Surface hardening of titanium by thermal oxidation

Abstract

Titanium and titanium alloys have been widely used in industrial and medical fields due to their excellent corrosion resistance and biocompatibility. Poor wear resistance, however, is one of the major problems of these alloys and restricts them from wider application. Nitriding has been the major surface hardening method used for titanium alloys so far [1–4]. Oxygen can also strengthen titanium by solid solution, but reports of the oxidation strengthening of titanium are lacking. In the present study, the oxygen diffusion layers of titanium, which were obtained after thermal oxidation at 700– 900 ◦C for 1–4 hr, were systemically investigated in order to explore the feasibility of an oxygen diffusion hardening process as an alternative surface modification technique for titanium. Specimens of size 10 mm × 10 mm × 2 mm were cut from commercially pure titanium sheet of TA2 grade (Northwest Nonferrous Metal Institute, China). All the specimens were annealed at 800 ◦C for 1 hr, and then pickled in HF aqueous solution to remove the surface oxide layers and the underlying oxygen diffusion layers. The thermal oxidation was carried out at 700–900 ◦C for 1–4 hr in an air atmosphere. The surface hardness of the oxygen diffusion layers was measured using an HV-1000 Vickers microhardness tester (Shanghai Materials Tester Machine Company, China) with 50–1000 g indenter loads. The cross-sectional hardness profile was measured using a 50 g indenter load. The crystal structure was analyzed using a PHILIPS X-ray diffractometer with Cu-Kα radiation at a scan speed of 4 ◦ min−1. The microstructure of the diffusion layers was observed using a LEICA MPS60 optical microscope and a HITACHI S-570 scanning electron microscope (SEM). The surface oxide on all the specimens was identified as rutile TiO2, as indicated by the X-ray diffraction (XRD) pattern of the oxide film on the 700 ◦C/1 hr treated specimen shown in Fig. 1. By quenching in water following oxidation, the oxide films on all of the specimens could be removed by gently grinding with SiC paper, exposing the underlying oxygen diffusion layers. Fig. 1 shows the XRD patterns of the oxygen diffusion layers of the specimens with various treatments. It can be seen that with increasing treatment temperature or period of time, all the diffraction peaks, corresponding to α-Ti, move to lower angles, suggesting increased lattice parameters. Detailed analyses reveal that the (002) peak, representing the c axis, shifts the most, while the (100) peak, representing the a axis, moves the least. This can be attributed to the fact that the solution of oxygen atoms in octahedral interstices of α-Ti causes mainly the elongation of the c axis and hardly affects the a axis [5, 6]. Table I lists the lattice parameters that were calculated from the 2θ values of (002) and (101) diffraction peaks and the estimates of the associated oxygen concentrations [6]. The higher the treatment temperature and the longer the period of time, the higher the oxygen concentration. However, under the treatment conditions involved in the present research, the oxygen concentration did not reach the highest value of 34 at.%, as reported in Reference [6]. The micrographs in Fig. 2 show the oxygen diffusion layers of the specimens after 800 ◦C/1 hr and 900 ◦C/1 hr treatments. It can be seen that in the region near to the surface there is a black line (indicated by an arrow) that runs across the grain boundaries. The hardness indenter marks in Fig. 2 indicate an obvious hardening effect, even on the interior side of the black line, suggesting that the black line is not the boundary between the diffusion layer and the metal substrate. In