Does order-disorder transition exist in near-stoichiometric Ti-Ni shape memory alloys?

Previous work by Honma et al. suggested that there is an order-disorder transition around 1090 °C in Ti-Ni alloys, but its existence has not been confirmed by independent investigation. In the present work, various techniques including TG-DTA, resistance measurement and TEM were employed to detect the possible signs of such a transition. We found interestingly that when strong measures were taken to prevent sample oxidation, no sign for order-disorder transition was detected; but when the sample was allowed to be oxidized during measurement, we did find a clear DTA peak at 1118 °C. With SEM-EPMA analysis this peak was identified as originating from a eutectic reaction in the oxidation-affected layer, where selective oxidation of Ti causes significant enrichment of Ni and hence drives the oxidation-affected layer into the eutectic regime. Therefore, our results disprove the existence of "order-disorder transition" in the B2 portion of the Ti-Ni phase diagram.