Materials Science & Engineering C 105 (2019) 110054

Contents lists available at ScienceDirect

Materials Science & Engineering C

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

Zirconia toughened hydroxyapatite biocomposite formed by a DLP 3D T

printing process for potential bone tissue engineering
Jiancheng Zhanga,1, Da Huangb,1, Shuifeng Liua, Xianming Donga, Yiheng Lic, Hongwu Zhangb,
⁎ ⁎
Zijun Yanga, Qisheng Sua, Wenhua Huangb, Wenxu Zhenga, , Wuyi Zhoua,
a
Research Center of Biomass 3D Printing Materials, College of Materials and Energy, South China Agricultural University, Guangzhou 510642, China
b
Department of Anatomy, Guangdong Provincial Key laboratory of Construction and Detection in Tissue Engineering, Southern Medical University, 1023# Shatai South
Road, Guangzhou 510515, China
c
Guangzhou Trauer Biotechnology Co., Ltd., 4F, A Building, U-Best Industrial Park, No. 17 Xiangshan Road, Science Town, Guangzhou 510663, China,

A R T I C LE I N FO A B S T R A C T

Keywords: The construction of ceramic components with UV curing is a developing trend by an additive manufacturing
Digital light processing (AM) technology, due to the excellent advantages of high precision selective fixation and rapid prototyping, the
Additive manufactured application of this technology to bone defect repair had become one of the hotspots of research. Hydroxyapatite
Ceramic (HAP) is one of the most popular calcium phosphate biomaterials, which is very close to the main ingredient of
Hydroxyapatite
human bones. Thus, hydroxyapatite biomaterials are popular as bone graft materials. In summary, the pre-
Bone tissue engineering
paration of HAP bioceramics by a 3D printing of digital light processing (DLP) is a promising work. However, the
Mechanical properties
preparation of HAP hybrid suspensions with high solid loading and good fluidity that can be printed by DLP
encountered some challenges. Therefore, the purpose of this work is to improve and develop a novel UV-curing
suspension with a high solids loading, which the suspension with the hydrodynamic properties and stability are
suitable for DLP printer, in order to compensate for the brittleness of HAP ceramics itself to a certain extent, a
low amount of zirconia was added in the suspension as an additive to fabricate a zirconia toughened HAP
bioceramic composite by a DLP of 3D printing. In this work, the HAP powder was pre-modified by two organic
modifiers to improve the compatibility in the acrylic resin system, and the addition of the castor oil phosphate
further reduced the shear stress of the suspension to ensure strong liquidity. The UV suspension with 60 wt%
powder particle loading had a minimum viscosity of 7495 mPa·s at 30 rpm, which was vacuum sintered at
1100 °C, 1200 °C, and 1250 °C, respectively. The composite ceramics (with 6 wt% ZrO2) at 1200 °C had a relative
density of 90.7%, while the sintered samples at 1250 °C had stronger tensile strength and bending strength. The
toughening effect of zirconia incorporation on HAP ceramics was also confirmed by the change of tensile
modulus and bending modulus, whereas the corresponding mechanical properties were also significantly en-
hanced.

1. Introduction
According to extensive research, it was known that hydroxyapatite
was the main inorganic component in the bone tissue material of the
organism [1–5], its excellent biocompatibility and osteoconductivity
made it become an essential biomaterials [6,7] in bone repair and bone
replacement and related medical fields, such as biological scaffolds
[2,8,9], drug carrier [10,11] and bone transplants and repair [12–16].
One of the calcium-phosphate-based biomaterials, hydroxyapatite
ceramics, was a bone graft and bone substitute with significant appli-
cation potential. Related studies have revealed that hydroxyapatite
ceramics adhere to cells in the body, deposits of minerals, protein ad-
sorption both have obvious positive effects [17–19]. In orthopaedic
surgery, the use of natural bone graft materials had been common in the
past. In particular, allograft bone grafts had dominated [20], but their
supply was far from meeting actual clinical needs, coupled with poor
healing due to rejection and the presence of diseases. Factors such as
the possibility of propagation of factors made the study of artificially
processed bone graft materials of concern [21,22]. Artificially prepared
calcium phosphate ceramic materials were known high bone tissue
affinities and low immunogenicity [5,23,24], Hydroxyapatite was a
calcium phosphate substance that could stably exist in an alkaline and

⁎
Corresponding authors.
E-mail addresses: wzheng@scau.edu.cn (W. Zheng), zhouwuyi@scau.edu.cn (W. Zhou).
1
Jiancheng Zhang and Da Huang contributed equally to this work.

https://doi.org/10.1016/j.msec.2019.110054
Received 25 March 2019; Received in revised form 25 July 2019; Accepted 3 August 2019
Available online 09 August 2019
0928-4931/ © 2019 Elsevier B.V. All rights reserved.
J. Zhang, et al.
neutral environment, which were more suitable for transplantation
applications in the human body environment.
Considering the requirements of tissue engineering, tissue en-
gineering scaffolds or bone substitutes suitable for clinical defect repair
must have some properties including osteoconductivity, osteoinduc-
tion, good biocompatibility, easily controllable shape size and micro-
porous structure [25]. In order to achieve the integrity of the bone
defect repair and maintain the original contour shape, it was necessary
to provide a customizable high-precision bone repair substitute. The
preparation method of the traditional bioceramic was limited in its
accuracy due to its processing difficulty. In contrast, additive manu-
facturing technology, known as 3D printing, provides a method for
preparing accurate medical clinical digital three-dimensional models in
combination with rapid stacking of CAD. At present, materials that
could be directly used in machines for additive manufacturing were
mainly high-molecular polymer, ceramics, metals, alloy and composite
materials [26–29], besides, materials used for tissue engineering such
as chitosan, fibrin, polylactic acid polymers, and hyaluronic acid, so-
dium alginate, calcium phosphate, titanium alloys, and numerous
composites had all been studied in the additive manufacturing appli-
cation, where new areas of research were devoted to mimicking the
natural regeneration of bone tissue and surface cell proliferation
[30–33]. As the most widely used type of bioceramics, calcium phos-
phate ceramics combined good rigid mechanical properties with bio-
degradability and bioactivity.
In the current study, there were few works in the literature on the
direct and swift preparation of inorganic ceramic matrix materials by
additive manufacturing techniques as bone repair substitutes.
Sionkowska et al. mixed collagen with hydroxyapatite suspension and
physically and chemically crosslinks to prepare a single-shaped hy-
droxyapatite composite scaffold which acted as a tissue-friendly im-
plant in mouse muscle [34,35]. Solaiman Tarafder et al. prepared tri-
calcium phosphate (TCP) green ceramic bodies using an additive
manufacturing method in which the powder was laid layer by layer
with water-based adhesive, while TCP scaffolds with 27%, 35% and
41% designed macro porosity having pore sizes of 500 μm, 750 μm, and
1000 μm, respectively, were fabricated by microwave sintering [36]. In
the area of additive manufacturing, selective laser sintering (SLS), li-
thography-based ceramic manufacturing (LCM), inkjet printing, and
three-dimensional printing were used in the preparation of ceramic
components [37–47]. Brian Derby [42] and Zhidong Pan [43] system-
atically introduced the recent advances in the research on the pre-
paration of ceramic components by ink-jet printing, in which com-
mercial ink-jet printing was still more applied to stereoscopic printing
and decoration, but its mature technology made it attractive to the
development of the manufacturing direction. Yingying Du et al. con-
structed a biocomposite ceramic scaffold composed of poly (PCL) mi-
crospheres and hydroxyapatite by selective laser sintering, meanwhile
excellent mechanical properties and cartilage regeneration of the mul-
tilayer scaffolds were confirmed in this research [37]. A. Licciulli et al.
[44] and Kehang Li et al. [45] explored the construction of zirconia and
alumina ceramics through stereo lithography apparatus (SLA). In both
researches, evolution of rheological behavior were studied which
caused by zirconia and alumina as powder particles, in result, selection
of acrylate as binder and control of the sintering process had a major
impact on the quality of the final ceramic part. In Johannes Maurath's
study, direct writing of ceramic green bodies were manufactured by
using capillary suspension based inks. By this way, the obtained
ceramic scaffold with honeycomb structure had excellent compression
performance. Digital light processing (DLP) technology has the ad-
vantages of fast forming speed and high precision in the method of
building new structural material. However, the printers of digital light
treatment were commonly used in organic polymerization. The con-
struction of ceramic powder-loaded organic monomer system suspen-
sion was one of the basic idea of this technology for the preparation of
ceramic components [47,48]. It was emphasized that the relationships
2
Materials Science & Engineering C 105 (2019) 110054
among various properties of ceramic suspensions, such as viscosity,
powder loading, reactivity and resin type, were the key facts to 3D
printing process [49].
The purpose of this work was to study the effect of organic long
chain active agent on the dispersion of hydroxyapatite in acrylic resin
medium, and to explore the improvement direction of mechanical
properties and the feasibility of preparation in zirconia/hydroxyapatite
composite ceramic made by digital light processing (DLP) technology.
In the experiment, the surface of hydroxyapatite was modified by oleic
acid (long-chain carboxylic acid) and γ-methacryloxypropyl tri-
methoxysilane-KH-570 (silane coupling agent) respectively, the effect
of castor oil phosphate (COPE) as a mixed anionic dispersant on the
stability of non-polar organic system was further discussed.
Hydroxyapatite composite ceramic green bodies were prepared by DLP
technology. The relative density, phase composition, microstructure
and mechanical properties of hydroxyapatite composite ceramics at
different sintering temperatures were studied and the optimum condi-
tions for preparing the composite ceramics were discussed.
2. Experimental
2.1. Materials
The commercial nano-hydroxyapatite powder (DK nano, Beijing,
China) was selected as the mineral filler for the UV photosensitive resin
paste and nano-zirconia (ZrO2) was purchased from Macklin (Shanghai,
China). 1,6-hexanediol diacrylate (HDDA), acryloylmorpholine
(ACMO), trimethylolpropane triacrylate (TMPTA), and hyperbranched
polyester acrylate (functionality: 15–18) as highly active oligomer were
purchased from BASF (Germany). Modified agent γ-(methacryloxy)
propyl trimethoxysilane (KH-570) and oleic acid were purchased from
Spark Chemical (Jiangxi, China). Castor oil phosphate (COPE) as an-
ionic surfactant was supplied by Reform Petro-Chemical (Jiangsu,
China). Phenylbis (2,4,6-trimethylbenzoyl) phosphine oxide (Omnirad
819) as photoinitiator was supplied by BASF (Germany). Mouse bone
mesenchymal stem cells (mBMSCs) were procured from the American
Type Culture Collection (ATCC, Manassas, VA).
2.2. Organic modification for hydroxyapatite powder
Add 100 mL of ethanol and 10 mL of KH-570 mixed with 50 mL of
acidified distilled water (pH = 3.8 ± 0.8) into 400 mL beaker, and
hydrolyzing at 25 ± 5 °C for 4 h, then remove water layer and re-
served. Add Hydroxyapatite powder (HAP 5% w/v) and ethanol into a
flask with three necks, add KH-570 during heating reflux step for
0.121 mol/L, and conduct suction filtration after reaction at 60 ± 2 °C
for 14 h. Adjust the pH value of the solution from 3.5 to 4.0 by am-
monia and glacial acetic acid during the grafting process, and then
exclude filter residue with ethanol, finally dry the powder by circula-
tion oven at 50 °C. In the same way, HAP could be modified by oleic
acid in acetone at 65 ± 1 °C for 12 h with pH 7.5, the content of oleic
acid was selected as 0.106 mol/L.
2.3. Preparation of UV-curable suspensions
HAP/ZrO2 suspensions were prepared through the combination of
two simple steps: 1) Mix monomers and UV oligomers according to a
certain proportion (HDDA 45 wt%; ACMO 35 wt%; TMPTA 15 wt%;
hyperbranched polyester acrylate 5 wt%) with simple magnetic stirring
for 30 min to form unloaded resin; 2) Add HAP powder (modified or
not, 10-60 wt% with respect to the mass of ceramics UV-curable sus-
pensions), ZrO2 commercial powder (average size:50–80 nm, 3 wt%),
antistatic agent castor oil phosphate ester (1 wt%), and photoinitiator
(0.5 wt%) to unloaded resin under mechanical agitation
(700–1200 rpm) by machine mixer for 6 h in dark.
J. Zhang, et al.
Fig. 1. Schematic illustration of the fabrication of cured ceramic green body basing on 3D extrusion printing.
2.4. Fabrication of the ceramic green body through DLP
In this work, a modification of conventional DLP 3D printers was
supplied by PRINT-RITE Solutions to print out the ceramic slurries
described previously. First, 3D Max software was used to build a 3D
solid model, and EZ layout software was used to slice the parts and
generate the supporting structures. The saved edg file was imported
into the DLP control 1.5 software to set up parameters for CoLiDo DLP
printer to fabricate the ceramic green body. The parameters of the DLP
printer were as follows: the exposure time was set as 35 s each layer of
curing, and the UV source used to carry out the research presented in
this work was a HDMI 1080P DLP projector developed by Vivitek,
emitting UV light at 405 nm. The generating module was constituted by
a metal plane with size of 148 mm × 80 mm and the acrylic trough with
size of 205 mm × 124 mm × 50 mm, and UV-curable suspensions
would be used 100–600 mL for printing process (Fig. 1).
2.5. Post-processing of green body
Proper post-treatment can effectively improve the result of de-
binding and sintering. Anhydrous ethanol was used to rinse the green
body which scraping from the metal plate cautiously. In this work,
before sintering and degreasing process, the sample was placed under
triple ultraviolet lamp tube (45 W), illuminated for 3 h to farther re-
inforce.
2.6. Debinding and sintering
The preparation of ceramic components was consisting of a vacuum
debinding step followed by vacuum sintering in order to remove the
organic resin in green body. The sample was put in a muffle furnace
connecting with vacuum pump, heated from 50 °C to 200 °C and then
600 °C with rate of 0.4 °C/min and 0.5 °C/min under vacuum condition,
3
Materials Science & Engineering C 105 (2019) 110054
degreasing in vacuum for 2 h devoted to reduce crack damage to
components from thermal decomposition of organic matter. Following
the debinding step, sample was sintered sequentially in vacuum muffle
furnace, and the sintering temperatures were set as 1100 °C, 1200 °C
and 1250 °C.
2.7. Characterization
In this work, the average particle size of HAP was determined by
laser particle size meter (Zetasizer Nano ZSE, Malvern, Britain).
Stability of UV ceramic suspension was evaluated by experimental re-
sults of natural settling, in which the test suspensions were prepared by
dispersing HAP powder in HDDA. Viscosity test was employed to
characterize hydromechanics of UV ceramic suspension with visc-
ometer (NDJ-8S, Shanghai pingxuan scientific instrument Co., ltd,
China) and the modular rheometer (MCR502, Anton Paar, Austria). The
Fourier transform infrared (FTIR, Nicolet iS10, Thermo Fisher,
America) spectra was used to characterize hydroxyapatite powder
modified by γ-(methacryloxy) propyl trimethoxysilane and oleic acid.
The phase composition of modified powder and prepared ceramic were
analyzed by a XRD (Ultima IV, Rigaku, Japan), in which 2θ scanning
mode was set as range of 10–80°. Microstructural features of sintered
ceramics were characterized by a scanning electron microscope (SEM,
EVO 18, Zeiss, Germany), while microscopic forms of modified HAP
powder were observe by transmission electron microscope (TEM,
Tecnai G220, FEI, America). Besides, mechanical property was con-
sidered as an important indicator in the field of bone tissue engineering,
thus tensile strength and bending strength of ceramic samples were
characterized by electronic universal testing machine (UTM-4204,
Shenzhen SANS Testing Machine Co., Ltd., China).
J. Zhang, et al.
Fig. 2. FTIR spectroscopy of modified hydroxyapatite sample with KH-570 and oleic acid.
2.8. Cytotoxicity
To investigate the biocompatibility of our biomaterials, mBMSCs
cells were seeded on the ceramic chip samples prepared by 3D printing.
Briefly, after sterilized by UV light for 2 h and soaking in 75% medical
alcohol, the ceramic wafers were taken out and washed with phosphate
buffered saline (PBS) for three times, after washing, samples were
soaked in PBS for 5 h, and then mBMSCs cells were seeded on these
wafers at a density of 3000 cells/cm2. After cultured for 24 h, cells on
each wafers were washed with PBS three times and treated with
Ethidium homodimer-1 (0.5 μM) and calcein AM (0.25 μM) (Live/Dead
viability kit, Molecular Probes) for 45 min. Cells were observed under
an inverted fluorescent microscope (IX53, Olympus).
In addition, mBMSCs cells were seeded on HAP ceramic wafers with
3 wt% ZrO2 at a density of 3000 cells/cm2, co-cultured onto a 24-well
plate for 24 h. The planted samples were washed by PBS for three times,
after 24 h immobilization with glutaraldehyde (2.5 wt%) solution, the
cells were cleaned with PBS to remove residual glutaraldehyde. The
cells were dehydrated with 30% 50% 70%, 85%, 90% (1 time) and
100% (3 time) ethanol, and the samples were freeze-dried at 20 °C,
40 °C, and 80 °C for 12 h, respectively. The morphologies of the
mBMSCa cells on the HAP ceramic wafers with ZrO2 were investigated
by a S-3000N scanning electron microscope (SEM; HITACHI, Japan).
2.9. Cell proliferation assay
The proliferation rate of mBMSCs cells was evaluated by
AlamarBlue assay (Molecular Probes) after cultured for 1, 3 and 5 days,
respectively. The ceramic wafers were sterilized by UV light for 1 h and
fixed in the bottom of 24 wells plate as previous mention. The mBMSCs
cells were seeded on wafers at a density of 3000 cells/cm2, and medium
was changed every two days. At each time point of assay, the cells were
incubated in medium containing 10% (v/v) AlamarBlue reagent at 37 °C
in 5% CO2 for 4 h, then 100 μL solution from each example was re-
moved into a 24 wells plate and read at 530/600 nm in a SpectraMax
fluorescence microplate reader (Molecular Devices). Medium con-
taining 10% (v/v) AlamarBlue.
3. Results and discussion
3.1. Characterization of pure hydroxyapatite (HAP)
As the inorganic filler in the resin slurry, the size and microscopic
4
Materials Science & Engineering C 105 (2019) 110054
morphology of the HAP powder have a significant effect on the dis-
persivity and the stability of the mixed slurry. On the other hand, the
mechanical properties of the sintered ceramic material are largely af-
fected by the nature of hydroxyapatite, such as crystallinity, composi-
tion, etc. Therefore, the morphology and crystallinity information of
powder raw material is fundamental. The XRD and SEM image per-
formed on prototype HAP powder were shown in Fig. S1 (a) and (b)
respectively. The XRD pattern indicated that the compound was a high
crystalline monophasic apatite. Fraction of crystallinity phase Xc of
HAP could be evaluated by following formula [50]:
V112/300
Xc ≈ 1 −
I300 (1)
where I300 is the intensity of (300) reflection and V112/300 is the lowest
intensity measured between (112) and (300) reflections. According to
Eq. (1), crystallinity of prototype HAP power was estimated to be nearly
94%. According to Fig. S1 (b), stick-like particles were stacked together
to show a coral-like reunion. Laser particle size meter report of powder
distribution was shown in Fig. S1 (c), while TEM image of prototype
powder was observed in Fig. S1 (d). Particle size distribution was un-
imodal and the mean particle diameter was confirmed about
160 ± 5 nm. Besides, microscopic forms of HAP could be determined
as rhabdolith morphology from TEM image, in which rod length was
counted to agree with laser particle size meter result.
3.2. Characterization of grafted hydroxyapatite
The HAP powder was modified in organic solvents by heating re-
flux, and grafting results was indicated in the infrared spectrum from
250 cm−1 to 4000 cm−1. In order to ensure that the organic modifier
physically adsorbed on the particles and did not affect the character-
ization results, the modified particles were repeatedly washed by
ethanol, DMSO and deionized water, and then vacuum filtered before
being characterized. From Fig. 2, compared with the curve of prototype
hydroxyapatite, in the spectrum of oleic acid modification, the new
weak peak at 1645 cm−1 was attributed to C]C of alkene in oleic acid
molecule, and the peaks located at 2920 cm−1 and 2850 cm−1 were
derived from stretching vibration adsorption of eCH2 belonging to
alkyl chain. Besides, the new peak band from 1550 cm−1 was attributed
to stretching vibration of carboxylate which represented that ester-
ification was happened between carboxyl group of oleic acid and hy-
droxyl of HAP surface [51]. From the spectrum of KH-570 modification,
the strong peaks at 1720 cm−1 and 1640 cm−1 were attributed to C]O
J. Zhang, et al.
bonds of KH-570, the peaks at 1300 cm−1 was assigned to the CeO in
ester group, which proved that HAP had been successfully modified by
coupling agent KH-570 [52].
By comparison of XRD patterns before and after modification, it was
confirmed that the main diffraction peak/intensity was not changed by
oleic acid and KH-570 grafting as shown in Fig. S2, which means that
the modification process has no influence on crystallinity and crystal-
line phases of prototype HAP powder. Disparities of dispersion behavior
shown in TEM images (Fig. S3) indicated that reunion probability of
HAP powder had reduced after organic grafting in ethanol. However,
the microcosmic form of stick was bound to distribution of surface
hydroxyl groups on powder, which reducing chance of modifier sila-
nols/carboxyl connecting with hydroxyl groups. Thus, the surface
coating of eOeSi(CH2)3COO(CH3)C]CH2/eCO(CH2)7CH]CH
(CH2)7CH3 were observed barely. So it seems that surface organic
modification was no obvious enough to improve suspension char-
acteristics effectively.
3.3. Hydromechanics characteristics of high solid loading of hydroxyapatite
suspensions
Viscosity and settlement rate were adopted to assessed hydro-
mechanics characteristics of solid suspensions. With increasing solid
content, viscosity of ceramic suspension risen sharply which was un-
favorable for paving of the suspension after each layer printing. It was
verified that index growth of suspension viscosity would emerge with
solid content when it overtake 60 wt% [25]. However, solid mass
fraction should be over 50 wt% to ensure the relative density of sintered
ceramic samples. Considering of this contradiction, our effort was de-
voted to maintain high solid content while limiting the increase in
viscosity. The effect of castor oil phosphate as dispersant in different
solid loading was illustrated in Fig. 3, in which the rotation rate of
viscosimeter was set as 30 rpm. Addition of castor oil phosphate and
modification of KH-570/oleic acid both could alleviate the increase of
viscosity caused by solid loading enhancement obviously. Dispersion
mechanism of long-chain organic modification was depended on sur-
face energy. In this work, nano-HAP particle was modified by excessive
modifiers, and modified powder was fed in photosensitive resin, how-
ever, different rheological behaviors appeared between the two dif-
ferent modifiers. Because of double (unsaturated) bonds in organic
chain, modified particles were made better compatibility in curable
acrylic resin, while oleic acid modification caused lower shear viscosity
(about 10,860 mPa·s) at 60 wt%, even before adding castor oil phos-
phate (COPE). To achieve better dispersion and stability, castor oil
Fig. 3. Influence of dispersants type on viscosities of each loads suspensions at
30 rpm rotational speed.
5
Materials Science & Engineering C 105 (2019) 110054
phosphate as antistatic agent was added to diminish interparticle
electrostatic attraction ulteriorly. In our preliminary work, various
concentrations of COPE (0.1 wt%–0.9 wt%) were added to estimate its
influence to the viscosity tendency and 0.7 wt% of COPE as treatment
concentration was considered appropriately in ceramic suspension
preparation with low viscosity. According to molecular structure of
COPE, carboxyl groups were considered to interact with remaining
hydroxyl on surface of particles after modified while carbon chain and
phosphate group at the other end have better compatibility with acrylic
resin. Thus， the force between particles was weaken and the ag-
glomeration rate was reduced by modification and addition of dis-
persants, which decreased the viscosity significantly (Table 1).
The hydrodynamic behavior of suspension could be studied more
fully through the characterization of suspension by modular rheometer.
By adjusting the shear rate from 0.1 s−1 to 100 s−1 viscosity flow ten-
dency of each solid content suspension was shown in Fig. 4 (a, b). In
this part, KH-570 (or oleic acid) modified HAP particles and 0.7 wt%
castor oil phosphate were added in the suspensions in determined test
temperature (25 °C). With the increasing of content of HAP, the varia-
tion of pseudoplastic behavior of high solid content suspension was
observed obviously from these curves. According to the result from a
related nanofiller researches [53], pseudoplastic was ascribed to large
rheological units which were disrupted at high shear rate. Zirconia
particles as extra filler caused stronger interaction between particles
and monomer by higher surface energy, thus ceramic suspension con-
taining HAP & zirconia displayed high viscosity. Even though, max-
imum solid content suspension has a viscosity lower than 25 Pa·s at low
shear rate which appropriates to recoating process of DLP printing. In
particular, the effects of different modifiers on hydroxyapatite suspen-
sion were significantly different, on this point, in order to further re-
presentation of the steady-state rheological behavior of suspension,
evolvement of shear stress (σ) and shear rate (γ) were analyzed in Fig. 4
(c, d), in which σ and γ were related by following formula:
σ = η(γ) ∙γ (2)
in which η(γ) is viscosity of suspension.
In this part, HAP suspensions with 0.7 wt% castor oil phosphate
were utilized to study rheological behavior, and evolution of shear yield
stress was generally surveyed by the curve composed of shear stress (σ)
and shear rate (γ) which influenced by increment of nano particles. The
influences of modification types and solid contents could be summar-
ized to pseudoplastic behavior and existence of yield stress from Fig. 4-
σ(γ) graph. With growing shear rate, in conformity with the above
conclusion, the suspensions displayed an obvious shear thinning be-
havior, which were confirmed as pseudoplastic fluid. However, the
growth of shear stress curves varies greatly under different modification
conditions of particles. Before adding ZrO2, the stress growth of KH-570
modified particles suspension was not obvious with the increase of
shear rate. On the other hand, the change of ZrO2 content and the in-
crease of HAP solid content led to much evolution of shear stress, the
shear stress grown exponentially with the increased of shear rate, which
represents a stronger particle interaction force. The transform of yield
stress was attributed to the reticular microstructures consisted of in-
terlinkage between particle-particle and particle-monomer. Thus, de-
struction of large rheological units in ceramic suspensions dominated
the variation of yield stress which caused the result of yield stress
greater than shear stress, and the pseudoplastic behavior of shear
thinning was exhibited in process of experiments. It was remarkable
that similar experimental results could be observed in oleic acid mod-
ified curve, while the growth rate of stress was much lower than that of
KH-570 modified curve, and the maximum shear stress is not more than
7 Pa. According to the rheological data, the acrylate group at the end of
KH-570 molecule chain made the modified hydroxyapatite particles
have better compatibility with acrylic resin and lower shear stress
change rate. However, with the increasing of solid content, the inter-
action strength between nanoparticles in suspension exceeded that
J. Zhang, et al. Materials Science & Engineering C 105 (2019) 110054

Table 1
Partial experimental reagent names and structures.
Reagent type Purpose Molecular structure

HDDA Reactive diluent

ACMO Reactive diluent

TMPTA Reactive diluent

KH-570 Modified agent

Oleic acid Modified agent

COPE Anionic surfactant

(a) (b )

(c) ( d)

Fig. 4. Rheological curves of HAP resin filled with different amounts of HAP modified by KH-570 (a) or oleic acid (b) with/without ZrO2, Shear stress (σ) versus shear
rate (γ) from 0.1 s−1 to 100 s−1 for different amounts of HAP modified by KH-570 (c) or oleic acid (d) with/without ZrO2.

6
J. Zhang, et al.
between nanoparticles and liquid resins, especially after the addition of
ZrO2 with larger surface energy, in which the shear stress increased
sharply when the rheological units was destroyed at high shear rate. In
contrast, when the solid content was less than 60 wt%, the shear stress
of oleic acid modified curves increased slightly higher than that of KH-
570 curve under the same conditions, but when the solid content
reached 60 wt% and ZrO2 was added, the shear stress of oleic acid
modified curve still tended to grow gently with the increase of shear
rate. This was attributed to the fact that long unsaturated carbon chains
of oleic acid were more conducive to weakening the interactions be-
tween particles-particles and particles-monomers in suspensions with
high solidification content. In fact, considering the strategy of reducing
the viscosity of ceramic suspension, the highest viscosity of the sus-
pension prepared by oleic acid modified HAP (20.05 Pa·s) was lower
than that of suspension prepared by KH-570 modified HAP (28.02 Pa·s).
Settling property of ceramic suspensions is one of the important
criteria for evaluating its stability and printing feasibility. Settlement
velocity of powder particles floating in the suspension could be de-
scribed by Stokes' law as following equation:
2 (ρp − ρf ) 2
ν= gR
9 μ (3)
in which μ refers to the dynamic viscosity; ρp and ρf are mass density of
particles and fluid mass density respectively, g refers to gravitational
acceleration constant and R is mean radius of particle. According to the
pre-observation, solid particles tended to settle downward rather than
floating on liquid resin, which could confirm ρp > ρf.
On the basis of the above conclusions, with the increasing of par-
ticles mass fraction, dynamic viscosity (μ) of the suspensions displayed
exponential type rising. For nanoscale hydroxyapatite powder disper-
sing in the acrylic resin liquid, the micro particle settlement caused by
gravity was not obvious. Thus, in the case of high content of particles
and high viscosity, the settling velocity of the suspensions was rela-
tively slow. According to the previous experiments, 30 wt% of HAP
suspension had apparent and easy-to-measure velocity of settlement,
and it was adopted to estimated influence of modification steps and
castor oil phosphate dispersant on colloidal stability of suspensions.
Sedimentation data of ceramic suspensions was recorded by settlement
distances according to Fig. S4 within a designed period of time and in
constant temperature condition. In 48 h of the experiments, prototype
suspension exhibited subsidence evidently, which was governed by
poor compatibility and Van der Waals' force between rich hydroxyl
groups on granular surfaces and acrylic resin monomer. Actually, or-
ganic surface grafting of particles and addition of dispersants had
contribution on stability improvement of suspension. From Fig. 5 (a, b,
c) of settlement column diagram, settlement velocity had been swin-
geing cut after adding of castor oil phosphate. This phenomenon was
owed to amphiphile groups which introduced the space hindrance layer
between dispersive medium and particles. Besides, modification of oleic
acid made further short settlement distance. It seems that the un-
saturated carboxylic acid had better modification effect.
3.4. DLP processing of HAP mixed resin
In this part, ceramic suspension was prepared according to the
method described in section 2.3 and tested by DLP photocurable printer
mentioned in section 2.4. Different from transparent resin, after mixed
solid particles, it displayed more obvious refraction and absorption of
ultraviolet light. Molding of cured ceramic green bodies depended on
acrylate UV curing, in which ceramic powder would be fixed on cured
solid resin. As shown in Fig. 6 (a), the ceramic printed pieces were
formed by bonding each cured two-dimensional planes by layer-by-
layer stacking. Therefore, single layer thickness which decided com-
pactness of interlayer combination was gauged in Fig. 6 (b) to estimate
curing sensitivity. The figure showed that only suitable exposure time
7
Materials Science & Engineering C 105 (2019) 110054
could get enough curing depth to ensure the shape of the sample. From
the curves of relationship between exposure time and thickness, the
addition of ceramic powders did have a significant effect on the pho-
tosensitivity of suspension. Curing sensitivity was greatly attenuated
along with the increase of particle content. In particular, under equal
conditions, cure depth of suspension decreased as the increase of zir-
conia particles. We tended to consider optical properties of suspensions
containing zirconia, such as index of refraction and shading coefficient.
Nevertheless, the ceramic suspension with high solid content still had a
high curing activity, which was due to the fact that the hyperbranched
polyester acrylate molecule as oligomer could provide a large number
of functional groups during the curing process to ensure sufficient re-
active activity. As for HAP powder, weakening effect of visible light
through it could be observed from the Fig. S5.
3.5. Characterization of debinding and sintering
In current work, debinding process and sintering process were
carried out in vacuum since air debinding process used in previous
work caused a large number of component cracks and internal defects.
Due to high de-carbonation rate of organic resin, pyrolytic vapor could
not escape completely from the sample in a relatively short period of
time. Internal partial pressure generated inside the sample led to for-
mation of heterogeneous phase region in the sintering process. On the
other hand, adjusting the heating rate was an important method to
control the pyrolysis conditions, but excessive pyrolysis rate was not
conducive to densification of samples and removal of residual carbon.
Thus, stepped temperature program was designed as Fig. S6 described,
in which heating rate was set as 0.4 °C/min and 0.5 °C/min with heating
from normal temperature to sintering temperature without cooling step
in the middle.
The X-ray diffraction spectrum shown in Fig. 7 was used to char-
acterize the phase composition of sintered ceramic parts. The samples
still maintained the HAP phase when sintering temperature was set as
1100 °C since the characteristic diffraction peaks of single phase apatite
could be seen. Thus, during the degreasing process, the organic matter
was decomposed without changing the crystal form of HAP, whereas
the grains of HAP were not sufficiently grown to make the parts dense,
and this result could be supplemented by the data in Table 3 and Fig. 8
Compared with the 1100 °C spectrum line, the formation of new phase-
β-TCP phase (Ca3(PO4)2) could be seen from the spectral lines of
1200 °C and 1250 °C, which was possibly due to decomposition of hy-
droxyapatite [54]. The increase in the content of new phase of whi-
tlockite led to a slight decrease in the relative density of ceramic
components when sintering temperature reached to 1250 °C. The in-
fluences of hydroxyapatite consumption and the formation of scheelite
phase on the mechanical properties of ceramic components need fur-
ther integration with the following data.
Linear shrinkage rate is one of the important indexes to evaluate the
sintering results, which can be calculated by the following formula:
l 0 − lt
η= *100%
l0 (4)
where l0 is the length of the green bodies and lt is the length of sintered
samples.
In this work, plate shape (Fig. S8) was chosen as test model for
measurement of three axis dimensions (x, y, z) to calculate linear
shrinkage rate which was recorded in Table 2. When the solid content
of suspension was less than 40 wt%, the sintered plate sample had ob-
vious cracks and large shrinkage rate which caused structural damages
of samples after sintering. Observed from the sintering result, in-
sufficient curing content of suspension got obvious relationship for
sintered samples to have surface defects. The sintered sample from
suspension with 60 wt% powder achieved acceptable shrinkage rate
which was controlled within 10% of each directions. In a series of
samples, a clear divergence in shrinkage was shown in each directions
J. Zhang, et al.
(a)
(c)
Fig. 5. Settling height contrast diagram (a, b, c) and sedimentation curves (d) of 30 wt% HAP with 3 wt% ZrO2 suspensions at different time.
which presenting a tendency that shrinkage rate grew with the increase
of axis length.
The microstructure of bending fracture was surveyed by a scanning
electron microscope as shown in Fig. 8. Compared to the prototype
particles with size of 160 ± 5 nm, significant growth of sintered grains
with different degrees could be observed in each sintering tempera-
tures. The sample sintered at 1100 °C displayed inadequate density and
loose grain distribution, while other two samples sintered at higher
temperatures shown fuller grain growth and more prominent particle
connections. In fact, the relevant mechanism of ceramic sintering had
been studied and reported in the field of ceramics [55,56]. Due to
atomic diffusion mechanism in high temperature, relative density in-
creased following with grain growth, thus, the sintered samples as
shown in Fig. 8 (b1, b2, c1, c2) achieved higher relative density and
stronger connectivity of grain-grain because of the effective tempera-
ture sintering treatment. However, with the increasing of sintering
temperature and growth of grains, hydroxyapatite began to decompose
and produce phase transformation (into β-TCP phase), leaving voids in
the sintered samples [58–61]. Compared with Fig. 8 (b1, b2), the denser
grain interconnecting and larger grain size could be observed in Fig. 8
(c1, c2) due to higher sintering temperature, and the voids around the
grain was became obvious. Besides, the holes and defects left by gases
produced by thermal decomposition of organic matter were the minor
factor for poor particle filling effect.
3.6. Mechanical property
Observing from Fig. S7, it was apparent that the insufficient particle
solidification of the green bodies would cause the sintered sample to
produce significant surface cracks and, provided irrevocable negative
influence on the mechanical properties. The width of surface cracks
could be reach to 10 to 30 μm, meanwhile, plate-like particles de-
taching was perceived from low rate microscopic image, which caused
8
Materials Science & Engineering C 105 (2019) 110054
(b)
(d)
by encountering of parallel and vertical cracks. This phenomenon could
be devoted by two main reasons: (1) the high content of organic matter
led violent thermal decomposition phenomenon during the sintering
process, and a large amount of pyrolytic steam effusion devoted ob-
vious rupture on the surface of the sample; (2) the organic resin sup-
porting ceramic particles was gradually decomposed during the sin-
tering process, and the differentiated shrinkage on surface and interior
of samples caused stress concentration enhancing cracks, grooves, holes
and other defects from inside and outside. We considered that the latter
one was the more important reason.
By using Archimedes drainage method, the relative densities of
sintered bodies were measured to be compared with mechanical
properties of the corresponding samples, which could be observed in
Table 3 and Figs. 9–10. Obviously, fragile samples were unable to
provide satisfactory density data. Fortunately, by adjustment of the
maximum sintering temperature and increasing of curing content, it
exhibited positive influences on the densification of samples, from
1100 °C to 1200 °C, the relative densities of samples leaped a con-
siderable improvement from 62.6% to 90.2%, and the mechanical
properties of corresponding samples were also improved by leaps and
bounds following the relative densities.
The mechanical properties of tensile strength and bending strength
were discussed from Figs. 9 and 10. It could be observed that the
strength of tensile and bending were improved with the increase of
sintering temperature and solid content, which was consistent with the
relative density variation. Furthermore, because of increasing of the
grain size with temperature, 15.4 MPa was the maximum value that
tensile strength reached, correspondingly, bending strength also im-
proved significantly during the advancement of sintering temperature.
On the other hands, the test results of mechanical properties also
showed that the 60 wt% of HAP sintered samples at 1200 °C had a
tensile modulus of 3760 ± 73 kPa and a bending modulus of
349 ± 16 MPa as shown in Table S1, which displayed a tendency of
J. Zhang, et al. Materials Science & Engineering C 105 (2019) 110054

(a)

(b)

Fig. 6. (a) The schematic diagram of DLP curing and (b) influence of increasing solid loading on curing depth.

fragile material. Moreover, the mechanical modulus of sintered samples
with solid content below 50 wt% (including 50 wt%) showed incon-
spicuous development regular pattern. Besides, the influence of dif-
ferent doping amount of ZrO2 on the mechanical properties of samples
was also reflected in the Figs. 9–10. When sintered at 1100 °C, the
addition of ZrO2 could not improve the tensile strength of the composite
ceramics, which indicated that ZrO2 had no obvious toughening effect
when sintering temperature could not guarantee the densification of
hydroxyapatite ceramics. When sintering temperature rose up to
1200 °C and 1250 °C, the mechanical properties of pure HAP ceramics

Fig. 7. The XRD of pattern of samples at each temperature.

9
J. Zhang, et al. Materials Science & Engineering C 105 (2019) 110054

(a1) (a2)

10 m 2 m

(b1) (b2)

10 m 2 m

(c1) (c2)

10 m 2 m

Fig. 8. The SEM images for HAP ceramics sintered surface fabricated from 60 wt% suspension containing oleic acid modified powder at (a1-3000×, a2-6000×)
1100 °C, (b1-3000×, b2-6000×) 1200 °C, (c1-3000×, c2-6000×) 1250 °C.

Table 2 were significantly improved, and the toughening effect of zirconia was
Sintered samples shrinkage rate. more obvious. Compared with pure HAP ceramics, with the increasing
Shrinkage rate in each X-axis (long) Y-axis (wide) Z-axis (thick)
of ZrO2 content to 1 wt%, 3 wt% and 6 wt%, the tensile strength of
direction l0 = 66 mm l0 = 10 mm l0 = 2.5 mm composite ceramics sintered at 1250 °C were improved by 3.03%,
22.2% and 29.4%, respectively. Under the same conditions, the bending
20 wt% 56.1% (fragile) 32.6% (fragile) 21.7% (fragile) strength increases by 8.9%, 17.9% and 23.9% respectively. The above
30 wt% 21.2% 27.6% 19.8%
40 wt% 16.8% 13.8% 12.3%
experimental data demonstrated that the addition of zirconia could
50 wt% 13.7% 6.9% 6.7% further improve the mechanical modulus of the HAP ceramics with
60 wt% 9.6% 5.8% 4.1% improved mechanical properties, and thus had a positive effect on
improving the toughness of the brittle ceramics. In order to further
study the toughening mechanism, SEM images of burnished section and
bending section were observed and shown in Fig. 11. It was proved that

Table 3
Relative densities of sintered samples with each conditions (without ZrO2) in Archimedes method result.
Relative density (%) 20 wt% 30 wt% 40 wt% 50 wt% 60 wt% 60 wt% + 1 wt% 60 wt% + 3 wt% 60 wt% + 6 wt%

1100 °C Fracture Fracture Fracture 55.4 62.6 63.4 61.8 62.7

1200 °C Fracture 53.8 56.3 75.6 84.2 87.6 90.7 92.7
1250 °C Fracture Fracture 52.7 74.8 91.8 93.5 95.2 96.8

10
J. Zhang, et al.
Fig. 9. Tensile strength of sintered HAP ceramics at each temperature.
Fig. 10. Bending strength of sintered HAP ceramics at each temperature.
the pure hydroxyapatite grain indicated a nonuniform size and shape
because of abnormal grain growth. But in contrast, with zirconia mi-
crocrystal on surface evenly, regular growth of grain was found on the
surface of HAP zirconia toughened ceramics. Because of the difference
of electron density, the ZrO2 grains were found with lighter color, and
the darker and larger in size were HAP grains. The smaller ZrO2 par-
ticles were evenly dispersed in the HAP matrix and attached to the
surface of bigger HAP particles, which had obvious “pinning” effect,
could inhibit the grain boundary migration and grain growth of HAP to
a certain extent during the sintering process. Generally, the hydro-
xyapatite is easy to be decomposed during the high temperature sin-
tering process, the addition of ZrO2 nanoparticles could reduce the
thermal decomposition effect caused by phase transformation (HAP to
β-TCP), resulting in the increasing of the relative density of composite
ceramics. In addition, quite a few part of ZrO2 particles in the composite
ceramics existed in the grain of HAP and an “intragranular” structure
was formed, which was caused by the fact that the diffusion rate of
nano-hydroxyapatite in sintering process was obviously higher than
that of ZrO2 particles. The grain boundary of hydroxyapatite particles
crossed smaller zirconia and coats it. For the thermal expansion coef-
ficients of HAP and ZrO2 were different, the residual stress would be
generated at the sub-interface of the matrix grain after sintering and
cooling, which played a key role in the crack propagation of ceramics
[62,63].
For the exploration of mechanical properties of hydroxyapatite
11
Materials Science & Engineering C 105 (2019) 110054
zirconia toughened ceramics, we tended to consider that zirconia had
approved degree of toughening effect on HAP ceramic samples and
made it applicable to bone tissue engineering 3D printing. According to
the trend of experimental results, the mechanical properties of com-
posite ceramics could be further improved by increasing the content of
zirconia (more than 6 wt%). In fact, based on the discussion in section
3.3, the increasing of zirconia content would promote the rapid growth
of the ceramic suspension viscosity. Under the method described in this
paper, the maximum addition of ZrO2 was limited to 6 wt% while the
content of HAP was kept at 60 wt%. When the content of ZrO2 was
higher than 6 wt%, the suspension had poor fluidity because of its ex-
cess viscosity, which was not suitable for DLP forming technology.
3.7. Phase evaluation and retention of sintered ceramics
The composite ceramics calcined at each temperatures were ground
to powders, the X-ray diffraction pattern was shown in Fig. 12. Zirconia
toughening agent used in raw materials was mainly tetragonal phase,
and contains a small amount of monoclinic phase, which is confirmed in
Fig. S1(e). Sintering with hydroxyapatite at three different tempera-
tures, obvious phase change could not be detected in XRD spectra of
ZrO2. On the one hand, when the sintering temperature was not more
than 1250 °C was not enough to induce the martensitic transformation
of ZrO2 (t to m) [64,65], which was observed that ZrO2 dispersed in the
form of small particles and embedded in HAP grains. On the other hand,
when the amount of zirconia was less than 10 wt%, it was difficult for
zirconia nanoparticles dispersed in hydroxyapatite to achieve phase
transformation and grain growth through transport of mass [63]. Thus,
in this work, due to the low additive content of ZrO2, the mechanism of
dispersing tensile residual stress by martensitic transformation zirconia
(t-m) was not the main toughening reason [66,67]. Importantly, com-
pared with XRD spectra of sintered hydroxyapatite, with the increase of
sintering temperature, the phase transformation ratio of HAP decreases
significantly. The diffraction peak of β-TCP could not be found under
the XRD detection of sintered composite ceramics at 1100 °C and
1200 °C, which is probably due to the diffraction peak was lower than
the detection limit, while the weak diffraction peaks of β-TCP were
found when it was sintered at 1250 °C. Therefore, comparing the X-ray
diffraction patterns of composite ceramics and pure HAP ceramics
sintered at 1250 °C, it could be found that the doping of ZrO2 and the
corresponding amount of doping have effectively inhibited on phase
transformation decomposition of HAP during the sintering process [68],
reduced the formation of HAP voids and improved its compactness
[69], which was consistent to the deduction of section 3.6. To sum-
marize, the addition of zirconia could improve the relative density of
the composite ceramics as shown in Table 3, and indicated positive
toughening effect on the mechanical properties of the ceramics.
3.8. In vitro cell viability
In vitro cell culture experiment was used to estimate the bio-
compatibility of the HAP/ZrO2 ceramics, in which mBMSCs were
seeded on the ceramic chip samples. Fig. 13 (a, b, c) shows the cell
culture in medium as control group of live/dead staining assay after
culturing 24 h The related culture result of sample wafers with same
conditions could be observed from Fig. 13 (d, e, f). After 24 h of culture,
the cell population of mBMSCs owned a high survival rate and grew
uniformly on the ceramics wafers without a large number of dead cells.
Compared with the control group, the survival rate of cells cultured on
the sample was not lower than that of cells cultured only in the
medium, which was attributed to the good biocompatibility of HAP/
tricalcium phosphate. Moreover, due to the roughness of the surface of
composite ceramics, the aggregation behavior of mBMSCs cells on the
samples was more obvious, and the growth morphology of mBMSCs
cells became more elongated. The porous nature of HAP ceramics was
conducive to adsorbing cell adhesion proteins and promoting cell
J. Zhang, et al.
Fig. 11. The SEM images for sintered ceramics of 60 wt% of HAP without (a)/with 1 wt% (b) 3 wt% (c) 6 wt% (d) zirconia toughened at 35000×, EDS of composite
ceramics spot constituents (e).
adhesion, growth and proliferation in cell culture. Therefore, the HAP
ceramics prepared by additive manufacturing technology had no sig-
nificant toxicity to the cells, and most of the cells grew in spindle shape
(non-circular), which proved that the cells could adhere to the ceramic
wafers well. In addition, doping of zirconia did not significantly affect
the bioactivity of the HAP ceramics according to the results of live/dead
staining assay. With the same culture times, clearer and more detailed
morphology of cell growth could be observed in Fig. 13 (g, h). The
mBMSCs cells adhered and spread over the rough surface of HAP
ceramics. Besides, the cells were able to grow obvious tentacles, which
indicated that the cells had a benign growth state on the surface of the
samples. Although a small amount of HAP granules adhered to the cell
body, it did not affect the cell growth morphology. The result of cell
multiplication on HAP ceramic was evaluated by AlamarBlue assay and
showed in Fig. 13 (i). After 1, 3, 5-day of culture, similar to the control
12
Materials Science & Engineering C 105 (2019) 110054
group, the number of mBMSCs cells per unit area increased significantly
and a rapid cell proliferation rate was showed either, which indicated
that stem cells could adhere, grow and proliferate on the ceramics. The
porous properties of HAP ceramics achieved high roughness surface for
the cells to climb up easily, and the toxic acrylic polymer was effec-
tively removed during the sintering process, leaving the inorganic HAP
(3 wt% ZrO2) ceramics with good biocompatibility.
4. Conclusions
In this work, it was concluded that composite ceramic suspension
with high-loading and qualified fluidity could be prepared by mixing
active liquid resin and modified powder with adding effective dis-
persant. This suspension had high reactivity and relatively low visc-
osity, so it was suitable for DLP printing equipment, and could
J. Zhang, et al. Materials Science & Engineering C 105 (2019) 110054

Fig. 12. The XRD of pattern of composite ceramics with 3 wt% ZrO2 at each temperature.

Fig. 13. The fluorescence microscopy of mBMSCs cells morphology after culture for 24 h in culture medium as control group (a-live, b-dead, c-merged) and the HAP
ceramic wafers with 3 wt% ZrO2 as study group (d-live, e-dead, f-merged), the SEM image displayed the microstructure of mBMSCs cells on the HAP ceramic wafers
with 3 wt% ZrO2 (g-200×, h-2000×) and, (i) proliferation rate of mBMSCs cells over 1, 3, 5 days cultivating on HAP ceramic wafers with 3 wt% ZrO2.

13
J. Zhang, et al.
guarantee the successful construction of ceramic green body. KH-570
and oleic acid were used to modify hydroxyapatite powder before
mixing, and the modification of oleic acid could obtain ceramic sus-
pension with higher load and lower viscosity. Besides, the improvement
of castor oil phosphate (0.7 wt%) on the rheological behavior of the
suspension was not negligible, which settlement had been minimal on
the suspension in the past 45 h of experimentation. In the dynamic
hydrodynamics, it could be observed that with the increase of the
ceramic particle content, the viscosity of different suspension were
obviously increasing, and the reaction activity decreases. Even though,
the ceramic suspension with maximum particle percentage (60 wt%)
displayed suitable for DLP equipment. Zirconia was incorporated for
1 wt%, 3 wt%, 6 wt% and sintered with hydroxyapatite in a vacuum
high-temperature furnace. Microstructure showed that smaller zirconia
grains were attached to fully grown calcium phosphate large grain.
Although because of the low content of zirconia, it was impossible to
resist the propagation of cracks by forming a compressive strain field
caused by volume expansion of martensitic transformation, which was
different from the toughening mechanism originally envisaged.
However, according to XRD phase analysis, the addition of ZrO2 had
effectively reduced the phase transformation decomposition of HAP
during sintering process, and promoted the densification during high
temperature sintering, and the mechanical properties were obviously
improved. According to the experiments of in vitro cell viability which
was adopted to evaluate the biocompatibility, HAP ceramics with 3 wt
% ZrO2 showed good cell viability and no cytotoxicity. Furthermore,
the mBMSCs cells continued to proliferate on the surface of the HAP
ceramics observed for 5 days. The HAP ceramics sintering with 3 wt%
ZrO2, prepared by digital light processing, could be considered as
promising biomaterials for bone tissue repair or other medical appli-
cations, including 3D printing.
Acknowledgements
This work was financially supported by the Science and Technology
Foundation of Guangdong Province (2016A050502044,
2016A010103024, 2015B020237009) and National Natural Science
Foundation of China (51603074), International Science and Technology
Cooperation Project of Guangzhou Economic Technological
Development Zone (2017GH09) and Guangzhou Tianhe District Science
and Technology Planning Project (201706YG083) and Natural Science
Foundation of Guangdong Province in China (2016A030313602).
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.msec.2019.110054.
References
[1] K.J.L. Burg, S. Porter, J.F. Kellam, Biomaterial developments for bone tissue en-
gineering[J], Biomaterials 21 (23) (2000) 2347–2359.
[2] K. Rezwan, Q.Z. Chen, J.J. Blaker, Aldo Roberto Boccaccini, Biodegradable and
bioactive porous polymer/inorganic composite scaffolds for bone tissue engineering
[J], Biomaterials 27 (18) (2006) 3413–3431.
[3] S. Bose, M. Roy, A. Bandyopadhyay, Recent advances in bone tissue engineering
scaffolds[J], Trends Biotechnol. 30 (10) (2012) 546–554.
[4] M. Šupová, Substituted hydroxyapatites for biomedical applications: a review[J],
Ceram. Int. 41 (8) (2015) 9203–9231.
[5] S. Samavedi, A.R. Whittington, A.S. Goldstein, Calcium phosphate ceramics in bone
tissue engineering: a review of properties and their influence on cell behavior[J],
Acta Biomater. 9 (9) (2013) 8037–8045.
[6] S. Saska, L. CarlaPires, M.A. Cominotte, L.S. Mendes, M.F. de Oliveira, I.A. Maia,
J.V.L. da Silva, S.J.L. Ribeiro, J.A. Cirelli, Three-dimensional printing and in vitro
evaluation of poly(3-hydroxybutyrate) scaffolds functionalized with osteogenic
growth peptide for tissue engineering[J], Mater. Sci. Eng. C 89 (1) (2018) 265–273.
[7] J. Zhang, D. Xiao, X. He, F. Shi, P. Luo, W. Zhi, K. Duan, J. Weng, A novel porous
bioceramic scaffold by accumulating hydroxyapatite spheres for large bone tissue
engineering. III: characterization of porous structure[J], Mater. Sci. Eng. C 89 (1)
(2018) 223–229.
14
Materials Science & Engineering C 105 (2019) 110054
[8] M. Elbadawi, Z.J. Wally, I.M. Reaney, Porous hydroxyapatite-bioactive glass hybrid
scaffolds fabricated via ceramic honeycomb extrusion[J], J. Am. Ceram. Soc. 101
(5) (2018) 3541–3556.
[9] B.J. Zhang, H.Y. Li, L. He, Z.W. Han, T. Zhou, W. Zhi, X.B. Lu, X. Lu, J. Weng,
Surface-decorated hydroxyapatite scaffold with on-demand delivery of dex-
amethasone and stromal cell derived factor-1 for enhanced osteogenesis[J], Mater.
Sci. Eng. C 89 (2018) 355–370.
[10] M. Ribeiro, M.H. Fernandes, M.M. Beppu, F.J. Monteiro, M.P. Ferraz, Silk fibroin/
nanohydroxyapatite hydrogels for promoted bioactivity and osteoblastic prolifera-
tion and differentiation of human bone marrow stromal cells[J], Mater. Sci. Eng. C
89 (1) (2018) 336–345.
[11] M. Sumathra, M.A. Munusamy, A.A. Alarfaj, M. Rajan, Osteoblast response to vi-
tamin D3 loaded cellulose enriched hydroxyapatite mesoporous silica nanoparticles
composite[J], Biomed. Pharmacother. 103 (2018) 858–868.
[12] L. Xia, Y. Xie, B. Fang, X. Wang, K. Lin, In situ modulation of crystallinity and nano-
structures to enhance the stability and osseointegration of hydroxyapatite coatings
on Ti-6Al-4V implants[J], Chem. Eng. J. 347 (1) (2018) 711–720.
[13] B. He, J. Zhao, Y. Ou, D. Jiang, Biofunctionalized peptide nanofiber-based com-
posite scaffolds for bone regeneration[J], Mater. Sci. Eng. C 90 (1) (2018) 728–738.
[14] M.G. Raucci, C. Demitri, A. Soriente, I. Fasolino, A. Sannino, L. Ambrosio, Gelatin/
nano-hydroxyapatite hydrogel scaffold prepared by sol-gel technology as filler to
repair bone defects[J], J. Biomed. Mater. Res. A 106 (2018) 2007–2019.
[15] S.T. Wei, Y. Wei Lin, Z.Y. Jie, C. Feng, Z.Y. Gang, J.Y. Ying, H.Y. Hua, Porous
nanocomposite comprising ultralong hydroxyapatite nanowires decorated with
zinc-containing nanoparticles and chitosan: synthesis and application in bone defect
repair[J], Chemistry 24 (2018) 8809–8821.
[16] E. Quinlan, A. López-Noriega, E. Thompson, H.M. Kelly, S.A. Cryan, F.J. O’Brien,
Development of collagen–hydroxyapatite scaffolds incorporating PLGA and alginate
microparticles for the controlled delivery of rhBMP-2 for bone tissue engineering
[J], J. Control. Release 198 (28) (2015) 71–79.
[17] H. Yuan, K. Kurashina, J.D. de Bruijn, Y. Li, K. de Groot, X. Zhang, A preliminary
study on osteoinduction of two kinds of calcium phosphate ceramics [J],
Biomaterials 20 (19) (1999) 1799–1806.
[18] A. Rimola, M. Corno, C. Marcelo, Zicovich-Wilson, P. Ugliengo, Ab initio modeling
of protein/biomaterial interactions: glycine adsorption at hydroxyapatite surfaces
[J], J. Am. Chem. Soc. 130 (48) (2008) 16181.
[19] N. Almora-Barrios, K.F. Austen, N.H. de Leeuw, Density functional theory study of
the binding of glycine, proline, and hydroxyproline to the hydroxyapatite (0001)
and (0110) surfaces[J], Langmuir 25 (9) (2009) 5018–5025.
[20] A.A. Jahangir, R.M. Nunley, S. Mehta, A. Sharan, Bone-graft substitutes in ortho-
paedic surgery[J], Am. Acad. Orthop. Surg. Bull. 1 (2008) 5–9.
[21] R.F. Heary, R.P. Schlenk, T.A. Sacchieri, D. Barone, C. Brotea, Persistent iliac crest
donor site pain: independent outcome assessment[J], Neurosurgery 50 (3) (2002)
516–517.
[22] Marie Françoise Moreau, Yves Gallois, Michel-Félix Baslé, Daniel Chappard,
Gamma irradiation of human bone allografts alters medullary lipids and releases
toxic compounds for osteoblast-like cells[J], Biomaterials 21 (4) (2000) 369–376.
[23] M. Azadi, A. Teimouri, G. Mehranzadeh, Preparation, characterization and bio-
compatible properties of beta-chitin/silk fibroin/nanohydroxyapatite composite
scaffolds prepared using a freeze-drying method[J], RSC Adv. 6 (9) (2016)
7048–7060.
[24] L. Chen, H.L. Liu, Y. Gu, Y. Feng, H.L. Yang, Lumbar interbody fusion with porous
biphasic calcium phosphate enhanced by recombinant bone morphogenetic protein-
2/silk fibroin sustained-released microsphere: an experimental study on sheep
model[J], J. Mater. Sci. Mater. Med. 26 (3) (2015) 1–12.
[25] F. Scalera, C. Esposito Corcione, F. Montagna, A. Sannino, A. Maffezzoli,
Development and characterization of UV curable epoxy/hydroxyapatite suspen-
sions for stereolithography applied to bone tissue engineering[J], Ceram. Int. 40 (
(10) (2014) 15455–15462.
[26] T.D. Ngo, A. Kashani, G. Imbalzano, K.T.Q. Nguyen, D. Hui, Additive manufacturing
(3D printing): a review of materials, methods, applications and challenges[J],
Compos. Part B 143 (15) (2018) 172–196.
[27] P. Jana, O. Santoliquido, A. Ortona, P. Colombo, G.D. Soraru, Polymer-derived SiCN
cellular structures from replica of 3D printed lattices[J], J. Am. Chem. Soc. 101 (7)
(2018) 2732–2738.
[28] P. Parandoush, D. Lin, A review on additive manufacturing of polymer-fiber com-
posites[J], Compos. Struct. 182 (2017) 36–53.
[29] W.S.W. Harun, M.S.I.N. Kamariah, N. Muhamad, S.A.C. Ghani, F. Ahmad,
Z. Mohamed, A review of powder additive manufacturing processes for metallic
biomaterials[J], Powder Technol. 327 (2018) 128–151.
[30] S. Bose, D. Ke, H. Sahasrabudhe, A. Bandyopadhyay, Additive manufacturing of
biomaterials[J], Prog. Mater. Sci. 93 (2018) 45–111.
[31] F. Krujatz, A. Lode, J. Seidel, T. Bley, M. Gelinsky, J. Steingroewer, Additive bio-
tech—chances, challenges, and recent applications of additive manufacturing
technologies in biotechnology[J], New Biotechnol. 39 (2017) 222–231.
[32] S. Singh, S. Ramakrishna, Biomedical applications of additive manufacturing: pre-
sent and future[J], Curr. Opin. Biomed. Eng. 2 (2017) 105–115.
[33] S. Bose, S.F. Robertson, A. Bandyopadhyay, Surface modification of biomaterials
and biomedical devices using additive manufacturing[J], Acta Biomater. 66 (15)
(2018) 6–22.
[34] A. Sionkowska, J. Kozłowska, Properties and modification of porous 3-D collagen/
hydroxyapatite composites[J], Int. J. Biol. Macromol. 52 (2013) 250–259.
[35] J. Kozlowska, A. Jundzill, A. Bajek, M. Bodnar, A. Marszalek, H. Witmanowski,
A. Sionkowska, Preliminary in vitro and in vivo assessment of modified collagen/
hydroxyapatite composite[J], Mater. Lett. 221 (15) (2018) 74–76.
[36] T. Solaiman, B. Vamsi, D. Neal, B. Amit, B. Susmita, Microwave sintered 3D printed
J. Zhang, et al.
tricalcium phosphate scaffolds for bone tissue engineering[J], J. Tissue Eng. Regen.
Med. 7 (8) (2013) 631–641.
[37] Du Yingying, Haoming Liu, Qin Yang, Shuai Wang, Jianglin Wang, Jun Ma,
Insup Noh, Antonios G. Mikos, Shengmin Zhang, Selective laser sintering scaffold
with hierarchical architecture and gradient composition for osteochondral repair in
rabbits[J], Biomaterials 137 (2017) 37–48.
[38] A. Danezan, G. Delaizir, N. Tessier-Doyen, G. Gasgnier, J.M. Gaillard, P. Duport,
B. Nait-Ali, Selective laser sintering of porcelain[J], J. Eur. Ceram. Soc. 38 (2)
(2018) 769–775.
[39] Loïc Ferrage, Ghislaine Bertrand, Pascal Lenormand. Dense yttria-stabilized zir-
conia obtained by direct selective laser sintering[J], Addit. Manuf. 21 (2018)
472–478.
[40] W. Li, A. Ghazanfari, D. McMillen, M.C. Leu, G.E. Hilmas, J. Watts, Characterization
of zirconia specimens fabricated by ceramic on-demand extrusion[J], Ceram. Int. 44
(11) (2018) 12245–12252.
[41] J. Maurath, N. Willenbacher, 3D printing of open-porous cellular ceramics with
high specific strength[J], J. Eur. Ceram. Soc. 37 (15) (2017) 4833–4842.
[42] B. Derby, Additive manufacture of ceramics components by inkjet printing[J],
Engineering 1 (1) (2015) 113–123.
[43] Z. Pan, Y. Wang, H. Huang, Z. Ling, Y. Dai, S. Ke, Recent development on pre-
paration of ceramic inks in ink-jet printing[J], Ceram. Int. 41 (10) (2015)
12515–12528.
[44] A. Licciulli, C. Esposito Corcione, A. Greco, V. Amicarelli, A. Maffezzoli, Laser
stereolithography of ZrO2 toughened Al2O3[J], J. Eur. Ceram. Soc. 25 (9) (2005)
1581–1589.
[45] hangLi Ke, Z. Zhao, The effect of the surfactants on the formulation of UV-curable
SLA alumina suspension[J], Ceram. Int. 43 (6) (2017) 4761–4767.
[46] E. Schwarzer, M. Götz, D. Markova, D. Stafford, U. Scheithauer, T. Moritz,
Lithography-based ceramic manufacturing (LCM) – viscosity and cleaning as two
quality influencing steps in the process chain of printing green parts[J], J. Eur.
Ceram. Soc. 37 (16) (2017) 5329–5338.
[47] S. Westbeek, J.A.W. van Dommelen, J.J.C. Remmers, M.G.D. Geers, Multiphysical
modeling of the photopolymerization process for additive manufacturing of cera-
mics[J], Eur. J. Mech. A/Solids 71 (2018) 210–223.
[48] Gerald Mitteramskogler, Robert Gmeiner, Ruth Felzmann, Simon Gruber,
Christoph Hofstetter, Jürgen Stampfl, Jörg Ebert, Wolfgang Wachter,
Jürgen Laubersheimer, Light curing strategies for lithography-based additive
manufacturing of customized ceramics[J], Addit. Manuf. 1-4 (2014) 110–118.
[49] P. Lima, A. Zocca, W. Acchar, J. Günster, 3D printing of porcelain by layerwise
slurry deposition[J], J. Eur. Ceram. Soc. 38 (9) (2018) 3395–3400.
[50] E. Landi, A. Tampieri, G. Celotti, S. Sprio, Densification behaviour and mechanisms
of synthetic hydroxyapatites[J], J. Eur. Ceram. Soc. 20 (14–15) (2000) 2377–2387.
[51] J. Huang, G. Liang, G. Lu, J. Zhang, Conservation of acidic papers using a dispersion
of oleic acid-modified MgO nanoparticles in a non-polar solvent[J], J. Cult. Herit.
34 (2018) 61–68 https://linkinghub.elsevier.com/retrieve/pii/
S1296207417308531.
[52] Y. Lu, X. Li, C. Wu, S. Xu, Comparison between polyether titanate and commercial
coupling agents on the properties of calcium sulfate whisker/poly(vinyl chloride)
15
Materials Science & Engineering C 105 (2019) 110054
composites[J], J. Alloys Compd. 750 (2018) 197–205.
[53] C.C. Esposito, G. Mensitieri, A. Maffezzoli, Analysis of the structure and mass
transport properties of nanocomposite polyurethane[J], Polym. Eng. Sci. 49 (2009)
1708–1718.
[54] M. Yetmez, Z.E. Erkmen, C. Kalkandelen, A. Ficai, F.N. Oktar, Sintering effects of
mullite-doping on mechanical properties of bovine hydroxyapatite[J], Mater. Sci.
Eng. C 77 (2017) 470–475.
[55] Zhijian Shen, Zhe Zhao, Hong Peng, Mats Nygren, Formation of tough interlocking
microstructures in silicon nitride ceramics by dynamic ripening[J], Nature 417
(6886) (2002) 266–269.
[56] I.W. Chen, X.H. Wang, Sintering dense nanocrystalline ceramics without final-stage
grain growth[J], Nature 404 (6774) (2000) 168.
[58] P.-Y. Chen, S.-F. Wang, R.R. Chien, C.-S. Tu, K.-C. Feng, C.-S. Chen, K.-Y. Hung,
V.H. Schmidt, Evolution of the microstructural and mechanical properties of hy-
droxyapatite bioceramics with varying sintering temperature[J], Ceram. Int.
(2019), https://doi.org/10.1016/j.ceramint.2019.05.144.
[59] T. Ebadzadeh, A. Behnamghader, R. Nemati, Preparation of porous hydroxyapatite
ceramics containing mullite by reaction sintering of clay, alumina and hydro-
xyapatite[J], Ceram. Int. 37 (7) (2011) 2887–2889.
[60] J. Song, Y. Liu, Y. Zhang, L. Jiao, Mechanical properties of hydroxyapatite ceramics
sintered from powders with different morphologies[J], Mater. Sci. Eng. A 528
(16–17) (2011) 5421–5427.
[61] K. Poovendran, K.S. Joseph Wilson, Amalgamation and characterization of porous
hydroxyapatite bio ceramics at two various temperatures[J], Mater. Sci. Semicond.
Process. 100 (2019) 255–261.
[62] S. Gali, K. Ravikumar, B.V.S. Murthy, B. Basu, Zirconia toughened mica glass
ceramics for dental restorations[J], Dent. Mater. 34 (3) (2019) e36–e45.
[63] P.K. Rao, P. Jana, M.I. Ahmad, P.K. Roy, Synthesis and characterization of zirconia
toughened alumina ceramics prepared by co-precipitation method[J], Ceram. Int.
45 (13) (2019) 16054–16061.
[64] R.-F. Guo, P. Shen, C. Sun, Y.-J. Fu, Y.-H. Liu, Z.-A. Ren, Q.-C. Jiang, Effects of
composition and sintering temperature on the structure and compressive property
of the lamellar Al2O3–ZrO2 scaffolds prepared by freeze casting[J], J. Mater. Sci. 50
(14) (2015) 5039–5046.
[65] S. Ramesh, N. Zulkifli, C.Y. Tan, Y.H. Wong, F. Tarlochan, S. Ramesh, W.D. Teng,
I. Sopyan, L.T. Bang, A.A.D. Sarhan, Comparison between microwave and con-
ventional sintering on the properties and microstructural evolution of tetragonal
zirconia[J], Ceram. Int. 44 (8) (2018) 8922–8927.
[66] Y. Wang, P. Xiao, The phase stability and toughening effect of 3Y-TZP dispersed in
the lanthanum zirconate ceramics[J], Mater. Sci. Eng. A 604 (2014) 34–39.
[67] M. Szutkowska, M. Boniecki, Subcritical crack growth in zirconia-toughened alu-
mina (ZTA) ceramics[J], J. Mater. Process. Technol. 175 (1–3) (2006) 416–420.
[68] B. Yilmaz, A.Z. Alshemary, Z. Evis, Co-doped hydroxyapatites as potential materials
for biomedical applications[J], Microchem. J. 144 (2019) 443–453.
[69] Satish Kanhed, Shikha Awasthi, Sneha Goel, Aditi Pandey, Rajeev Sharma,
Anish Upadhyaya, Kantesh Balani, Porosity distribution affecting mechanical and
biological behaviour of hydroxyapatite bioceramic composites[J], 43 (13) (2017)
10442–10449.