Comparative study on electrocrystallization of calcium phosphate ceramics on commercially pure titanium and selective laser melting titanium

Accepted Manuscript Comparative study on electrocrystallization of calcium phosphate ceramics on commercially pure titanium and selective laser melting titanium Xuetong Sun, Huaishu Lin, Xianshuai Chen, Peng Zhang PII: DOI: Reference:

S0167-577X(16)31939-5 http://dx.doi.org/10.1016/j.matlet.2016.12.051 MLBLUE 21863

To appear in:

Materials Letters

Received Date: Revised Date: Accepted Date:

1 October 2016 12 December 2016 18 December 2016

Please cite this article as: X. Sun, H. Lin, X. Chen, P. Zhang, Comparative study on electrocrystallization of calcium phosphate ceramics on commercially pure titanium and selective laser melting titanium, Materials Letters (2016), doi: http://dx.doi.org/10.1016/j.matlet.2016.12.051

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.

Comparative study on electrocrystallization of calcium phosphate ceramics on commercially pure titanium and selective laser melting titanium Xuetong Suna,*, Huaishu Linb, Xianshuai Chena , Peng Zhangc a

Center for Precision Engineering, Shenzhen Institutes of Advanced Technology Chinese Academy of Sciences, Shenzhen 518055, China. b

Guangdong Technical College of Water Resources and Electric Engineering, Guangzhou 510925, China c

Oral subsidiary Foshan University Hospital, 528000, China

Abstract The present study investigates the electrocrystallization processes of calcium phosphate ceramics on commercially pure titanium (CP-Ti) and selective laser melting produced pure titanium (SLM-Ti) parts. The microstructure investigations show that SLM-Ti parts have martensitic α՛ microstructure with more surface defects, whereas CP-Ti samples exhibit equiaxed α microstructure. The comparative electrochemical measurements reveal that the CaP depositions on both substrates exhibit the diffusion-controlled 3D instantaneous nucleation mechanism. However, the polarization potential recorded with SLM-Ti sample is obviously lower than that with CP-Ti sample at the initial stage of CaP deposition process, during which concentration polarization can be neglected. This result suggests that the overpotential caused by the slow electrocrystallization process was decreased on SLM-Ti samples since the SLM process develops more surface defects that can act as nucleation centers in the electrocrystallization process. Keywords: Electrocrystallization; Defects; Selective laser melting; Pure titanium 1. Introduction Commercially pure titanium is widely used in the field of biomedical implants, such as dental and orthopedic implants, but there are still disadvantages, such as poor osseointegration and low wear *Corresponding author. Tel.:+86 20 22912758; fax: +86 20 22912601 E-mail address: [email protected]

1

resistance [1]. Nevertheless, the production of highly complex CP-Ti parts, which could be used as bone substitutes or implants, is challenging due to the poor machinability. Selective Laser Melting is a powder-based additive manufacturing capable to produce parts layer-by-layer from a 3D CAD model [2]. Currently, there is a growing interest in biomedical applications for applying this technology for generating implants and prostheses with complex final shape [2,3]. Surface modification has involved the frequent use of apatite minerals (CaP) as a coating material on Ti implants, because CaP has chemical and crystallographic similarity to the inorganic component of hard dental and bone tissues [4]. Currently, the most common method for applying such coatings on Ti substrates is electrochemical deposition due to its inherent advantages [5]. As for electrochemical deposition process, the initial stages of the electrochemical phase formation processes are strongly dependent on the atomic structure and surface inhomogeneities of the substrate [6]. It is known that the resulting microstructures and phase compositions of SLM parts are different from those of conventionally manufactured parts because of special processing route [7]. Besides, SLM process is prone to develop defects due to improper choice of process parameters or process disturbances [8]. A number of studies are available on the influence of microstructures or structure defects on mechanical properties of SLM parts [9]. This work aimed at investigating and comparing the electrocrystallization of CaP deposition on CP-Ti and SLM-Ti parts in order to better understand of the influence of surface defects on the nucleation and growth of CaP ceramics on titanium electrodes. 2. Material and methods Commercially pure Ti grade 2 cold rolled-sheets after annealing at 650 °C were supplied by Baoji 2

Titanium Industry Co., Ltd, China. The SLM-Ti was fabricated using SLM 125HL machine supplied by SLM Solutions GmbH, Germany. Argon atomized commercially pure Ti powder (99.7% purity, d50 ~ 39 µm) with spherical morphology was used. The SLM processing parameters were as follow: the laser power of 100 W, the scanning speed of 200 mm/s, the spot size of 83 µm, the layer thickness of 30 µm and the hatch spacing of 60µm. The working electrodes, having the surface of 1 cm2 , were sealed into thermosetting resin. Electrodeposition of CaP was carried out in a solution containing 0.61 mM CaCl2 and 0.36 mM NH4H2PO4 at a constant temperature of 60 ºC. The cyclic voltammograms, chronoamperometric and chronopotentiometric measurements were performed in a deposited solution in a standard three-electrode cell. The counter electrode was a larger area platinum electrode and the reference electrode was a SCE. Five repeated experiments were always performed to check the reproducibility of the experimental results. A CHI660D potentionstat was used for electrochemical control and measurements. For microstructure analysis, the 1mm thickness sheet samples were cross-sectioned and etched using Kroll’s reagent for about 30 s. The microstructure was observed by optical microscope (Leitz Wetzlar MM5). The phase composition of 1mm thickness sheets measuring 10 mm × 10 mm was identified by XRD analysis (X' Pert Pro MRD). 3. Results and discussion Fig.1

3

Fig. 1 displays the typical microstructures and XRD patterns of CP-Ti and SLM-Ti samples. As evident, SLM-Ti sample shows acicular α' martensitic microstructure (Fig.1(a)) which results from diffusionless phase transition at high cooling rates (103-108 K/s) [7,8]. In contrast, the CP-Ti sample exhibits an equiaxed α-phase grain associated with diffusion phase transition at moderate cooling rate (Fig. 1(b)). In the XRD patterns (Fig. 1(c)), only peaks corresponding to the hexagonal close-packed titanium (hcp-Ti) are observed, since the α' and α phases have the same lattice structure and almost the same lattice parameters [7]. The main difference between the two phases is that the diffusionless transformation leads to a mass of crystallographic defects inside the α՛

martensitic grains. It can be seen that the peak intensities of SLM-Ti sample declines

obviously compared to that of CP-Ti sample, indicating the finer microstructure [10]. The smaller the grain size, the greater grain boundaries and structure defects appear. It was reported that the nucleation of HAP is much influenced by the crystallographic structure and orientation of the metal substrate [11]. Besides, it is known that the nucleation process of metal crystallites is energetically favored on surface defects such as atomic disorder, emergence points of dislocations, monatomic steps and grain boundaries. [6]. Fig.2

4

The typical cyclic voltammograms curves recorded with the CP-Ti and SLM-Ti samples immersed in the deposition electrolyte at different scan rates are presented in Fig. 2. It appears that the voltammograms are characterized by two cathodic (A and B) peaks, which are associated with the reduction of water and the reduction of H2PO4- to HPO42-, respectively [12]. In either case, there is a linear relationship between the magnitude of the current density of the peak B and the square root of the scan rate, indicating that the processes are diffusion controlled [12,13]. Additionally, it should be noted that at -1.2 V a current loop between the forward and the reverse direction suggests that depositions on both substrates are initiated by an overpotential driven nucleation process [14]. Fig. 3

To obtain a deeper insight into the nucleation processes, the potentiostatic transients were recorded with both samples at potential of -1.4 V versus SCE (see Fig.3). Concerning the absolute current, it follows an initial sharp decay and then increases progressively to reach a maximum (tmax, Imax) at about 1000 seconds before slowly decreasing. Scharifker and Hills derived analytical expressions for multiple nucleation phenomena followed by diffusion-controlled growth of three-dimensional islands [15]. The resulting expressions for the normalized current densities allow to distinguish between instantaneous nucleation (Eq.(1)) and progressive nucleation (Eq.(2)). ௜మ

మ ௜೘

= 1.9542

௧೘ ௧

ቂ1 − ݁‫ ݌ݔ‬ቀ−1.2564

௧

௧೘

ଶ

ቁቃ

(1) 5

௜మ

మ ௜೘

= 1.2254

௧೘ ௧

ቂ1 − ݁‫ ݌ݔ‬ቀ−2.3367

௧మ

మ ௧೘

ଶ

ቁቃ

(2)

Fig. 3(b,c) shows the none-dimensional i2/i2m versus t/tm plot for data from Fig. 3a in comparison to the theoretical curves for instantaneous and progressive nucleation under diffusion control. It is evident that the experimental data for both samples are close to those of an instantaneous nucleation growth model. In the case of instantaneous nucleation, the growth rate of a new phase is high but the number of active nucleus site is low, while for the progressive nucleation is slow, but occurs on a large number of active sites. As mentioned above, CaP electrodepositions on both substrates are initiated by an overpotential driven nucleation process. Therefore, the surface density of active nucleus site should play an important role in the electrocrystallization of CaP deposition. Fig. 4

Fig. 4 shows the chronopotentiometric curves obtained with both substrates at the current density of 1 mA·cm-2. The curves exhibit the expected typical shape. When the current is applied, there is an instantaneous voltage difference increase corresponding to the ohmic voltage drop. After the initial ohmic voltage drop, the electrochemcial reaction starts to occur and the main reason for the voltage drop changes to the electrochemical polarization. As is evident from Fig. 4 within the time scale from 2 s to 7 s, the potentials for both substrates are almost constant with time. However, the potential recorded with SLM-Ti sample is about 200 mV lower than that with CP-Ti sample. In this initial stage, the concentration polarization can be neglected. Thus, the polarization potential 6

difference between CP-Ti and SLM-Ti samples should be caused by the slow electrocrystallization kinetic step. The lower electrocrystallization overpotential for SLM samples can be attributed to the more active nucleation centers induced by the more surface defects resulting from SLM processing route. Experiments in this work only involves the SLM-Ti samples produced with the optimum set of parameters. A further research on the SLM samples with more defects due to improper parameters or process disturbances is valuable. 4. Conclusions In this work, the electrocrystallization of CaP deposition on CP-Ti and SLM-Ti substrates are comparatively studied. The microstructure and phase structure investigations show that CP-Ti sample had an equiaxed α microstructure whilst SLM-Ti sample had a martensitic α՛ microstructure with more surface defects. Based on the comparative electrochemical measurements, it is found that the electrocrystallization of CaP deposition on both substrates occurs as an instantaneous 3D nucleation mechanism and is initiated by the nucleation overpotential. However, the polarization potential on SLM-Ti sample is about 200 mV lower than that on CP-Ti sample at the same current when the concentration polarization is negligible. This indicates that the larger number of nucleation centers induced by surface defects on SLM samples can decrease the overpotential caused by the slow electrocrystallization process. Acknowledgments This research was supported by the Basic Research Foundation of Shenzhen City, China (No. JCYJ20150521094519494); the Key Laboratory of Guangzhou City, China (No. 201509010015); and the Science and technology projects of Guangzhou City, China (No. 201604020147). 7

References [1] D. Banerjee, J.C. Williams, Acta Mater., 61 (2013), pp. 844-879. [2] S. A. Yavari, R. Wauthle, A.J. BŐttger, J. Schrooten, H. Weinans, Appl. Surf. Sci., 290 (2014), pp. 287-294. [3] T. Habijan, C. Haberland, H. Meier, J. Frenzel, J. Wittsiepe, C. Wuwer, Mater. Sci. Eng. C., 33 (2013), pp. 419-426. [4] U. Ripamonti, L. C. Roden, L. F. Renton, Biomaterials., 33 (2012), pp. 3813-3823. [5] Q.Y Zhang, Y. Leng, R.L. Xin, Biomaterials., 26 (2005), pp. 2857-2865. [6] E. Budevski, G. Staikov, W.J. Lorenz, Electrochim. Acta., 45 (2000), pp. 2559-2574. [7] J.W. Elmer, T. A. Palmer, S. S. Babu, W. Zhang, J. Appl. Phys., 95 (2004), pp. 8327-8339. [8] H.J. Gong, R. Khalid, H.F. Gu, G.D. Janaki Ram, Mater. Des., 86 (2015), pp. 545-554. [9] H. Attar, K. G. Prashanth, A. K. Chaubey, M. Calin, L. C. Zhang, Mater. Lett., 142 (2015), pp. 38-41. [10] C. Yan, L. Hao, A. Hussein, P. Young, J. Mech. Behav. Biomed., 51(2015), pp.61-73. [11] N. Eliaz, M. Eliyahu, J. Biomed. Mater. Res. A., 80 (2007), pp. 621-634. [12] Z. Grubač, M. Metikoš-Huković, R. Babić, Electrochim. Acta., 109 (2013), pp. 694-700. [13] J. C. Ballesteros, E. Chaînet, P. Ozil, G. Trejo, Electrochim. Acta., 56 (2011), pp. 5443-5451. [14] A. Zimmer, L. Broch, C. Boulanger, N. Stein. Electrochim. Acta., 174 (2015), pp. 376-383. [15] B. Scharifker, G. Hills. Electrochim. Acta., 28 (1983), pp. 879-889. Figure Captions Fig.1 Microstructure of (a) SLM-Ti and (b) CP-Ti; (c) XRD patterns. Fig.2 (a, b) Cyclic voltammetric curves obtained at different scanning rate and (c,d) variation of the cathodic peak current with the square root of the scan rate. Fig.3 (a) Current against time transients recorded with SLM-Ti and CP-Ti at the potential of -1.4V in a deposition solution; (b,c) The normalized experimental transients are respectively compared to the simulated transients for instantaneous and progressive nucleation models. Fig.4 Voltage against time curves recorded with SLM-Ti and CP-Ti at the current of 1mA·cm-2 in a deposition solution. 8

Highlights


Electrocrystallization of CaP electrodeposition on CP-Ti and SLM-Ti.




Both samples exhibit diffusion controlled 3D instantaneous nucleation mechanism.




Lower overpotential was obtained on SLM-Ti in initial electrodeposition stage.




Lower overpotential on SLM-Ti is attributed to more surface defects.

9