Corrosion Resistance of Ultrafine-grained Titanium Alloys in Different Corrosive Environments

Corrosion Resistance of ultrafine-grained (UFG) tanium alloys fabricated by equal channel angular pressing (ECAP) was inves gated in this study. Electrochemical measurements of pure Ti and Ti-6Al-7Nb alloy were conducted in 3.5 wt.% NaCl and Ringer’s solu on separately. Results indicated that both ultrafine-grained pure Ti and Ti6Al-7Nb alloy had much lower corrosion current density than annealed coarse-grained counterparts in the specified corrosive environment.


Introduc on
Titanium alloy is widely applied in the field of chemical engineering, biomedical and marine industries due to high specific strength and corrosion resistance [1][2][3] . Ultrafine-grained (UFG) and nanocrystalline (NC) materials with high strength and duc lity processed by severe plas c deforma on are widely studied in the recent decades [4][5][6] . Compared with large number researches on mechanical behavior of nanocrystalline materials, corrosion resistance is rarely studied and results indicated inconsistent, even within the same alloy system [7,8] .
Nie [9] et al. inves gated corrosion behavior of UFG-Ti processed by HPT in 3.5 wt.% NaCl solu on, which found coarse grain (CG) Ti possess be�er corrosion resistance than UFG Ti. Garbacz [10] studied corrosion behavior of UFG-Ti fabricated by hydrosta c extrusion (HE) by Auger electron spectroscopy (AES) and Ar + spu�ering technology in 0.15 mol/L NaCl solu on, results shown that corrosion resistance decrease for UFG-Ti and thickness of passive film formed on the surface was no much more different. However, Balyanov [11] founded that UFG-Ti processed by ECAP possess be�er corrosion resistance than CG-Ti in the HCl and H 2 SO 4 solu ons, which could be a�ributed to the rapid forma on of passive film for UFG-Ti. Fa�ah-Alhosseini [12,13] fabricated UFG-Ti by the method of fric on s r processed (FSP) and accumula ve roll bonding (ARB), results indicated that corrosion current and passive current density decreased with decreasing grain size and corrosion poten al increased, which revealed that corrosion resistance of UFG-Ti is be�er than CG-Ti, the same conclusion was also founded in researches of Gurao [14] and Raducanu [15] , who studied corrosion behavior of UFG-Ti and UFG-Ti-10Zr-5Nb-5Ta in simulated body fluid separately.
Obviously, different corrosive environment and UFG microstructure by various process methods could be responsible for inconsistent conclusion. In the present work, corrosion behavior of UFG-Ti and UFG-Ti-6Al-7Nb processed by ECAP was inves gated in 3.5 wt.% NaCl and Ringer's solu on respec vely. The aim of our studies is trying to clarify the effect of grain size on corrosion behavior.

Experimental
Commercial purity Ti Grade 4 (CP Ti) and Ti-6Al-7Nb alloy were used in the present study, ECAP was conducted in Ufa State Aviation Technical University. Chemical composition is listed in Table 1 and Table 2. Microstructure characteriza on of samples was performed by the method of electron backsca�ered diffrac on (EBSD). Before performing the EBSD inves ga on, the samples were mechanically polished with abrasive paper and polishing paste, then subjected to electro-polishing (solu on: HClO 4 : C 2 H 5 OH =19:1 (volume ra o); voltage: 40V; polishing me: 90s). Automated EBSD scans were using Flamenco data acquisi on so�ware. All the data were processed with HKL Technology Channel 5 so�ware.
Corrosion behavior of samples was investigated by potentiodynamic polarization. electrochemical test was performed using CHI660E electrochemical station in 3.5 wt% NaCl solution for CG-Ti, potentiodynamic polarization data were acquired at a sweep rate of 2 mV/s from -250 mV to 800 mV/SCE. On the other hand, electrochemical test was performed using CHI660E electrochemical station in Ringer's solution for Ti-6Al-7Nb, potentiodynamic polarization data were acquired at a sweep rate of 2 mV/s from -800 mV to 800 mV/SCE. The electrolyte was a naturally aerated Ringer's solution (an aqueous solution of the chlorides of sodium, potassium, and calcium that is isotonic to animal tissue and is used topically as a physiological saline), of which composition is 9g NaCl, 0.42 g KCl and 0.25 g CaCl 2 . The open circuit potential variation with time was measured in Ringer's solution at 37℃,the pH is adjusted to 7.

Microstructure evolu on
Microstructure of the CG-Ti and Ti-6Al-7Nb is illustrated in Figure 1 and Figure 2, It can be seen fairly equiaxed and homogeneous grains were consisted in the microstructure, the grain size distribution calculated by Technology Channel 5 software indicates that the average grain size of CG-Ti is 142.9µm and large volume fraction concentrates (∼8.5%) on the grain size interval 50~60μm (Figure 1b), at the same time, average grain size of the Ti-6Al-7Nb is 13.8µm and large volume fraction concentrates (∼45%) on the grain size interval 5~10μm (Figure 2b). Microstructure of the UFG-Ti and UFG-Ti-6Al-7Nb processed by ECAP is illustrated in Figure 3 and Figure 4, which displays a remarkable grain refinement resulted from the large volume of plastic deformation applied during ECAP process. Microstructure of UFG-Ti and UFG-Ti-6Al-7Nb sample also reveals equiaxed and homogeneous grains, the grain size distribution calculated by Technology Channel 5 software indicates that the average grain size of Ti reduced to 0.567µm and large volume fraction concentrates (∼20%) on the grain size interval 300~400 nm (Figure 3b), at the same time, average grain size of the UHG-Ti-6Al-7Nb reduced to 0.4µm and large volume fraction concentrates (∼55%) on the grain size interval 200-400nm (Figure 4b).

Corrosion behavior
Potentiodynamic polarization curves of CG-Ti and UFG-Ti processed by ECAP in 3.5 wt.% NaCl solution is illustrated in Fig.5. The results show that both CG-Ti and UFG-Ti have similar polarization behaviors all samples exhibited obvious pitting corrosion phenomenon, potential from -0.276 V to 0.091 V is passivation region for CG-Ti, passivation range is about 0.367V, pitting corrosion happened when the potential reached to 0.091 V. On the other hand, the passivation range of UFG-Ti is 0.315 V, passive film cracked when the potential reached to 0.117 V. The major reaction of the Tafel cathodic regions is O 2 +4H + +4e − = 2H 2 O, while the major reaction of the Tafel anodic regions is Ti-4e -= Ti 4+ . The mean corrosion potential (E corr ) and corrosion current density (I corr ) values obtained from all the tested polarization curves are listed in Table 3. The E corr and I corr values were determined by extrapolating the linear Tafel segments of the anode and cathode polarization curves. According to Table 3, the I corr of initial CG-Ti is 4.8 times higher than UFG-Ti processed by ECAP, from which can be concluded that the corrosion resistance of UFG-Ti is superior to the CG-Ti.  Potentiodynamic polarization curves of CG-Ti-6Al-7Nb and UFG-Ti-6Al-7Nb processed by ECAP in Ringer's solution is illustrated in Fig.6. All of the samples exhibited the similar polarization behavior and reached their respective stable passive current densities as the potential increased. Potential from -0.228 V to 0.191 V is passivation region for CG-Ti-6Al-7Nb, passivation range is about 0.419V, pitting corrosion happened when the potential reached to 0.191 V. On the other hand, passivation range increased to 0.478 V for UFG-Ti-6Al-7Nb, much higher than CG one, passive film cracked when the potential reached to 0.230 V. The mean corrosion potential and corrosion current density values obtained from all the tested polarization curves are listed in Table 4. Results indicated that corrosion current density of CG-Ti-6Al-7Nb is also higher than the UFG-Ti-6Al-7Nb, which means that the corrosion resistance of UFG-Ti-6Al-7Nb is better than CG-Ti-6Al-7Nb.  Titanium alloy has excellent corrosion behavior due to the forma on of TiO2 passive film on the surface. However, the exhibi on of high concentra on of aggressive ion can penetrate passive film and reach to the interface between passive film and matrix tanium, which could form corrosive galvanic cell and make rapid reac on of matrix with aggressive ion, then lead to pi ng corrosion [16] . On the other hand, pi ng corrosion could be restrained due to its self-repairing ability, which means a fresh passive film could be developed to provide further protec on.
Pi ng corrosion is difficult to happen for UFG tanium alloy in resent research, because ECAP could introduce UFG metals with plenty of disloca on in grain interior and boundaries simultaneous, nuclea on of passive film could be formed in grain interior and boundaries at the same me, which made passive film more dense and resisted penetra on of aggressive ion. Sotniczuk [17] revealed that corrosion behavior improve for UFG-Ti due to slight increasing roughness of surface, which made the passive film more stable. At the same me, drama c increase of defec ve promote more dense of passive film suppressed the occurrence of pi ng corrosion. On the other side, passive current density of UFG Ti is much lower than CG Ti, which indicated that passiva on film of UFG tanium is more prone to self-repairing a�er pi ng corrosion. This is mainly a�ributed to a high storage energy in the disloca ons and grain boundaries in UFG tanium alloy processed by ECAP, which is beneficial to the forma on of the passiva on film, thereby enhancing the self-repairing ability of the passiva on film.

Conclusions
Corrosion behavior of ultrafine-grained Ti and Ti-6Al-7Nb fabricated by ECAP was investigated in 3.5 wt.% NaCl and Ringer's solution separetely. Electrochemical results indicated that UFG-Ti and UFG-Ti-6Al-7Nb alloy exhibits much lower corrosion current density than CG counterparts, which could be attributed to more dense and rapid self-repairing of passivation film of UFG titanium alloy.