With rising carbon dioxide emissions there has been increasing interest in schemes for converting the green house gas to valuable fuels such as methanol. Various attempts at the electrochemical and photoelectrochemical reduction of carbon dioxide to methanol have been described in the literature in the past few decades, however, all approaches operated under high overpotentials making them highly inefficient in energy conversion. Specifically, photoelectrochemical schemes have shown minimal or no actual light to chemical energy conversion. We have previously reported on a stable system containing a soluble pyridinium component at hydrogenated Pd and Pt electrodes that operated very close to the thermodynamic potential to yield near 30% faradaic efficiency for methanol. We have recently shown that we can export this chemistry to yield upwards of 100% faradaic efficiency for methanol at an illuminated p-GaP semiconducting electrode with large underpotentials, thus driven by light energy. In effect, our catalyzed photoelectrochemical system offers the prospect of pure energy conversion requiring no input of electrical energy. In this work we will expand on our efforts in the photoelectrochemical conversion of carbon dioxide to methanol. We will also explore our recent efforts to fully understand the mechanism of conversion, not previously examined, at both metal and semiconducting electrodes.