Corrosion behavior of titanium boride composite coating fabricated on commercially pure titanium in Ringer's solution for bioimplant applications

Materials Science and Engineering C 48 (2015) 243–255

Contents lists available at ScienceDirect

Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec

Corrosion behavior of titanium boride composite coating fabricated on commercially pure titanium in Ringer's solution for bioimplant applications Bose Sivakumar, Raghuvir Singh ⁎, Lokesh Chandra Pathak CSIR-National Metallurgical Laboratory, Jamshedpur 831007, India Academy of Scientific and Innovative Research, New Delhi 110025, India

a r t i c l e

i n f o

Article history: Received 12 August 2014 Received in revised form 20 October 2014 Accepted 2 December 2014 Available online 4 December 2014 Keywords: Titanium Bioimplant Boride coating Corrosion resistance Ringer's solution

a b s t r a c t The boriding of commercially pure titanium was performed at 850 °C, 910 °C, and 1050 °C for varied soaking periods (1, 3 and 5 h) to enhance the surface properties desirable for bioimplant applications. The coating developed was characterized for the evolution of phases, microstructure and morphology, microhardness, and consequent corrosion behavior in the Ringer's solution. Formation of the TiB2 layer at the outermost surface followed by the TiB whiskers across the borided CpTi is unveiled. Total thickness of the composite layer on the substrates borided at 850, 910, and 1050 °C for 5 h was found to be 19.1, 26.4, and 18.2 μm respectively which includes b 3 μm thick TiB2 layer. The presence of TiB2 phase was attributed to the high hardness ~2968 Hv15 gf of the composite coating. The anodic polarization studies in the simulated body fluid unveiled a reduction in the pitting corrosion resistance after boriding the CpTi specimens. However, this value is N 0.55 VSCE (electrochemical potential in in-vivo physiological environment) and hence remains within the safe region. Both the untreated and borided CpTi specimens show two passive zones associated with different passivation current densities. Among the CpTi borided at various times and temperatures, a 3 h treated shows better corrosion resistance. The corrosion of borided CpTi occurred through the dissolution of TiB2. © 2014 Published by Elsevier B.V.

1. Introduction Surface modification is a proven alternative to minimize the surface damage of metallic materials including bioimplants. Titanium and its alloys are extensively preferred bioimplants due to their low density, excellent corrosion resistance, and acceptable biocompatibility. The Ti based bioimplants, however, suffer from the poor fretting corrosion and sliding wear resistance when subjected to the two body contact such as in the hip and knee joints [1–3]. The naturally formed passive oxide layer on the titanium is easily removed from the contact zone, due to poor mechanical properties of the oxide, and as a result fresh metallic surface is exposed to the physiological fluid [4]. This increases corrosion of metals/alloys and subsequent release of metallic ions (such as Ti, Al and V from Ti–6Al–4V alloy) along with the wear debris to the physiological fluid. A number of surface modification techniques and processes such as plasma based, laser based and assisted, thermal oxidation, organic and inorganic coatings, diffusion coatings have been developed to overcome the inferior tribological resistance of the titanium and its alloys [4–9]. Over the past several years, boriding, a diffusion based ⁎ Corresponding author at: CSIR-National Metallurgical Laboratory, Jamshedpur 831007, India. E-mail address: [email protected] (R. Singh).

http://dx.doi.org/10.1016/j.msec.2014.12.002 0928-4931/© 2014 Published by Elsevier B.V.

coating method has been increasingly used to improve the tribological and corrosion behavior of metallic materials. Titanium borides (TiB2 and TiB) have shown the remarkable properties such as low electrical resistivity (similar to metals), high melting points, high hardness and excellent wear, and corrosion resistance [10,11]. Boriding of the titanium, therefore, can be an extremely viable method to improve its tribological behavior. Interestingly, TiB2 and TiB coatings can be developed in-situ on the surface of titanium by heating the Ti specimen encapsulated by the mixture containing boron source, an activator, and filler material. The coating by reinforcement of the borides (a non-diffusive ex-situ method) performed on the surfaces of several metals/alloys in which coating adherence may be lost can, thus, be avoided. The other important reason for choosing the boride coating is the comparable thermal expansion coefficient of TiB2/TiB and Ti, which may ensure much less thermal stresses and so the distortion at the interface. The coefficients of thermal expansion of TiB2, TiB, and Ti are 8.1 × 10−6, 8.5 × 10−6 and 8.6 × 10− 6/°C respectively. The undesirable thermal mismatch between Ti matrix and several other ceramics such as Ti5Si3, CrB, B4C, SiC, and TiC is reported to cause high residual stresses that degrade the coating–matrix interface [12–14]. The boriding by thermochemical diffusion process (namely pack boriding) has been extensively investigated on the ferrous substrates for various industrial applications. The pack boriding is a very simple

244

B. Sivakumar et al. / Materials Science and Engineering C 48 (2015) 243–255

and cost effective process compared to the liquid and gas boriding methods [15–17]. Recently, boriding process has been attempted on the titanium and its alloys for orthopedic implant applications [10, 18–20]. It is reported that the TiB and TiB2 intermetallic phases formed under the different boriding conditions have average hardness of 850 and 3300 Hv. The boriding process involves diffusion of boron (B) atoms from the high concentration boron region (source) to a low or no boron region (sink such as titanium) through the surface defects at elevated temperatures N 800 °C. Boride coating on the titanium and its alloys has been reported to enhance the wear resistance and showed a low coefficient of friction against the sapphire ball when subjected to sliding under the dry and lubricated conditions [10,11, 21]. The growth kinetics of boriding process on the surface of titanium has been studied by Sarma et al.; they have shown that the growth of TiB2/TiB greatly governed by the β-transition temperature and subsequent B diffusion in the α and β phases differs significantly [22]. Tikekar et al. have showed that the boriding at ~900 °C for 24 h (near β transition temperature) produced a higher coating thickness (~60 μm) than the samples borided at 850 °C (~ 28 μm) or at N 1050 °C (~ 50 μm) for 24 h [23]. Most of the studies on TiB2/TiB coatings are aimed to investigate the kinetics of boron diffusion and evaluation of physical and microstructural properties. Attempts are also made to establish the biocompatibility of boride coatings. An in-vitro test on the plasma processed titanium boride coatings on α-Ti matrix has demonstrated good cytocompatibility compared to that of the untreated titanium specimens, indicating boride coated titanium a strong candidate for the biomedical applications [24]. A favorable cell growth rate on the plasma sintered composite TiB2–Ti within the 48 h duration along with an excellent blood compatibility with a low hemolytic level (b 0.12) has been demonstrated [25]. From the literature it appears that the due attention has rarely been given to understand or study the corrosion behavior of TiB2/TiB coating on titanium and its alloys. Needless to mention, the corrosion is an important criterion to establish the biocompatibility of a newly proposed implant material or modification thereon (such as boride coating) and deserves extensive investigations. Among a few earlier studies, Covino et al. have focused on the corrosion of pure TiB2 ceramic in the acid solutions and compared the performance of TiB2 made from the various routes such as press and sintered (PAS) and an electro deposition process [26]. The study concluded that the PAS titanium diboride corrodes at the rate 50 to 100 times higher than that of the electrodeposited and does not passivate in sulfuric acid solution. A relatively recent study has evaluated the corrosion performance of a sintered TiB2 mixed with the nickel powder (as a sintering additive) in 3.5% NaCl solution. The study showed protection to the TiB2 by the formation of passive film at 25 and 45 °C but the resistance of the film is reduced with further increase in the temperature of solution up to ~65 °C. The corrosion attack was found concentrating along the nickel rich grain boundaries [27]. In most of such studies, the current density associated with the passive region of a TiB2 is reported to be in the order of 10−3 A/cm2 or higher indicating pseudo passivation. The corrosion behavior of a pure and sintered TiB2 is shown to influence by the porosity present in the specimens or the segregation along the grain boundaries [27]. However, no corrosion studies, to the best of our knowledge, appear in the literature on TiB2/TiB composite coatings fabricated on the titanium or titanium alloys developed by the pack boriding method. The corrosion behavior of such coated specimens may actually differ largely from those described earlier [26,27] where TiB2 is coated by the other methods (ex-situ) due to obvious reasons. The corrosion study of the borided titanium in SBF solution is thus necessary in order to establish its biocompatibility for the bioimplant application. The present study, thus, aimed to develop the boride composite coatings on CpTi by using a pack boriding process. It was performed at temperatures in the range 850 to 1050 °C for varied soaking durations such as 1, 3 and 5 h (hours). Corrosion studies on the fabricated coatings were made in the Ringer's solution using the open circuit potential

(OCP), anodic polarization, and the electrochemical impedance spectroscopy (EIS) techniques. The dissolution behavior of the boride coating was also investigated and compared with the untreated titanium by performing the long term exposure at OCP in the Ringer's solution; analysis of the post exposure solution was done for measuring the metal/B ion emission using the ICP-OES (Inductively Coupled PlasmaOptical Emission Spectroscopy). The study also includes the coating characterizations for the phase analysis by the X-ray diffraction (XRD) technique and Boron accumulation in TiO2 passive film by X-ray photoelectron spectroscopy (XPS). The coating morphology was investigated by the scanning electron microscope (SEM), mechanical properties by the microhardness tester, and coating roughness by the surface profilometer. 2. Materials and methods 2.1. Materials A commercially pure titanium, CpTi (wt.%: N-0.01; C-0.03; H-0.01; Fe-0.20; O-0.18 and Ti-balance) was used as a base substrate for the coating development. The CpTi in the form of a rod of 20 mm diameter was cut in the disks of 3 mm thick for the fabrication of coatings and subsequent corrosion and other characterizations. Prior to boriding, the specimens were polished using SiC abrasive papers ranging from 100 to 2/0 grades followed by cleaning with the distilled water. Thus polished substrates were ultrasonicated in acetone to remove the contaminants from the surface of the substrate. A mixture consisting of 50% boron source (B powder, Merck), 15% of an activator (NaCO3, Merck) and 35% of filler (charcoal activated, Merck) (in wt.%) was used for boriding the polished CpTi specimens. 2.2. Boriding process The boriding was performed by encapsulating the polished and cleaned CpTi specimens into the above-mentioned boriding mixture in a sealed alumina crucible. Sealing is required to prevent the oxygen ingress into the mixture, which otherwise may interrupt the titanium boride formation by forming a TiO2 layer in preference to boride. The boriding process was carried out at temperatures 850, 910, and 1050 °C with varied soaking periods (1–5 h) in an argon environment. After the completion of pre-decided soaking periods, the borided CpTi specimens were furnace cooled to room temperature. The borided specimens were then cleaned by the ultrasonication in acetone to remove the loosely adhered boriding mixtures and these samples were used for the further characterizations. 2.3. Coating characterizations The coating on the CpTi specimen was characterized for phase evolution, microhardness, microstructure, and surface morphology. The XRD analysis was performed using a Cu-Kα source (λ—1.5426 A°) for the phase identification in both the untreated and borided CpTi specimens using D8 Discover, Bruker make X-ray diffractometer. The SEM and energy dispersive X-ray analysis (EDAX) (Nova Nano SEM 430 make) were used to investigate the coating morphology, thickness, and composition at the cross section of the borided specimens. The microhardness and surface roughness (a 2D surface profilometer) measurements were carried out using the Leica and Tayor Hobson 2 makes respectively. Adhesion test of the coating (TiB2) was carried out by a Rockwell C diamond indenter (120° angle, 200 μm radius) using computer controlled scratch tester (Model TR101, DUCON). The specimen was mounted on the test table and slid at a speed of 0.2 mm/s. The load on the indenter was increased at a rate of 2 N/mm to know the critical load (as per the ASTM standard C1624-2010) for the coating damage. The change in the slope of the normal load versus displacement

B. Sivakumar et al. / Materials Science and Engineering C 48 (2015) 243–255

curve indicates the coating damage/disbondment and the corresponding load recorded as critical load.

245

3. Results 3.1. XRD pattern of borided and untreated CpTi

2.4. Electrochemical polarization studies

The XRD analysis performed on the untreated and CpTi specimens borided at 850, 910, and 1050 °C for 5 h is shown in Fig. 1a–b. The XRD analysis revealed the presence of TiB, TiB2, and Ti3B4 in all the specimens borided at various temperatures. The TiB (PDF#732148) and Ti3B4 (PDF#191368) possess orthorhombic crystal structures while TiB2 (PDF#350741) is hexagonal structure similar to α-Ti (PDF#895009). According to the Ti–B phase diagram, TiB and TiB2 formation begins when the boron reaches the approximate concentration (by wt.%) of 18–18.5 and 30.1–31.1 respectively. Both the TiB and TiB2 phases possess reasonable temperature–composition variation within the range of 0.5 and 1 wt.% respectively, whereas Ti3B4 forms at a critical boron concentration of 22.4 wt.% [21]. The peak intensities related to untreated CpTi, are decreased after the boriding at the temperature from 850 to 1050 °C; the peaks at 2θ values of 35.22, 40.12, and 62.91° corresponding to crystallographic planes (101), (110), and (100) (designated in Fig. 1a) indicate the formation of increased volume fraction of the borides. The crystallographic planes (001), (100), and (101) at 2θ values of 27.5, 34.21 and 44.4° correspond to the formation of TiB2 respectively, while (201), (200), and (102) at 2θ values of 29.17, 35.36 and 42.42° correspond to the TiB formation. The Ti3B4 formation is indicated by the presence of its prominent peak at 35.43o (Fig. 1b). The low peak intensity implying the small volume fraction of Ti3B4 at 35.43° is clearly shown in a magnified XRD pattern in Fig. 1b. The XRD studies also indicate that the amount of Ti3B4 phase is increasing with temperatures. The Ti3B4 phase still remained small compared to other Ti–B phases due to the fact that the Ti3B4 is a stoichiometric compound, formed only at 22.4 wt.% of boron. Generally, in such a varying diffusive process, probability of meeting a specific (sharp) composition is rarely expected.

The electrochemical investigations of the untreated and borided CpTi specimens were carried out by exposing them to the Ringer's solution (a SBF) using Potentiostat/Galvanostat/FRA, GAMRY, Reference 600™. The Ringer's solution was used as an electrolyte for all the electrochemical experiments, which was prepared by adding (in g/l) the AR grades NaCl — 9.0, KCl — 0.43, NaHCO3 — 0.20 and CaCl2 — 0.24 to the requisite quantity of distilled water. A three electrode cell consisting of the saturated calomel electrode (SCE) as reference, graphite as counter, and CpTi disk (untreated and borided specimens) as a working electrode were used for electrochemical studies. All the electrochemical experiments were performed at 25 ± 2 °C. Before the onset of anodic polarization measurements, OCP of the untreated and borided CpTi specimens was monitored for ~1 h immersion to ensure the attainment of a stabilized OCP. The anodic polarization experiments were performed by scanning the specimen from −250 mV with respect to OCP up to the potential of + 3000 mV at a sweep rate of 1.67 mV/min. All the potential measurements were made with respect to the Saturated Calomel Electrode. The electrochemical impedance spectroscopy tests were conducted at OCP with 10 mV amplitude of the alternating current signal in the frequency range from 0.01 to 10,000 Hz.

2.5. XPS analysis A passive film formed on the untreated and borided CpTi specimens was analyzed by the XPS technique under the vacuum of N10−9 Torr using the SPECS Spectrometer with Al monochromator (1486.6 eV) and Mg Kα (1253.6 eV) X-ray source. The CASA XPS software was used for the de-convolution of the XPS spectrum. The specimens (both untreated and borided) prior to the XPS analysis were subjected to immersion in the Ringer's solution for 24 h in order to corrode and grow a passive film at the OCP.

3.2. Coating morphology, growth, and strength The morphology of the CpTi borided at 850 to 1050 °C for 5 h demonstrates the growth of TiB whiskers and TiB2 compound, shown in Fig. 2a–c. The XRD analysis along with the morphology of the coating fabricated at different boriding conditions confirms the formation of a composite layers consisting of TiB2 and TiB whiskers. The outermost layer in the figure is TiB2 followed by the TiB whiskers grown in the preferred direction (along the thickness). The compositional analysis across the TiB2 layer and along the TiB whisker, as shown in Fig. 2d (at 850 °C/5 h), illustrates that the TiB2 contains boron and Ti in the range from 20 to 40% and 40–60% respectively while these in the TiB whiskers varied from 10 to 25 and 75–90% respectively. The elemental

2.6. Long term immersion analysis using ICP-OES A few untreated and borided specimens were exposed (2 cm2 area) to 150 ml Ringer's solution for 15 and 30 days at OCP to observe the possible release of the constituent elements in the solution. After exposure, the resulting solutions were analyzed by the ICP-OES using Varian Inc., Vista MPXTM make for the quantification of Ti and B as a result of the corrosion reaction.

2000 Absolute intensity (cps)

1600

a

1050oC-5h

0 910oC-5h

2000 0 4000

850oC-5h

2000 0 4000 2000 0

CpTi

20

30

40

50 60 2θ (degree)

70

80

1400

Absolute intensity (cps)

4000

TiB

b

Bare Ti B-Ti-850C-5h B-Ti-910C-5h B-Ti-1050C-5h

1200

Ti3B4

1000 800 600 400 200

90

0 35.2

35.3

35.4

35.5

35.6

2 θ (degree)

Fig. 1. (a) Shows the XRD pattern of CpTi borided at 850, 910 and 1050 °C for 5 h and compared with the untreated CpTi and (b) magnified peaks showing the formation of Ti3B4.

246

B. Sivakumar et al. / Materials Science and Engineering C 48 (2015) 243–255

a

b

20 μ m

Elemental composition (wt%)

d

20 μ m

c

100 Ti across TiB whiskers

90 80 70 60

Ti along TiB2

50 40

B along TiB2

30 20 10

B across TiB whiskers

20 μm

0 0

2

4 6 Scan distance (μm)

8

Fig. 2. The cross sectional morphology of CpTi borided at (a) 850 °C, (b) 910 °C, and (c) 1050 °C for 5 h and (d) change in boron and titanium concentration across the TiB2 and along TiB whiskers formed at 850 °C for 5 h.

concentration (of B and Ti) remains almost constant across the TiB2 and along the length of TiB whisker. The CpTi specimen borided at 850 °C contains thin (line shaped) TiB whiskers, which are densely arranged in the direction transverse to the specimen surface (Fig. 2a) compared to those obtained by boriding at 910 and 1050 °C where the whiskers are relatively thicker and a few in numbers (Fig. 2b and c). The numbers of whiskers are found to grow linearly in the form of fine needles when borided at a low temperature (850 °C). Whereas, for the samples borided at N 850 °C the whiskers have plate like shapes with significant lengths and widths. However, a few whiskers that have inclined growth with respect to the surface of the specimen and cut cross-sectionally, during the specimen fabrication, reveal polygonal morphology. Such irregular hexagon shaped whiskers have also been observed by the earlier researcher [28]. Apart from the formation of continuous whiskers/

needles, islands of the TiB precipitates are also prevalent ahead or adjacent to the needle like morphology. Such isolated islands may also be considered as continuous three dimensional TiB phase that grew from the other directions linearly or angularly and eventually appears like island in the two dimensional planes. This is confirmed as encircled in Fig. 2a on the cross section region; it shows an explicit orientation to the surface at certain angle. The total coating thicknesses are measured considering both the continuous TiB2 layer and the TiB whiskers as demarcated by the arrows (Fig. 2a–c); the shorter lines in the figure show TiB2 layer thickness. The coating dimensions of the TiB whiskers and TiB2 as a function of temperature are shown in Table 1. The total average coating thickness is found to be ~19.11, 26.36, and 18.18 μm for the specimens borided at 850, 910, and 1050 °C for 5 h respectively. As seen from Table 1 and Fig. 2a–c, the

Table 1 The coating dimension of TiB whiskers and TiB2 across CpTi borided at various temperatures and times and the aspect ratio of the TiB whiskers. Borided conditions

Average coating thickness (μm)

Average TiB whiskers width (μm)

Average TiB2 thickness (μm)

Aspect ratio (length/width)

850 °C-5 h 910 °C-5 h 1050 °C-5 h 850 °C-3 h 910 °C-3 h 1050 °C-3 h 850 °C-1 h 910 °C-1 h 1050 °C-1 h

19.21 26.36 18.18 11.06 11.66 8.66 7.23 6.8416 8.5853

0.82 ± 0.39 1.2 ± 0.6 1.74 ± 0.54 NDa

1.704 ± 0.30 1.77 ± 0.28 2.78 ± 0.55

23.3 21.3 9.16

a

ND—means ‘not done’.

± ± ± ± ± ± ± ± ±

4.69 8.15 6.36 3.10 1.87 1.87 0.87 1.33 1.49

B. Sivakumar et al. / Materials Science and Engineering C 48 (2015) 243–255

thickness of the TiB2 layer is much less compared to the TiB layer. While the average thickness (length) of the TiB whiskers is 17.5, 24.6 and 15.22 μm, the TiB2 layer has thickness of 1.70, 1.77 and 2.78 μm on the samples borided for 5 h at 850, 910, and 1050 °C respectively. The table also includes the thickness data of the composite layers formed by boriding for 1 and 3 h at various temperatures. The thickness of the individual TiB2 and TiB layers were not distinctly measured (Table 1) as the former is insignificant as compared to the layers containing TiB whiskers. These results indicate that (i) the TiB whiskers do not grow monotonously with the temperature as it is the longest at 910 °C compared to that at 850 and 1050 °C, (ii) the TiB2 layer acts as a diffusion barrier that slows down the diffusion process and hinders further growth and (iii) the TiB2 layer grows uniformly with the increase in temperature and thus follow a law of diffusion. The length of the whiskers is found to be larger in the samples borided at 910 °C than at 850 and 1050 °C which may be due to the accelerated growth kinetics of B diffusion near α to β phase transition temperature of CpTi. The reason for an increased diffusion near the phase transition temperature is not though clear, a few have reported that the diffusivity of the boron is (~ 1 × 10−13 m2/s) as high as selfdiffusivity (anomalous diffusion) of the Ti at 910 °C (near transition temperature) [22]. This is expected to be the consequence of large concentration of substitutional vacancies, at this temperature, and appears to promote B diffusion extraordinarily [22]. Such enhancement in the diffusivity of B at 910 °C is not followed in the TiB and TiB2. At 850 and 1050 °C, the diffusivity of boron is higher (6.13 × 10−15 and 1.34 × 10−13 m2/s respectively) in the TiB than that in the TiB2 [29–31]. A lower diffusivity of boron in TiB2, by approximately 1–2 orders of magnitude, than in the TiB at 850–1050 °C is responsible for the smaller thickness of the TiB2 compared to the TiB. This is further complemented by measuring the volume fraction of the various phases in the coating and shown in Table 2. This indicates a much smaller volume of TiB2 (less than by an order of magnitude) as compared to the TiB and thus implies an extremely slow rate of growth of TiB2, as also evidenced by the thickness data. A continuous increase in the volume fraction of TiB2 with temperature, similar to the thickness, follows a classical diffusion law. Contrarily, TiB grew to a larger volume at 910 °C compared to that at the other two temperatures (850 and 1050 °C). The measured aspect ratios of the TiB whiskers are provided in Table 1, which clearly indicates the reduction of aspect ratio with the increase in boriding temperatures. Thickness of whiskers is increased with the temperature which implies that there exists a competition between the diffusion paths (preferential growth direction) as the temperature varies from 850 to 1050 °C. This is also attributed to the difference in the diffusivity of boron in the TiB and Ti that varies differently with the temperature [22]. The TiB whiskers act as fiber reinforcement in the Ti matrix, which is expected to provide good adhesive strength to the composite surface. These acicular particles (TiB) have good mechanical properties and it will minimize the thermal stresses generated at the interface. The growth of the TiB in either direction occurs at the cost of other; therefore increase in thickness with the temperature lowers the whisker's length. The aspect ratio of TiB is, thus, an important parameter and it is related to the strength of the TiB–titanium composites by the following equation [32]

σ c ¼ Vσ Ti

s þ σ Ti 2

ð1Þ

247

where the σc and σTi are the yield strengths of Ti–TiB composite and Ti matrix respectively, V is the volume fraction of the TiB and S is an aspect ratio of the whiskers (length/diameter). The increase in strengthening efficiency, R, due to the TiB whiskers can be determined by the following equation R¼

σC : σ Ti

ð2Þ

The ‘R’ strengthening efficiency, is the ratio of increase in the yield strength of the composite to that of the titanium. The strength and strengthening efficiency of the composite layer are worked out using the above-mentioned equations (Eqs. (1) & (2)) considering the yield strength of CpTi ~170 MPa and are presented in Table 2. It can be seen from Table 2 that the yield strength of the composite layer fabricated by boriding at 910 °C is higher than the samples borided at 850 and 1050 °C. Despite having similar volume fraction of the TiB whisker formed on the samples borided at 850 and 1050 °C, a significant decrease in the strengthening efficiency at a later temperature is attributed to the lower aspect ratio of the TiB whisker. The strengthening efficiency of the titanium is also presented in Table 2, which indicates the highest value of R for the samples fabricated by boriding at 910 °C. 3.3. Micro hardness measurement Micro hardness measurements, performed on the cross section of the specimens borided at 850–1050 °C for 5 h, are plotted with the distance from the outermost surface and are shown in Fig. 3a and b. The figure reveals a large increase in the hardness of CpTi, borided at different temperatures, which ranges from ~2422 to 2968 Hv15gf on the outermost surface; the hardness of an untreated CpTi is ~250 Hv15gf. In the hardness profile, vertical lines are drawn, which demarcate the approximate hardness zones resulted from the formation of TiB2 and TiB coating containing whiskers. The high hardness at the outermost surface (Fig. 3a) is due to the formation of TiB2 and decreases further with the emergence of the TiB whiskers (N5 μm distance from the outermost surface in Fig. 3a). An overall highest hardness is obtained for the CpTi specimens borided at 1050 °C (~ 2968 Hv15gf) while the lowest value (~2422 Hv15gf) is noticed when borided at 850 °C. This clearly indicates that the highest hardness of the composite coating is primarily dependent upon the thickness of the TiB2 and not on the combined (TiB2 + TiB) thickness. The hardness of the pure TiB2 is reported to be as high as 3400 Hv while that of the pure TiB is 2000 Hv [33]. The hardness values were also measured ahead of the tip of the whiskers (in the Ti matrix at approximately N35 μm from the outermost surface as shown in Fig. 3b) which were significantly low compared to that at the outermost surface or at the whiskers, yet higher than that the base or at a far away distance from the tip of the whiskers. This indicates the presence of Boron in the solid solution that causes a hardness of the matrix to increase beyond ~250 Hv15gf of CpTi. Fig. 3b (enclosed in box) also evidenced a smaller initial slope of the boride coating at 910 °C as compared to that borided at 850 and 1050 °C; which signify that the coating formed at the later temperatures may less likely to accommodate the operational stresses as compared to that at 910 °C. During indentations in the borided specimens, small cracks at the edges of indentation pyramid were observed which are typical of the brittle ceramics with a low fracture toughness. Such indents were

Table 2 The volume fractions of different phases and calculated strengthening efficiency of TiB whiskers in the CpTi specimens borided at 850, 910 and 1050 °C for 5 h. Borided specimens

Volume fraction of TiB + TiB2

Volume fraction of TiB2 coating

Volume fraction of TiB coating

Aspect ratio

Strength of composite (σc)

Strengthening efficiency

850 °C-5 h 910 °C-5 h 1050 °C-5 h

0.21 ± 0.02 0.32 ± 0.03 0.23 ± 0.02

0.01 ± 0.001 0.016 ± 0.003 0.035 ± 0.006

0.20 ± 0.02 0.30 ± 0.03 0.19 ± 0.02

23.3 21.3 9.16

566 713 318

2.3 3.2 0.8

B. Sivakumar et al. / Materials Science and Engineering C 48 (2015) 243–255

where Kc (MPa √m) is the fracture toughness, E is the young modulus, Hv is micro hardness (MPa), g is the microcrack length (m), l is half diagonal length of the micro indentation (m), and ϕ is constraint factor ~ 3. The crack length was measured after the indentation made at TiB2/TiB interface (under the load of 15 gf) and on the TiB whiskers (under 10 gf) and viewed under SEM as shown in Fig. 4. This was found to be significantly low 1.99–2.13 MPa √m (at TiB2/TiB interface and on TiB whiskers respectively). Adhesion strength, which is proportional to the critical load, was observed to be 10, 13, and 13.7 N for coatings fabricated at 850, 910, and 1050 °C for 5 h respectively. These values correspond to the deflection in the linearity of the curve plotted between scratch length and the load. The corresponding scratches after the test viewed under the optical microscope also indicate the damage on the coated surface. The slight increase in critical load with the temperature of the coating indicates improvement in the adhesion strength of the coating.

Microhardness, Hv15gf

248

a

Case depth, μm

-0.06

a Potential (V) vs SCE

-0.08 -0.10 -0.12

Bare Ti B-850oC-1h

-0.14 -0.16

B-850oC-3h B-850oC-5h

-0.18 -0.20 -0.22 -0.24

b

0

used to measure the fracture toughness value of the coatings. The fracture toughness values of the specimen borided at 1050 °C for 5 h that showed the highest thickness of the TiB2 layer, is estimated using the following equation [34] !
g1=2 H −2=5 H l1=2 v v Kc ¼ 0:048 Eϕ ϕ l

ð3Þ

b Potential (V) vs SCE

Fig. 3. (a) Shows the micro hardness profile of across the CpTi borided at 850, 910 and 1050 °C for 5 h, (b) the cross section with qualitative indentation size.

500 1000 1500 2000 2500 3000 3500 Time (s)

0.00 -0.05 -0.10 Bare Ti B-910C-1h B-910C-3h B-910C-5h

-0.15 -0.20 -0.25 0

c

0.00 Bare Ti B-1050C-1h B-1050C-3h B-1050C-5h

Potential (V) vs SCE

-0.05

15 gf

10 gf

20 μm Fig. 4. SEM images showing the crack at TiB2/TiB interface and at TiB whiskers after subjected to indentation made under 15 and under 10 gf respectively.

500 1000 1500 2000 2500 3000 3500 Time (s)

-0.10 -0.15 -0.20 -0.25 0

500 1000 1500 2000 2500 3000 3500 Time (s)

Fig. 5. Variation in OCP of CpTi borided specimens at (a) 850 °C, (b) 910 °C and (c) 1050 °C for 1, 3 and 5 h compared with the untreated CpTi in Ringer's solution.

B. Sivakumar et al. / Materials Science and Engineering C 48 (2015) 243–255

3.4.2. Anodic polarization behavior of untreated and borided CpTi The anodic polarization curves obtained on CpTi borided at 850 to 1050 °C for 1 to 5 h including untreated in the Ringer's solution are shown in Fig. 6a–c. The corresponding electrochemical parameters derived, from the anodic polarization curves, are presented in Table 3. The untreated CpTi corrodes with a corrosion current density, icorr, of ~ 0.12 × 10−6 A/cm2 and specimen borided at various conditions corrode with icorr values ranging from ~0.07 to 0.48 × 10−6 A/cm2. These values, though, are of the similar order, CpTi appears to have slightly lower icorr than most, except a few, of the borided specimens. Among the icorr obtained for the specimens borided at different temperatures and durations, a CpTi borided for 3 h generally produced the lowest current density (Table 3 and Fig. 6a–c) and seems performing marginally better than the untreated CpTi. This is possibly due to a relatively uniform coating of TiB2 on the TiB–Ti formed at 3 h (specifically at 910 and 1050 °C) as compared to that at 1 and 5 h. The specimens during the anodic polarization were scanned up to +3.0 V with respect to SCE; it may be noticed that the untreated CpTi did not encounter a characteristic breakdown potential until this regime. The borided specimens, however, indicate a sharp breakdown potential below +3.0 V, and are thus less resistant to the pitting corrosion as compared to that of the untreated CpTi. On comparing the pitting/ break down potential of various borided CpTi, specimen borided for 5 h (at all the temperatures) shows a lower pitting potential (b2.52 V)

3.0

Potential (V) vs SCE

2.5

850oC-1h 850oC-3h

2.0

850oC-5h bare Ti

1.5 1.0 0.5 0.0

-0.5 -10

10

b

-9

10

-8

10

-7

10

-6

10

i (A/cm2)

-5

10

-5

10

-5

10

10

-4

10

-3

10

-4

10

-4

10

-2

-3

10

-3

10

3.0 2.5

Potential (V) vs SCE

3.4.1. Open circuit potential (OCP) measurements The changes in the OCP of the CpTi untreated and borided at 850, 910 and 1050 °C for 1 to 5 h as a function of immersion time in the Ringer's solution are displayed in Fig. 5a–c. All the borided CpTi specimens seem approaching a nobler (anodic) OCP as compared to that by an untreated CpTi, suggesting a change in the nature of electrochemical activities on the borided specimens. The OCP of an untreated CpTi starts from −0.215 V and reaches to −0.235 V, nearly a stabilized potential within an hour of the immersion time. Shift in the OCP is due to the progressive increase in the thickness of the oxide layer. The uniform coverage provided by the oxide layer to the underneath substrate causes a − ve shift in the OCP with respect to the bare substrate (without oxide). It is observed that all the specimens, except those borided for 3 h, show a negative shift in the potential values after an hour of immersion (Fig. 5a–c). While the CpTi borided for 3 h indicates a positive shift, for instance by about 0.029, 0.044, and 0.027 V for boriding performed at 850, 910, and 1050 °C; relatively large fluctuations in the potentials are noticed for these specimens. Such fluctuations are often characteristics of metastability in the passive film. The CpTi borided for 5 h shows overall the most + ve while the one borided for 1 h a most − ve OCP at 850 and 910 °C; the OCP of specimen borided for 3 h stays in between 1 and 5 h borided. This trend may be related to the thickness of the coating; the thicker the coating (with increase in boriding time) the more noble is the OCP. The trend, however, is different for the specimens borided for similar durations at 1050 °C, specimen borided for 3 h results in the most positive OCP followed by specimens borided for 1 and 5 h. The OCP values after an hour of immersion are −0.123, 0.047 and −0.173 V for 1, 3 and 5 h (at 1050 °C) respectively in Ringer's solution. This could be explained based on the fact that the coating at 1050 °C has relatively large stresses and numerous cracks as illustrated in Fig. 2c, which possibly could be the conducting paths for the corrosive solution to access the subsurface layer (TiB/Ti surface) and follow the OCP of untreated CpTi. While the specimen borided for 1 h did not result in a uniform coating and therefore solution could easily contact the base substrate to yield a more negative OCP than the borided at 3 h. From these results, the boriding CpTi for 3 h appears to provide a better resistant and impermeable oxide film against the penetration of the corrosive specie reaching the underlying substrate.

a

910oC-1h 910oC-3h

2.0

910oC-5h bare Ti

1.5 1.0 0.5 0.0 -0.5 -10

10

c

-9

10

-8

10

-7

10

-6

10

i (A/cm2)

10

-2

3.0 2.5

Potential (V) vs SCE

3.4. Electrochemical behavior of borided and untreated CpTi

249

Bare Ti 1050oC-1h

2.0

1050oC-3h 1050oC-5h

1.5 1.0 0.5 0.0 -0.5 -10

10

-9

10

-8

10

-7

10

-6

10 10 i (A/cm2)

-2

Fig. 6. Anodic polarization behavior of CpTi untreated and borided specimens at (a) 850 °C, (b) 910 °C and (c) 1050 °C for 1, 3 and 5 h in Ringer's solution.

than that of the borided for 1 and 3 h. However, in view of the reported in-vivo electrochemical potential of the titanium and its alloy in the range + 0.450 to + 0.550 V (with respect to SCE), the borided CpTi seems to be in the safe region [35,36]. A lower breakdown potential of the borided compared to the CpTi, will, thus, not affect its performance as a bioimplant. From the anodic polarization, it has been further observed that the untreated CpTi shows the two passive regions with a sharp departure from the Ist passive zone and differ in the characteristic current density significantly. For instance, the Ist passive zone of the untreated CpTi is extended from −0.295 to +1.57 V and the second passive zone lies beyond + 1.57 V which corresponds to the passive current densities of ~15.32 × 10−6 and 463.8 × 10−6 A/cm2 (average values) respectively. The average passive current densities during the Ist and IInd passive zone are designated by the ipass1 and ipass2 respectively. The higher passive current density during the IInd passive region could be due to the

250

B. Sivakumar et al. / Materials Science and Engineering C 48 (2015) 243–255

Table 3 The electrochemical parameters derived from the anodic polarization curves on CpTi untreated and borided at 850, 910 and 1050 °C for 1, 3 and 5 h in Ringer's solution. Untreated and borided specimens

icorr (μA/cm2)

ipass1 (μA/cm2)

ipass2 (μA/cm2)

Average transition potential (V)

Pitting potential (V)

Untreated CpTi 850 °C-1 h 850 °C-3 h 850 °C-5 h 910 °C-1 h 910 °C-3 h 910 °C-5 h 1050 °C-1 h 1050 °C-3 h 1050 °C-5 h

0.12 0.33 0.21 0.48 0.35 0.12 0.15 0.61 0.07 0.25

15.32 51.53 21.19 56.5 55.6 15.05 52.72 60.9 14.49 39.11

463.8 109.22 97.7 133.2 102.5 60.1 80.34 175.6 119 89.9

1.57 1.29 1.40 1.51 1.47 1.37 1.51 1.47 1.39 1.43

N3.0 2.51 ± 0.02 2.58 ± 0.19 2.48 ± 0.05 2.63 ± 0.05 2.60 ± 0.05 2.52 ± 0.01 2.68 ± 0.02 2.51 ± 0.03 2.5 ± 0.01

± ± ± ± ± ± ± ± ± ±

0.01 0.02 0.07 0.04 0.07 0.02 0.02 0.10 0.01 0.05

± ± ± ± ± ± ± ± ± ±

0.73 6.6 3.4 5.9 5.5 1.5 7.3 5.0 3.0 2.6

± ± ± ± ± ± ± ± ± ±

43 7 9.7 8.9 4.5 6.5 3 10 11 6.3

± ± ± ± ± ± ± ± ± ±

0.01 0.06 0.10 0.08 0.07 0.09 0.08 0.12 0.08 0.11

ipass1 — current density in the Ist passive zone; ipass2 — current density in the IInd passive zone.

partial transformation of the most resistant titanium dioxide (TiO2) film to another variant, titanium trioxide (TiO3) as follows: −

þ

−

2TiO2 þ 2OH →2TiO3 þ 2H þ 2e :

ð4Þ

The E-pH diagram for the Ti–H2O system also shows that at a potential higher than +1.7 V (with respect to SCE) in the pH range of 6–7, a TiO3 is thermodynamically stable [37]. The potential, however, may

a

10

40 1

20 0

0.1

-5

10

-4

10

-3

-2

-1

10 10 10 Log Frequency (kHz)

0

10

Bode phase angle (-deg)

Uncorroded surface

Bare Ti 80 B-850oC-1h B-850oC-3h 60 B-850oC-5h

100

Log mod Z (Kohm.cm2)

a

-20 1 10

500 m

b

40 1

20 0

0.1

-5

2 mm

-4

10

-3

-2

-1

10 10 10 Log Frequency (kHz)

Fig. 7. The surface appearance (a) before and (b, c) after the anodic polarization of borided specimen at 850 °C for 3 h (c) shows the selective corrosion of TiB whiskers after the dissolution of TiB2 layer.

Log mod Z (Kohm.cm2)

-20 1 10

10

40

1

20 0

0.1

-5

10

-4

10

-3

-2

-1

10 10 10 Log Frequency (kHz)

0

10

Bode phase angle (-deg)

5 μm

0

10

Bare Ti B-1050C-1h 80 B-1050C-3h B-1050C-5h 60

c 100

Marks of dissolved TiB whiskers

60

10

10

c

80 Bode phase angle (-deg)

Corroded surface

Bare Ti B-910C-1h B-910C-3h B-910C-5h

100

Log mod Z (Kohm.cm2)

b

-20 1 10

Fig. 8. Bode impedance and associated phase angle plots of CpTi borided specimens at (a) 850 °C, (b) 910 °C and (c) 1050 °C for 1–5 h with the untreated CpTi in Ringer's solution.

B. Sivakumar et al. / Materials Science and Engineering C 48 (2015) 243–255

TiO2

Porous layer/outer Compact layer/inner

CPE1 Rs

CPE2

TiB2 layer

R1

Ti matrix Pores on TiB2 layer

R2

TiB whiskers

Fig. 9. Schematic illustration of an interface and corresponding EIS equivalent circuit fitted model.

further be reduced by the additional corrosive specie such as chloride (e.g. in Ringer solution). The TiO3 has a higher solubility than the TiO2 in an aqueous solution and therefore it is less stable and protective as a passive film than the TiO2. For instances, solubility of the TiO3 and TiO2 is nearly 1 and 0.0001 g-mol/l respectively in an acidic solution [37]. The current density associated with the Ist passive zone (ipass1) of the most boride specimens (except borided at 910 and 1050 °C for 3 h) is more than that of the untreated CpTi; while it is smaller (b463.8 × 10− 6 A/cm2) during the IInd passive zone of the borided specimens than the ipass2 of the untreated CpTi. The ipass1 of the CpTi borided at 910 and 1050 °C for 3 h is ~ 15.05 × 10− 6 and 14.49 × 10−6 A/cm2 respectively which are the lowest values among all the borided specimens as shown in Table 3. The reduction in the pitting resistance of the borided specimens is investigated by subjecting the post-anodic polarized specimens (up to the transpassive region) to optical and scanning electron microscope for the nature and sites of the failures on the surface of borided CpTi at 850 °C for 3 h. These are shown in Fig. 7a–c. The selective corrosion of localized sites on the coated surface is responsible for the breakdown of the passive film which is explicit from the surface topography (Fig. 7b and c). Interestingly, the corroded sites on the surface of borided CpTi were not a typical pit like geometry as commonly observed in other alloys/materials when polarized to beyond the break down potential. These, rather, appear as large white scars/patches (shown in Fig. 7b), in contrast to the absence of similar features on the uncorroded surface (in Fig. 7a). All the borided specimens that were subjected to the anodic polarization revealed similar morphological features. The EDAX analysis of the corroded regions indicate large contents of titanium (~94%) with low boron while a relatively smaller Ti (~82%) and larger boron (~10%) is revealed at uncorroded surface. This is possibly due to the dissolution of TiB2 coating from the as-coated surface during polarization and is responsible for lower breakdown potential. The higher Ti contents after the corrosion (at corroded spot) is indicative of the Ti content of a bare surface. Fig. 7b shows the marks of an underlying TiB whisker which is unveiled subsequent to the dissolution and removal of an outer TiB2 layer.

251

3.4.3. Electrochemical impedance analysis of untreated and borided CpTi The electrochemical impedance behavior of CpTi borided at 850 to 1050 °C for 1 to 5 h along with the untreated specimen in Ringer's solution at OCP is shown in Fig. 8a–c. Figure shows the Bode impedance and associated phase angle plots. The Bode impedance curves of all the borided and untreated CpTi specimens depict two distinct regions at a low (b 100 Hz) and high frequency (N100 Hz). The plot at higher frequencies (100 to 10,000 Hz) displays a flat regime with a slope nearly zero which is due to the electrolyte resistance; the frequency dependent linear region with a slope ~1 at a frequency b100 Hz indicates a capacitive nature of the passive film [38]. The equivalent circuit for the data fitting is illustrated in Fig. 9 along with its corresponding schematic interface formed in the Ringer's solution. Both the untreated and borided CpTi specimens fit to a similar equivalent circuit (Fig. 9) represented as Rs(CPE1 [R1(R2 CPE2)]) for the duplex film in the titanium materials, as described elsewhere [3,39,40]. This may also be justified as the TiB2/ TiB coating does not produce a typical barrier type effect, like several other ceramics, as substantiated by the anodic polarization behavior (Fig. 6a–c). The barrier type behavior of the TiB2/TiB during anodic polarization is absent as its electrical resistivity (10–30 Ω·cm) is close to the metals [41]. The EIS fitting considers Bode plots exhibiting the two time constants for all the borided and untreated CpTi specimens as shown in Fig. 8a–c. The equivalent circuit consists of an electrolyte resistance (Rs), resistance of the inner layer (R2), CPE2, a constant phase element of an inner layer, resistance of pores on the outer passive layer (R1) and CPE1 is a constant phase element of the outer passive film. This is typically a ladder type structure with characteristic impedance described as follows: Z ¼ Rs þ

1 Q 1 ð JωÞ

n1

:

1

þ R1

ð5Þ

1 Q 2 ð JωÞn2 þ

1 R2

The impedance of CPE is represented by the ZCPE = [Q (jω)n]−1 and describes a limiting behavior of a capacitor by ‘n’ which is the CPE exponent (footnotes 1 and 2 are the parameters of an outer and inner layer); n = 1 indicates a pure capacitor, ‘0’ a pure resistor and −1 an inductor. Q is the magnitude of CPE (in Ω−1 cm−2 s−n), ω is an angular frequency (ω = 2πf rad s−1) and j = (−1)1/2. The various EIS parameters obtained from the curve fitting, based on the above model, are presented in Table 4. The low Chi-square values obtained in the range ~ 10− 4 to 10−5 illustrate a good agreement between the experimental and modeled values. Table 4 shows that the untreated CpTi has higher resistance on the outer passive layer (1.56 × 105 Ω·cm2) as compared to that of the inner (7.379 × 104 Ω·cm2), suggesting defect/pore free oxide formation on the outer oxide layer of the CpTi in Ringer's solution. This appears to be a good protection strategy for the underlying layer and to minimize the charge transfer occurring at the inner oxide/metal interface. The borided specimens, however, show wide variations in the resistance

Table 4 The electrochemical parameters obtained from the impedance curve fittings for CpTi untreated and borided at various temperatures and times in Ringer's solution. Untreated and borided specimens

Rs (Ω·cm2)

CPE1 (Q1 × 10−5) (S·sn)

n1

R1 (105 × Ω·cm2)

CPE2 (10−5 × Q2) (S·sn)

n2

R2 (105 × Ω·cm2)

Chi-square

Untreated CpTi 850 °C-1 h 910 °C-1 h 1050 °C-1 h 850 °C-3 h 910 °C-3 h 1050 °C-3 h 850 °C-5 h 910 °C-5 h 1050 °C-5 h

52.38 50.03 42.3 49.25 53.26 56.9 42.7 48.7 52 50.67

6.21 4.50 3.87 6.06 2.59 2.78 2.42 2.92 4.21 4.2

0.89 0.86 0.89 0.80 0.89 0.90 0.91 0.95 0.91 0.89

1.56 0.82 0.04 0.83 0.62 1.23 0.003 0.001 0.003 0. 43

8.25 4.01 4.2 8.13 0.6 3.98 0.07 5.15 2.45 7.7

0.96 0.76 0.81 0.82 0.85 0.96 0.89 0.83 0.77 0.96

0.74 1.30 1.27 0.88 1.34 2.22 8.36 1.14 2.9 0.59

1.08 2.39 4.08 2.34 9.43 1.07 8.20 5.4 4.15 5.96

× × × × × × × × × ×

10−3 10−3 10−3 10−2 10−4 10−2 10−4 10−4 10−4 10−5

252

B. Sivakumar et al. / Materials Science and Engineering C 48 (2015) 243–255

Table 5 The roughness of borided CpTi as a function of boriding time and temperature. Specimens

Average surface roughness (μm) Ra

Bare Ti (2/0 grade polished) 850 °C-1 h 850 °C-3 h 850 °C-5 h 910 °C-1 h 910 °C-3 h 910 °C-5 h 1050 °C-1 h 1050 °C-3 h 1050 °C-5 h

0.16 0.24 0.26 0.30 0.27 0.41 0.45 0.53 0.55 0.68

± ± ± ± ± ± ± ± ± ±

0.01 0.03 0.02 0.02 0.02 0.02 0.01 0.07 0.01 0.01

(both R1 and R2) values, depending upon the temperature and time of the boriding process. A few of these indicates a relatively superior resistance to that of the untreated CpTi. The boriding of CpTi for 1 h, performed at three different temperatures, seems to reduce R2 (inner layer resistance) from 1.30 × 105 to 0.88 × 105 Ω·cm2 while the R1 (outer layer resistance) did not change significantly (0.082 × 105 to 0.083 × 105 Ω·cm2) as temperature is raised from 850 to 1050 °C. The reverse is, however, noticed in the R2 values of CpTi specimens borided for 3 h. The R2 of specimens borided for 3 h increases from 1.34 × 105 to 8.36 × 105 Ω·cm2 with the temperature from 850 to 1050 °C. For specimens borided at 5 h, it is increased until a temperature of 910 °C and decreased further at 1050 °C, and thus does not follow a specific trend. The variations in the resistance of the oxide layer on the specimens borided at various temperatures and times may be correlated with the composite coating thickness (TiB2 and TiB whiskers) and defects present in the coatings. The phase angle versus frequency curves in Fig. 8a–c demonstrates the maxima for specimens borided at 1 h corresponds to the phase angle at lower (64–71°) than that borided at 3 and 5 h (N75°), at almost all the temperatures. The lower maxima indicates a lower phase angle and so the large deviation from an ideal capacitor. It is possibly due to the large discontinuity that exists in the TiB2 coating carried out for 1 h duration. This is similar to as demonstrated by the earlier researcher where the large variations in the oxide coating thickness (large ratio of maximum to minimum thickness) on the same surface attributed to the lower maxima in the phase angle versus frequency curve [42]. To elucidate more on the deviation of a capacitor from its ideal behavior, a roughness in the coating was measured. This is particularly important as the dispersion of capacitance with increase in the frequency is generally attributed to (i) a microscopic/macro roughness present on electrode/electrolyte interphase (causes coupling of the solution

resistance with the surface capacitance) and (ii) due to the slow adsorption of ions and chemical inhomogeneities on the surface. The roughness on the boride coatings produced at varied time and temperature is shown in Table 5. The table indicates that both increase in the boriding temperature and the time lead to greater undulations and surface roughness in the coating. An average roughness, ‘Ra’ of the coating is observed to increase from ~0.242 to 0.524 μm with the boriding temperature from 850 to 1050 °C during 1 h treatment. The corresponding ‘n2’ value can be noticed to enhance from 0.76 to 0.82. A similar increase in the ‘n2’ values from 0.82 to 0.96 with corresponding increase in the roughness from ~ 0.524 to 0.683 is noticed with the boriding time from 1 to 5 h at 1050 °C. The ‘n2’ values are generally observed to be higher at higher boriding temperature or time which corresponds to higher roughness. This can be elaborated by the following equation that correlates a parallel plate capacitor with surface area and therefore roughness as: Cdl ¼

εεo Atrue d

ð6Þ

where Cdl in farads is a double layer capacitance, d is separation of the double layer (in cm), ε is the dielectric constant, and εo is the permittivity of the free space (8.85 × 10−14 F/cm), and Atrue is the true surface area in cm2 which is related to the surface roughness as Atrue ¼ rAgeometric

ð7Þ

where ‘r’ is surface roughness, Ageometric and Atrue are the geometric and the true surface area of specimen, Eq. (8) becomes

Cdl ¼

εεo rAgeometric : d

ð8Þ

From Eq. (8), the double layer capacitance is directly proportional to the surface roughness. The observed increase in the ideality (‘n’ values) of a capacitor with the roughness may thus be justified in a few situations. It is, however, not a straight forward and do not follow the same trend due to several other reasons; one to one correspondence with roughness is thus not observed in the several other cases. Daikhin et al. [43] has demonstrated that the capacitive behavior at the electrolyte–metal interface is largely governed by the interplay between the characteristic roughness and the Debye length. The capacitor behavior thus may be expected to vary with the magnitude of difference between the apparent surface roughness and the Debye length.

b

a

20 μm

20 μm

Fig. 10. Showing the coating uniformity at cross section of CpTi borided at 850 °C for 1 and 3 h.

B. Sivakumar et al. / Materials Science and Engineering C 48 (2015) 243–255

4. Discussion The corrosion behavior of an implant alloy is important to describe at OCP as this is the actual state that the metal/alloy encounters in a physiological environment. The corrosion of CpTi borided for 3 h at the OCP is low as compared to that at 1 and 5 h, at all the temperatures. The icorr decreases from ~ 0.21 × 10−6 to 0.07 × 10−6 A/cm2 with the boriding temperature varied from 850 to 1050 °C for 3 h. The increase in resistance of the inner passive layer as obtained from the EIS agrees well with the decrease in icorr for specimen borided for 3 h with the temperature increase from 850 to 1050 °C. In most situations, the R2 from obtained from the EIS investigation corresponds well with the changes in icorr values of the borided CpTi (Tables 3 and 4). Possible reasons for a better corrosion resistance of boride coating at 3 h than that at 1 and 5 h and with temperature could be (i) the outermost coating, TiB2 at 3 h is more uniformly developed and covered the underneath surface unlike a segmented and crack containing coating formed during 1 and

Intensity (a.u.)

a

Survey scan uncoated Survey scan coated

O1s Ti2s Ti2p

C1s Na2s and Cl2p

B1s

Ca2s K2s K2p 600

500

400

300

200

100

0

B.E.(eV)

b Intensity (a.u.)

Cl2p B1s

240

220

200

180

160

B.E.(eV)

Intensity(a.u.)

c

Table 6 The release of the coating constituents on long term exposure to the Ringer's solution at OCP for 15 and 30 days. Substrate

15 days

Untreated CpTi Pure TiB2 pellet Ti-850 °C-3 h Ti-850 °C-5 h Ti-910 °C-3 h Ti-910 °C-5 h Ti-1050 °C-5 h a

192

190

188

186

184

182

180

B.E.(eV) Fig. 11. (a) XPS survey spectrum and its (b) magnified spectrum distinguishing the presence of boron on CpTi borided specimen at 910 °C for 3 h from the untreated one and (c) the presence of B1s position after 24 h immersion in Ringer's solution.

30 days

Ti (ppm)

B (ppm)

Ti (ppm)

B (ppm)

Nil Nda Nda Nil 0.02 Nil Nil

– Nda Nda Nil 0.03 0.54 Nil

Nil 0.012 0.02 Nda 0.02 Nil Nil

– 1.74 0.04 Nda 0.05 0.77 0.96

Nd means ‘not done’.

5 h boriding time (ii) the increase in the corrosion resistance with the temperature at 3 h boriding time is due to the increase in uniform coating thickness as listed in Table 1 and shown in Fig. 10b. As opposed to this, the TiB2 layer grew segment-wise and in a non-uniform manner over the surface during 1 h boriding treatment, as explicit from Fig. 10a (one such borided surface at 850 °C). This is due to the fact that the parametric combination (time and temperature) is not favorable for the uniform growth of a TiB2 layer. Increase in the temperature (up to 1050 °C) though thickened the coating, yet remained a nonuniform. These, in unison, reduced the R2 value of CpTi borided for 1 h with the increase in the temperature. However, a reduction in the resistance of an inner layer can be taken care by the subsequent passive film grown on the outer side and thus may restore overall resistance of the borided CpTi. The lower resistance of the coating obtained at 5 h may be due to the presence of large number of cracks/pores in it as apparent in Fig. 2c. The resistance of the outer oxide (R1) does not follow a specific trend with increase in temperature, however, the lower R1 value for the boride coating at 5 h as compared to 1 and 3 h also implies the lesser pore resistance of the oxide which may be related again to the more porosity and cracks in the coating. The overall coating resistance of the borided specimen is thus appears relying upon the outermost TiB2 layer. The pitting corrosion may be a concern when the physiological environment changes drastically, specifically during the post-surgery period, consequently a large shift in the electrochemical potential for a short duration is possible. The pitting resistance of the CpTi or titanium alloys is generally attributed to the formation of impermeable TiO2 film on the surface which is insoluble and resistant to the aggressive chemical species. The apparent reduction in the pitting potential of the borided CpTi, however, indicates a change in the nature and composition of the TiO2 film. This has been substantiated by carrying out the XPS analysis of a few selected borided (at 910 °C for 3 h) and untreated CpTi specimens after exposure to the Ringer's solution for 24 h at OCP. The XPS spectra obtained are presented in Fig. 11a–c. The survey spectrum of the borided specimen displays the presence of the boron in addition to Ti (from the coating) and O (from the solution); the film on untreated CpTi revealed only the presence of Ti and O as shown in Fig. 11a and magnified survey spectrum in Fig. 11b. A slow scan spectrum of the B1s with the binding energy of 187.2 eV on the borided specimen clearly indicates the presence of boron, illustrated in Fig. 11c. The boron appeared to interfere by generating defects in the impermeable passive film (TiO2) and provided the localized sites for the passivity break down in the borided specimens. The corrosion of TiB2 is assumed to proceed such that it dissolved and forms a boric acid (as H3BO3) in the solution and titanium dioxide (TiO2) which is an insoluble compound [26,37]. The reaction may be shown as below þþ

194

253

TiB2 þ 6H2 O → Ti þþ

Ti

þ 2H3 BO3 þ 3H2 þ

−

þ 2H2 O→TiO2 þ 4H þ 2e :

ð9Þ ð10Þ

This has been further investigated by exposing the uncoated and coated titanium specimens in the Ringer's solution (75 ml/cm2) for 15

254

B. Sivakumar et al. / Materials Science and Engineering C 48 (2015) 243–255

and 30 days at OCP. The solution after exposure was analyzed by the ICP-OES (resolution up to 0.01 ppm) to detect the traces of titanium and boron; the results are shown in Table 6. There was no trace of Ti detected in the solution after untreated CpTi was exposed to the Ringer's solution. A passive film (TiO2) is formed on the surface of CpTi that adhered to the surface of the specimen; the film is impermeable and insoluble in the near neutral (aqueous) solutions and does not allow the Ti ions to go into the solution. The solution after the exposure of borided specimens, however, contained both the titanium (b 0.02 ppm) and boron indicating the dissociation/degradation of the TiB2. The boron content was as high as 0.96 ppm as a result of corrosion reaction of the borided CpTi during its exposure to the Ringer's solution for 15 and 30 days. An accumulation of B in the TiO2 film is responsible for a lower pitting resistance of the borided CpTi. Considering the point defect model for the film nucleation and growth, the inherently low cation vacancy diffusion is one of the important reasons for the high pitting potential, generally, observed for the metals such as titanium, tantalum, and zirconium [44]. This is primarily because the high potential is required to generate the critical concentration of vacancies for the breakdown of the passive film. By assuming the segregation of B in the TiO2 passive film, as supported by the XPS analysis (by showing the presence of boron in the film), and a lower pitting potential of the borided CpTi, larger flux of the gross cation vacancies at the film–solution interface has occurred. This is a result of the enhanced boron vacancy diffusion (across the B modified TiO2 film) compared to the diffusion of Ti4+ vacancy in the pure TiO2 and seems to be responsible for the early breakdown of passive film on the borided CpTi specimens. A similar decrease in pitting potential of the boron alloyed Ti–6Al–4V (studied up to 1% B) in NaCl solution is also reported earlier [45]. Another objective of the ICP-OES analysis of the solution after the specimen exposure was to investigate the possible release of coating (composite boride) constituents to the human body. The results indicate that, though, there is insignificant quantity of titanium and boron leached to the solution during the studied durations, the latter may, however, increase with the time of exposure (Table 6). Comparison has also been made with the leaching from a pure TiB2 porous pellet which shows to contain a highest quantity of boron in the solution. 5. Conclusions 1. Multilayered composite coating of TiB2/TiB whiskers is fabricated on the surface of CpTi by boriding at temperatures ranging from 850 to 1050 °C. The coating thickness is varied from ~ 18.18 to 26.36 μm with the processing temperatures; the largest thickness is obtained at 910 °C. Despite the highest aspect ratio of the TiB whiskers (0.23) obtained on the specimen borided at 850 °C, a maximum strengthening efficiency of 3.2 is achieved for the specimen borided at 910 °C due to the presence of large TiB volume fraction. 2. The high hardness value of ~ 2968 Hv15gf on the outer layer of the coating is attributed to the formation of TiB2 which decreased further with the emergence of TiB. 3. Anodic polarization studies of borided CpTi unveiled the reduction in the pitting corrosion resistance, yet remain much higher than the 0.450–0.550 VSCE (electrochemical potential in in-vivo) in the safe region. Both untreated CpTi and after boriding show two different passive zones leading to the different passivation current densities. The icorr and passive current density, among the specimens borided at various conditions, indicates a better corrosion performance of the specimen borided for 3 h. 4. EIS studies also imply the highest polarization resistance (of inner layer) for the CpTi borided for 3 h at all the temperatures (850, 910 and 1050 °C) which is in agreement with the anodic polarization results. The increasing roughness of the borided specimen with both temperature and time influenced the capacitor behavior of double layer.

5. The corrosion of the boride coated CpTi in Ringer's solution appears to occur through the dissolution of TiB2.

Acknowledgments B. Sivakumar, Ph.D. Scholar of AcSIR sincerely acknowledges the CSIR, India for providing the senior research fellowship. The authors express their sincere thanks to the Director, CSIR-NML, Jamshedpur for his kind support during this work. Authors also thank Mr. Himansu Bapari for his help in experimental work. The present work is partly supported by the iPSG of CSIR-NML under the project no. OLP-174. References [1] J. Jacobs Joshua, L. Gilbert Jeremy, M. Urban Robert, Current concepts review — corrosion of metal orthopaedic implants, J. Bone Joint Surg. 80 (1998) 268–282. [2] U. Kamachi Mudali, T.M. Sridhar, Baldev Raj, Corrosion of bioimplants, parts 3 & 4, Sadhana 28 (2003) 601–637. [3] S.L. Assis, S. Wolynec, I. Costa, Corrosion characterization of titanium alloys by electrochemical techniques, Electrochim. Acta 5 (2006) 1815–1819. [4] B. Sivakumar, S. Kumar, T.S.N. Sankara Narayanan, Fretting corrosion behaviour of Ti–6Al–4V alloy in artificial saliva containing varying concentrations of fluoride ions, Wear 270 (2011) 317–324. [5] F. Galliano, E. Galvanetto, S. Mischler, D. Landolt, Tribocorrosion behavior of plasma nitrided Ti–6Al–4V alloy in neutral NaCl solution, Surf. Coat. Technol. 145 (2001) 121–131. [6] Rahman Mahfujur, I. Reid, P. Duggan, D.P. Dowling, G. Hughes, M.S.J. Hashmi, Structural and tribological properties of the plasma nitrided Ti-alloy biomaterials: influence of the treatment temperature, Surf. Coat. Technol. 201 (2007) 4865–4872. [7] R. Singh, A. Kurella, N.B. Dahotre, Laser surface modification of Ti–6Al–4V: wear and corrosion characterization in simulated biofluid, J. Biomater. Appl. 21 (2006) 49–73. [8] S. Kumar, T.S.N. Sankara Narayanan, S. Ganesh Sundara Raman, S.K. Seshadri, Fretting corrosion behaviour of thermally oxidized CP-Ti in Ringer's solution, Corros. Sci. 52 (2010) 711–721. [9] T.M. Manhabosco, S.M. Tamborim, C.B. dos Santos, I.L. Muller, Tribological, electrochemical and tribo-electrochemical characterization of bare and nitrided Ti6Al4V in simulated body fluid solution, Corros. Sci. 53 (2011) 1786–1793. [10] C. Lee, A. Sanders, N. Tikekar, K.S. Ravi Chandran, Tribology of titanium boridecoated titanium balls against alumina ceramic: wear, friction, and micromechanisms, Wear 265 (2008) 375–386. [11] B. Sarma, K.S. Ravi Chandran, Recent advances in surface hardening of titanium—review, J.O.M. 63 (2011) 85–92. [12] G. Liu, D. Zhu, J.K. Shang, Temperature dependence of fracture toughness in TiCparticulate reinforced Ti–1b6Al–1b4V matrix composite, Scr. Mater. 28 (1993) 729–732. [13] S. Gorsse, J.P. Chaminade, Y.L. Petitcorps, In situ preparation of titanium base composites reinforced by TiB single crystals using a powder metallurgy technique, Compos. Part A 29 (1998) 1229–1234. [14] M.Y. Koo, J.S. Park, M.K. Park, K.T. Kim, S.H. Hong, Effect of aspect ratios of in situ formed TiB whiskers on the mechanical properties of TiBw/Ti–6Al–4V composites, Scr. Mater. 66 (2012) 487–490. [15] V. Jain, G. Sundararajan, Influence of the pack thickness of the boronizing mixture on the boriding of steel, Surf. Coat. Technol. 149 (2002) 21–26. [16] I. Campos, J. Oseguera, U. Figueroa, J.A. Garcıa, O. Bautista, G. Kelemenis, Kinetic study of boron diffusion in the paste-boriding process, Mater. Sci. Eng. A 352 (2003) 261–265. [17] M. Keddam, M. Kulka, A. Pertek, L. Maldzinski, A kinetic model for estimating the boron activation energies in the FeB and Fe2B layers during the gas-boriding of Armco iron: effect of boride incubation times, Appl. Surf. Sci. 298 (2014) 155–163. [18] A.P. Sanders, N. Tikekar, C. Lee, K.S. Ravi Chandran, Surface hardening of titanium articles with titanium boride layers and its effects on substrate shape and surface texture, J. Manuf. Sci. Eng. 131 (2009) 031001-1. [19] G. Kartal, S. Timur, M. Urgen, A. Erdemir, Electrochemical boriding of titanium for improved mechanical properties, Surf. Coat. Technol. 204 (2010) 3935–3939. [20] L.I. Fenghua, Y.I. Xiaohong, Z. Jinglei, F. Zhanguo, G. Dianting, X.I. Zhengping, Growth kinetics of titanium boride layers on the surface of Ti6Al4V, Acta Metall. Sin. (Engl. Lett.) 23 (2010) 293–300. [21] Erdem Atar, E. Sabri Kayali, Huseyin Cimenoglu, Characteristics and wear performance of borided Ti6Al4V alloy, Surf. Coat. Technol. 202 (2008) 4583–4590. [22] B. Sarma, K.S. Ravi Chandran, Accelerated kinetics of surface hardening by diffusion near phase transition temperature: mechanism of growth of boride layers on titanium, Acta Mater. 59 (2011) 4216–4228. [23] N. Tikekar, K.S. Ravi Chandran, A. Sanders, Nature of growth of dual titanium boride layers with nanostructured titanium boride whiskers on the surface of titanium, Scr. Mater. 57 (2007) 273–276. [24] A. Miklaszewski, M.U. Jurczyk, K. Jurczyk, M. Jurczyk, Plasma surface modification of titanium by TiB precipitation for biomedical applications, Surf. Coat. Technol. 206 (2011) 330–337. [25] F.M. Makau, K. Morsi, N. Gude, R. Alvarez, M. Sussman, K.M. Newman, Viability of titanium–titanium diboride composite as a biomaterial, ISRN Biomater. 2013 (2013) Article ID 970535, 8 pages.

B. Sivakumar et al. / Materials Science and Engineering C 48 (2015) 243–255 [26] B.S. Covino Jr., S.D. Cramer, J.P. Carter, D. Schlain, Corrosion of titanium diboride, J. less common met. 41 (1975) 211–224. [27] C. Monticelli, A. Frignani, A. Bellosi, G. Brunoro, G. Trabanelli, The corrosion behavior of titanium diboride in neutral chloride solution, Corros. Sci. 43 (2001) 979–992. [28] H. Feng, Y. Zhou, D. Jia, Q. Meng, Microstructural characterization of spark plasma sintered in situ TiB reinforced Ti matrix composite by EBSD and TEM, Mater. Trans. 46 (2005) 575–580. [29] Z. Fan, Z.X. Guo, B. Cantor, The kinetics and mechanism of interfacial reaction in sigma fibre reinforced Ti MMCs, Compos. Part A 28 (1997) 131–140. [30] U. Kohler, C.H. Herzig, On the anomalous self-diffusion in BCC-titanium, Phys. Status Solidi B 144 (1987) 243–251. [31] H. Nakajima, M. Koiwa, Diffusion in titanium, ISIJ Int. 31 (1991) 757–766. [32] H.J. Ryu, S.I. Cha, S.H. Hong, Generalized shear-lag model for load transfer in SiC/Al metal–matrix composites, J. Mater. Res. 18 (2003) 2851–2858. [33] Z. Zachariev, Metal, ceramic and polymeric composites for various uses, in: John Cuppoletti (Ed.), New Superhard Ternary Borides in Composite Materials, In Tech, Croatia, 2011, pp. 61–78 (Chap. 3). [34] K. Niihara, R. Morena, D.P.H. Hasselman, Evaluation of KIc of brittle solids by the indentation method with low crack-to-indent ratios, J. Mater. Sci. Lett. 1 (1982) 13–16. [35] T.P. Hoar, D.C. Mears, Corrosion-resistant alloys in chloride solutions: materials for surgical implants, Proc. R. Soc. A 294 (1966) 486–510. [36] R. Singh, M. Martin, N.B. Dahotre, Influence of laser surface modification on corrosion behaviour of stainless steel 316L and Ti–6Al–4V in simulated biofluid, Surf. Eng. 21 (2005) 297–306.

255

[37] M. Pourbaix, Atlas of electrochemical equilibriums in aqueous solutions, Pergamon Press, Oxford, New York, 1966. 213–222. [38] A.W.E. Hodgson, Y. Mueller, D.C. Forster, S. Virtanen, Electrochemical characterization of passive films on Ti alloys under simulated biological conditions, Electrochim. Acta 47 (2002) 1913–1923. [39] J. Pan, D. Thierry, C. Leygraf, Electrochemical impedance spectroscopy study of the passive oxide film on titanium for implant application, Electrochim. Acta 41 (1996) 1143–1153. [40] J.E.G. Gonzalez, J.C. Mirza-Rosca, Study of the corrosion behavior of titanium and some of its alloys for biomedical and dental implant applications, J. Electroanal. Chem. 471 (1999) 109–115. [41] A.D. McLeod, J.S. Haggerty, D.R. Sadoway, Electrical resistivities of monocrystalline and polycrystalline TiB2, J. Am. Ceram. Soc. 67 (1984) 705–708. [42] S. Cattarin, N. Comisso, M. Musiani, B. Trobollet, Electrode coating by passive oxide films with a thickness profile: modeling and measurement of impedance, J. Electrochem. Soc. 155 (2008) 543–549. [43] L.I. Daikhin, A.A. Kornyshev, M. Urbakh, Double layer capacitance on a rough metal surface: surface roughness measured by Debye ruler, Electrochim. Acta 42 (1997) 2853–2860. [44] D.D. Macdonald, The Point Defect Model for the Passive State, J. Electrochem. Soc 139 (1992) 3434–3449. [45] V. Ravi, A. Schissler, W. Chantrjaroen, C. Beecher, A. Razzak, R. Urak, S. Alas, Corrosion behaviour and biocompatibility of boron containing titanium alloys in simulated physiological environments, Corros. Eng. Sci. Technol. 47 (2012) 383–387.