Materials Science and Engineering C 77 (2017) 1078–1087

Contents lists available at ScienceDirect

Materials Science and Engineering C

journal homepage: www.elsevier.com/locate/msec

Carbon nanotube-reinforced mesoporous hydroxyapatite composites

with excellent mechanical and biological properties for bone
replacement material application
Haipeng Li a, Xiaoqing Song a, Baoe Li a, Jianli Kang b,c,⁎, Chunyong Liang a, Hongshui Wang a,
Zhenyang Yu d, Zhijun Qiao d
a
School of Materials Science and Engineering, Hebei University of Technology, Tianjin 300130, China
b
State Key Laboratory of Hollow Fiber Membrane Materials and Processes, Tianjin Polytechnic University, Tianjin 300387, China
c
School of Materials Science and Engineering, Tianjin Polytechnic University, Tianjin 300387, China
d
School of Mechanical Engineering, Tianjin Polytechnic University, Tianjin 300387, China

a r t i c l e i n f o a b s t r a c t

Article history: Carbon nanotube (CNT)-reinforced mesoporous hydroxyapatite (HA) composites with excellent mechanical and
Received 30 December 2016 biological properties were fabricated successfully by the in situ chemical deposition of mesoporous HA on homo-
Received in revised form 3 April 2017 geneously dispersed CNTs. The CNTs are first synthesized in situ on HA nanopowders by chemical vapor deposi-
Accepted 6 April 2017
tion, and then, the HA particles with mesoporous structures are deposited in situ onto the as-grown CNTs by
Available online 7 April 2017
using cetyl trimethyl ammonium bromide as templates to form mesoporous HA encapsulated CNTs (CNT@
Keywords:
meso-HA). The modification of CNTs by mesoporous HA leads to strong CNT-HA interfacial bonding, resulting
Biocomposite in efficient load transfer between CNT and HA and improved mechanical properties of CNT/HA composites.
Carbon nanotubes More importantly, the mesoporous HA structure has a high specific surface area and large surface roughness
Mesoporous hydroxyapatite that greatly promote the cell adhesion and proliferation, resulting in better biocompatibility and improved oste-
Mechanical property oblast viability (MC3T3-E1) compared to those fabricated by traditional methods. Therefore, the obtained CNT@
Biological property meso-HA composites are expected to be promising materials for bone regeneration and implantation
applications.
© 2017 Elsevier B.V. All rights reserved.

1. Introduction
As a bone mineral component, hydroxyapatite (HA) has long been
an attractive choice for bone grafting materials due to its similar chem-
ical composition to natural apatite, excellent biocompatibility with
human tissues, and distinguished bioactivity and osteoconductivity
[1–4]. Due to its similar crystal structure as that of apatite in the
human skeletal system, HA is suitable for new bone growth and integra-
tion [5–8]. Therefore, it is recognized as one of the most promising or-
thopedic biomaterials.
However, the poor mechanical properties of HA restrict its clinical
applications, such as its low strength and toughness. To improve the
mechanical properties of HA, many reinforcing phases (e.g., carbon
fiber, zirconia, glass and alumina) have been used to fabricate HA matrix
composites [9–13]. Among these reinforcements, carbon nanotubes
(CNTs) have many outstanding properties, such as low density and re-
markable mechanical, electronic and thermal properties [14,15],
⁎ Corresponding author at: State Key Laboratory of Hollow Fiber Membrane Materials
and Processes, Tianjin Polytechnic University, Tianjin 300387, China.
E-mail address: kangjianli@tjpu.edu.cn (J. Kang).
which will endow CNTs with ideal enhancements for HA composites.
Hence, a considerable number of studies have been devoted to the prep-
aration of CNT/HA composites to substantially enhance the mechanical
strength of HA by methods such as laser surface alloying, spark plasma
sintering and physical blending [16–20]. However, the toxicity issue of
CNTs for use in biomaterials is still controversial, as there are many re-
ports highlighting their toxic effects. For example, Ding et al. found
that CNTs induced cell cycle arrest and increased necrosis/apoptosis of
human skin fibroblasts [21]. Lam et al. showed that CNTs can cause
granulomas and lung lesions [22]. Fortunately, it has been proven that
the toxicity of CNTs can be eliminated by coating them with modified
materials, such as HA [23,24]. Zhang et al. found that apatite-modified
CNTs exhibited significantly increased cell viability and lower cytotoxic-
ity [25]. Our recent study also indicated that the tight HA encapsulation
layer on CNTs could improve the mechanical properties, cell prolifera-
tion and biocompatibility of CNT/HA composites [26]. Unfortunately,
the surfaces of CNTs modified with HA usually possess a smooth layered
HA structure and lack a large specific surface area and sufficient poros-
ity, which is not good for tissue fluid circulation and nutrition supply. In
addition, due to the lack of focal adhesion points, the smooth HA surface
is not favorable for the adhesion, proliferation and growth of osteoblast

http://dx.doi.org/10.1016/j.msec.2017.04.048
0928-4931/© 2017 Elsevier B.V. All rights reserved.
 H. Li et al. / Materials Science and Engineering C 77 (2017) 1078–1087
cells on CNT/HA composites. Thus, to mimic the real bone tissues, it is of
vital importance to develop CNT/HA composites with both excellent
mechanical and biological properties for the applications of potential
bone replacement material.
Herein, novel CNT-reinforced mesoporous HA composites were fab-
ricated by the in situ chemical deposition of mesoporous HA on homo-
geneously dispersed CNTs. First, the in situ chemical vapor deposition
(CVD) method was used to achieve a homogeneous dispersion of
CNTs within HA nanopowders. Then, the mesoporous HA particles
were deposited onto the as-grown CNTs by the in situ chemical deposi-
tion process to obtain mesoporous HA encapsulated CNTs (CNT@meso-
HA), which were consequently employed to fabricate CNT@meso-HA
bulk composites. CNT@meso-HA in situ composites obtained by chemi-
cal combination achieved tight tube-matrix interface bonding, leading
to improvements of the mechanical properties of the HA composites.
Meanwhile, the porous structure and high specific surface area of the
composites can not only contribute to the circulation of body fluids
but also exhibit a wettable and roughened surface, which are favorable
properties for cell adhesion, proliferation and growth. Thus, the novel
CNT@meso-HA composites with excellent mechanical and biological
properties possess great potential for use in bone replacement applica-
tions and highly active tissue scaffold preparations.
2. Materials and methods
2.1. Functionalization of CNTs
Pristine CNTs were synthesized by the in situ CVD method reported
in detail in our previous papers [26,27]. The synthesized CNTs were
functionalized with oxygen containing functional groups, specifically,
a 12 h treatment with H2O2 at room temperature, to create hydrophilic
functional groups, such as –OH, –COOH and –C = O. Briefly, the synthe-
sized CNT/HA was dispersed in ethanol using ultrasonication for 0.5 h,
and then, a H2O2 solution was added under continuous ultrasonic dis-
persion for 12 h. After that, the oxidized CNT/HA was filtered by a
millipore polycarbonate filter membrane, washed with deionized
water until a neutral pH was reached and dried at 80 °C under vacuum
for 24 h.
2.2. Preparation of CNT@meso-HA composite powders
The CNT@meso-HA composite powders were fabricated by an in situ
chemical precipitation method, as follows: First, the functionalized CNT/
HA composite powders were dispersed into deionized water by
ultrasonication for 20 min and then mixed with an aqueous solution
of calcium nitrate (Ca(NO3)2·4H2O) (0.5 mol L−1) under magnetic stir-
ring for 1 h at room temperature to obtain Ca2+ modified CNTs. Second,
an aqueous solution of diammonium phosphate ((NH4)2HPO4) and
cetyl trimethyl ammonium bromide (CTAB: CH3(CH2)15N(CH3)3Br)
was added drop-wise to the former mixture solution at a rate of
2 mL min−1 under vigorous stirring. Monitored with a pH meter, the
pH value (10−11) of the suspension was adjusted by an ammonia hy-
droxide solution (NH4OH). The amorphous HA was synthesized in situ
on the CNTs according to Eq. (1).
10CaðNO3 Þ2  4H2 O þ 6ðNH4 Þ2 HPO4 þ 8NH4 OH→Ca10 ðPO4 Þ6 ðOHÞ2
þ 46H2 O þ 20NH4 NO3 ð1Þ
Subsequently, the precipitates of amorphous meso-HA precursor
micelles, encapsulating the CNTs by the combination of Ca2 + and
PO3−
4 , were obtained after aging for 72 h and washed repeatedly with
deionized water until a neutral pH was reached. After drying at 80 °C
for 12 h, the amorphous precipitates were calcined in an argon atmo-
sphere at 450 °C for 2 h to remove the CTAB template and obtained crys-
talline CNT@meso-HA composite powders. According to the content of
CNTs, the samples of CNT@meso-HA were denoted as 1 wt%, 2 wt%,
1079
3 wt% and 4 wt% CNT@meso-HA, respectively. For comparison, the sam-
ples of non-mesoporous HA encapsulated CNTs (CNT@HA) synthesized
without the CTAB template were also prepared and were denoted as
1 wt%, 2 wt%, 3 wt% and 4 wt% CNT@HA, respectively.
2.3. Preparation of CNT@meso-HA composites
CNT/HA bulk composites were prepared by the powder metallurgy
method. First, CNT@meso-HA composite powders were pressed to a
bulk material under a pressure of 550 MPa. The obtained bulk samples
were then sintered at 1100 °C for 3 h under a moist Ar atmosphere.
For comparison, bulk materials composed of pure HA, CNT@HA and in
situ synthesized CNT/HA composite powders were also fabricated
with the same process.
2.4. Cell adhesion test and cell proliferation assay
Mouse osteoblasts (MC3T3-E1, Cell Bank of the Chinese Academy of
Sciences) were used to investigate the cytocompatibility of CNT/HA
composites. Before cell cultivation, the composites (pure HA, in situ syn-
thesized CNT/HA, CNT@HA and CNT@meso-HA) were ground and
polished to eliminate the influences arising from surface roughness.
For the cell culture experiment, the osteoblasts were first cultured in
Dulbecco's modified Eagle's medium (DMEM) with fetal bovine serum
(FBS, 10%) and antibiotic/antimycotic (1%). The medium was refreshed
every other day, and then, confluent cells were detached by a trypsin-
EDTA solution (0.1%) and counted in a hemocytometer under an optical
microscope to calculate the cell density. Then, cell suspension (1 mL, 1.0
× 104 cells mL−1) was seeded on the surface of sterile samples. After
culturing at 37 °C in a humidified atmosphere of CO2 (5%) for 1, 3 and
5 days, the samples were washed twice with PBS and fixed with glutar-
aldehyde (2.5%) for 2 h. Subsequently, the samples were dehydrated
through a graded alcohol solution (30, 50, 70, 80, 90 and 100% v/v)
followed by critical point drying (CPD), and the cells attached to the
composite surface were observed with scanning electron microscopy
(SEM; JEOL JSM-6700F).
The cell proliferation was quantified using the CCK-8 (2-(2-
methoxy-4-phenyl)-3-(4-phenyl)-5-(2,4-sulfo benzene)-2h-tetrazoli-
um monosodium salt, Shanghai Yisheng Bio-Technology Co., Ltd)
assay. Briefly, the samples (pure HA, in situ synthesized CNT/HA,
CNT@HA and CNT@meso-HA) were placed into a 96-well plate and
seeded with a density of 10,000 cells (MC3T3-E1) per well for 1, 3 and
5 days. After incubation, the supernatant was discarded and fresh medi-
um (100 μL) containing a sterile CCK-8 solution (10 μL) was added to
each well and kept for a further 4 h. Then, the plate was shaken for
5 min and the absorbance of the colored formazan products was mea-
sured at 450 nm with a microplate reader (Tecan ECOM-F6124, Aus-
tria). Only the culture medium was added to the blank group, whereas
the cells, culture medium and pure HA sample were added to the con-
trol group. The data were averaged from three parallel experiments,
and the values were represented as the average ± standard deviation
(SD).
2.5. Characterizations
The phase composition of the CNT/HA composite powders was ana-
lyzed by X-ray diffraction (XRD; Rigaku D/Max-2500). The morphology
and microstructure of the composite powders and bulk composites
were investigated by field emission scanning electron microscopy (FE-
SEM; JEOL JSM-6700F) and high-resolution transmission electron mi-
croscopy (HRTEM; Philips Tecnai G2 F20). The pore size of the deposited
meso-HA on CNTs was measured by a nitrogen adsorption-desorption
isotherm (BET and BJH). Qualitative analysis of functional groups was
accomplished by Fourier-transform infrared spectrometry (FTIR).
Raman spectroscopy of the CNTs was measured on a Thermo Fisher
DXR Raman microscope equipped with an argon ion laser with a
1080 H. Li et al. / Materials Science and Engineering C 77 (2017) 1078–1087

Fig. 1. Schematic illustration of the fabrication process of CNT@meso-HA composite powders. a) In situ synthesized CNTs on HA nanoparticles by CVD. b) Oxidation treatment of CNTs with
H2O2. c) Modification of CNTs with Ca2+. d, e) Formation of spherical micelles of CTAB. f) Formation of CTAB-PO43− micelles by electrostatic attraction. g) Formation of meso-HA precursor
micelles on CNTs by in situ chemical precipitation. h, i) Formation of meso-HA particles on CNTs.

1064 nm line as the excitation source. The flexural strength of the bulk aqueous solution of calcium nitrate (Ca(NO3)2·4H2O). The CTAB (Fig.
samples was determined through three-point bending tests in accor- 1d), as a soft template agent, can form CTAB micelles in aqueous solu-
dance with the ASTM C1161-13 standard. For each material, three rect- tion by self-assembly (Fig. 1e). When an aqueous solution of
angular specimens (3 × 4 × 45 mm3) were tested on a Shimadzu diammonium phosphate ((NH4)2HPO4) was mixed with CTAB micelles,
Universal Testing Machine model AG-50kNG (M/s Shimadzu, Kyoto, CTAB-PO3− 4 micelles were formed (Fig. 1f). When this solution (Fig. 1f)
Japan) at room temperature with a 40 mm outer span three-point fix- was mixed with the previous mixture solution (Fig. 1c), precursor ions
ture and a crosshead speed of 0.50 mm min−1. The fracture toughness (CTAB-PO34 − micelles) loaded onto CNTs and reconstructed in situ to
(KIC) was assessed by the indentation method (following ASTM C1327 form HA via the combination of PO34 − ions with Ca2 + ions (Fig. 1g).
standard), using a TI-Premier Nanoindenter (Hysitron, America) with After drying, the precipitates, amorphous mesoporous HA precursor
different indentation loads (0.1, 0.3, 0.5 and 1 kgf) and a dwell time of
15 s. KIC was calculated by Eq. (2) [28].
 0:5  
E P
K IC ¼ 0:016 1:5
ð2Þ
H C

Here, E is the elasticity modulus, H is the Vickers hardness, P is the

indentation load and C is the half-radial crack length. Six measurements
were conducted for each specimen to obtain the average value of frac-
ture toughness.

3. Results and discussion

3.1. Design principles

Fig. 1 shows a schematic illustration of the fabrication process of the

CNT@meso-HA composite powders by an in situ chemical precipitation
method. The synthesized CNTs (Fig. 1a) were treated with H2O2 to cre-
ate hydrophilic functional groups, such as \\OH, \\COOH and \\C_O
(Fig. 1b). These anionic groups lead to CNTs repulsion, maintain their
dispersion in the solution, and also attract positively charged Ca2 + Fig. 2. XRD spectra of a) in situ synthesized CNT/HA, b) (3 wt%)CNT@HA and c) (3
ions to attach on the surface of CNTs (Fig. 1c) when mixed with an wt%)CNT@meso-HA composite powders.
 H. Li et al. / Materials Science and Engineering C 77 (2017) 1078–1087 1081

micelles encapsulating CNTs, were obtained (Fig. 1h). Subsequently, the
amorphous precipitates were calcined in an argon atmosphere to re-
move the CTAB template, and crystalline CNT@meso-HA composite
powders were obtained (Fig. 1i).
3.2. Synthesis and characterizations of CNT@meso-HA composite powders
Fig. 2 shows the XRD patterns of in situ synthesized CNT/HA, (3
wt%)CNT@HA and (3 wt%)CNT@meso-HA composite powders. For in
situ synthesized CNT/HA (Fig. 2a), the diffraction peaks at 2θ = 26.4°
and 44.6° can be associated with the graphite (002) and (013) planes,
respectively. The characteristic peak at 2θ = 44.6° and 49.5° can be as-
sociated with the Fe (110) and (101) planes, respectively. The other
peaks can be indexed to the HA phase based on their position and rela-
tive intensity [16,29]. After modifying CNTs with non-mesoporous HA
(Fig. 2b) and mesoporous HA (Fig. 2c), there is no significant difference
in the XRD patterns, suggesting that the HA deposited on the surface of
CNTs has high crystallization and the surfactant (CTAB) has no obvious
influence on the crystallization of the deposited HA.
Fig. 3 shows the SEM and TEM images of the in situ synthesized CNT/
HA, CNT@HA and CNT@meso-HA composite powders. As shown in Fig.
3a, the CNTs are synthesized and distributed uniformly within HA
nanopowders by the CVD process, forming an initial molecular-level
mixture between CNTs and HA. This molecular-level contact arising
from the in situ synthesis process improves the wettability between
CNTs and HA and forms an initial interfacial bonding [27]. Even after un-
dergoing a strong ultrasonic treatment during the preparation of the
TEM sample, there are still several HA nanoclusters attached to the
CNTs, as shown in Fig. 3b. Meanwhile, it can be observed that the in
situ synthesized CNTs possess the characteristic hollow and tubular
graphite structure. The interlayer spacing of the sheets is 0.341 nm, sim-
ilar to the ideal (002) interlayer spacing (0.34 nm) of graphite, indicat-
ing the high degree of graphitization of CNTs. However, these
uncovered CNTs may be toxic or pose other potential dangers to the

Fig. 3. SEM and TEM images of CNT/HA composite powders. a, b) In situ synthesized CNT/HA. c–e) (3 wt%)CNT@meso-HA. f–h) (3 wt%)CNT@HA.
1082 H. Li et al. / Materials Science and Engineering C 77 (2017) 1078–1087

human body [21,22]. Further functional modifications are necessary be-

fore their application in biomedical materials.
Fig. 3c–e display the SEM and TEM images of typical (3 wt%)CNT@
meso-HA composite powders. Due to the encapsulation of the deposited
HA, most of the CNTs are embedded into HA powders and hard to be ob-
served. The visible CNT in Fig. 3c shows the obviously increased diame-
ter (approximately 100 nm) compared with those in Fig. 3a
(approximately 60 nm). The typical TEM image (Fig. 3d) of CNT@
meso-HA composite powders reveals that the rod-shaped HA particles
wrap CNTs tightly, with dimensions of approximately 20 nm in width
and 50–80 nm in length, similar to the apatite crystals in living bone
[30]. The EDX analysis result (inset in Fig. 3d) shows that the main
chemical elements include Ca, P, O, C and so on and the average Ca/P
molar ratio is 1.66, close to the stoichiometric value (1.67) of HA pre-
senting in bone tissue [31]. Furthermore, the selected area electron dif-
fraction (SAED) pattern (inset in Fig. 3d) shows three strong concentric
rings, which can be indexed as the (002), (110) and (310) planes of HA
with high crystallinity [32]. The diffraction arc corresponding to the
(002) plane indicates that the crystal growth of HA is along the c-axis di-
rection, which is consistent with the results of the TEM observation. Fig. 4. FTIR spectra of (a) in situ synthesized CNT/HA, (b) oxidatively treated CNT/HA, (c)
(3 wt%)CNT@meso-HA and (d) (3 wt%)CNT@HA composite powders.
Some clear lattice fringes of HA can be observed in Fig. 3e, with d-spac-
ing values of 0.82 and 0.32 nm, corresponding to the (100) and (102)
planes of hexagonal HA, respectively. Interestingly, the HRTEM image of CNT@HA and CNT@meso-HA have no obvious change (Fig. 4c and d),
in Fig. 3e reveals that these rod-shaped HA particles possess many except that two characteristic absorption bands of υas (C\\H) at
mesoporous structures (indicated with arrows in the inset of Fig. 3e), 2920 cm−1 and υa (C\\H) at 2848 cm−1 assigned to the CTAB template
with an average aperture of 6–10 nm. More particularly, after removing can be observed in Fig. 4c of CNT@meso-HA.
the CTAB template, these remaining mesopores disperse homogeneous- Fig. 5 shows the Raman spectra of in situ synthesized CNT/HA, (3
ly in the deposited HA on CNTs and endow the CNTs with a highly po- wt%)CNT@meso-HA and (3 wt%)CNT@HA composite powders. All sam-
rous and rough HA coating layer, which will improve the bioactivity of ples display two typical Raman bands associated with CNTs at
CNT/HA composites and the mechanical fixation of the living tissues ~ 1348 cm−1 (D band) and ~ 1592 cm− 1 (G band). The calculated
on HA composites. In addition, different from the strong acid treatment, value of the ID/IG ratio for in situ synthesized CNT/HA is 0.83, indicating
weak oxidative functionalization of H2O2 maintains the perfect struc- a high degree of graphitization of the CNTs. After modification by HA,
ture of CNTs, as well as introduces functional groups (\\OH, \\COOH the values of the ID/IG ratio for CNT@meso-HA and CNT@HA are approx-
and \\C_O) onto the defect sites of CNTs with high surface energy imately 0.85, suggesting that the CNTs maintain their well-crystallized
and reactivity. At these active sites with functional groups, Ca2+ ions graphite structure after modification with HA.
and CTAB-PO3− 4 micelles are guided onto the surface of CNTs in an or- Fig. 6 and Table 1 show the N2 adsorption-desorption isotherms,
derly fashion by electrostatic attraction, inducing oppositely charged pore size distribution and textural parameters of in situ synthesized
ions to nucleate and grow into HA particles. As a result, HA particles CNT/HA, CNT@HA and CNT@meso-HA composite powders, respective-
are successfully grafted onto CNTs and form a tight combination with ly. In Fig. 6A, the irreversibility between the adsorption and desorption
them. As shown in the inset of Fig. 3e, no microscale cavity and delam- processes leads to the formation of hysteresis loops of these isotherms
ination appears at the interface, ensuring a tight tube-matrix interface [35]. As can been seen from Fig. 6A(a, b), the N2 adsorption-desorption
and good adherence between the CNTs and HA, which will contribute isotherms of both in situ synthesized CNT/HA and CNT@HA exhibit a
to improving the mechanical properties of the composites and prevent faint hysteresis loop, implying that they barely have mesoporous struc-
the CNTs from stripping from the matrix and contacting human tissue. tures. Fig. 6B(a, b) shows that the two samples have certain weak peaks
For comparison, samples of CNT@HA were also prepared with the
same process used for CNT@meso-HA, except that no CTAB template
was applied, that is, only PO3−4 ions were introduced to the CNTs and
reacted with Ca2+. SEM and TEM images of the typical (3 wt%)CNT@
HA composite powders are shown in Fig. 3f–h. Compared with CNT@
meso-HA, the obvious difference is that the synthesized HA layer out-
side CNTs exhibits no mesoporous structure.
Fig. 4 shows the FTIR spectra of the in situ synthesized CNT/HA, ox-
idatively treated CNT/HA, (3 wt%)CNT@meso-HA and (3 wt%)CNT@HA
composite powders. The in situ synthesized CNT/HA sample (Fig. 4a)
displays the characteristic absorption peaks of natural HA, that is,
bands at 563 cm−1 and 605 cm−1 attributed to υ4 asymmetrical P\\O
stretching in PO3−4 and those at 1038 cm−1 and 1085 cm−1 associated
with the υ3 asymmetrical O\\P\\O of the phosphate ions of the hydrox-
yl site [33]. The peak at 1650 cm−1 is assigned to C_C stretching, which
originates from the inherent structure of CNTs. This observation con-
firms the presence of CNTs and HA in the in situ synthesized CNT/HA
composite powders. After oxidative functionalization with H2O2, C_O
and C\\O stretching exhibits characteristic peaks at 1736 cm− 1 and
1275 cm−1 (Fig. 4b), respectively, indicating the successful introduction
of hydrophilic functional groups (\\C_O and \\COOH) on the surface Fig. 5. Raman spectra of (a) in situ synthesized CNT/HA, (b) (3 wt%)CNT@meso-HA and (c)
of CNTs [34]. After the deposition of HA on CNTs, the functional groups (3 wt%)CNT@HA composite powders.
 H. Li et al. / Materials Science and Engineering C 77 (2017) 1078–1087
Fig. 6. (A) N2 adsorption-desorption isotherms and (B) pore size distributions of (a) in situ synthesized CNT/HA, (b) (3 wt%)CNT@HA and (c) (3 wt%)CNT@meso-HA composite powders.
Table 1
The textural parameters of the samples obtained by the nitrogen adsorption-desorption
method.
Samples SBET [m2 g−1] Vp [cm3 g−1] Dp [nm]
In situ synthesized CNT/HA 65.4 0.36 2.56
(3 wt%)CNT@HA 64.2 0.32 2.85
(3 wt%)CNT@meso-HA 131.7 0.67 6.21
that are attributed to the clearance (2.56 and 2.85 nm) of HA particles.
They also have a smaller specific surface area (SBET) and pore volume
(Vp), as shown in Table 1. The isotherm of CNT@meso-HA, as shown
in Fig. 6A(c), exhibits a type IV curve with a distinct hysteresis loop ac-
cording to IUPAC classification, indicating that the sample is a typical
mesoporous material [36]. A broad peak ranging from 2 to 10 nm and
centered at approximately 6 nm can be seen in Fig. 6B(c), and most
mesoporous HA samples have pore sizes (Dp) of approximately
6.21 nm, as shown in Table 1, consistent with the TEM observations.
The large SBET (131.7 m2 g−1) and Vp (0.67 cm3 g−1) values also demon-
strate the successful acquisition of mesoporous structures in CNT@
meso-HA, which exert a particular effect on its bioactivity.
3.3. Mechanical properties of CNT@meso-HA composites
Fig. 7 shows the flexural strength and fracture toughness of the three
HA composites composed of in situ synthesized CNT/HA, CNT@HA and
CNT@meso-HA composite powders. It is found that the flexural strength
and fracture toughness of the composites present a similar variation
trend; they increase at first and achieve a maximum with the addition
of 3 wt% CNTs. The flexural strength (79.0 MPa) and fracture toughness
Fig. 7. a) Flexural strength and b) fracture toughness of three HA composites.
1083
(2.37 MPa m1/2) of the CNT@meso-HA composite are improved by ap-
proximately 52% and 216% compared with pure HA, respectively, and
significantly outperform those of in situ synthesized CNT/HA compos-
ites. Though lower than those of CNT@HA composites, the performances
of CNT@meso-HA composites are still in the range of the mechanical
properties of cortical bone (flexural strength: 50–150 MPa, fracture
toughness: 2–12 MPa m1/2) and can meet the requirements for implant
materials [37]. Moreover, the moderate mechanical properties of CNT@
meso-HA composites can achieve successful matching with the sur-
rounding soft tissues [16,38,39]. In addition, CNT@meso-HA composites
containing mesoporous structures are conducive to the circulation of
body fluids and the mechanical fixation of living tissues [16,40–42],
thus possessing higher bone-forming bioactivity.
The obvious improvement of the mechanical properties of CNT@
meso-HA composites is due to the effective load transfer at the interface
of CNT-HA and the reinforcement roles of CNTs to the HA matrix. As
mentioned above, mesoporous HA is synthesized in situ on the surface
of CNTs, forms a strong chemical bonding with CNTs and maintains
the perfect structure of the CNTs. This endows the CNT-HA interface
with a high interfacial bonding strength and an efficient load transfer
capacity. During the fracture process, a load can be transferred effective-
ly from the HA matrix to CNT reinforcements. Thus, more fracture ener-
gy is consumed by the pulled-out CNTs (Fig. 8a) and bridged CNTs (Fig.
8b), resulting in the improvement of the mechanical properties. Fur-
thermore, compared with traditional CNT/HA composites [20,43], the
mesoporous structures of CNT@meso-HA composites are likely to gen-
erate micro crack initiations, but they can change the orientation of
crack propagation to some extent, consequently improving the fracture
toughness. The tight bonding interface between CNTs and HA was fur-
ther confirmed by TEM analysis. As shown in Fig. 8c, CNTs are well-
1084 H. Li et al. / Materials Science and Engineering C 77 (2017) 1078–1087

Fig. 8. (a, b) SEM images of a fracture surface and (c, d) TEM images of the interface of the (3 wt%)CNT@meso-HA bulk composite.

dispersed in the bulk composite and maintain their intact structure, crystalline mesoporous HA. The diffraction arcs can be ascribed to the
which is derived from the in situ synthesis and functionalization of (110), (002), (301), (213) and (420) planes of HA. Some mesoporous
CNTs. The SAED in the inset of Fig. 8c confirms the existence of highly HA particles are closely attached to CNTs (Fig. 8d), leading to strong

Fig. 9. SEM images of osteoblast cells on a) pure HA, b) in situ synthesized CNT/HA, c) CNT@HA and d) CNT@meso-HA composites after 3 days of culture.
 H. Li et al. / Materials Science and Engineering C 77 (2017) 1078–1087
interfacial bonding between CNTs and HA. Consequently, the load trans-
fer efficiency of the CNT-HA interface can be increased significantly,
which improves the integrative mechanical properties of HA
composites.
3.4. Biological properties of CNT@meso-HA composites
When biomaterials are implanted into the body, successful incorpo-
ration of the implant with the body depends on tissue integration by the
adhesion of the cells to the implant surface and the subsequent adsorp-
tion of certain proteins on the implant surface, which will mediate cell
attachment and biological responses [44,45]. Generally, osteoblasts at-
tach to the implant surface first and then deposit crystalline salts on
the collagen matrix, leading to mineralization and bone formation. HA
nanomaterials can enhance the formation of new bone tissue by in-
creasing osteoblast adhesion and bone bonding on their surface [46–
50]. Thus, the adhesion of osteoblasts to the implant surface is extreme-
ly important for their proliferation and growth, as well as the formation
of mineral deposits. To assess the potential osteo-application of CNT@
meso-HA composites, in vitro cell culture tests were performed. Fig. 9
shows SEM images of the osteoblast cells (MC3T3-E1) on pure HA, in
situ synthesized CNT/HA, CNT@HA and CNT@meso-HA composites
after 3 days of culture. The cells on pure HA substrate have an extended
and flattened morphology (Fig. 9a), which suggests that pure HA is non-
toxic and has good cytocompatibility. However, for the in situ synthe-
sized CNT/HA sample (Fig. 9b), the osteoblasts decrease significantly
and exhibit a relatively round morphology, implying that uncoated
CNTs have some negative effects on osteoblast attachment, spreading
and proliferation. In contrast, as shown in Fig. 9c, when exposed CNTs
were encapsulated by a HA layer, the number of osteoblasts on CNT@
HA was significantly increased. Many filopodia (marked with arrows
in the inset of Fig. 9c) cling to the substrate, indicating that the toxicity
of CNTs has been greatly eliminated. Superior to the three former sam-
ples, as shown in Fig. 9d, the number of anchored cells on the CNT@
meso-HA sample increases markedly. The cells connect to each other
and almost cover the entire substrate surface, presenting even higher
cellular activity than those on pure HA and CNT@HA. A high magnifica-
tion view (inset of Fig. 9d) of the osteoblast cells demonstrates the an-
choring of long filopodia on the CNT@meso-HA sample. The cells with
large and thin cytoplasmic features attach firmly to the substrate and
develop a flattened and elongated morphology. Additionally, numerous
filopodia stretch out from the cell body and enable the cells to tightly
adhere to the surface of the composite. Consequently, CNT@meso-HA
composites are more conducive to promoting the growth and adhesion
of osteoblasts.
The viability of osteoblast cells (MC3T3-E1) on samples of pure HA
(control group), in situ synthesized CNT/HA, CNT@HA and CNT@
meso-HA composites was further evaluated using a CCK-8 assay. As
shown in Fig. 10, the cell viability on the in situ synthesized CNT/HA
sample decreased with extended culture time, further confirming that
this sample has cytotoxicity due to the presence of exposed CNTs. In
contrast, for the CNT@HA and CNT@meso-HA samples, MC3T3-E1 cells
increase with increasing culture time and exhibit higher cell viability
after 3 and 5 days of incubation, inferring that these samples are non-
toxic. Interestingly, after 5 days of culture, cell viability on CNT@meso-
HA is even higher than that on pure HA and CNT@HA, implying that
CNT@meso-HA composites are more conducive to accelerating cell pro-
liferation, which may be ascribed to their unique mesoporous structure.
It has been reported that mesoporous HA has better biocompatibility
and bioactivity than smooth HA and the rough surfaces can enhance os-
teoblastic attachment, growth and proliferation [51–55]. In addition,
due to their increased specific surface area and surface wettability,
CNT@meso-HA composites can provide more sites for cell adhesion
and proliferation, endowing the composites with excellent bioactivity
and biocompatibility. The reasons for the better adhesion and prolifera-
tion of osteoblasts on CNT@meso-HA composites compared with those
1085
Fig. 10. Viability of osteoblast cells on pure HA, in situ synthesized CNT/HA, CNT@HA and
CNT@meso-HA composites after 1, 3 and 5 days of culture.
on pure HA bulk and the other two CNT/HA composites are related to
the mechanism of cell attachment. The cell adhesion process of
MC3T3-E1 on the substrate involves the identification and connection
of integrin in the cell membrane to the RGD (Arginine-glycine-aspar-
tate) site of proteins adsorbed from the culture medium. Thus, the
cell-substrate interactions are mediated by the proteins adsorbed on
the substrate. Furthermore, adsorption of proteins on the substrate
will be affected by the interactions among proteins, water and the inter-
face. Mesoporous HA has a high specific surface area and good hydro-
philicity, so the CNT@meso-HA substrate can accommodate an
increased amount of cell suspension because of the capillary effect and
the formation of a water molecular layer. Then, the water molecule
layer combines with proteins in the culture medium by hydrogen bond-
ing. In addition, the formation of the water molecular layer also helps to
protect the conformation of the adsorbed proteins. The integrity of the
protein molecular conformation will ensure the integrity of its function
[56], which can provide a suitable surface for the growth of MC3T3-E1
cells on the CNT@meso-HA substrate. Zhang et al. reported that
nanomaterials with good cell-substrate characteristics promoted a
large number of protein interactions and adhesion, thus stimulating
bone growth [57]. Cell spreading on the biomaterial is an important
step for the essential biological properties of the osteoblasts. Once the
adhesion to sample substrate occurs, cells probe the surroundings and
move using filopodia, followed by the establishment of focal adhesion
points [58]. Especially for the CNT@meso-HA sample, MC3T3-E1 cells
are cross-linked by climbing and stretching and show spindle shaped
extensions and more filopodia, indicating the formation of successful
cell-substrate interactions. Therefore, the CNT@meso-HA composites
with a high specific surface area and good hydrophilicity that facilitate
nutrient exchange between the cells and the surrounding environment,
especially the adsorption of proteins, are favorable for cell adhesion and
further promote cell proliferation, growth and bone formation, achiev-
ing excellent cell viability.
4. Conclusion
CNT@meso-HA composites with excellent mechanical and biological
properties were successfully fabricated by the in situ chemical deposi-
tion of mesoporous HA on homogeneously dispersed CNTs using CTAB
as a template. The hydrophilic functional groups (\\OH, \\COOH and
\\C_O) were introduced onto the surface of well-crystallized CNTs
by an oxidation treatment with H2O2. By electrostatic attraction, Ca2+
ions and CTAB-PO34 − micelles were guided onto the surface of CNTs
and reconstructed in situ to form rod-shaped mesoporous HA particles
with the help of CTAB, which resulted in chemical bonding and a tight
1086 H. Li et al. / Materials Science and Engineering C 77 (2017) 1078–1087
tube-matrix interface between the CNTs and HA. The CNT@meso-HA
composites exhibited enhanced flexural strength and fracture tough-
ness of approximately 52% and 216% compared with pure HA, respec-
tively. More importantly, the structural features of the CNT@meso-HA
composites endowed the material with good biocompatibility and in-
creased the viability of osteoblast cells (MC3T3-E1) compared with
even pure HA. Combined with their excellent mechanical property,
the CNT@meso-HA composites are greatly promising for the applica-
tions in bone regeneration and replacement, as well as in the formation
of highly active tissue scaffolds.
Acknowledgments
This work was supported by Natural Science Foundation of Hebei
Province of China (E2015202037), Tianjin Research Program of Applica-
tion Foundation and Advanced Technology (14JCYBJC20900,
14JCYBJC19600), Science and Technology Correspondent Project of
Tianjin (14JCTPJC00496), and Program for Changjiang Scholars and In-
novative Research Team in University of Ministry of Education of China
(IRT13084).
References
[1] T. Wu, L. Tan, N. Cheng, Q. Yan, Y.F. Zhang, C.J. Liu, B. Shi, PNIPAAM modified meso-
porous hydroxyapatite for sustained osteogenic drug release and promoting cell at-
tachment, Mater. Sci. Eng. C 62 (2016) 888–896.
[2] B. Palazzo, M. Iafisco, M. Laforgia, N. Margiotta, G. Natile, C.L. Bianchi, D. Walsh, S.
Mann, N. Roveri, Biomimetic hydroxyapatite-drug nanocrystals as potential bone
substitutes with antitumor drug delivery properties, Adv. Funct. Mater. 17 (2007)
2180–2188.
[3] S.P. Uswatta, I.U. Okeke, A.C. Jayasuriya, Injectable porous nano-hydroxyapatite/chi-
tosan/tripolyphosphate scaffolds with improved compressive strength for bone re-
generation, Mater. Sci. Eng. C 69 (2016) 505–512.
[4] W. Suchanek, M. Yoshimura, Processing and properties of hydroxyapatite-based
biomaterials for use as hard tissue replacement implants, J. Mater. Res. 13 (1998)
94–117.
[5] N. Ignjatović, V. Wu, Z. Ajduković, T. Mihajilov-Krstev, V. Uskoković, D. Uskoković,
Chitosan-PLGA polymer blends as coatings for hydroxyapatite nanoparticles and
their effect on antimicrobial properties, osteoconductivity and regeneration of osse-
ous tissues, Mater. Sci. Eng. C 60 (2016) 357–364.
[6] L.G. Yu, K.A. Khor, H. Li, P. Cheang, Effect of spark plasma sintering on the micro-
structure and in vitro behavior of plasma sprayed HA coatings, Biomaterials 24
(2003) 2695–2705.
[7] J.G. Liao, Y.Q. Li, Q. Zou, X.Z. Duan, Z.P. Yang, Y.F. Xie, H.H. Liu, Preparation, character-
ization and properties of nano-hydroxyapatite/polypropylene carbonate
biocomposite, Mater. Sci. Eng. C 63 (2016) 285–291.
[8] R. Cholas, S.K. Padmanabhan, F. Gervaso, G. Udayan, G. Monaco, A. Sannino, A.
Licciulli, Scaffolds for bone regeneration made of hydroxyapatite microspheres in
a collagen matrix, Mater. Sci. Eng. C 63 (2016) 499–505.
[9] J.A. Delgado, L. Morejón, S. Martínez, M.P. Ginebra, N. Carlsson, E. Fernández, J.A.
Planell, M.T. Clavaguera-Mora, J. Rodríguez-Viejo, Zirconia-toughened hydroxyapa-
tite ceramic obtained by wet sintering, J. Mater. Sci. 10 (1999) 715–719.
[10] X. Zhang, G.H.M. Gubbels, R.A. Terpstra, R. Metselaar, Toughening of calcium hy-
droxyapatite with silver particles, J. Mater. Sci. 32 (1997) 235–243.
[11] M. Zakeri, E. Hasani, M. Tamizifar, Mechanical properties of TiO2-hydroxyapatite
nanostructured coatings on Ti-6Al-4V substrates by APS method, Int. J. Miner.
Metall. Mater. 20 (2013) 397–402.
[12] M.A. Lopes, F.J. Monteiro, J.D. Santos, Glass-reinforced hydroxyapatite composites:
fracture toughness and hardness dependence on microstructural characteristics,
Biomaterials 20 (1999) 2085–2090.
[13] W.P.S.L. Wijesinghe, M.M.M.G.P.G. Mantilaka, K.G.C. Senarathna, H.M.T.U. Herath,
T.N. Premachandra, C.S.K. Ranasinghe, R.P.V.J. Rajapakse, R.M.G. Rajapakse, M.
Edirisinghe, S. Mahalingam, I.M.C.C.D. Bandara, S. Singh, Preparation of bone-im-
plants by coating hydroxyapatite nanoparticles on self-formed titanium dioxide
thin-layers on titanium metal surfaces, Mater. Sci. Eng. C 63 (2016) 172–184.
[14] A. Peigney, Composite materials: tougher ceramics with nanotubes, Nat. Mater. 2
(2003) 15–16.
[15] F.M. Blighe, K. Young, J.J. Vilatela, A.H. Windle, I.A. Kinloch, L. Deng, R.J. Young, J.N.
Coleman, The effect of nanotube content and orientation on the mechanical proper-
ties of polymer-nanotube composite fibers: separating intrinsic reinforcement from
orientational effects, Adv. Funct. Mater. 21 (2011) 364–371.
[16] Y. Chen, Y.Q. Zhang, T.H. Zhang, C.H. Gan, C.Y. Zheng, G. Yu, Carbon nanotube rein-
forced hydroxyapatite composite coatings produced through laser surface alloying,
Carbon 44 (2006) 37–45.
[17] K. Yang, J. He, Z. Su, J.B. Reppert, M.J. Skove, T.M. Tritt, A.M. Rao, Inter-tube bonding,
graphene formation and anisotropic transport properties in spark plasma sintered
multi-wall carbon nanotube arrays, Carbon 48 (2010) 756–762.
[18] K.N. Sun, A.M. Li, Y.S. Yin, Z.Y. Yu, Study on the fabrication of hydroxyapatite and car-
bon nanotube composites, Chin. J. Biomed. Eng. 23 (2004) 573–578.
[19] R. Rastogi, R. Kaushal, S.K. Tripathi, A.L. Sharma, I. Kaur, L.M. Bharadwaj, Compara-
tive study of carbon nanotube dispersion using surfactants, J. Colloid Interface Sci.
328 (2008) 421–428.
[20] S. Mukherjee, B. Kundu, S. Sen, A. Chanda, Improved properties of hydroxyapatite-
carbon nanotube biocomposite: mechanical, in vitro bioactivity and biological stud-
ies, Ceram. Int. 40 (2014) 5635–5643.
[21] L.H. Ding, J. Stilwell, T.T. Zhang, O. Elboudwarej, H.J. Jiang, J.P. Selegue, P.A. Cooke,
J.W. Gray, F.Q.F. Chen, Molecular characterization of the cytotoxic mechanism of
multiwall carbon nanotubes and nano-onions on human skin fibroblast, Nano
Lett. 5 (2005) 2448–2464.
[22] C.W. Lam, J.T. James, R. McCluskey, R.L. Hunter, Pulmonary toxicity of single-wall
carbon nanotubes in mice 7 and 90 days after intratracheal instillation, Toxicol.
Sci. 77 (2004) 126–134.
[23] X.Y. Lu, T. Qiu, X.F. Wang, M. Zhang, X.L. Gao, R.X. Li, X. Lu, J. Weng, Preparation of
foam-like carbon nanotubes/hydroxyapatite composite scaffolds with
superparamagnetic properties, Appl. Surf. Sci. 262 (2012) 227–230.
[24] J. Venkatesan, Z.J. Qian, B. Ryu, N.A. Kumar, S.K. Kim, Preparation and characteriza-
tion of carbon nanotube-grafted-chitosan-natural hydroxyapatite composite for
bone tissue engineering, Carbohydr. Polym. 83 (2011) 569–577.
[25] F.M. Zhang, A. Weidmann, J.B. Nebe, E. Burkel, Osteoblast cell response to surface-
modified carbon nanotubes, Mater. Sci. Eng. C 32 (2012) 1057–1061.
[26] H.P. Li, Q.Y. Zhao, B.E. Li, J.L. Kang, Z.Y. Yu, Y.X. Li, X.Q. Song, C.Y. Liang, H.S. Wang,
Fabrication and properties of carbon nanotube-reinforced hydroxyapatite compos-
ites by a double in situ synthesis process, Carbon 101 (2016) 159–167.
[27] H.P. Li, N.Q. Zhao, Y. Liu, C.Y. Liang, C.S. Shi, X.W. Du, J.J. Li, Fabrication and properties
of carbon nanotubes reinforced Fe/hydroxyapatite composites by in situ chemical
vapor deposition, Compos. Part A 39 (2008) 1128–1132.
[28] G.R. Anstis, P. Chantikul, B.R. Lawn, D.B. Marshall, A critical evaluation of indentation
technique for measuring fracture toughness: I, direct crack method, J. Am. Ceram.
Soc. 64 (1981) 533–538.
[29] A.C. Tas, Synthesis of biomimetic Ca-hydroxyapatite powders at 37 degrees C in syn-
thetic body fluids, Biomaterials 21 (2000) 1429–1438.
[30] H.L. Zhang, Z.Q. Chen, Fabrication and characterization of electrospun PLGA/
MWNTs/hydroxyapatite biocomposite scaffolds for bone tissue engineering, J.
Bioact. Compat. Polym. 25 (2010) 241–259.
[31] A.O. Lobo, M.A.F. Corat, S.C. Ramos, J.T. Matsushima, A.E.C. Granato, C. Pacheco-
Soares, E.J. Corat, Fast preparation of hydroxyapatite/superhydrophilic vertically
aligned multiwalled carbon nanotube composites for bioactive application, Lang-
muir 26 (2010) 18308–18314.
[32] C.M. Zhang, J. Yang, Z.W. Quan, P.P. Yang, C.X. Li, Z.Y. Hou, J. Lin, Hydroxyapatite
nano- and microcrystals with multiform morphologies: controllable synthesis and
luminescence properties, Cryst. Growth Des. 9 (2009) 2725–2733.
[33] Y.P. Guo, Y.B. Yao, Y.J. Guo, C.Q. Ning, Hydrothermal fabrication of mesoporous car-
bonated hydroxyapatite microspheres for a drug delivery system, Microporous
Mesoporous Mater. 155 (2012) 245–251.
[34] G.M. Neelgund, K. Olurode, Z.P. Luo, A. Oki, A simple and rapid method to graft hy-
droxyapatite on carbon nanotubes, Mater. Sci. Eng. C 31 (2011) 1477–1481.
[35] G. Verma, K.C. Barick, N. Manoj, A.K. Sahu, P.A. Hassan, Rod-like micelle templated
synthesis of porous hydroxyapatite, Ceram. Int. 39 (2013) 8995–9002.
[36] M. Kruk, M. Jaroniec, Gas adsorption characterization of ordered organic-inorganic
nanocomposite materials, Chem. Mater. 13 (2001) 3169–3183.
[37] L.L. Hench, An Introduction to Bioceramics, second ed. Imperial College Press, Lon-
don, 2013.
[38] L.C. Gerhardt, A.R. Boccaccini, Bioactive glass and glass-ceramic scaffolds for bone
tissue engineering, Materials 3 (2010) 3867–3910.
[39] S.K. Yadav, T. Bera, P.S. Saxena, A.K. Maurya, R.S. Garbyal, R. Vajtai, P.
Ramachandrarao, A. Srivastava, MWCNTs as reinforcing agent to the Hap-Gel
nanocomposite for artificial bone grafting, J. Biomed. Mater. Res. A 93A
(2010) 886–896.
[40] Y.F. Zhu, S. Kaskel, Comparison of the in vitro bioactivity and drug release property
of mesoporous bioactive glasses (MBGs) and bioactive glasses (BGs) scaffolds, Mi-
croporous Mesoporous Mater. 118 (2009) 176–182.
[41] Y.K. Jun, W.H. Kim, O.K. Kweon, S.H. Hong, The fabrication and biochemical evalua-
tion of alumina reinforced calcium phosphate porous implants, Biomaterials 24
(2003) 3731–3739.
[42] Y.P. Guo, Y. Zhou, D.C. Ha, Fabrication of hydroxycarbonate apatite coatings with hi-
erarchically porous structures, Acta Biomater. 4 (2008) 334–342.
[43] D.Y. Kim, Y.H. Han, J.H. Lee, I.K. Kang, B.K. Jang, S. Kim, Characterization of
multiwalled carbon nanotube-reinforced hydroxyapatite composites consolidated
by spark plasma sintering, Biomed. Res. Int. 2014 (2014) 768254.
[44] W. Pon-On, P. Suntornsaratoon, N. Charoenphandhu, J. Thongbunchoo, N.
Krishnamra, I.M. Tang, Hydroxyapatite from fish scale for potential use as bone scaf-
fold or regenerative material, Mater. Sci. Eng. C 62 (2016) 183–189.
[45] L.K. Randeniya, A. Bendavid, P.J. Martin, M.S. Amin, E.W. Preston, F.S.M. Ismail, S.
Coe, Incorporation of Si and SiOx into diamond-like carbon films: impact on surface
properties and osteoblast adhesion, Acta Biomater. 5 (2009) 1791–1797.
[46] S. Sinlapabodin, P. Amornsudthiwat, S. Damrongsakkul, S. Kanokpanont, An axial
distribution of seeding, proliferation, and osteogenic differentiation of MC3T3-E1
cells and rat bone marrow-derived mesenchymal stem cells across a 3D Thai silk fi-
broin/gelatin/hydroxyapatite scaffold in a perfusion bioreactor, Mater. Sci. Eng. C 58
(2016) 960–970.
[47] T.J. Webster, C. Ergun, R.H. Doremus, R.W. Siegel, R. Bizios, Enhanced functions of os-
teoblasts on nanophase ceramics, Biomaterials 21 (2000) 1803–1810.
[48] C. Sharma, A.K. Dinda, P.D. Potdar, C.F. Chou, N.C. Mishra, Fabrication and character-
ization of novel nano-biocomposite scaffold of chitosan–gelatin–alginate–hydroxy-
apatite for bone tissue engineering, Mater. Sci. Eng. C 64 (2016) 416–427.
 H. Li et al. / Materials Science and Engineering C 77 (2017) 1078–1087
[49] Z.L. Shi, X. Huang, Y.R. Cai, R.K. Tang, D.S. Yang, Size effect of hydroxyapatite nano-
particles on proliferation and apoptosis of osteoblast-like cells, Acta Biomater. 5
(2009) 338–345.
[50] J.X. Yan, Y.T. Miao, H.P. Tan, T.L. Zhou, Z.H. Ling, Y. Chen, X.D. Xing, X.H. Hu, Inject-
able alginate/hydroxyapatite gel scaffold combined with gelatin microspheres for
drug delivery and bone tissue engineering, Mater. Sci. Eng. C 63 (2016) 274–284.
[51] X.Y. Ye, G.H. Xu, S. Cai, Y. Dou, Y.R. Zhu, H.T. Hu, Preparation, surface properties, and
MC3T3-E1 cell response of mesoporous hydroxyapatite thin films, J. Mater. Sci. 47
(2012) 3763–3769.
[52] W. Xiao, B.S. Bal, M.N. Rahaman, Preparation of resorbable carbonate-substituted
hollow hydroxyapatite microspheres and their evaluation in osseous defects in
vivo, Mater. Sci. Eng. C 60 (2016) 324–332.
[53] K.A. Bhadang, C.A. Holding, H. Thissen, K.M. McLean, J.S. Forsythe, D.R. Haynes, Bio-
logical responses of human osteoblasts and osteoclasts to flame-sprayed coatings of
hydroxyapatite and fluorapatite blends, Acta Biomater. 6 (2010) 1575–1583.
1087
[54] O. Gryshkov, N.I. Klyui, V.P. Temchenko, V.S. Kyselov, A. Chatterjee, A.E. Belyaev, L.
Lauterboeck, D. Iarmolenko, B. Glasmacher, Porous biomorphic silicon carbide ce-
ramics coated with hydroxyapatite as prospective materials for bone implants,
Mater. Sci. Eng. C 68 (2016) 143–152.
[55] W.L. Yang, W.Q. Han, W. He, J.L. Li, J.R. Wang, H.T. Feng, Y. Qian, Surface topography
of hydroxyapatite promotes osteogenic differentiation of human bone marrow mes-
enchymal stem cells, Mater. Sci. Eng. C 60 (2016) 45–53.
[56] J.W. Shen, T. Wu, Q. Wang, H.H. Pan, Molecular simulation of protein adsorption and
desorption on hydroxyapatite surfaces, Biomaterials 29 (2008) 513–532.
[57] L.J. Zhang, T.J. Webster, Nanotechnology and nanomaterials: promises for improved
tissue regeneration, Nano Today 4 (2009) 66–80.
[58] S. Fujita, M. Ohshima, H. Iwata, Time-lapse observation of cell alignment on
nanogrooved patterns, J. R. Soc. Interface 6 (2009) S269–S277.