardening cement. The promising 11

results presented here in terms of compression strength enhancement must be confirmed with 12

larger sample sizes and a comprehensive study of the influence of CNT surface modification and 13

CPC pre-treatment (e.g. upon soaking in SBF) on the mechanical properties (fracture toughness, 14

elastic modulus, shear strength and elongation at break) to confirm the suitability of the 15

composites for the intended application as bone substitutes. Due to concerns about the 16

biocompatibility and toxicity of CNT for their use as biomaterials (Boccaccini and Gerhardt, 17

2010), further research should also focus on in vitro cell-biological investigations and on the in 18

vivo performance of the novel MWCNT and BSA containing CPC developed here. Future 19

experimental work should also include investigations on the optimisation of blending 20

properties/characteristics and on investigating the setting time of the present CPC. 21

22

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

19

Acknowledgements 1

The financial support provided by the British Council through the UK’s Prime Minister’s 2

Initiative for International Education (PMI2) Connect scheme is gratefully acknowledged. Kean 3

Khoon Chew and Kah Ling Low also acknowledge the USM Fellowship for the support for their 4

studies. 5

6

References 7

Boccaccini, A. R., Gerhardt, L.C., 2010. Carbon nanotube composite scaffolds and coatings for 8

tissue engineering applications. Key Engineering Materials: Advanced Bioceramics for 9

Medical Applications. 441, 31-52. 10

Boccaccini, A.R., Chicatun, F., Cho, J., Bretcanu, O., Roether, J.A., Novak, S., Chen, Q.Z., 2007. 11

Carbon nanotube coatings on bioglass-based tissue engineering scaffolds. Adv. Eng. Mater. 12

17(15), 2815-2822. 13

Bohner, M., 2000. Calcium orthophosphates in medicine: from ceramics to calcium phosphate 14

cements. Injury. 31, 37-47. 15

Bohner, M., Baroud, G., 2005. Injectability of calcium phosphate pastes. Biomaterials 26(13), 16

1553-1563. 17

Brown, W.E., Chow, L.C., 1985. Dental restorative cement pastes. American Dental Association 18

Health Foundation (Washington, DC): USA. US patent no: 4518430. 19

Burguera, E.F., Xu, H.H.K., Sun, L.M., 2008. Injectable calcium phosphate cement: Effects of 20

powder-to-liquid ratio and needle size. J. Biomed. Mater. Res. Part B: Appl. Biomater. 21

84B(2), 493-502. 22

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

20

Burke, E.M., Guo, Y., Colon, L., Rahima, M., Veis, A., Nancollas, G,H., 2000. Influence of 1

polyaspartic acid and phosphophoryn on octacalcium phosphate growth kinetics. Coll. 2

Surfaces: Biointerfaces 17(1), 49-57. 3

Carey, L.E., Xu, H.H., Simon, Jr. C.G., Takagi, S., Chow, L.C., 2005. Premixed rapid-setting 4

calcium phosphate composites for bone repair. Biomaterials 26(24), 5002-5014. 5

Casagrande, T., Lawson, G., Li, H., Wei, J., Adronov, A., Zhitomirsky, I., 2008. 6

Electrodeposition of composite materials containing functionalized carbon nanotubes. Mater. 7

Chem. Phys. 111(1), 42-49. 8

Cho, J., Boccaccini, A.R., Shaffer, M.S.P., 2009. Ceramic matrix composites containing carbon 9

nanotubes. J. Mater. Sci. 44(8), 1934-1951. 10

Chow, L.C., 2009. Next generation calcium phosphate-based biomaterials. Dent. Mater. J. 28(1), 11

1-10. 12

Combes, C., Rey, C., Freche, M., 1996. In vitro crystallization of octacalcium phosphate on type 13

I collagen: influence of serum albumin. J. Mater. Sci. Mater. Med. 10(3), 153-160. 14

Correa-Duarte, M.A., Wagner, N., Rojas-Chapana, J., Morsczeck, C., Thie, M., Giersig, M., 15

2004. Fabrication and biocompatibility of carbon nanotube-based 3D networks as scaffolds 16

for cell seeding and growth. Nano Letters 4(11), 2233-2236. 17

Costantino, P.D., Friedman, C.D., Jones, K., Chow, L.C., Sisson, Sr., 1992. Experimental 18

hydroxyapatite cement cranioplasty. Plast. Reconstr. Surgr. 90(2), 174-185. 19

Dos Santos, L.A., De Oliveira, L.C., Rigo, E.C.D., Carrodeguas, R.G., Boschi, A.O., De Arruda, 20

A.C.F., 2000. Fiber reinforced calcium phosphate cement. Artificial Organs 24(3) 212-216. 21

Fernandez, E., Gil, F.J., Best, S.M., Ginebra, M.P., Driessens, F.C., Planell, J.A., 1998. 22

Improvement of the mechanical properties of new calcium phosphate bone cements in the 23

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

21

CaHPO4-alpha-Ca3(PO4)2 system: compressive strength and microstructural development. J. 1

Biomed. Mater. Res. Part A. 41(4), 560-567. 2

Friedman, C.D., Costantino, P.D., Takagi, S., Chow, L.C., 1998. BoneSource (TM) 3

hydroxyapatite cement: A novel biomaterial for craniofacial skeletal tissue engineering and 4

reconstruction. J. Biomed. Mater. Res. Part B: Appl. Biomater. 43(4) 428-432. 5

Gheith, M.K., Sinani, V.A., Wicksted, J.P., Matts, R.L., Kotov, N.A., 2005. Single-walled 6

carbon nanotube polyelectrolyte multilayers and freestanding films as a biocompatible 7

platform for neuroprosthetic implants. Adv. Mater. 17(22), 2663-2670. 8

Gojny, F.H., Nastalczyk, J., Roslaniec, Z., Schulte, K., 2003. Surface modified multi-walled 9

carbon nanotubes in CNT/epoxy-composites. Chem. Phys. Lett. 370(5-6), 820-824. 10

Hassan, N.H.A., Mohamed, A.R., Zein, S.H.S., 2007. Study of hydrogen storage by 11

carbonaceous material at room temperature. Diamond Relat. Mater. 16(8), 1517-1523. 12

Hong, K.B., Ismail, A.A.B., Mahayuddin, M.E.M., Mohamed, A.R., Zein, S.H.S., 2006. 13

Production of high purity multi-walled carbon nanotubes from catalytic decomposition of 14

methane. J. Nat. Gas. Chem. 15(4), 266-270. 15

Iijima, S., 1991. Helical microtubules of graphitic carbon. Nature 354(6348), 56-58. 16

Janusz, W., Skwarek, E., Pasieczna-Patkowska, S., Slosarczyk, A., Paszkiewicz, Z., Rapacz-17

Kmita, A., 2008. A study of surface properties of calcium phosphate by means of 18

photoacoustic spectroscopy (FT-IR/PAS), potentiometric titration and electrophoretic 19

measurements. Eur. Phys. J. - Special Topics 154, 329-333. 20

Kaya, C., Singh, I., Boccaccini, A.R., 2008. Multi-walled carbon nanotube-reinforced 21

hydroxyapatite layers on Ti6Al4V medical implants by electrophoretic deposition (EPD). 22

Adv. Eng. Mater. 10(1-2), 131-138. 23

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

22

Kokubo, T., Kim, H.M., Kawashita, M., 2003. Novel bioactive materials with different 1

mechanical properties. Biomaterials 24(13), 2161-2175. 2

Komath, M., Varma, H.K., Sivakumar, R., 2000. On the development of an apatitic calcium 3

phosphate bone cement. Bull. Mater. Sci. 23(2), 135-140. 4

Leonor, I.B., Alves, C.M., Azevedo, H.S., Reis, R.L., 2009. Effects of protein incorporation on 5

calcium phosphate coating. Materials Science & Engineering C-Biomimetic and 6

Supramolecular Systems 29(3), 913-918. 7

MacDonald, R.A., Laurenzi, B.F., Viswanathan, G., Ajayan, P.M., Stegemann, J.P., 2005. 8

Collagen-carbon nanotube composite materials as scaffolds in tissue engineering. J. Biomed. 9

Mater. Res. Part A 74A(3), 489-496. 10

Mahabole, M.P., Aiyer, R.C., Ramakrishna, C.V., Sreedhar, B., Khairnar, R.S., 2005. Synthesis, 11

characterization and gas sensing property of hydroxyapatite ceramic. Bull. Mater. Sci. 28(6), 12

535-545. 13

Marrs, B., Andrews, R., Rantell, T., Pienkowski, D., 2006. Augmentation of acrylic bone cement 14

with multiwall carbon nanotubes. J. Biomed. Mater. Res. Part A 77(2), 269-276. 15

Matthews, F.L., Rawlings, R.D., 2006. Fracture mechanics and toughening mechanisms, in 16

Composite materials: Engineering and science. Matthews, F.L., and Rawlings, R.D., Editors. 17

Woodhead Publishing Limited, CRC Press. Cambridge, England. p. 326-62. 18

Mattson, M.P., Haddon, R.C., Rao, A.M., 2000. Molecular functionalization of carbon nanotubes 19

and use as substrates for neuronal growth. J. Mol. Neurosci. 14(3), 175-182. 20

Meng, J., Song, L., Meng, J., Kong, H., Zhu, G., Wang, C., Xu, L., Xie, S., Xu, H., 2006. Using 21

single-walled carbon nanotubes nonwoven films as scaffolds to enhance long-term cell 22

proliferation in vitro. J. Biomed. Mater. Res. Part A 79A(2), 298-306. 23

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

23

Mermelstein, L.E., McLain, R.F., Yerby, S.A., 1998. Reinforcement of thoracolumbar burst 1

fractures with calcium phosphate cement - A biomechanical study. Spine. 23(6), 664-670. 2

Niyogi, S., Hamon, M.A., Hu, H., Zhao, B., Bhowmik, P., Sen, R., Itkis, M.E., Haddon, R.C., 3

2002. Chemistry of single-walled carbon nanotubes. Acc. Chem. Res. 35(12), 1105-1113. 4

Peters Theodore, Jr., 1996. All about albumin: biochemistry, genetics, and medical applications. 5

CA: Academic Press. San Diego. 6

Ratier, A., Freche, M., Lacout, J.L., Rodriguez, F., 2004. Behaviour of an injectable calcium 7

phosphate cement with added tetracycline. Int. J. Pharm. 274(1-2), 261-268. 8

Schmitz, J.P., Hollinger, J.O., Milam, S.B., 1999. Reconstruction of bone using calcium 9

phosphate bone cements: A critical review. J. Oral. Maxillofac. Surg. 57(9), 1122-1126. 10

Shi, X.F., Hudson, J.L., Spicer, P.S., Tour, J.M., Krishnamoorti, R.M., Mikos, A.G., 2006. 11

Injectable nanocomposites of single-walled carbon nanotubes and biodegradable polymers 12

for bone tissue engineering. Biomacromolecules 7(7), 2237-2242. 13

Suchanek, W., Yoshimura, M., 1998. Processing and properties of hydroxyapatite-based 14

biomaterials for use as hard tissue replacement implants. J. Mater. Res. 13(1), 94-117. 15

Treacy, M.M.J., Ebbesen, T.W., Gibson, J.M., 1996. Exceptionally high Young's modulus 16

observed for individual carbon nanotubes. Nature 381(6584), 678-680. 17

Tsuchiya, H., Macak, J.M., Muller, L., Kunze, J., Muller, F., Greil, P., Virtanen, S., Schmuki, P., 18

2006. Hydroxyapatite growth on anodic TiO2 nanotubes. J. Biomed. Mater. Res. Part A. 19

77A(3), 534-541. 20

Wang, X.P., Ye, J.D., Wang, Y.J., Chen, L., 2007. Reinforcement of calcium phosphate cement 21

by bio-mineralized carbon nanotube. J. Am. Ceram. Soc. 90(3), 962-964. 22

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

24

Wang, X.P., Ye, J.D., Wang, Y.J., Wu, X.P., Bo, B., 2007. Control of crystallinity of hydrated 1

products in a calcium phosphate bone cement. J. Biomed. Mater. Res. Part A. 81(4), 781-2

790. 3

Watson, K.E., Tenhuisen, K.S., Brown, P.W., 1999. The formation of hydroxyapatite-calcium 4

polyacrylate composites. J. Mater. Sci. Mater. Med. 10(4), 205-213. 5

White, A.A., Best, S.M., Kinloch, I.A., 2007. Hydroxyapatite-carbon nanotube composites for 6

biomedical applications: A review. Int. J. Appl. Ceram. Tech. 4(1), 1-13. 7

Wong, E.W., Sheehan, P.E., Lieber, C.M., 1997. Nanobeam mechanics: elasticity, strength, and 8

toughness of nanorods and nanotubes. Science 277(5334), 1971-1975. 9

Xu, H.H.K., Weir, M.D., Burguera, E.F., Fraser, A.M., 2006. Injectable and macroporous 10

calcium phosphate cement scaffold. Biomaterials 27(24), 4279-4287. 11

Xu, H.H.K., Weir, M.D., Simon, C.G., 2008. Injectable and strong nano-apatite scaffolds for 12

cell/growth factor delivery and bone regeneration. Dent. Mater. 24(9), 1212-1222. 13

Zhao, B., Hu, H., Mandal, S.K., Haddon, R.C., 2005. A bone mimic based on the self-assembly 14

of hydroxyapatite on chemically functionalized single-walled carbon nanotubes. Chem. 15

Mater. 17(12), 3235-3241. 16

Zhao, L.P., Gao, L., 2004. Novel in situ synthesis of MWNTs-hydroxyapatite composites. 17

Carbon 42(2), 423-426. 18

Zieba, A., Sethurman, G., Perez, F., Nancollas, G.H., Cameron, D., 1996. Influence of organic 19

phosphonates on hydroxyapatite crystal growth kinetics. Langmuir 12(11), 2853-2858. 20

21

22

23

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

25

Captions 1

Figures legends 2

Figure 1. The comparison of compressive strength values of CPC/MWCNTs/BSA composites 3

containing different percent by weight of pristine MWCNTs. All composites have a BSA content 4

of 15 wt%. Data are presented as mean ± 1 standard deviation (n = 2). 5

6

Figure 2. The comparison of compressive strength values of the materials investigated. The 7

composites have a MWCNT loading of 0.5 wt % and a BSA content of 15 wt %. Data are 8

presented as means ± 1 standard deviation (n = 2). *Note that the compressive strength of the 9

CPC/BSA composite could not be measured because the composite was too weak to form the 10

required shape for the compressive test purpose. 11

12

Figure 3. Typical SEM images of (a) CPC/MWCNTs-OH/BSA composite, (b) CPC/MWCNTs-13

COOH/BSA composite (white arrows are considered to be hydroxyapatite crystals), which was 14

confirmed by FTIR and XRD (Sections 3.3 and 3.4). 15

16

Figure 4. Typical SEM images of (a) CPC, (b) CPC/MWCNTs/BSA composite, (c) 17

CPC/MWCNTs-COOH/BSA composite (solid line circle: longer needle-like HA crystals 18

exhibiting so-called cauliflower morphology, dashed line circle: shorter plate-like HA crystals) 19

and (d) CPC/MWCNTs-OH/BSA composite showing compact microstructure of fine HA 20

crystals. 21

22

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

26

Figure 5. An FTIR spectrum of the investigated calcium phosphate-based ceramic cements (a) 1

CPC/MWCNTs/BSA, (b) CPC/MWCNTs-OH/BSA and (c) CPC/MWCNTs-COOH/BSA 2

(Discussion in Section 3.4). 3

4

Figure 6. X-ray diffraction patterns of the investigated samples: (a) CPC/MWCNTs/BSA, (b) 5

CPC/MWCNTs-OH/BSA and (c) CPC/MWCNTs-COOH/BSA. The composites have a 6

MWCNT loading of 0.5 wt % and a BSA content of 15 wt %. 7

8

9

10

11

12

13

14

15

16

17

18

19

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

27

1

Figure 1. Comparison of the compressive strength value of CPC/MWCNTs/BSA composites 2

containing different percent by weight of pristine MWCNTs. All composites have a BSA content 3

of 15 wt%. Data are presented as mean ± 1 standard deviation (n = 2). 4

5

6

7

8

9

10

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

28

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

Figure 2. The comparison of the compressive strength values of the materials investigated. The 18

composites have a MWCNT loading of 0.5 wt % and a BSA content of 15 wt %. Data are 19

presented as means ± 1 standard deviation (n = 2). *Note that the compressive strength of the 20

CPC/BSA composite could not be measured because the composite was too weak to form the 21

required shape for the compressive test purpose. 22

23

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

29

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

Figure 3. Typical SEM images of (a) CPC/MWCNTs-OH/BSA composite, (b) CPC/MWCNTs-20

COOH/BSA composite (white arrows are considered to be hydroxyapatite crystals), which was 21

confirmed by FTIR and XRD (Sections 3.3 and 3.4). 22

23

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

30

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

31

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

Figure 4. Typical SEM images of (a) CPC, (b) CPC/MWCNTs/BSA composite, (c) 19

CPC/MWCNTs-COOH/BSA composite (solid line circle: longer needle-like HA crystals 20

exhibiting so-called cauliflower morphology, dashed line circle: shorter plate-like HA crystals) 21

and (d) CPC/MWCNTs-OH/BSA composite showing compact microstructure of fine HA 22

crystals. 23

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

32

1

2

3

4

5

6

7

8

9

10

11

12

13

Figure 5. An FTIR spectrum of the investigated calcium phosphate-based ceramic cements (a) 14

CPC/MWCNTs/BSA, (b) CPC/MWCNTs-OH/BSA and (c) CPC/MWCNTs-COOH/BSA 15

(Discussion in Section 3.4). 16

17

18

19

20

21

22

23

ACC

EPTE

DM

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

33

1

Figure 6. X-ray diffraction patterns of the investigated samples: (a) CPC/MWCNTs/BSA, (b) 2

CPC/MWCNTs-OH/BSA and (c) CPC/MWCNTs-COOH/BSA. The composites have a 3

MWCNT loading of 0.5 wt % and a BSA content of 15 wt %. 4

5

6

7

8