PVD process as a thin film coating method is highly applicable for both metallic and ceramic materials, which is faced with the necessity of choosing the correct parameters to achieve optimal results. In the present study, a GEP-based model for the first time was proposed as a safe and accurate method to predict the adhesion strength and hardness of the Nb PVD coated aimed at growing the mixed oxide nanotubular arrays on Ti67. Here, the training and testing analysis were executed for both adhesion strength and hardness. The optimum parameter combination for the scratch adhesion strength and micro hardness was determined by the maximum mean S/N ratio, which was 350W, 20 sccm, and a DC bias of 90V. Results showed that the values calculated in the training and testing in GEP model were very close to the actual experiments designed by Taguchi. The as-sputtered Nb coating with highest adhesion strength and microhardness was electrochemically anodized at 20V for 4h. From the FESEM images and EDS results of the annealed sample, a thick layer of bone-like apatite was formed on the sample surface after soaking in SBF for 10 days, which can be connected to the development of a highly ordered nanotube arrays. This novel approach provides an outline for the future design of nanostructured coatings for a wide range of applications.
Recently, the robust optimization and prediction models have been highly noticed in district of surface engineering and coating techniques to obtain the highest possible output values through least trial and error experiments. Besides, due to necessity of finding the optimum value of dependent variables, the multi-objective metaheuristic models have been proposed to optimize various processes. Herein, oriented mixed oxide nanotubular arrays were grown on Ti-6Al-7Nb (Ti67) implant using physical vapor deposition magnetron sputtering (PVDMS) designed by Taguchi and following electrochemical anodization. The obtained adhesion strength and hardness of Ti67/Nb were modeled by particle swarm optimization (PSO) to predict the outputs performance. According to developed models, multi-objective PSO (MOPSO) run aimed at finding PVDMS inputs to maximize current outputs simultaneously. The provided sputtering parameters were applied as validation experiment and resulted in higher adhesion strength and hardness of interfaced layer with Ti67. The as-deposited Nb layer before and after optimization were anodized in fluoride-base electrolyte for 300min. To crystallize the coatings, the anodically grown mixed oxide TiO2-Nb2O5-Al2O3 nanotubes were annealed at 440°C for 30min. From the FESEM observations, the optimized adhesive Nb interlayer led to further homogeneity of mixed nanotube arrays. As a result of this surface modification, the anodized sample after annealing showed the highest mechanical, tribological, corrosion resistant and in-vitro bioactivity properties, where a thick bone-like apatite layer was formed on the mixed oxide nanotubes surface within 10 days immersion in simulated body fluid (SBF) after applied MOPSO. The novel results of this study can be effective in optimizing a variety of the surface properties of the nanostructured implants.