Combinatorial synthesis of solid state electronic materials for renewable energy applications

Abstract

We report on the development of new combinatorial capabilities for both the synthesis of new solid state opto-electronic materials and optimization of existing materials for renewable energy applications especially photovoltaic devices. Some of the important materials for these renewable energy applications include semiconductors (such as for absorber layers), transparent conductors, energy storage materials and more. In this paper, we focus on the application of a combinatorial approach to two specific material areas, transparent conducting oxides and thin film Si. In the former case, libraries are generated by sputtering and in the latter by the hot-wire chemical vapor deposition. The application of a combinatorial approach to these materials areas can greatly accelerate the rate of discovery and optimization of new materials and the optimization of devices. We report on the development of tools for the production and characterization of libraries and on initial important results in both of these areas.

Introduction

Increasing the experimental efficiency, i.e. the experimental throughput, is a central piece of the motivation for going to a combinatorial approach to material and device research. The initial move to parallel experimentation started with Hanak [1] almost 30 years ago. He was interested in the possibility of investigating composition space during sputtering by looking at more than one composition at a time. However, it was not until recently that the deposition tools, computing power and, more importantly, the analytical techniques became available to allow true combinatorial approaches [2], [3], [4], [5]. The combinatorial approach utilizes the basic principle that you deposit a “library” of the materials or devices of interest, subsequently process the library, if needed, and then analyze the library rapidly using an appropriate analytical technique. Analytical techniques must be chosen look specifically at the variables of the most interest. It is this parallel process approach that defines combinatorial science and potentially makes it much more efficient than the traditional linear method of optimizing materials and devices one point at a time. It is also the massively parallel nature of combinatorial science that has really only become possible with the advent of highly automated deposition and characterization tools.

Most of the conventional and commercially important transparent conducting oxides (TCOs) are n-type materials such as InSnO_x, ZnO, and SnO₂. They are widely used in the solar cells, low-e windows, flat screen TV, and flat panel display applications. The recent demonstration of p-type TCOs [6], [7], [8], [9], [10] has stimulated a flurry of new activities in the TCO field. With viable p-type materials it becomes possible to make functional transparent electronics. p-Type materials are also potentially important in a range of existing and new applications such as transparent contacts and heterojunction partners for solar cells. There is also considerable basic science interest in the nature of the electronic structure and doping in p-type oxides. The key process necessary to achieve p-type TCOs is that of being able to dope them efficiently. To become truly useful, their resistivities must be in the range of 10 Ω/sq with a visible transparency above 80% for PV applications. Oxides by their nature tend to have oxygen vacancies and interstitial metal atoms which act as n-type dopants. To get good p-type materials requires compensation of these intrinsic dopants and the introduction of new intentional dopants. Recent reports in the literature indicate a variety of possible mechanisms for this process from co-doping to direct substitution. In the ZnO system, nitrogen or nitrogen oxide species have been reported to be p-type dopants [6], [7], [8]. There is considerable controversy as to the nature of the mechanism of nitrogen incorporation and of the relative importance of the presence of Al or Ga to the doping process [9]. The combinatorial approach whereby you can create a library of ZnO with a variable Al or Ga profile is ideal to begin to answer questions with regard to the nature of the p-type doping.

The optimization of thin film Si materials and devices for photovoltaic and related applications has become an economically important and a significant technical challenge in recent years [12], [13], [14], [15], [16], [17]. Although most laboratories can grow reasonably high quality thin films of thick undoped and doped hydrogenated amorphous silicon (a-Si:H) by a variety of robust techniques, the ability to produce high-efficiency devices does not necessarily follow. Industrial and research solar cells utilize thin layers of ternary (Si, H, B and Si, Ge, H) and quaternary (Si, C, H, B) materials that are grown very close to a structural phase transition from amorphous to microcrystalline Si material (μc-Si). Growth of these desirable phase-boundary materials occurs in a region with a highly non-linear dependence on the multi-parameter space of the deposition equipment for such parameters as gas mixtures, pressures and flows. Small changes in parameter space can produce gross alterations in the sample structure and resultant electronic and optical properties. The best a-Si:H solar cells are no longer simply p–i–n or n–i–p sandwiches of an intrinsic collector layer between two doped layers. Increases in the cell stable efficiency records from 6 to 13% over the last 15 years has involved the optimization of complex interface layers which are very sensitive to deposition parameters and differ from lab-to-lab, especially at the crucial p–i interface. A few general principles are well established but the details of why certain schemes are more successful than others is as yet unknown. The optimization of solar cell structures that are based upon compositionally and phase-graded interface layers has required years of slow, empirical experimentation and is far from complete. Thus, the optimization of amorphous Si solar cells to reduce cost and improve efficiency is an excellent candidate for a combinatorial approach to materials and device development.

We will report on recent results on the application of the combinatorial approaches to both the TCOs and to amorphous Si solar cells. These results demonstrate the utility of this approach as applied to solid state electronic materials.