Sie sind auf Seite 1von 15

Acta Biomaterialia 71 (2018) 200–214

Contents lists available at ScienceDirect

Acta Biomaterialia
journal homepage: www.elsevier.com/locate/actabiomat

Full length article

In vitro and in vivo studies on zinc-hydroxyapatite composites as novel


biodegradable metal matrix composite for orthopedic applications
Hongtao Yang a,1, Xinhua Qu b,1, Wenjiao Lin c, Cong Wang d, Donghui Zhu e, Kerong Dai b,⇑,
Yufeng Zheng a,f,⇑
a
Department of Materials Science and Engineering, College of Engineering, Peking University, Beijing 100871, China
b
Department of Orthopedics, Shanghai Key Laboratory of Orthopedic Implant, Shanghai Ninth People’s Hospital, Shanghai Jiaotong University School of Medicine,
Shanghai 200011, China
c
R&D Center, Lifetech Scientific (Shenzhen) Co Ltd, Shenzhen 518057, China
d
Department of Interventional Radiology and Vascular Surgery, Peking University Third Hospital, Beijing 100191, China
e
Department of Biomedical Engineering, College of Engineering, University of North Texas, Denton, TX 76207, USA
f
International Research Organization for Advanced Science and Technology, Kumamoto University, 2-39-1 Kurokami, Chuo-Ku, Kumamoto 860-8555, Japan

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

Article history: Recent studies indicate that there is a great demand to optimize pure Zn with tunable degradation rates
Received 17 December 2017 and more desirable biocompatibility as orthopedic implants. Metal matrix composite (MMC) can be a
Received in revised form 6 February 2018 promising approach for this purpose. In this study, MMC with pure Zn as a matrix and hydroxyapatite
Accepted 3 March 2018
(HA) as reinforcements were prepared by spark plasma sintering (SPS). Feasibility of novel Zn-HA com-
Available online 9 March 2018
posites to be used as orthopedic implant applications was systematically evaluated. After sintering, HA
distributed in the Zn particle boundaries uniformly. Corrosion tests indicated that the degradation rates
Keywords:
of Zn-HA composites were adjustable due to the biphasic effects of HA. Zn-HA composites showed signif-
Zinc
Hydroxyapatite
icantly improved cell viability of osteoblastic MC3T3-E1 cells compared with pure Zn. Both pure Zn and
Composite composites exhibited a low thrombosis risk and hemolysis rates while a Zn ion concentration-dependent
Degradation effect was found on coagulation time. An effective antibacterial property was observed as well. The vol-
Biocompatibility ume loss of pure Zn and Zn-5HA composite was 1.7% and 3.2% after 8 weeks’ implantation. Histological
analysis found newly formed bone surrounding pure Zn and Zn-5HA composite at week 4 and increased
bone mass over time. With prolonged implantation time, Zn-5HA composite was more effective on stim-
ulating new bone formation than pure Zn. In summary, MMC is a feasible way to design Zn based mate-
rials with adjustable degradation rates and improved biocompatibility.

Statement of Significance

Biodegradable zinc materials are promising candidates for the new generation of orthopedic implants.
However, the slow degradation rates and unsatisfactory cytocompatibility of pure Zn in bone environ-
ments limit its future clinical applications. Generally, alloying is a common way to improve the perfor-
mance of pure Zn. In this study, metal matrix composite was chosen as a novel strategy to solve the
problems. Hydroxyapatite, as a bioactive component, was added into Zn matrix via spark plasma sinter-
ing. We find that Zn-HA composites exhibited adjustable degradation rates and improved biocompatibil-
ity both in vitro and in vivo. This study provides exhaustive and significant information including
microstructure, mechanical performance, degradation behavior, biocompatibility, hemocompatibility
and antibacterial property for the future Zn based implants design.
Ó 2018 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

1. Introduction

⇑ Corresponding authors at: Department of Materials Science and Engineering, Biodegradable metals have been viewed as ideal materials for
College of Engineering, Peking University, Beijing 100871, China (Y. Zheng). applications in orthopedics due to their desirable biodegradability,
E-mail addresses: krdai@163.com (K. Dai), yfzheng@pku.edu.cn (Y. Zheng). excellent mechanical property and good biocompatibility. Zinc, as
1
These two authors contributed equally to this study.

https://doi.org/10.1016/j.actbio.2018.03.007
1742-7061/Ó 2018 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
H. Yang et al. / Acta Biomaterialia 71 (2018) 200–214 201

a novel candidate, has attracted increasing interests recently. to pure iron [19]. Therefore, improvements in degradation and bio-
Unlike magnesium, the degradation of zinc does not create hydro- compatibility can be expected by incorporating HA into pure Zn.
gen gas pockets caused by fast corrosion which may interfere with In the present study, Zn-HA composites was fabricated by spark
the bone healing process, resulting in callus formation and cortical plasma sintering (SPS). The microstructure, mechanical properties,
defects [1]. Also, unlike iron, zinc exhibits a better in vivo degrada- in vitro corrosion behavior, cytotoxicity, hemocompatibility,
tion behavior without leaving voluminous corrosion products that antibacterial property and in vivo test have been systematically
are hard to be eliminated by human body [2]. Apart from the favor- investigated to evaluate the performance of Zn-HA composites as
ite degradation behavior, the importance of zinc in human nutri- biodegradable implants for orthopedic applications.
tion has also been proved. Zinc, as an essential element in the
human body, participates in various basic biological functions 2. Materials and methods
including nucleic acid metabolism, signal transduction, apoptosis
regulation and gene expression in addition to interacting with a 2.1. Material preparation
lot of organic ligands [3]. More importantly, zinc plays an impor-
tant role in the growth and mineralization of bone tissue. Zinc Pure Zn powder (99.9%, 140+325 mesh, Alfa) and hydroxyap-
stimulates bone formation and mineralization; the metal directly atite (HA) powder (99%, 150 nm in length, Beijing Daoking) were
activates aminoacyl-tRNA synthetase in osteoblastic cells and used as raw materials. Pure Zn powders were mixed with HA pow-
stimulates cellular protein synthesis. In addition, zinc inhibits ders (1, 5 and 10 wt.%) in planet grinding machine (XM-4, KELI
osteoclastic bone resorption by inhibiting osteoclast-like cell for- CERAMICS) under argon atmosphere and ball milling for 1 h at
mation from marrow cells [4]. Zinc also plays a role in the preser- 300 rpm. Then, pure Zn powders, the mixed Zn-HA powders were
vation of the bone mass. However, the reported degradation rates put into a 30 mm diameter graphite die and sintered under vac-
of pure Zn in bone environments are relatively slow. The in vitro uum by SPS-1050 (Sumitomo Coal Mining Company, Ltd.) with sin-
and in vivo degradation rates of pure Zn were approximately tering temperature of 380 °C, sintering pressure of 40 MPa, and
0.06–0.08 mm/year and 0.2 mm/year [5,6], respectively. For bone holding time of 6 min. Pure Zn prepared by SPS served as control.
fracture fixation, the mechanical support of biodegradable The as-sintered pure Zn and Zn-HA composites were cut into disk
implants should sustain for 12–24 weeks depending on the clinical samples (U10  1 mm) and cylinders (U2  4 mm and U2  5
conditions [7]. Thereafter, while the mechanical integrity mm), respectively. Disk samples were used for density measure-
decreases, the degradation should progress at a sufficient rate ment, microstructure characterization, microhardness, electro-
without causing an intolerable host response. Considering a typical chemical tests, immersion tests, biocompatibility,
screw 5 mm in diameter and 50 mm in length for the fixation of a hemocompatibility and antibacterial tests. Cylinders were used
fracture bone. A simple calculation reveals that the complete for compressive (U2  4 mm) and animal tests (U2  5 mm). Each
degradation of this screw takes more than a decade. In addition, sample was mechanically polished to 2000 grit, then ultrasonically
the type of environmental tissue plays an important role in the cleaned in acetone and ethanol, then dried in the open air. Before
degradation process of biodegradable implants. Soft tissue may cytotoxicity, antibacterial and animal tests, samples were sterilized
cause more rapid degradation due to the superior vascularization with ultraviolet radiation for at least 4 hours each side.
or the avoidance of changes in local pH value [8,9]. For example,
a faster degradation of intramedullary areas compared to cortical
2.2. Microstructure characterization
areas was observed in Mg alloys, which attributed to the enhanced
vascularization of the bone marrow [1]. Therefore, it is necessary to
An electronic balance (XS105, METTLER TOLEDO) attached with
develop Zn based materials with adjustable degradation rates. Pure
density accessories was used to measure the density of all samples.
Zn also showed significant toxicity to normal human osteoblast
Samples were further polished by 0.1 lm diamond paste and
cells. Cell viability dropped to lower than 50% after 7 days’ culture
cleaned in acetone. After samples were etched with a 4% HNO3/
with pure Zn extracts. Cells showed a decrease in both the size and
alcohol solution, an optical microscope (Olympus BX51 M) was
the actin microfilament intensity after exposing to Zn. Moreover,
used to observe the metallographic structure. X-ray diffractometer
significant DNA damage was detected in cells after 7 days incuba-
(XRD, Rigaku DMAX 2400, Japan) using CuKa radiation was
tion with extracts [10]. Pure Zn exhibited toxicity in human
adopted to identify the constituent phases of pure Zn and Zn-HA
endothelium-derived cells and L292 cells as well [11,12].
composites with scanning range from 20° to 80° and scan rate of
Metal matrix composite (MMC) can be a feasible approach to
6° min1. Energy dispersive spectrometer (EDS, Bruker QUANTAX,
improve the performance of pure Zn. The advantages of MMC are
Germany) was used for the analysis of chemical compositions.
adjustable corrosion properties as well as biocompatibility by
choosing the appropriate composites. The interaction of MMC with
surrounding tissues can be optimized by varying the constituents, 2.3. Mechanical tests
the type and distribution of the reinforcing phase. Hydroxyapatite
(HA) is a well-known bioceramic with bioactivity that supports cell Microhardness and compressive tests were conducted to evalu-
proliferation, bone ingrowth and osseointegration [13,14]. HA has ate the mechanical properties of pure Zn and Zn-HA composites.
similar chemical and crystallographic structures to bone, which Microhardness test was carried out by a microhardness tester
can form a chemical bond with osseous tissue and act like nucle- (SHIMADZUHMV-2t) measuring Vickers hardness with loading
ations for new bone [15]. Its main components, Ca and P, are both force of 0.1 kN and dwell time of 15 s. Instron 5969 universal test
essential elements that maintain physiological homeostasis pro- machine was used for the compressive test. The specimens were in
cesses [16]. Gu et al. [17] prepared Mg/HA composite and found the form of cylinder 2 mm in diameter and 4 mm in length. Com-
that HA created more anodic sites in Mg matrix forming galvanic pression strain rate was 2  104 s1. An average of at least three
coupling which accelerated the corrosion rate of Mg. Witte et al. measurements was taken for each group.
[18] fabricated the AZ91D/HA metal matrix composite by powder
metallurgy method and found that the composite was cytocompat- 2.4. Electrochemical measurements
ible with better corrosion behavior compared with AZ91D alloy.
Addition of bioceramics (HA, TCP, BCP) into iron resulted in The electrochemical tests were conducted with an electrochem-
increased corrosion rates and improved cellular activity compared ical workstation (Autolab, Metrohm, Switzerland) at room temper-
202 H. Yang et al. / Acta Biomaterialia 71 (2018) 200–214

ature in Hank’s solution (NaCl 8.00 g L1, KCl 0.40 g L1, CaCl2 0.14 2.6.2. Hemolysis and platelet adhesion
g L1, NaHCO3 0.35 g L1, glucose 1.00 g L1, MgCl26H2O 0.10 g Healthy human blood (anticoagulant with 3.8% citric acid
L1, MgSO47H2O 0.06 g L1, Na2HPO412H2O 0.06 g L1 and KH2- sodium) extracted from volunteers was diluted by physiological
PO4 0.06 g L1, pH 7.4). A three-electrode cell with a platinum saline with a volume ratio 4:5. Pure Zn and Zn-HA composites were
counter electrode and a saturated calomel electrode (SCE) as the installed in centrifugal tubes with 10 mL physiological saline and
reference electrode was utilized for electrochemical tests. The water bathed at 37 °C for 30 min. Then, 0.2 mL diluted blood was
open-circuit potential (OCP) of each sample was monitored for added to each tube and incubated at 37 °C for 60 min, 10 mL deion-
5400 s. Electrochemical impedance spectroscopy (EIS) measure- ized water with 0.2 mL diluted blood as the positive control and 10
ment was carried out by applying 10 mV perturbation with the mL physiological saline with 0.2 mL diluted blood as the negative
measuring frequency ranging from 105 Hz down to 102 Hz. After- control. Afterwards, samples were removed and tubes ware cen-
wards, potentiodynamic polarization test was carried out at a scan- trifuged at 3000 rpm for 5 min. Supernatant was transferred to
ning rate of 1 mV s1. The measurement area was 0.2826 cm2 (U 6 96-well plates, the absorbance (OD) was determined by a micro-
mm). At least five measurements were taken for each sample plate reader (Bio-RAD680) at a wavelength of 545 nm. Hemolysis
group. Corrosion parameters including open-circuit potential of samples was calculated by the formula:
(OCP), corrosion potential (Ecorr) and corrosion current density
ODðtestÞ  ODðnegative controlÞ
(icorr) were analyzed by linear fit and Tafel extrapolation to the Hemolysis ¼  100%
cathodic and anodic parts of polarization curves. ODðpositive controlÞ  ODðnegative controlÞ

For platelet adhesion, whole blood from healthy human was


2.5. Immersion tests centrifuged at 1000 rpm for 10 min. Platelet rich plasma (PRP)
was obtained from the upper fluid. Samples were moved into 24-
Immersion tests were carried out in Hank’s solution at 37 °C for well plates and 0.2 mL PRP was added to each well, incubated at
50 days. The pH value of the solution was recorded by a pH meter 37 °C for 1 h. After gently flushed by phosphate buffered saline
(Mettler FiveEasy pH FE20K) during immersion. After 50 day’s (PBS), platelets on samples were fixed with 2.5% glutaraldehyde
immersion, samples were removed from Hank’s solution, gently solution at room temperature for 2 h, then dehydrated with gradi-
rinsed with distilled water and dried in air. The corroded surface ent alcohol solution (50%, 60%, 70%, 80%, 90%, 95% and 100%) for 10
morphology was characterized by scanning electron microscopy min, and finally dried in desiccator for 2 days. The morphology of
(SEM, Hitachi S-4800, Japan) and optical microscope (Olympus platelet adhered on the specimens was coated by gold and
BX51 M). X-ray diffractometer (XRD, Rigaku DMAX 2400, Japan) observed by SEM (Hitachi S-4800, Japan).
using CuKa radiation was adopted to identify the constituent
phases of corrosion products. Samples were embedded in epoxy 2.6.3. Coagulation time
and the cross sections were polished for SEM observation and Platelet poor plasma (PPP) was prepared by centrifuging the
EDS analysis. The weight loss of the samples was measured on citrated whole blood at 3000 rpm for 15 min and then the super-
an electronic balance (XS105, METTLER TOLEDO) with a measuring natant was collected. The plasma was added into samples with a
sensitivity of 0.1 mg after removal of corrosion products. An aver- ratio 500 lL cm2 and water bathed at 37 °C for 30 min. Then the
age of five measurements were taken for each group. The in vitro assays including activated partial thromboplastin time (APTT), pro-
corrosion rate was calculated according to the equation: C = Dm/ thrombin time (PT) and thrombin time (TT) were measured at the
qAt, where C is the corrosion rate in mm year1, Dm is the weight Third Hospital of Peking University with corresponding reagents
loss, q is the density of the material, A is the initial implant surface added. Zn ion concentrations in PPP were measured by ICP-OES.
area and t is the implantation time. Related ion concentrations in
solution were measured by Inductively Coupled Plasma Optical
2.6.4. Antibacterial test
Emission Spectroscope (ICP-OES, iCAP6300, Thermo).
S. aureus (ATCC 25922) was used as a model bacterium to deter-
mine the antibacterial properties of pure Zn and Zn-HA composites
2.6. In vitro tests with Ti6Al4V as control. S. aureus was cultured in a 37 °C shaker
incubator at 220 rpm for 16 h in 20 mL of the Luria-Bertani (LB)
2.6.1. Cell viability medium containing 10 g L1 peptone, 10 g L1 NaCl and 5 g L1
MC3T3-E1 cells were cultured in Alpha-minimum essential beef extract. The density of the bacteria cultured in the medium
medium (MEM), 10% fetal bovine serum (FBS), 100 U mL1 peni- was quantified by absorption spectroscopy at a wavelength of
cillin and 100 lg mL1 streptomycin at 37 °C in a humidified atmo- about 600 nm (OD600) after performing colony counts on different
sphere of 5% CO2 (Alpha-minimum essential medium (MEM), fetal concentration bacterial suspensions. OD600 was converted to col-
bovine serum (FBS), penicillin and streptomycin were purchased ony forming units (CFUs) as follows: OD600 = 0.1 contained 108
from Gibco BRL (Gaithersburg, MD, USA)). The cytotoxicity test CFUs. Serial dilution of the bacteria was performed in the LB med-
was evaluated using the 100% extracts of metal plates, prepared ium to achieve a final bacterial cell suspension of 1  106 and
with a surface area to extraction medium ratio of 1.25 mL cm2 1  107 CFUs per mL, respectively. Then, 1 mL of the bacterial cell
in a humidified atmosphere, with 5% CO2 at 37 °C for 24 h. The suspension at a concentration of 107 CFU per ml was added to each
supernatant fluid was withdrawn, centrifuged and kept at 4 °C sample in 24-well plates. The plates were incubated at 37 °C in a
prior to use. Cells were incubated in 96-well culture plates at a humidified atmosphere with 5% CO2 for 24 h. After the incubation
density of 3  104 cells mL1 in each well and incubated for 24 h procedure, 100 lL bacterial suspension were plated in duplicate on
to allow attachment. The medium was then replaced with 100 lL the agar plates. CFUs were determined on the plates after incuba-
sample extracts. The culture medium was used as the negative tion for another 24 h. Meanwhile, samples were gently rinsed three
control and the culture medium added with 10% dimethylsulfoxide times with PBS in order to eliminate the non-adherent bacteria and
(DMSO, Invitrogen, USA) as the positive control. After 1, 2 and 4 then transferred to new tubes containing 5 mL of PBS. The samples
days of incubation, 10 lL Cell Counting Kit-8 (CCK-8, Dojindo, in the tubes were then sonicated for 5 min in a 40 W ultrasonic
Kumamoto, Japan) was added to each well and incubated for 1 h. bath at a nominal frequency of 43,000 Hz followed by vortexing
The spectrophotometrical absorbance of samples was measured for 30 seconds to remove the adherent bacteria into the PBS solu-
at 450 nm using a microplate reader (Bio-RAD680). tions. The viable organisms in the PBS buffers were quantified by
H. Yang et al. / Acta Biomaterialia 71 (2018) 200–214 203

plating serial dilutions on the LB agar plates. The plates were incu- frame averaging of 2 degree, 360 degree rotation) at week 4 and 8,
bated for 24 h at 37 °C in a humidified atmosphere with 5% CO2, respectively. CT images were reconstructed using Skyscan NRecon
and the colony-forming units were counted with the naked eye. software (Bruker micro-CT N.V., Kontich, Belgium). The volumes of
The antibacterial rates for planktonic bacteria and adhered bacteria remained metallic implants were determined in CTAn software
were calculated based on the formulas: (1) Antibacterial rate for and their 3D images were yield by CTVol and CTVox software (Bru-
planktonic bacteria (Rp) (%) = (B  A)/B  100% and (2) Antibacte- ker micro-CT N.V., Kontich, Belgium).
rial rate for adherent bacteria (Ra) (%) = (D  C)/D  100%. Here,
A indicates the average number of viable bacteria in the culture 2.7.4. Histomorphometric analysis
medium incubated with pure Zn and Zn-HA composites, B is the The retrieved implants were processed for hard tissue (n = 16)
average number of viable bacteria in the culture medium incu- and soft tissue (n = 16) staining. After fixation, the implants were
bated with Ti6Al4V, C is the average number of viable bacteria rinsed in water, dehydrated in ethanol, cleared in xylene, and
adhered on pure Zn and Zn-HA composites, D is the average num- embedded in methyl methacrylate. Each implant was cut along
ber of viable bacteria adhered on Ti6Al4V. its vertical axis giving 5–6 sections per implant. The three most
central sections from each specimen, which correspond to full-
2.7. In vivo tests length sections, were kept for grinding to 100 lm thickness, pol-
ishing, staining and histological analysis. Selected sections were
2.7.1. Surgical implantation ground to a thickness of 100 lm, polished and stained with van
All the animal procedures and experiments were approved by Gieson Picrofuschin. At least 8 sections were selected for hard tis-
the Animal Ethical Committee at the Ninth People’s Hospital affil- sue histological analysis at each time point in each group. For soft
iated to Shanghai Jiaotong University, School of Medicine (Shang- tissue staining, the samples were decalcified in 10% EDTA for 6
hai, China). A rat femur condyle model was used and the surgical weeks, and then embedded in paraffin. Histological sections were
procedures were conducted under sterile conditions. The rats were prepared for Hematoxylin & eosin (H&E), Masson’s trichrome and
anesthetized by intraperitoneal injection of ketamine (10 mg/kg, Tartrate-resistant acid phosphatase (TRAP) staining. At least 8 sec-
Shanghai Ziyuan Pharmaceutical Co. Ltd, Shanghai, China) and 2% tions per staining were selected for soft tissue histological analysis
xylazine (5 mg/kg, BayerAG, Leverkusen, North Rhine-Westphalia, at each time point in each group. The specimens were then exam-
Germany). Buprenorphine (Temgesic, Reckitt & Cloman, Hull, UK) ined and photographed under a high-quality microscope (Olympus
was given subcutaneously by a dose of 0.3 mg/kg for postoperative CKX41, Olympus Co. Ltd., Tokyo, Japan) and automatic digital slide
analgesia. Each rat was immobilized in maximally flexed position scanner (Pannoramic MIDI, 3D HISTECH, Budapest, Hungary).
with the knee joint, while both legs were shaved and depilated.
Afterwards, a 10 mm longitudinal incision was made along the lat-
2.7.5. Cross sectional analysis
eral side of the extensor mechanism around the knee joint. The
The hard tissue blocks were cut to create sections of 1–1.5 mm
femoral lateral condyle in each rat was sufficiently exposed with
thickness. Cross sections were prepared by grinding the exposed
the knee in flexion, and a hole (2 mm in diameter and 5 mm in
implants in a metallographic fashion with 1200, 2000 and 7000 grit
depth) was drilled through the distal of the epiphysis gap in the
SiC paper and then polishing with a water soluble 0.5 lm diamond
bone. After washing the bone cavity with normal saline to remove
slurry on microfiber. At least 8 sections were produced at each
debris, pure Zn or Zn-5HA composite were inserted into the left or
time point in each group. The polished sections were coated with
right femur condyle of rats in a blind manner. A total of 32
a thin layer of gold to improve conductivity prior to imaging with
implants were used. Two time points were selected (week 4 and
the scanning electron microscope (SEM). A SEM (Hitachi S-4800,
8) and 8 implants were implanted at each time point in each group.
Japan) equipped with energy dispersive spectrometry (EDS) was
Each rat received one implant. The wound was closed carefully.
used for examining the cross sections.
Thirty-two male Sprague Dawley rats aged 3 months and weighed
by an average of 200 g (180–235 g) were randomized to either
pure Zn or Zn-5HA group. Just after the surgery, an X-ray scan 2.8. Statistical analysis
was performed to control the position of the implant. After surgery,
the rats were housed in groups of four in clear plastic cages on Statistical analysis was performed with SPSS 18.0 software
standard bedding in ventilated rooms with a 12 h day/night cycle package (SPSS Inc. Chicago. USA). Statistical significance of differ-
and an ambient temperature of 21 °C, and given access to water ence between groups was assessed by one-way analysis of variance
and standard laboratory pellets. At 4 or 8 weeks, the rats were sac- (ANOVA) followed by post hoc Turkey’s multiple comparison test.
rificed and the rat condyles were explanted and fixed in 10% neu- P values less than 0.05 were considered statistically significant.
tral formalin buffer for micro-CT analysis and histomorphometric
observation. 3. Results

2.7.2. X-ray scanning 3.1. Microstructures


The X-ray machine (Digital Diagnost, Philips, Amsterdam,
Netherlands) at the operation conditions was set as follows: 52 The relative density of as-sintered pure Zn and Zn-HA compos-
kV and 3.2 mAs. The X-ray scanning was then performed at week ites are listed in Table 1. The densification of all materials was
0 (n = 32), week 4 (n = 32) and week 8 (n = 16) to evaluate the higher than 90%. Addition of HA caused a decrease of the relative
implants. density. The microstructure of pure Zn and Zn-HA composites is
shown in Fig. 1. Pure Zn consisted of particles with a large variety
2.7.3. Micro-CT analysis of sizes and shapes while the particle size distribution was more
Micro-CT analysis of pure Zn and Zn-HA composite implants uniform in Zn-HA composites (Fig. 1a). HA was found homoge-
was made using Computer Tomographic data obtained with a Sky- neously distributed along the Zn particle boundaries. With increas-
scan 1172 micro-CT system (Bruker micro-CT N.V., Kontich, Bel- ing content of HA, more conglomerations were formed in particle
gium). A total of 8 implants were scanned at each time point in boundaries. The phase constituents were determined by X-ray
each group. Explanted rat condyles were examined with a 20 lm diffraction (Fig. 1b). Based on the diffraction pattern, Zn was iden-
resolution protocol (100 kV, Al + Cu filter, 0.4 degree rotation step, tified as the major phase and the characteristic peaks of HA were
204 H. Yang et al. / Acta Biomaterialia 71 (2018) 200–214

Table 1 domain was associated with the corrosion products formed on


Compositions, SPS conditions and relative densities of pure Zn and Zn-HA composites. the uniform corrosion regions on the sample surface. The capaci-
Materials SPS Conditions Density (gcm3) Relative density (%) tive loop at low frequency domain was related to the interfacial
Pure Zn 40 MPa 7.106 (0.036) 99.524 (0.505) charge transfer process and electrochemical double-layer effects
Zn-1HA 380 °C 6.806 (0.019) 96.503 (0.262) at the sample/electrolyte interface. The diameters of semicircles
Zn-5HA 6 min 6.385 (0.022) 94.925 (0.312) obtained from Zn-HA composites were smaller than that of pure
Zn-10HA 5.912 (0.039) 92.995 (0.546) Zn, indicating a decrease in corrosion resistance.
Numbers in the brackets represent the standard deviation.

3.4. Immersion tests

The degradation property of pure Zn and Zn-HA composites was


evaluated by immersion tests in Hank’s solution at 37 °C for 50
days (Fig. 4). The corrosion morphologies of experimental samples
after immersion are shown in Fig. S1 and Fig. 4a1. From a macro
level, pure Zn kept almost intact with a uniform corrosion mode
while corrosion pits populated with increasing HA contents. Local-
ized corrosion was dominant in Zn-HA composites and severe cor-
rosion attack was found in Zn-10HA composite sample. The SEM
images indicated a continuous and compact corrosion layer of pure
Zn. Fine precipitates distributed homogenously on the sample sur-
face. The scratches formed during polishing were still visible indi-
cating the corrosion layer was relatively thin. Similar morphology
was seen in Zn-1HA surface with some cracks showed up. An
Fig. 1. Microstructure of pure Zn and Zn-HA composites. (a) Optical micrographs of increasing amount of corrosion products and bigger spherical pre-
pure zinc, Zn-1HA, Zn-5HA andZn-10HA composites. (b) X-ray diffraction patterns cipitates were generated in Zn-5HA and Zn-10HA composites. After
of pure Zn and Zn-HA composites and (c) EDS analysis of area A. The inserted removal of corrosion products, the fresh surface of pure Zn was
diffraction pattern is the magnified XRD pattern of Zn-10HA in the 2h range of 25–
35°.
smooth and integral. Some corrosion pits and particle boundaries
were observed in Zn-1HA composite sample. The surface of Zn-
5HA and Zn-10HA composites were corroded severely. Lamellar
identified in the 2h range of 25–35°. Moreover, Ca and P peaks were corrosion morphology was found typically in Zn matrix. Corrosion
detected from the HA agglomerates in area A (Fig. 1c). pits distributed in Zn matrix and particle boundaries. The agglom-
eration of HA can be clearly seen in a white feature due to poor
conductivity.
3.2. Mechanical properties
The corrosion rates calculated from the weight loss was dis-
played in Fig. 4a2. Addition of HA resulted in a wide range distribu-
The mechanical properties of pure Zn and Zn-HA composites are
tion of corrosion rates compared with pure Zn. Zn-1HA composite
shown in Fig. 2. The addition of HA showed little influence on the
exhibited a slight decreased corrosion rate while addition of HA
yield strength of pure Zn (Fig. 2a). Unfortunately, a deterioration in
higher than 5 wt.% accelerated the corrosion rates significantly
tensile strength was seen with increasing HA contents. Hydroxya-
(P < 0.05). The corrosion rate of Zn-10HA composite was about 5
patite showed little impacts on the microhardness of Zn matrix
times as that of pure Zn (24.88 ± 2.33 lm/y versus 4.97 ± 1.32
(Fig. 2b). The compression stress-strain curves (Fig. 2c) indicated
lm/y). The trend of released Zn2+ concentrations in Hank’s solution
excellent plasticity of pure Zn with more than 90% compression
was in consistent with the results of weight loss (Fig. 4a3). The
strain. However, a significant reduction in plasticity could be seen
released Ca2+ concentrations followed an opposite trend that may
in composites with HA contents higher than 5 wt.%.
due to the increasing amounts of corrosion products and precipi-
tates formed with increasing HA contents (Fig. 4a4). The evolution
3.3. Electrochemical measurements of pH values during immersion is shown in Fig. 4a5. Pure Zn and
Zn-HA composites shared a similar trend of pH evolution during
The corrosion of the experimental samples in Hank’s solution immersion in Hank’s solution. A rapid increase in pH values was
was evaluated using a short-term polarization measurement seen in the first 7 days after which a relatively steady state was
including potentiodynamic polarization curves and Nyquist plots achieved. The degradation of Zn-HA composites induced higher
(Fig. 3). The corrosion current density (Icorr) and corrosion poten- pH values than that of pure Zn. Their pH values reached to approx-
tials (Ecorr) were determined by the Tafel extrapolation method, imately 8.6 after 50 day’ immersion.
and the calculated electrochemical parameters are listed in Table 2. Fig. 4b shows the composition analysis of experimental speci-
A significant increase of Icorr with HA contents was observed in con- mens after immersion. The cross sections revealed distinct struc-
trast to pure Zn (p < 0.05). The Icorr of Zn-10HA composite was tures of corrosion products formed on pure Zn and Zn-HA
more than 10 times as that of pure Zn (51.04 ± 1.80 lA cm2 versus composites (Fig. 4b1). The SEM image was composed by epoxy,
4.90 ± 2.81 lA cm2). The Ecorr of Zn-HA composites was decreased corrosion products and composite matrix. A compact, roughly 5-
as well. The polarization curves of Zn-HA composites showed a dis- lm-thick, corrosion layer was seen covering the pure Zn surface.
tinct features compared with that of pure Zn (Fig. 3a). Current pla- EDS mapping revealed Zn, O and P as the main components of cor-
teaus could be seen in the curves of Zn-HA composites which rosion products. The fresh Zn surface was relatively smooth and
indicated a passivation behavior during corrosion. The electro- continuous. In contrast, irregular and non-uniform corrosion prod-
chemical impedance spectroscopy (EIS) responses of Zn-HA com- ucts were seen for Zn-10HA composite. Some rod-like products
posites characteristically appeared two semicircle-like curves seemed to grow from the substrate. The corrosion products mainly
which corresponded to two time constants, mainly including two consisted of Zn, O and P. Moreover, HA agglomerations rich in Ca, P
capacitive loops (Fig. 3b). The capacitive loop at high frequency and O was visible in Zn particle boundaries. The Zn matrix beneath
H. Yang et al. / Acta Biomaterialia 71 (2018) 200–214 205

Fig. 2. Mechanical properties of pure Zn and Zn-HA composites. (a) Compressive strength, (b) microhardness and (c) compressive stress-strain curves, *p < 0.05.

Fig. 3. Electrochemical measurements of pure Zn and Zn-HA composites in Hank’s solution. (a) Polarization curves, (b) Nyquist plots.

Table 2 extracts was 51.99 ± 8.5 at the first day and continuously
Electrochemical parameters for pure Zn and Zn-HA composites in Hank’s solution. decreased over time. According to the ISO 19003-5 [20], pure Zn
Materials Icorr (lAcm2) Ecorr (V) Corrosion rate (mmyear1) was cytotoxic to MC3T3-E1 cells. In contrast, significant improve-
Pure zinc 4.900 (2.810) 0.942 (0.075) 0.073 (0.042)
ments in cell viability can be observed in Zn-HA composite groups
Zn-1HA 21.076 (3.251)* 1.281 (0.037)* 0.327 (0.050)* (p < 0.05). A good biocompatibility to MC3T3-E1 cells was found
Zn-5HA 39.127 (0.661)* 1.274 (0.005)* 0.630 (0.011)* for all Zn-HA composites.
Zn-10HA 51.044 (1.803)* 1.290 (0.010)* 0.856 (0.031)*

Numbers in the brackets represent the standard deviation, *p < 0.05, compared with 3.5.2. Hemocompatibility
pure zinc. The morphologies of adhered human platelets on the experi-
mental samples are shown in Fig. 5a. Most platelets adhered on
pure Zn stayed round with no pseudopodia spreading. However,
the corrosion products showed a jagged edge. No clear boundary
platelets on the surface of Zn-1HA and Zn-5HA samples stretched
was found between HA and corrosion products.
a small amount of pseudopodia, which may indicate a beginning
EDS and XRD were used to further investigate the chemical
of early dendritic. As for Zn-10HA composite, the number of
compositions of corrosion products (Fig. 4b2, b3). Zn, O and P were
adhered platelets decreased greatly and most of them showed a
major elements in corrosion products of pure Zn. With increasing
round morphology. It is important to note that platelets adhered
HA, the proportion of Zn decreased while O, P and Ca increased.
in a homogeneous way without aggregation on all the samples.
Moreover, the Ca/P atom ratio of corrosion products increased
Therefore, the thrombosis of pure Zn and Zn-HA composites was
from 0.54 of pure Zn to 1.41 of Zn-10HA composite. XRD spectra
expected to be low. The hemolysis rates of all experimental sam-
of pure Zn identified ZnO and CaCO3 as main crystallized products.
ples were far below the safe value of 5% (Fig. 5b), suggesting a
As for Zn-HA composites, clear Bragg peaks of Zn3(PO4)24H2O
low risk of hemolysis according to ASTM F 756-08 [21].
were identified as well. The results of EDS and XRD provided clear
Activated partial thromboplastin time (APTT), prothrombin
evidence of a change in surface condition with different HA
time (PT) and thrombin time (TT) are critical parameters in clinical
contents.
blood examinations and in safety evaluations of biomaterials
[22,23]. The effects of pure Zn and Zn-HA composites on APTT,
3.5. In vitro tests PT and TT are shown in Fig. 5c. In contrast to the control group
(normal value of human blood plasma), pure Zn and Zn-HA com-
3.5.1. Biocompatibility posites induced a significant increase in PT and APTT while a
Fig. S2 illustrates the cell viability of MC3T3-E1 cells after 1, 2 decrease in TT (p < 0.05). APTT and PT prolonged from 37.37 ±
and 4 days’ incubation in pure Zn and Zn-HA composite extraction 1.76 s and 11.4 ± 0.17 s for the control to 58.15 ± 0.92 s and
mediums. The viability of MC3T3-E1 cells cultured in pure Zn 13.00 ± 0.28 s for the Zn-10HA composite, respectively. TT was
206 H. Yang et al. / Acta Biomaterialia 71 (2018) 200–214

Fig. 4. In vitro immersion test. (a) Degradation behavior of pure Zn and Zn-HA composites in Hank’s solution for 50 days. (a1) Representative images of corrosion morphology
of samples before and after removal of corrosion products. (a2) Corrosion rates calculated by weight loss. (a3) Zn ion and (a4) Ca ion concentrations in Hank’s solution. (a5)
Evolution of pH values with immersion time, *p < 0.05. (b) Corrosion products analysis of experimental samples. (b1) EDS mapping of corrosion products of Zn and Zn-10HA
composite, maps of carbon, calcium, zinc, oxygen and phosphate are in green, yellow, red, blue and light blue, respectively. (b2) Elemental compositions and (b3) X-ray
diffraction patterns of corrosion products. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
H. Yang et al. / Acta Biomaterialia 71 (2018) 200–214 207

Fig. 5. Hemocompatibility of pure Zn and Zn-HA composites. (a) Platelets adhesion on experimental samples, (b) hemolysis percentage, (c) coagulation time including
activated partial thromboplastin time (APTT), prothrombin (PT) and thrombin time (TT) and (d) Zn ion concentrations in platelet poor plasma (PPP). Control group represents
PPP without incubation of experimental samples, *p < 0.05, compared with pure Zn, #p < 0.05, compared with control group.

shorten from 12.10 ± 0.26 for the control to 7.45 ± 0.07 for the Zn- experimental samples. The Zn2+ concentrations and pH values in
10HA composite. Moreover, both APTT and PT were prolonged the LB medium cultured with experimental samples for 24 h are
while TT was shortened significantly compared with pure Zn (p illustrated in Fig. 6d. The medium cultured with Ti-6Al-4V con-
< 0.05). For a better understanding of this effect, Zn2+ concentra- tained almost no Zn2+. In contrast, the Zn2+ concentrations were
tions in platelet poor plasma (PPP) are showed in Fig. 5d. The pro- significantly increased in the medium cultured with pure Zn and
longation of APTT, PT and reduction of TT in test groups were Zn-HA composites (p < 0.05). The pH values of culture medium
induced by a synchronously increase in Zn2+ concentrations in were ranging from 7.05 to 7.15, no statistic significant was found
PPP compared with control group (p < 0.05). between test and control groups.

3.5.3. Antibacterial property 3.6. In vivo tests


In order to examine the antibacterial effects of pure Zn and Zn-
HA composites, S. aureus were incubated for 24 h with experimen- 3.6.1. X-ray scanning and micro-CT analysis
tal samples. After incubation, SEM was performed to investigate Zn-5HA composite was chosen for animal tests with pure Zn as
the effects of materials on the bacterial morphology and mem- control. Fig. 7a displays the X-ray radiographs of rat femoral con-
brane integrity (Fig. 6a). A mass of smooth and intact S. aureus bac- dyle with implants (red arrow). The images showed that pure Zn
terial cells on the surface of Ti-6Al-4V sample are found which and Zn-5HA composite implants were all located well at the sur-
clustered into grapelike colonies, indicating the formation of bio- gery sites without dislocation or loosening. The distinct profiles
film [24]. In contrast, only a small amount of bacterial cells were of implants indicated good radiopacity of Zn based materials.
observed adhered on the surface of pure Zn and Zn-HA composites, Moreover, no gas shadow was observed around the implants. No
indicating an inhibition of bacteria adhesion and biofilm formation. significant degradation was seen for both pure Zn and Zn-5HA
Antibacterial rates for S. aureus in the medium (Rp) and on the composite during the entire implantation period. There was no
specimen surface (Ra) were evaluated by the spread plate method obvious inflammatory reactions or adverse effects around the
using Ti-6Al-4V as control (Fig. 6b, c). It can be seen that the Rp val- implants. New bone formation and osseointegration are studied
ues of all the specimens reached more than 90% and the Ra values at week 4 and week 8 using micro-CT (Fig. 7b). Newly formed bone
were 80–90% as well, indicating a good antibacterial effect of was visualized in both groups after 4 weeks’ implantation, indicat-
208 H. Yang et al. / Acta Biomaterialia 71 (2018) 200–214

Fig. 6. Antibacterial properties of pure Zn and Zn-HA composites to S. aureus examined by the spread plate method with Ti6Al4V as control. (a) SEM morphology of S. aureus
seeded on the sample surfaces. Antibacterial rates estimated from the amount of living bacteria (b) in the medium and (c) on the sample surfaces. (d) Zn ion concentrations
and pH values in LB medium after culturing with experimental samples for 24 h, #p < 0.05, compared with control group.

Fig. 7. Micro-CT analysis of femoral condyle containing pure Zn and Zn-5HA composite implants at week 0, 4 and 8. (a) Radiographs and (b) micro-CT reconstruction images
of femoral condyle with implants. (c) 3D reconstruction models and (d) volume of metallic part of implants remained.

ing a good osteogenic property. An increase in bone volume could However, pure Zn, with tighter contact between implant and
be seen for both implants from week 4 to week 8. In contrast, the new bone tissue, exhibited better osseointegration than the com-
bone mass in Zn-5HA composite was higher than that of pure Zn. posite. The integrity was maintained for both implants during 8
H. Yang et al. / Acta Biomaterialia 71 (2018) 200–214 209

weeks’ implantation. The three-dimensional analysis of implants 4. Discussion


showed a homogeneous and slow degradation progress over the
observation period (Fig. 7c). The 3D images of Zn-5HA composite 4.1. In vitro and in vivo degradation mechanism of Zn-HA composites
exhibited a rougher surface morphology with more severe corro-
sion attack than pure Zn. After 8 weeks, the volume of Zn-5HA The basic degradation process of Zn-HA composites is similar to
composite dropped to 96.80 ± 0.56% compared to 98.30 ± 1.02% that of pure Zn and the schematic graphs are shown in Fig. S5. After
on the pure Zn implant (Fig. 7d). immersed in Hank’s solution, Zn generally corrodes by the oxida-
tion of Zn to Zn2+ (Eq. (1)) and the electrons generated from the
anodic reaction are consumed by the corresponding cathodic reac-
3.6.2. Degradation products analysis tion via reduction of oxygen (Eq. (2)), which give rise to generation
Fig. 8 shows the morphology of the bone-implant cross sections of OH. With corroding of Zn, pH value increased rapidly at early
and the elemental mappings of bone-implant interface after 4 and stage as shown in Fig. 4a5. Meanwhile, the released Zn2+ reacts
8 weeks’ implantation. Generally, pure Zn went through a macro- with the OH to form degradation products like oxide and hydrox-
scopic uniform corrosion mode without severely localized corro- ide (Eqs. (3)–(4)) which was found at the early stage of immersion
sion (Fig. 8a). No significant degradation was observed during (Fig. S4).
implantation. Only small amounts of degradation products were
visible surrounding the implant. The bone-implant interface of Zn ! Zn2þ þ 2e ð1Þ
pure Zn was composed by Zn matrix, degradation products, fibrous
tissue and bone. The EDS mapping illustrated that the degradation O2 þ 2H2 O þ 4e ! 4OH ð2Þ
products presented two-layer structure. The inner layer which
adjacent to the metal surface was rich in C, O and Zn, while the Zn2þ þ 2OH ! ZnðOHÞ2 ð3Þ
outer layer consisted of more P. Bone was rich in Ca and P. The
specific atomic percentages of degradation products was shown ZnðOHÞ2 ! ZnO þ H2 O ð4Þ
in Fig. S3 by select point analysis. In contrast to pure Zn, localized
corrosion attack was observed in Zn-5HA composite and its degra- Moreover, Zn(OH)2 can transform into more stable ZnO with
dation was more noticeable (Fig. 8b). Similar to pure Zn, the integ- more exposure time [25]. However, the surface oxides thermody-
rity of the composite implant was maintained at week 8. For Zn- namically predicted to generate in neutral solution do not form
5HA composite, the degradation products were accumulated an effective corrosion protection barrier [26]. The Cl in the simu-
locally in a diffused way. At week 4, C, O and Zn was detected as lated body fluid readily converts ZnO and Zn(OH)2 to the soluble
the predominant elements in the degradation products. In contrast, ZnCl2, damaging the corrosion layer and promoting further corro-
Ca and P were detected in the outer most layer of degradation sion. Nevertheless, a dense and compact corrosion layer mainly
products at week 8. Besides, intense signal of Ca and P was consisted of Zn, O and P was observed covering the Zn matrix after
detected mainly in newly formed bone and HA. 50 days’ immersion (Fig. 4b1). This can be explained by the subse-
quently formation of more thermodynamically stable phase of zinc
phosphate (Eq. (5)) [27].
3.6.3. Histological analysis
3Zn2þ þ 2HPO4 2 þ 2OH þ 2H2 O ! Zn3 ðPO4 Þ2  4H2 O ð5Þ
Fig. 9 shows the tissue response to pure Zn and Zn-5HA com-
posite after 8 week’s implantation using van Gieson staining. The absence of zinc phosphate in the XRD pattern (Fig. 4b3) of pure
New bone formation could be observed surrounding pure Zn and Zn may cause by its amorphous structure [28]. Zinc phosphate was
composite at week 4 and plenty of osteocytes were seen in the reported as major degradation products both in vitro and in vivo
new bone tissue. However, a thin layer of fibrous connective tissue [27,29–31]. Therefore, ZnO and Zn(OH)2 are able to be converted
was seen separating the bone tissue from the implants. For Zn-5HA to zinc phosphate depending on the shift in the precipitation disso-
composite, the localized corrosion and fast degradation rate caused lution equilibrium. The zinc phosphate contained layer is protec-
a thicker fibrous tissue layer than that of pure Zn at week 4. An tive and remains intact with immersion time, leading to a
increased bone mass, extending outward from the implants, could relative stable pH in Hank’s solution at late stage. As a result, pure
be observed in both implants over time, which indicated a benefi- Zn showed little degradation throughout the immersion time. The
cial role of pure Zn and Zn-5HA composite in promoting new bone incorporation of HA into Zn matrix results in a great influence on
formation. Zn-5HA composite was observed to stimulate larger the microstructure of pure Zn. Consequently, a difference in degra-
amounts of mineralized and newly formed bone than pure Zn at dation behavior between pure Zn and Zn-HA composites can be
week 8. In addition, the newly formed bone gradually replaced expected. HA in the composites plays biphasic roles during degra-
the fibrous connective tissue and reduced the gap between com- dation: On the one hand, HA is characterized by a poor conductiv-
posite and new bone tissue. To further investigate the inflamma- ity and low solubility under physiological condition. Thus, the
tory response, collagen formation and osteoclasts, H&E, Masson’s distribution of HA along the particle boundary creates a network
trichrome and TRAP stainings are shown in Fig. 10. At week 4, both structure that impedes the charge transfer among Zn particles. As
implants were surrounded by a thin layer of connective tissue a results, addition of HA improves the corrosion resistance of Zn
mainly containing fibroblasts. Local infiltration of lymphocytes matrix. On the other hand, the relatively low sintering temperature
and macrophages indicated a mild inflammatory response caused and pressure and lack of interface reaction between Zn and HA
by implants. Moreover, Masson’s trichrome staining illustrated during SPS process contribute to the increased porosity with HA
that collagen fiber build-up was evident and small islands of contents [32]. Therefore, pores and voids which form in the inter-
osteoid were also detected surrounding the implants. After 8 face of Zn particles and HA agglomerates promote the penetration
weeks, inflammatory reaction with increased fibrous tissue thick- of corrosive fluid in the composite matrix. This can be proved by
ness was still found in pure Zn. In contrast, less inflammatory cells the decreased relative density of Zn-HA composites (Table 1) and
and reduced fibrous tissue layer was seen surrounding the Zn-5HA cracks in the particle boundary of samples after removal of corro-
composite. Moreover, more collagen-rich bone tissue was formed sion products (Fig. 4a1). Moreover, more sites that are vulnerable
around the composite than pure Zn. Besides, no TRAP marked to corrosion attack are generated due to the physical presence of
osteoclasts was found in both groups at the selected time points. HA in the particle boundary [33]. These factors lead to the acceler-
210 H. Yang et al. / Acta Biomaterialia 71 (2018) 200–214

Fig. 8. SEM images of the cross sections. (a) Pure Zn and (b) Zn-5HA composite at week 4 and week 8 with magnified images of yellow rectangles and corresponding
elemental distribution of view in images. (Carbon (green), oxygen (purple), zinc (red), phosphate (light blue), calcium (yellow)). (For interpretation of the references to colour
in this figure legend, the reader is referred to the web version of this article.)

ated corrosion of Zn-HA composites. In this way, the degradation a corresponding increase in corrosion layer thickness (Fig. 4). The
rates of Zn-HA composites are adjustable by tuning the volume formation of this protective layer can retard the corrosion of com-
fraction of HA. In contrast to pure Zn, Zn-1HA composite showed posites. As a result, a relatively stable stage can be expected in the
a slight decreased degradation rate while addition of 5 wt.% and evolution of pH at late period (Fig. 4a5). In contrast to pure Zn, the
10 wt.% HA accelerated the degradation significantly. As the degra- apparent compositional nature of degradation products began to
dation proceeds, HA agglomerates play a positive role in sponta- change after adding HA. Elements including O, P and Ca accounted
neously absorbing and inducing the accelerated deposition of for more proportions of degradation products while percentage of
Ca2+ and PO3
4 in the solution, thereby the compact Ca, P contained Zn decreased with more HA contents. Moreover, the Ca/P atom
layer which can slow down the corrosion effectively forms on the ratio of degradation products increased from 0.54 of pure Zn to
sample surface [34–37]. A significant reduction in Ca2+ concentra- 1.41 of Zn-10HA composite, which is near the Ca/P ratio of HA
tion with increased HA contents was seen in the Hank’s solution (HA, Ca/P = 1.6). Previous studies have reported that HA in the
after immersion (Fig. 4a4). Meanwhile, Zn-HA composites showed composites could act as the hydroxyapatite nuclei and stimulated
H. Yang et al. / Acta Biomaterialia 71 (2018) 200–214 211

Fig. 9. Histological characterization of hard tissue sections at implant sites. Van Gieson staining of (a) pure Zn and (b) Zn-5HA composite at week 4 and 8. The red rectangles
correspond to the magnified bone-implant interface. The red triangle indicates newly formed bone. (For interpretation of the references to colour in this figure legend, the
reader is referred to the web version of this article.)

the growth of hydroxyapatite, which is beneficial to new bone rosion reaction proceeds as fast as dissolved oxygen reaches the
growth [34,36,38]. However, XRD is hard to confirm the newly metal surface. Nevertheless, the availability of oxygen is limited
formed HA after immersion due to the interference of HA phase. in the surrounding tissue of implants investigated here (34
Diffraction pattern of Zn-HA composites identified a mixture of mmHg) [39]. As a result, the cathodic oxygen reduction reaction
zinc oxide, zinc phosphate and calcium carbonate as degradation is inhibited which leads to the slow degradation rates of samples.
products (Fig. 4b3). Similar results have been reported in a 52-week in vivo study of
To further investigate the degradation behavior of Zn-HA com- Fe-based alloys implanted in femoral mid-diaphyseal region of rats
posites in vivo, pure Zn and Zn-5HA composite were implanted in [40]. The limited availability of oxygen is a critical factor causing
the femoral condyle of rats. Micro-CT indicated that little degrada- the low degradation rates of Fe-based implants. In our previous
tion of pure Zn took place throughout the implantation time study, we found a discrepancy in degradation rates of Zn stents
(Fig. 7). Moreover, the degradation proceeded in a relatively uni- in blood fluid and neointima, which partly due to the difference
form mode. However, the compact and protective corrosion layer of oxygen partial pressure in different degradation environments
failed to form under in vivo conditions (Fig. 8). In this respect, [40]. In contrast to pure Zn, addition of HA induced accelerated
the slow degradation rate seems to be contradicted with the degradation and a localized corrosion mode (Figs. 7 and 8). Firstly,
incomplete corrosion layer. This can be explained by the availabil- corrosion attack preferentially happens in the interface of Zn par-
ity of oxygen in the surrounding tissue, which influences the ticles and HA agglomerates. Then, the metallic parts are gradually
degradation rates of Zn-based implants in vivo. The underlying cor- corroded and replaced by degradation products. The voluminous
212 H. Yang et al. / Acta Biomaterialia 71 (2018) 200–214

Fig. 10. Histological characterization of soft tissue sections at implant sites. H&E, Masson’s trichrome and TRAP staining of (a) pure Zn and (b) Zn-5HA composite after 4 and 8
weeks’ implantation. The red, light blue and black rectangles correspond to the inserted magnified areas in H&E, Masson’s trichrome and TRAP staining, respectively.

corrosion products act as wedges along the interface allowing the the outer region induced by the release of Ca and P ions from HA
corrosive fluid to penetrate inside and keep reacting with the fresh or local tissues and a local rise in the pH value. However, the for-
matrix [41]. Similar to that of pure Zn, the dense and integral cor- mation of zinc carbonate is hard to confirm because the contami-
rosion layer formed in vitro is absent under in vivo conditions, nation of carbon during sample fabrication. The discrepancy in
which facilitates the continuous degradation of Zn-HA composites. morphology and composition of degradation products found
It is worth to mention that both pure Zn and Zn-HA composites are in vitro and in vivo suggests that the degradation of the Zn-based
able to maintain the mechanical integrity for more than 8 weeks composites is greatly influenced by the surrounding environments.
which is critical important for bone healing process [7]. Analysis The in vivo degradation process is more complex due to the expo-
of the degradation products can help us understand the in vivo cor- sure of materials to an environment rich in a diversity of ions and
rosion process of implants. The EDS investigation revealed that the organic substances such as proteins and cells.
degradation products of both pure Zn and Zn-5HA composite were
characterized by layered structure (Figs. 8, S3). The inner layer 4.2. Biocompatibility
adjacent to the matrix mainly consisted of oxidized constituents
of the material, i.e. C, O and Zn. On top, there is an outer layer that, Formation of new bone was visualized by micro-CT, SEM and
besides C, O and Zn, also contained elements that originate from histological analysis in both implants at week 4 (Figs. 7–10). How-
the surrounding tissue, i.e. P, Ca. Several works have reported the ever, both implants showed a fibrotic and collagenous encapsula-
formation of zinc oxide, zinc carbonate and Ca/P phase as major tion response separating the new bone tissue and materials. In
degradation products of pure Zn wire in abdominal aorta of rats contrast, Zn-5HA composite was surrounded by a thicker fibrous
[2,42], but we fail to find any references related to the in vivo connective tissue than pure Zn. An increased bone mass could be
degradation of Zn based materials in bone environments. In this seen with prolonged implantation time in both implants. In addi-
study, zinc oxide is believed to be one of the main components tion, Zn-5HA composite showed a more pronounced effect in stim-
of degradation products. Besides, Ca/P phase could generate in ulating new bone formation at week 8. And fibrous layer was
H. Yang et al. / Acta Biomaterialia 71 (2018) 200–214 213

gradually replaced by new bone tissue over time. Nevertheless, the 5. Conclusions
lack of direct bone bonding to the implants at week 8 indicates a
delayed osseointegration in both pure Zn and composite. Gener- Incorporation of HA into pure Zn is an effective approach to
ally, for biodegradable materials, the tissue response is closely adjust its degradation rate and improve its biocompatibility both
related to the degradation process of materials. The in vivo degra- in vitro and in vivo. A wide range of degradation rates can be
dation of pure Zn creates solid products like oxide, hydroxide and achieved by changing the concentrations of HA. Cell test indicated
Ca/P compound and releases chemical species like Zn ions and a significant better cytocompatibility of Zn-HA composites. In vivo
hydroxide ions. Apart from this, the disintegration and dissolution tests showed that addition of HA resulted in a better performance
of hydroxyapatite generates Ca and P ions from Zn-HA composites in osteogenesis with prolonged implantation time. Further studies
[43–45]. Zn, as an essential element for basic biological function, will investigate the combinations of Zn alloys with better mechan-
shows biphasic effects on new bone formation. Previous studies ical performance as matrix and more bioactive reinforcements.
have reported enhanced mineralization of extracellular matrix Moreover, a better interface bonding between matrix and addi-
and human bone marrow mesenchymal cells (hMSC) osteogenic tions is necessary to improve the corrosion uniformity. This sys-
differentiation and increased expression of bone-related genes tematic study suggests that Zn based MMC can be a promising
including alkaline phosphatase, collagen I and osteopontin when strategy to improve the performance of pure Zn for applications
hMSC were cultured with Zn or Zn ions [46–48]. Zn also exhibited in orthopedic implants.
an inhibitory effect on osteoclast differentiation [49,50]. In vivo
animal studies showed efficient promotion on new bone formation
Acknowledgements
and bone contact induced by Zn incorporated materials [48,51,52].
However, excessive release of Zn ions can induce a cytotoxicity and
This work was supported by the National Key Research and
DNA damage on human osteoblast cells and may even cause bone
Development Program of China (Grant No. 2016YFC1102402),
resorption in vivo [10,51,53]. Hydroxyapatite is one of the fre-
National Natural Science Foundation of China (Grant No.
quently used bioceramics for bone tissue reconstitution. It pos-
51431002), NSFC/RGC Joint Research Scheme (Grant No.
sesses excellent biocompatibility with hard tissue, high
51361165101 and 5161101031) and NSFC-RFBR Cooperation Pro-
osteoconductivity and bioactivity [54]. Hydroxyapatite scaffold
ject (Grant No. 51611130054). Hongtao Yang and Xinhua Qu con-
has shown an enhanced effect on the proliferation rate and osteo-
tributed equally to this study.
genic differentiation of hMSC [55]. Moreover, promotion of new
bone formation by hydroxyapatite was confirmed in animal and
clinical tests [43,56]. As a result, HA is a favorable choice to incor- Appendix A. Supplementary data
porate into matrix like magnesium or polymer in order to increase
their biocompatibility and bioactivity [18,57,58]. In the present Supplementary data associated with this article can be found, in
study, the in vitro cell test showed a great toxicity of pure Zn on the online version, at https://doi.org/10.1016/j.actbio.2018.03.007.
MC3T3-E1 cells while addition of HA improved the cytocompatibil-
ity significantly (Fig. S2). In comparison, the effects of HA are more
References
complicated under in vivo conditions. In the early stage of implan-
tation, an increased amount of fluid accumulates in the cancellous [1] T. Kraus, S.F. Fischerauer, A.C. Hänzi, P.J. Uggowitzer, J.F. Löffler, A.M. Weinberg,
tissue as a result of tissue trauma. Meanwhile, a haematoma is Magnesium alloys for temporary implants in osteosynthesis: in vivo studies of
generated and consisted of cells from blood and bone marrow, fol- their degradation and interaction with bone, Acta Biomater. 8 (3) (2012) 1230–
1238.
lowed by an inflammatory response [59]. Therefore, a relatively [2] P.K. Bowen, J. Drelich, J. Goldman, Zinc exhibits ideal physiological corrosion
rapid degradation with large amounts of Zn ions released from behavior for bioabsorbable stents, Adv. Mater. 25 (18) (2013) 2577–2582.
implants is expected at this time. As a result, the inhibitory effect [3] C.J. Frederickson, K. Jae-Young, A.I. Bush, The neurobiology of zinc in health
and disease, Nat. Rev. Neurosci. 6 (6) (2005) 449–462.
of Zn ions is predominant and a decreased Zn ion gradient can form [4] M. Yamaguchi, Role of zinc in bone formation and bone resorption, J. Trace
with increasing distance from the implant. In the vicinity of Elem. Exp. Med. 11 (1998) 119–135.
implant, Zn ion concentration is too high to allow new bone forma- [5] H.F. Li, X.H. Xie, Y.F. Zheng, Y. Cong, F.Y. Zhou, K.J. Qiu, X. Wang, S.H. Chen, L.
Huang, L. Tian, L. Qin, Development of biodegradable Zn-1X binary alloys with
tion. Consequently, a fibrous connective tissue instead of new bone
nutrient alloying elements Mg, Ca and Sr, Sci. Rep. 5 (2015) 10719.
forms, which causes delayed osseointegration. The addition of HA [6] D. Vojtěch, J. Kubásek, J. Šerák, P. Novák, Mechanical and corrosion properties
induced accelerated degradation and localized corrosion which of newly developed biodegradable Zn-based alloys for bone fixation, Acta
Biomater. 7 (9) (2011) 3515–3522.
led to a thicker fibrous tissue layer than pure Zn (Figs. 8 and 9).
[7] Y.F. Zheng, X.N. Gu, F. Witte, Biodegradable metals, Mater. Sci. Eng., R 77
As bone healing progresses, the haematoma is replaced by a (2014) 1–34.
fibrin-rich granulation tissue, followed by formation of cartilagi- [8] I.S. Berglund, B.Y. Jacobs, K.D. Allen, S.E. Kim, A. Pozzi, J.B. Allen, M.V. Manuel,
nous tissue and mineralization. Less fluid and reduced inflamma- Peri-implant tissue response and biodegradation performance of a Mg-1.0Ca-
0.5Sr alloy in rat tibia, Mater. Sci. Eng., C 62 (2016) 79–85.
tion accompanied this process. As a result, the degradation of [9] N. Erdmann, N. Angrisani, J. Reifenrath, A. Lucas, F. Thorey, D. Bormann, A.
implants slows down and less Zn ions are released over time. Then, Meyer-Lindenberg, Biomechanical testing and degradation analysis of
the beneficial role of Zn ions in stimulating new bone formation MgCa0.8 alloy screws: a comparative in vivo study in rabbits, Acta Biomater.
7 (3) (2011) 1421–1428.
becomes more prevailed. In contrast to pure Zn, Zn-5HA composite [10] N.S. Murni, M.S. Dambatta, S.K. Yeap, G.R. Froemming, H. Hermawan,
exhibited a better performance in osteogenesis at week 8, which Cytotoxicity evaluation of biodegradable Zn-3Mg alloy toward normal
induced by the combined promotion effects of Zn and HA. Despite human osteoblast cells, Mater. Sci. Eng., C 49 (2015) 560–566.
[11] J. Kubásek, D. Vojtěch, E. Jablonská, I. Pospíšilová, J. Lipov, T. Ruml, Structure,
the larger release of Zn ions at early stage, Zn-HA composites pro- mechanical characteristics and in vitro degradation, cytotoxicity, genotoxicity
duced superior biological responses in bone growth than pure Zn and mutagenicity of novel biodegradable Zn–Mg alloys, Mater. Sci. Eng., C 58
with prolonged implantation time. According to the degradation (2016) 24–35.
[12] Z. Tang, H. Huang, J. Niu, L. Zhang, H. Zhang, J. Pei, J. Tan, G. Yuan, Design and
rates, the released Zn for pure Zn and Zn-5HA composite was about
characterizations of novel biodegradable Zn-Cu-Mg alloys for potential
0.035 and 0.059 mg/day, both of which are far below the daily biodegradable implants, Mater. Des. 117 (2017) 84–94.
allowance of zinc (15 mg/day) [2]. Therefore, the biosafety of Zn- [13] C. Ohtsuki, M. Kamitakahara, T. Miyazaki, Bioactive ceramic-based materials
with designed reactivity for bone tissue regeneration, J. R. Soc., Interface 6 (3)
HA composites will be guaranteed.
(2009) S349–S360.
214 H. Yang et al. / Acta Biomaterialia 71 (2018) 200–214

[14] H. Kawahara, Bioceramics for hard tissue replacements, Clin. Mater. 2 (3) [38] S.Z. Khalajabadi, M.R.A. Kadir, S. Izman, M. Marvibaigi, The effect of MgO on
(1987) 181–206. the biodegradation, physical properties and biocompatibility of a Mg/HA/MgO
[15] J. Edwards, J. Brunski, H. Higuchi, Mechanical and morphologic investigation of nanocomposite manufactured by powder metallurgy method, J. Alloys Compd.
the tensile strength of a bone-hydroxyapatite interface, J. Biomed. Mater. Res., 655 (2016) 266–280.
Part A 36 (4) (1997) 454–468. [39] T. Kiaer, B. Dahl, G. Lausten, Partial pressures of oxygen and carbon dioxide in
[16] K.Y. Renkema, R.T. Alexander, R.J. Bindels, J.G. Hoenderop, Calcium and bone and their correlation with bone-blood flow: effect of decreased arterial
phosphate homeostasis: concerted interplay of new regulators, Ann. Med. 40 supply and venous congestion on intraosseous oxygen and carbon dioxide in
(2) (2008) 82–91. an animal model, J. Orthop. Res. 10 (6) (1992) 807–812.
[17] X. Gu, W. Zhou, Y. Zheng, L. Dong, Y. Xi, D. Chai, Microstructure, mechanical [40] T. Kraus, F. Moszner, S. Fischerauer, M. Fiedler, E. Martinelli, J. Eichler, F. Witte,
property, bio-corrosion and cytotoxicity evaluations of Mg/HA composites, E. Willbold, M. Schinhammer, M. Meischel, Biodegradable Fe-based alloys for
Mater. Sci. Eng., C 30 (6) (2010) 827–832. use in osteosynthesis: outcome of an in vivo study after 52 weeks, Acta
[18] F. Witte, F. Feyerabend, P. Maier, J. Fischer, M. Störmer, C. Blawert, W. Dietzel, Biomater. 10 (7) (2014) 3346–3353.
N. Hort, Biodegradable magnesium-hydroxyapatite metal matrix composites, [41] W.A. Ferrando, Review of corrosion and corrosion control of magnesium alloys
Biomaterials 28 (13) (2007) 2163–2174. and composites, J. Mater. Eng. 11 (4) (1989) 299–313.
[19] M. Ulum, A. Arafat, D. Noviana, A. Yusop, A. Nasution, M.A. Kadir, H. [42] A. Drelich, S. Zhao, R.J. Guillory, J.W. Drelich, J. Goldman, Long-term
Hermawan, In vitro and in vivo degradation evaluation of novel iron- surveillance of zinc implant in murine artery: surprisingly steady
bioceramic composites for bone implant applications, Mater. Sci. Eng., C 36 biocorrosion rate, Acta Biomater. 58 (2017) 539–549.
(2014) 336–344. [43] J. Lu, M. Descamps, J. Dejou, G. Koubi, P. Hardouin, J. Lemaitre, J.P. Proust, The
[20] ISO 10993-5:2009(E), Biological evaluation of medical devices—Part 5: Tests biodegradation mechanism of calcium phosphate biomaterials in bone, J.
for in vitro cytotoxicity, International Organization for Standardization, 2009. Biomed. Mater. Res., Part A 63 (4) (2002) 408–412.
[21] ASTM F 756-08, Standard Practice for Assessment of Hemolytic Properties of [44] P. Ducheyne, Q. Qiu, Bioactive ceramics: the effect of surface reactivity on bone
Materials, ASTM International, Philadelphia West Conshohocken, PA, 2009. formation and bone cell function, Biomaterials 20 (23–24) (1999) 2287–2303.
[22] J. Wang, Y. He, M.F. Maitz, B. Collins, K. Xiong, L. Guo, Y. Yun, G. Wan, N. Huang, [45] M.T. Fulmer, I.C. Ison, C.R. Hankermayer, B.R. Constantz, J. Ross, Measurements
A surface-eroding poly(1,3-trimethylene carbonate) coating for fully of the solubilities and dissolution rates of several hydroxyapatites,
biodegradable magnesium-based stent applications: toward better Biomaterials 23 (3) (2002) 751–755.
biofunction, biodegradation and biocompatibility, Acta Biomater. 9 (10) [46] D. Zhu, Y. Su, M.L. Young, J. Ma, Y. Zheng, L. Tang, Biological responses and
(2013) 8678–8689. mechanisms of human bone marrow mesenchymal stem cells to Zn and Mg
[23] A. Mochizuki, H. Kaneda, Study on the blood compatibility and biodegradation biomaterials, ACS Appl. Mater. Interfaces 9 (33) (2017) 27453–27461.
properties of magnesium alloys, Mater. Sci. Eng., C 47 (2015) 204–210. [47] K. Yusa, O. Yamamoto, M. Fukuda, S. Koyota, Y. Koizumi, T. Sugiyama, In vitro
[24] V. Antoci, C.S. Adams, J. Parvizi, H.M. Davidson, R.J. Composto, T.A. Freeman, E. prominent bone regeneration by release zinc ion from Zn-modified implant,
Wickstrom, P. Ducheyne, D. Jungkind, I.M. Shapiro, N.J. Hickok, The inhibition Biochem. Biophys. Res. Commun. 412 (2) (2011) 273–278.
of Staphylococcus epidermidis biofilm formation by vancomycin-modified [48] X. Shen, Y. Hu, G. Xu, W. Chen, K. Xu, Q. Ran, P. Ma, Y. Zhang, J. Li, K. Cai,
titanium alloy and implications for the treatment of periprosthetic infection, Regulation of the biological functions of osteoblasts and bone formation by Zn-
Biomaterials 29 (35) (2008) 4684–4690. incorporated coating on microrough titanium, ACS Appl. Mater. Interfaces 6
[25] A. Goux, T. Pauporté, J. Chivot, D. Lincot, Temperature effects on ZnO (18) (2014) 16426–16440.
electrodeposition, Electrochim. Acta 50 (11) (2005) 2239–2248. [49] M. Hie, I. Tsukamoto, Administration of zinc inhibits osteoclastogenesis
[26] S. Thomas, N. Birbilis, M.S. Venkatraman, I.S. Cole, Corrosion of zinc as a through the suppression of RANK expression in bone, Eur. J. Pharmacol. 668
function of pH, Circ. Res. 68 (1) (2012) 160–166. (2) (2011) 140–146.
[27] H. Yang, C. Wang, C. Liu, H. Chen, Y. Wu, J. Han, Z. Jia, W. Lin, D. Zhang, W. Li, [50] M. Roy, G. Fielding, A. Bandyopadhyay, S. Bose, Effects of zinc and strontium
Evolution of the degradation mechanism of pure zinc stent in the one-year substitution in tricalcium phosphate on osteoclast differentiation and
study of rabbit abdominal aorta model, Biomaterials 145 (2017) 92–105. resorption, Biomater. Sci. 1 (1) (2013) 74–82.
[28] L. Herschke, J. Rottstegge, I. Lieberwirth, G. Wegner, Zinc phosphate as [51] H. Kawamura, A. Ito, S. Miyakawa, P. Layrolle, K. Ojima, N. Ichinose, T. Tateishi,
versatile material for potential biomedical applications Part 1, J. Mater. Sci.: Stimulatory effect of zinc-releasing calcium phosphate implant on bone
Mater. Med. 17 (1) (2006) 81–94. formation in rabbit femora, J. Biomed. Mater. Res., Part A 50 (2) (2000) 184–
[29] L. Zhao, Z. Zhang, Y. Song, S. Liu, Y. Qi, X. Wang, Q. Wang, C. Cui, Mechanical 190.
properties and in vitro biodegradation of newly developed porous Zn scaffolds [52] Y. Qiao, W. Zhang, P. Tian, F. Meng, H. Zhu, X. Jiang, X. Liu, P.K. Chu, Stimulation
for biomedical applications, Mater. Des. 108 (2016) 136–144. of bone growth following zinc incorporation into biomaterials, Biomaterials 35
[30] Y. Chen, W. Zhang, M.F. Maitz, M. Chen, H. Zhang, J. Mao, Y. Zhao, N. Huang, G. (25) (2014) 6882–6897.
Wan, Comparative corrosion behavior of Zn with Fe and Mg in the course of [53] K. Ishikawa, Y. Miyamoto, T. Yuasa, A. Ito, M. Nagayama, K. Suzuki, Fabrication
immersion degradation in phosphate buffered saline, Corros. Sci. 111 (2016) of Zn containing apatite cement and its initial evaluation using human
541–555. osteoblastic cells, Biomaterials 23 (2) (2002) 423–428.
[31] K. Torne, M. Larsson, A. Norlin, J. Weissenrieder, Degradation of zinc in saline [54] C.V.M. Rodrigues, P. Serricella, A.B.R. Linhares, R.M. Guerdes, R. Borojevic, M.A.
solutions, plasma, and whole blood, J. Biomed. Mater. Res. Part B 104 (6) Rossi, M.E.L. Duarte, M. Farina, Characterization of a bovine collagen–
(2016) 1141–1151. hydroxyapatite composite scaffold for bone tissue engineering, Biomaterials
[32] R. Kumar, K.H. Prakash, P. Cheang, K.A. Khor, Microstructure and mechanical 24 (27) (2003) 4987–4997.
properties of spark plasma sintered zirconia-hydroxyapatite nano-composite [55] T. Mygind, M. Stiehler, A. Baatrup, H. Li, X. Zou, A. Flyvbjerg, M. Kassem, C.
powders, Acta Mater. 53 (8) (2005) 2327–2335. Bünger, Mesenchymal stem cell ingrowth and differentiation on coralline
[33] C.A. Nunez-Lopez, P. Skeldon, G.E. Thompson, P. Lyon, H. Karimzadeh, T.E. hydroxyapatite scaffolds, Biomaterials 28 (6) (2007) 1036–1047.
Wilks, The corrosion behaviour of Mg alloy ZC71/SiC p metal matrix [56] C. Schopper, D. Moser, A. Sabbas, G. Lagogiannis, E. Spassova, F. König, K.
composite, Corros. Sci. 37 (5) (1995) 689–708. Donath, R. Ewers, The fluorohydroxyapatite (FHA) FRIOS Algipore is a suitable
[34] X.N. Gu, X. Wang, N. Li, L. Li, Y.F. Zheng, X. Miao, Microstructure and biomaterial for the reconstruction of severely atrophic human maxillae, Clin.
characteristics of the metal–ceramic composite (MgCa-HA/TCP) fabricated by Oral Implants Res. 14 (6) (2003) 743–749.
liquid metal infiltration, J. Biomed. Mater. Res., Part B 99 (1) (2011) 127–134. [57] X. Ye, M. Chen, M. Yang, J. Wei, D. Liu, In vitro corrosion resistance and
[35] L.T. de Jonge, S.C. Leeuwenburgh, J.G. Wolke, J.A. Jansen, Organic-inorganic cytocompatibility of nano-hydroxyapatite reinforced Mg-Zn-Zr composites, J.
surface modifications for titanium implant surfaces, Pharm. Res. 25 (10) Mater. Sci.: Mater. Med. 21 (4) (2010) 1321–1328.
(2008) 2357–2369. [58] S.S. Kim, S.P. Min, O. Jeon, C.Y. Choi, B.S. Kim, Poly(lactide-co-glycolide)/
[36] K. Yu, L. Chen, J. Zhao, S. Li, Y. Dai, Q. Huang, Z. Yu, In vitro corrosion behavior hydroxyapatite composite scaffolds for bone tissue engineering, Biomaterials
and in vivo biodegradation of biomedical b-Ca3(PO4)2/Mg-Zn composites, 27 (8) (2006) 1399–1409.
Acta Biomater. 8 (7) (2012) 2845–2855. [59] H.M. Frost, The biology of fracture healing, Injury 42 (6) (1989) 551–555.
[37] R.D. Campo, B. Savoini, L. Jordao, A. Muñoz, M.A. Monge, Cytocompatibility,
biofilm assembly and corrosion behavior of Mg-HAP composites processed by
extrusion, Mater. Sci. Eng., C 78 (2017) 667–673.

Das könnte Ihnen auch gefallen