Debunking the effect of water content on anodizing current: Evidence against the traditional dissolution theory