Oxide degradation precedes additively manufactured Ti-6Al-4V selective dissolution: An unsupervised machine learning correlation of impedance and dissolution compared to Ti-29Nb-21Zr

Abstract

Additively manufactured (AM) Ti-6Al-4V devices are implanted with increasing frequency. While registry data report short-term success, a gap persists in our understanding of long-term AM Ti-6Al-4V corrosion behavior. Retrieval studies document β phase selective dissolution on conventionally manufactured Ti-6Al-4V devices. Researchers reproduce this damage in vitro by combining negative potentials (cathodic activation) and inflammatory simulating solutions (H2O2-phosphate buffered saline). In this study, we investigate the effects of these adverse electrochemical conditions on AM Ti-6Al-4V impedance and selective dissolution. We hypothesize that cathodic activation and H2O2 solution will degrade the oxide, promoting corrosion. First, we characterized AM Ti-6Al-4V samples before and after a 48 h −0.4 V hold in 0.1 M H2O2/phosphate buffered saline. Next, we acquired nearfield electrochemical impedance spectroscopy (EIS) data. Finally, we captured micrographs and EIS during dissolution. Throughout, we used AM Ti-29Nb-21Zr as a comparison. After 48 h, AM Ti-6Al-4V selectively dissolved. Ti-29Nb-21Zr visually corroded less. Structural changes at the AM Ti-6Al-4V oxide interface manifested as property changes to the impedance. After dissolution, the log-adjusted constant phase element (CPE) parameter, Q, significantly increased from −4.75 to −3.84 (Scm−2(s)α) (p =.000). The CPE exponent, α, significantly decreased from.90 to.84 (p =.000). Next, we documented a systematic decrease in oxide polarization resistance before pit nucleation and growth. Last, using k-means clustering, we established a structure–property relationship between impedance and the surface's dissolution state. These results suggest that AM Ti-6Al-4V may be susceptible to in vivo crevice corrosion within modular taper junctions.