Design and fabrication of graduated porous Ti-based alloy implants for biomedical applications

Accepted Manuscript Design and fabrication of graduated porous Ti-based alloy implants for biomedical applications Jianping Shi, Jiquan Yang, Zongan Li, Liya Zhu, Lan Li, Xingsong Wang PII:

S0925-8388(17)32919-5

DOI:

10.1016/j.jallcom.2017.08.190

Reference:

JALCOM 42947

To appear in:

Journal of Alloys and Compounds

Received Date: 23 November 2016 Revised Date:

9 August 2017

Accepted Date: 19 August 2017

Please cite this article as: J. Shi, J. Yang, Z. Li, L. Zhu, L. Li, X. Wang, Design and fabrication of graduated porous Ti-based alloy implants for biomedical applications, Journal of Alloys and Compounds (2017), doi: 10.1016/j.jallcom.2017.08.190. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Design and fabrication of graduated porous Ti-based alloy implants for biomedical applications Jianping Shia,b, Jiquan Yangb, Zongan Lia,b, Liya Zhub, Lan Lic, Xingsong Wang*a

RI PT

a School of Mechanical Engineering, Southeast University, Nanjing, 211189, China b Jiangsu Key Laboratory of 3D Printing Equipment and Manufacturing, Nanjing Normal University, Nanjing, 210042, China

SC

c Nanjing Drum Tower Hospital, Nanjing University, Nanjing, 210093, China

M AN U

Abstract

In this study, a porous implant model with controllable pores was created, where the pores were distributed with a gradient change from the surface of each pore inwards. The aim was to develop an implant with an elastic modulus of gradient change. The models were subjected to 3D finite element

TE D

analysis in order to achieve the optimal design parameters. A direct metal laser sintering process was used to print the implant models. Investigations on the physical and mechanical properties revealed

EP

that the fabricated implants had a porosity of 65.8–88.2% and an elastic modulus of 12–18 GPa. The property of the sample was close to that of cortical bone. Therefore, the stress shielding effect of the

AC C

implant and human bone could be reduced. An in vitro cell culture experiment conducted on the samples after surface modification demonstrated that the printed porous parts had good biocompatibility.

Keywords: Direct metal laser sintering, Gradient porous structure, Stress shielding effect, 3D printing 1. Introduction

*Corresponding author Xingsong Wang. E-mail:[email protected] 1 / 14

ACCEPTED MANUSCRIPT The human joint often experiences lesions and damage due to several factors, such as disease, aging, and sports injuries, which seriously affect people and their quality of life. Therefore, the repair or replacement of damaged joints with artificial joints has become an important step in the treatment

RI PT

of joint disease in the human body [1]. Presently, artificial joint replacement has become one of the biggest and most prominent areas in plastic surgery. Reports suggest approximately 1 million cases

SC

of hip implants in the world each year [2]. Currently used implant components comprise mainly of standard parts whose structure and size do not match the lesion parts well. Since the elastic modulus

M AN U

of these implants was larger than that of human bone it was relatively easy to produce the stress shielding effect [3]. Therefore, it is necessary to create a customized implant, having good mechanical properties and biocompatibility, to ensure effective biological fixation and to reduce the stress shielding effect.

TE D

Because of its excellent mechanical properties and biocompatibility, titanium (Ti) and its alloys have been widely used in human bone tissue replacement surgery. The elastic modulus of titanium

EP

alloy is approximately 110 GPa. In contrast, the elastic modulus of human bone tissue is about 10–30 GPa. The conventional titanium alloy implants (solid implant part) inevitably, still exhibit stress

AC C

mismatch. Therefore, Ti alloy parts with different components or with controlled pores are fabricated in order to control the elastic modulus of the implant [4–6]. It is more convenient to study the fabrication of implant parts with porous structures than it is to study the fabrication of same parts with different components. The elastic modulus of the implant was not only determined by the density of the parts, but also with other factors, such as the distribution of the density and the shape and size of the internal pores [7–9]. In terms of artificial implant modeling, the present study mainly focused on the design of an implant model with good 2 / 14

ACCEPTED MANUSCRIPT mechanical properties and fewer concerns regarding the biological fixation of implants via the infiltrating growth of bone tissue [10–12]. In the manufacture of porous artificial implants, the traditional porous metal forming process

RI PT

has a large variety where more commons are: solid metal sintering, liquid metal solidification, metal deposition, and corrosion. However, these processes are generally suitable for the preparation of

SC

porous parts, which do not exercise any control over the distribution of pores.

The emergence of 3D printing technology has made the fabrication of model porous parts with

M AN U

controllable porosity reality [13–17]. 3D printing technology is more commonly used in the manufacture of implant parts based on the discrete and bulk forming principles. Currently, implants from many 3D printing manufacturers have obtained FDA licenses in the recent years, such as 3D printed titanium posterior lumbar intervertebral disc by Stryker, 3D printed titanium alloy bone plate

TE D

by MedShape, and 3D printed titanium craniometrical implant by BioArchitects. However, these models mainly focused on coating porous features onto the surface of the implant, which did not the

EP

harness the true potential of the elastic modulus of the porous parts with gradient change. In this paper, a porous implant with a characteristic and variable elasticity modulus was

AC C

designed. Direct metal laser sintering (DMLS) technology was used to fabricate the implant. The biological and mechanical properties of the implant were tested and analyzed to achieve the best possible outcome that ultimately improved the service life of implant and the quality of life of patients after surgery. 2. Materials and Methods 2.1 Materials and equipment 3 / 14

ACCEPTED MANUSCRIPT The commercially available Ti6Al4V powder (EOS GmbH Electro Optical systems, Germany) was used in the experiment. The chemical composition of the Ti alloy powder is listed in Table 1. The SEM image in Fig. 1, photographed by scanning electron microscope (SEM, S-4800, Hitachi,

Ti

Content

Bal

Al

V

O

Fe

C

N

H

6.00 4.00 0.15 0.10 0.03 0.01 0.01

M AN U

Materials

SC

Table 1. Chemical composition of Ti6Al4V (mass fraction %)

RI PT

Japan), shows the spherical nature of the Ti6Al4V particles with an average particle size of 35 µm.

Fig. 1. SEM image of Ti6Al4V

The 3D printing machine used was the EOS M290 (EOS GmbH Electro Optical Systems, Germany) that employed DMLS and had a build volume of 250 × 250 × 325 mm. The following

TE D

parameters were chosen to obtain the best printing quality: the layer thickness and scanning speed were fixed at 20 µm and 300 mm/s respectively; the laser spot diameter was 100 µm; and the

EP

Yb-fiber laser power ranged between 380–400 W. At a melting temperature of 1050°C, the laser could melt the Ti6Al4V powder.

AC C

2.2 Model design

Four kinds of porous structures were designed with the help of the commercial CAD software, 3-matics (Materialise, Belgium). The structure shown in Fig. 2a is a porous box (10 × 10 × 6 mm), divided into three levels: the first layer (D1), the second layer (D2), and the solid layer, with each layer having a height of 2 mm. The side view of the structure depicted in Fig. 2b, shows that the diameters of D1 and D2 are not necessarily consistent (diameter difference values are presented in Table 2). Therefore, gradient based structures can be designed with the different values. The average 4 / 14

ACCEPTED MANUSCRIPT pore size of the samples ranged between 500–800 µm. The structure would be converted to STL format after the design was completed.

Sample 1

sample 2

sample 3

sample 4

D1（mm）

0.2

0.2

0.3

0.3

D2（mm）

0.2

0.3

0.3

0.4

SC

Diameter

RI PT

Table 2. Different diameter values of D1 and D2

M AN U

Fig. 2. Design model of the porous structure (a). The side view of the porous structure (b). 2.3 Virtual testing

Commercial 3D- finite element analysis (FEA) software (ABAQUS, Dassault Systems, USA) was used for virtual testing of the designed porous structure. The FEA model was assumed to be

TE D

linear, elastic, and homogeneous. The elastic modulus was assumed to be 105 GPa and the Poisson’s ratio to be 0.31. The density of the model was fixed at 4430 kg/m3 and the yield strength was 830

EP

MPa. The loading conditions and boundary constraints were based on a previous study [18]. The sample was loaded with a top pressure of 180 MPa, while the bottom boundary was fixed. Through

AC C

these parameter settings, the model was meshed and solved to investigate the deformation and stress under pressure. 2.4 3D printing

The sample model was fabricated on the EOS DMLS machine. A schematic diagram of the DMLS process is shown in Fig. 3. The printing model was first designed in a CAD environment using the 3-matics software, following which the CAD data of the model was converted into the input file prior to printing. Next, the Ti6Al4V powder was placed in the supply chamber of the 5 / 14

ACCEPTED MANUSCRIPT machine and spread evenly across the supplier and builder. During the printing process, the roller spread a layer of powder, 20-µm thick, onto the builder and the graphics trajectory was printed via a laser galvanometer scanning system, which sintered each layer of the powder on the builder. The

RI PT

process was looped until the samples were eventually formed. After the DMLS process was completed, the samples remained in the building chamber at high temperature in an argon

SC

atmosphere for 2 hours to improve the chemical homogeneity. After that, the forming parts were allowed to slowly cool to 100°C for 4 h in the machine. This was to realize a slow cooling process,

M AN U

and then the forming parts were then taken out of the part printing chamber and allowed to cool to room temperature in flowing air without any additional treatment [19].

2.5 Mechanical testing

TE D

Fig. 3. Schematic of the direct metal laser sintering process

Compression tests were conducted in an Electro-mechanical Universal Testing Machine (C44,

EP

MTS, Shenzhen) using a 10 kN load. The loading speed was determined as 1 mm/min. To obtain the relationship between stress and strain, the values of Young’s modulus and yield stress of the samples

AC C

were calculated from the available data. 2.6 Cell culture test

The Ti alloy samples needed to be modified before the cell culture experiment was conducted [20–21]. All the samples were washed using pure water in an ultrasonic cleaner for 10 h, followed by soaking in an NaOH solvent with a concentration of 10 mol/l for 24 h at 60°C. The samples were further washed in pure water for 48 h at 40°C and then immersed in an HCl (0.5 mol/l ) solution for 24 h. Next, the samples were placed in an electric furnace for 1 h that gradually increased to a 6 / 14

ACCEPTED MANUSCRIPT temperature of 600°C at a rate of 5°C/min, and then finally cooled down to room temperature in the furnace. The Bone Mesenchymal Stem Cells (BMSC) used in these cell culture experiments were

RI PT

provided by Nanjing Drum Tower Hospital. The porous samples were immersed in the cell culture medium so that the cells could attach to the alloy samples. During this process, the growth medium

SC

was replaced every other day. In 4 days, the samples were then removed from the medium culture and washed after incubation. The live BMSC cells on the porous samples were showered in a

M AN U

tetrazolium dye for 4 h at 37°C. The morphology of the BMSC cells on the surface of the porous samples was imaged using a laser scanning confocal microscope (LSCM; LEXT OLS4100, Olympus, Japan).

3.1.Physical characteristics

TE D

3. Results and Discussion

Fig. 4a shows the porous samples printed in the EOS M290 machine using the parameters stated

EP

in Section 2.1. Fig.4b shows the SEM image of side view of the printed sample2. It can be seen that

AC C

the pore size of the first layer was larger than that of the second layer; sample 4 exhibits similar properties. The structures of the printed samples were observed to be similar in design. It can also be stated that models with complex porous structures can be printed by DMLS printing technology. Fig. 4c shows the 3D view of a porous sample via SEM. The pore size of the sample can be measured from the SEM image that is shown in Table 3. Since both the inner and outer surfaces of the sample were rough within the 550–800 µm range, increasing the bone cell growth into the porous coating was determined to be beneficial and conducive to the biological fixation of the implants. The porosity 7 / 14

ACCEPTED MANUSCRIPT of the samples was calculated by their mass and volume and was determined to be within the range of 65.8–88.2%. The average porosity of different layers are shown in Table 3. Table 3. Different diameter values corresponding to different pore sizes Pore size of designed

Average pore size of

Average porosity of

of D (mm)

model(µm)

printed part (µm)

each layer (%)

0.2

800

780

0.3

700

690

0.4

600

550

RI PT

Diameter value

88.23

SC

72.84

M AN U

65.76

Fig. 4. The fabricated sample (a). Side view of a porous sample through SEM (b). 3D view of a porous sample through SEM (c). 3.2.FEA result

TE D

The relationship between the stress and strain fields could be obtained from the FEA simulation. The loading conditions and boundary constraints were also defined. Fig. 5 shows the FEA result for

EP

the von Mises stress field of the samples. The maximum von Mises stress value of samples 1, 2, 3, and 4 were 377 MPa, 376 MPa, 192 MPa, and 190 MPa respectively. The von Mises stress for

AC C

samples 1 and 2 was observed to be larger than the others under the same equivalent load pressure. However, the ability of samples 1 and 2 to withstand pressure was weaker compared to the others. Moreover, the von Mises stress in the first and second layer was almost the same in samples 1 and 3. On the contrary, the von Mises stress was mainly concentrated in the first layer for samples 2 and 4 as well as the second layer being smaller than the first layer. Additionally, the deformation of the model from 1 to 4 gradually became smaller. The generation of this phenomenon was due to the variation of the model design in the different layers containing different pore sizes. The stress 8 / 14

ACCEPTED MANUSCRIPT concentration layer was proportional to the pore size of the design layer in the model. The gradient of the level was formed in the suffering stress. Therefore, it can be argued that the design of such a

RI PT

gradient structure can effectively reduce the stress shielding effect.

Fig. 5. FEA results for the von Mises stress field of the four kinds of porous samples, sample 1 (a),

SC

sample 2 (b), sample 3 (c) and sample 4 (d) 3.3. Mechanical properties

M AN U

The compressive strength test of these four samples was conducted in the Electro-mechanical Universal Testing Machine and the obtained stress-strain curve of the samples is shown in Fig. 6. The compressive yield strength of porous samples was determined as 165±10 MPa and the elastic modulus of the samples was 15±3 GPa. The compressive yield strength and elastic modulus of the

TE D

human bone are close to 190 MPa and 10–30 GPa respectively. Interestingly, the elastic modulus of the common solid titanium alloy parts is observed to be 110 GPa. The stress shielding effect would

EP

be generated between the implant and human bone if the titanium alloy implant was not porous. From the test, we observed that the mechanical properties of the testing samples were similar to those

AC C

of human bone tissue, reducing the mismatch of the elastic modulus and the stress shielding effect. Fig. 6 shows varying compressive yield strength of the samples, where the value of sample 1 was observed to be the least, while the value of sample 4 was determined to be the most. Samples 2 and 4 separately demonstrated higher compressive strength than samples 1 and 3. The compressive strength improved with increasing D1 and D2 values, which were stated in Section 1.2. The failure of the sample was divided into two stages with respect to specific performance: the first layer was destroyed followed by the subsequent layer. The slope of the stress-strain curve for samples 2 and 4 9 / 14

ACCEPTED MANUSCRIPT were observed to be steeper after the initial crush. This phenomenon also indicated that the elastic modulus of samples became larger which was observed by the slope of the figure. The elastic modulus of the sample surface in contact with bone tissue was relatively small and was close to that

RI PT

of the bone. On the other hand, the solid layer of the sample had a higher modulus of elasticity, which ensured the strength of the implant. Moreover, there was an additional layer between them,

SC

which reduced the effect of mutation of the elastic modulus. It could also be indicated that the yield strength of the sample presented a gradient change that was more favorable to the biological fixation

Fig. 6 Stress-strain curve of samples 3.4. Bioactive properties

M AN U

of the implants with less mechanical mismatch.

TE D

Fig. 7a shows the LSCM image of the live BMSC cells on porous samples. The blue stain represents the nuclear site of cells and the green stain represents actin. Fig. 7b shows the SEM image

EP

of the BMSC cell morphology on the Ti alloy samples. Several polygonal cells with ‘tentacles’ were observed attached and spread on the surface of the porous sample. The images showed that the

AC C

BMSC cells survived well on the porous samples, which indicated that the porous samples facilitated an ideal growth environment for BMSC. The high biological attraction of BMSC cells towards the implant suggested that the fabricated sample possibly biomimicked the human bone scaffold.

Fig. 7 The LSCM images of BMSC on the implant sample (a). SEM image of BMSC cell morphology cultured on implant sample (b).

10 / 14

ACCEPTED MANUSCRIPT 4. Conclusions In this study, a porous implant with controllable pores was designed and a series of samples

RI PT

were successfully printed using DMLS technology. CAD simulations and mechanical tests, applied to the samples, basically revealed similar experimental results. The results showed that the composite elastic modulus of the sample surface was close to the elastic modulus of the human bone and that

SC

the gradient of the composite elastic modulus of the sample increased from surface inwards with a change in pore distribution.

M AN U

Therefore, by controlling the size and distribution of the pores in the sample, the change in the elastic modulus of the implant could be manipulated, which could effectively reduce the stress shielding effect of the implant and the human bone.

TE D

Additionally, SEM microanalyses revealed that the pores of the Ti alloy samples were approximately 500–800 µm in size, which was beneficial to the growth of bone cells. It was shown that porous Ti alloy parts successfully facilitated the growth of BMSC cells in vitro and was more

EP

conducive to the biological fixation of implants.

AC C

The composite elastic modulus of the fabricated porous sample proved to be a gradient mutation in this study. Our future work will focus on fabricating a porous implant with linear gradient change instead of the composite elastic modulus. Acknowledgments

This study was supported by the National Natural Science Foundation of China (Grant No. 61273243), University Nature Science Foundation of Jiangsu (Grant No. 15KJB470010), and the Program of Natural Science Foundation of Jiangsu Province (Grant BK21050973). The authors 11 / 14

ACCEPTED MANUSCRIPT would like to thank all the people who provided advice and support to the work, like Tony of Materialise, Tian Zhongjun of Nanjing University of Aeronautics & Astronautics.

RI PT

References [1] J Jiang, L Zhu, et al. Review of 3 dimensional printing technology in the medical field. Journal of Mechanical Design and Manufacturing Engineering, 2014(11):5-9.

SC

[2] Pivec R, Johnson A J, Mears S C, et al. Hip arthroplasty[J]. Lancet, 2012,

M AN U

380(9855):1768-1777.

[3] Aouni E H, Kolos E C, Jun Kit W, et al. Physical and mechanical characterisation of 3D-printed porous titanium for biomedical applications. Journal of Materials Science Materials in Medicine, 2014, 25(11):2471-2480.

TE D

[4] Shen H, Brinson L C. A numerical investigation of porous titanium as orthopedic implant material. Mechanics of Materials, 2011, 43(8):420-430. [5] Shen H, Brinson L C. A numerical investigation of porous titanium as orthopedic implant

EP

material. Mechanics of Materials, 2011, 43(8):420-430.

AC C

[6] Yan L, Yuan Y, Ouyang L, et al. Improved mechanical properties of the new Ti-15Ta- x Zr alloys fabricated by selective laser melting for biomedical application. Journal of Alloys & Compounds, 2016, 688:156-162.p

[7] Asadi-Eydivand M, Solati-Hashjin M, Farzad A, et al. Effect of technical parameters on porous structure and strength of 3D printed calcium sulfate prototypes. Robotics and Computer-Integrated Manufacturing, 2015, 37(C):57-67. [8] Vanstreels K, Wu C, Gonzalez M, et al. Effect of pore structure of nanometer scale porous 12 / 14

ACCEPTED MANUSCRIPT films on the measured elastic modulus[J]. Langmuir the Acs Journal of Surfaces & Colloids, 2013, 29(38):12025-35. [9] Dong J L, Jung J M, Latypov M I, et al. Three-dimensional real structure-based finite element

RI PT

analysis of mechanical behavior for porous titanium manufactured by a space holder method[J]. Computational Materials Science, 2015, 100:2-7.

SC

[10] SR Johnston，DW Rosen，HV Wang. Design of a Graded Cellular Structure for an Acetabular Hip Replacement Component. Georgia Institute of Technology, 2006; 8(2):181-187.

M AN U

[11] KS Arabnejad，D Pasini. Multiscale design and multiobjective optimization of orthopedic hip implants with functionally graded cellular material. Journal of Biomechanical Engineering, 2012, 134(3):65-66.

2015:329-350.

TE D

[12] I Gibson，D Rosen，B Stucker. Additive Manufacturing Technologies. Springer New York,

[13] Chang C H, Lin C Y, Liu F H, et al. 3D Printing Bioceramic Porous Scaffolds with Good

EP

Mechanical Property and Cell Affinity. Plos One, 2015, 10(11). [14] Shanjani Y, De Croos J N, Pilliar R M, et al. Solid freeform fabrication and characterization

AC C

of porous calcium polyphosphate structures for tissue engineering purposes. Journal of Biomedical Materials Research Part B Applied Biomaterials, 2010, 93B(2):510–519. [15] Parthasarathy J. A design for the additive manufacture of functionally graded porous structures with tailored mechanical properties for biomedical applications. Journal of Manufacturing Processes, 2011, 13(2):160-170. [16] Petrovic V, Haro J V, Blasco J R, et al. Additive Manufacturing Solutions for Improved Medical Implants// Biomedicine. 2012. 13 / 14

ACCEPTED MANUSCRIPT [17] Shanjani Y, Croos J N A D, Pilliar R M. Solid freeform fabrication and characterization of porous calcium polyphosphate structures for tissue engineering purposes. Journal of Biomedical Materials Research Part B Applied Biomaterials, 2010, 93(2):510-9.

RI PT

[18] Mircheski I, Gradišar M. 3D finite element analysis of porous Ti-based alloy prostheses. Computer Methods in Biomechanics & Biomedical Engineering, 2016:1-10.

SC

[19] Li X, Wang C, Wang L, et al. Fabrication of Bioactive Titanium with Controlled Porous Structure and Cell Culture in Vitro. Rare Metal Materials & Engineering, 2010, 39(10):1697-1701.

M AN U

[20] Subramani K. Titanium surface modification techniques for implant fabrication – From microscale to nanoscale. Journal of Biomimetics Biomaterials & Tissue Engineering, 2010, 5:39-56. [21] Takemoto M, Fujibayashi S, Neo M, et al. Osteoinductive porous titanium implants: effect of

AC C

EP

TE D

sodium removal by dilute HCl treatment.[J]. Biomaterials, 2006, 27(13):2682-2691.

14 / 14

ACCEPTED MANUSCRIPT

SC

RI PT

Fig1.png

AC C

EP

TE D

M AN U

Fig.1 SEM image of Ti6Al4V power

1/7

ACCEPTED MANUSCRIPT

SC

RI PT

Fig2.png

AC C

EP

TE D

M AN U

Fig.2 Design model of the porous structure (a). The side view of the porous structure (b).

2/7

ACCEPTED MANUSCRIPT

M AN U

SC

RI PT

Fig3.png

AC C

EP

TE D

Fig.3 Schematic of the direct metal laser sintering process

3/7

ACCEPTED MANUSCRIPT

RI PT

Fig4.png

SC

Fig.4 The fabricated sample (a). Side view of a porous sample through SEM (b). 3D view of a porous

AC C

EP

TE D

M AN U

sample through SEM (c).

4/7

ACCEPTED MANUSCRIPT

TE D

M AN U

SC

RI PT

Fig5.png

Fig.5 FEA results for the von Mises stress field of the four kinds of porous samples, sample 1 (a),

AC C

EP

sample 2 (b), sample 3 (c) and sample 4 (d)

5/7

ACCEPTED MANUSCRIPT

AC C

EP

TE D

Fig.6 Stress-strain curve of samples

M AN U

SC

RI PT

Fig6.png

6/7

ACCEPTED MANUSCRIPT

M AN U

SC

RI PT

Fig7.tif

Fig.7 The LSCM images of BMSC on the implant sample (a). SEM image of BMSC cell morphology

AC C

EP

TE D

cultured on implant sample (b).

7/7

ACCEPTED MANUSCRIPT Highlights •

A pore controllable porous implant is designed and 3D printed to reduce the stress

shielding effect. The composite elastic modulus of the sample is gradient changed with the change

RI PT

•

of pore distribution.

In vitro experiment shows that the porous sample has good biocompatibility.

AC C

EP

TE D

M AN U

SC

•