Semi-Conductor Photocells and Rectifiers

Publisher Website
Google Scholar
Abstract
One of the most interesting products of photoelectric researches during recent years is the semi‐conductor cuprous oxide cell. Numerous attempts have been made in the past to satisfactorily interpret the phenomena involved, but so far, due largely to variables not under proper control experimentally, all attempts have failed. One of these variables is the cuprous oxide film. Time and again this has been prepared thermally under apparently identical conditions and yet the behavior of the oxide and the performance of the cells in which this oxide was incorporated showed wide variations. To date, the best method of producing the photoactive cuprous oxide film or layer has been by thermal oxidation of a copper metal surface. In view of the ease with which cuprous oxide can be produced electrolytically (the red “antifoul;ing“ pigment is thus produced commercially) it naturally followed that the electrolytic method with its close control of operating conditions deserved more intensive study in the preparation of light‐sensitive layers on copper. However, all electrolytic experiments in the past carried out at Columbia and elsewhere failed to evolve an active cuprous oxide in spite of the wide physical variety of cuprous oxide produced. Recently a patent was published which described deposits composed primarily of cuprous oxide. Upon investigation these deposits proved to be photoactive—in other words these were the first electrolytic cuprous oxide deposits which showed appreciable response when exposed to radiation. One of the fundamental advantages of the electrolytic cuprous oxide film as compared with that produced thermally is the ease with which the thickness of the electrolytic cuprous oxide film or layer can be reproduced and developed to any predetermined desired thickness. Furthermore, films of micron thickness can be produced and reproduced electrochemically but not thermally. This is of particular importance from a theoretical as well as a commercial point of view. Heretofore with the thermally produced oxide the control of film thickness was practically impossible. Finally, in the case of the electrolytically prepared cuprous oxide films we can deposit these on metals other than copper. This would be very difficult by thermal methods. With the aid of the closely controlled thickness of the cuprous oxide layer the present experiments produced results, a number of which are contrary to those previously recorded or predicted with thermally prepared cuprous oxide layers. For example, in the present experiments it was found that rectification is independent of the basis metal whether it be copper or any other metal and that the so‐called funnel effect of Teichmann does not exist. It has also been definitely established in the present research that the nature of the metal at the back‐wall, whether copper, gold, silver, nickel, brass, etc., does not appreciably affect the photo‐e.m.f. This finding is new and quite contrary to the assumption and belief of previous investigators. Our experiments further indicate that the presence of insulating material is not necessary for either the rectification effect or the photo‐effect. Nor is the rectification effect an essential component of the photo‐effect. The seat of the photo‐e.m.f. in “back‐wall” cells was located in the front surface of the cuprous oxide layer. This is also contrary to previous assumptions. On the basis of the new experimental data submitted, a theoretical analysis of the rectification effect and photo‐effect is presented. We believe this analysis not only more readily accounts for our own observations than do older theories, but likewise accounts for the observations of previous investigators who did not have available the quantitatively controlled cuprous oxide films used in our experiments.

This publication has 0 references indexed in Scilit: