Larry C. Olsen

High efficiency monochromatic GaAs solar cells

Efforts to develop GaAs solar cells for coupling to laser beams in the wavelength range of 800 to 840 nm are described. The work has been motivated primarily by interest in space-to-space power beaming applications. Modeling calculations of GaAs cell performance have been carried out using PC-1D to determine an appropriate design for p/n and n/p cell structures. Epitaxial layers were grown by MOCVD and cells were fabricated. Measurements under laser beam and simulated conditions yield an efficiency of 52.1% for a cell coupled to 806 nm light at 100 mW/cm/sup 2/.<<ETX>>

CIS solar cells with ZnO buffer layers

This paper describes investigations of CIS solar cells with ZnO buffer layers. Studies concentrate on determining optimum ZnO buffer layer properties that will yield the maximum ZnO/CIS solar cell efficiency. Buffer layers are grown by chemical vapor deposition (CVD) by reacting a zinc adduct with tetrahydrofuran to grow ZnO. Substrate temperatures (T/sub sub/) have been varied from 100/spl deg/C to 350/spl deg/C, with the best device performance obtained for T/sub sub//spl ap/225/spl deg/C to 250/spl deg/C. ZnO/CIS solar cells were fabricated by first depositing a ZnO buffer layer, followed by deposition of a low resistivity ZnO top contact layer and a Al/Ag collector grid. Several cells have been fabricated with an area of 0.44 cm/sup 2/ that have total area efficiencies greater than 11%. In particular, a cell was fabricated with a Siemens substrate that exhibited a total area efficiency of 11.3%, and one based on a NREL substrate that has a total area efficiency of 12%. This work has focused on determining optimum deposition parameters for CVD ZnO buffer layers. It is clear that in order to achieve efficient ZnO/CIS cells, the buffer layer must be deposited with a high resistivity.

Model calculations for silicon inversion layer solar cells

Abstract Theoretical models for the current versus voltage characteristics of inversion layer (IL) solar cells are developed and used in modelling calculations. Two-dimensional carrier transport from the base region to the inversion layer and collector grid is taken into account using two approaches: by modelling the cell as a series of infinitesimally thin slices which supply current to the inversion layer and by solution of the two-dimensional diffusion equation for minority carriers. The latter approach, although more involved, is found to be necessary in order to achieve accurate modelling in many cases of interest. Significant device parameters are identified and model calculations are carried out over a range covering most physically realizable values for these parameters for p-type silicon-based devices. Conditions necessary for good device performance are thus identified and the results of poor device parameters are discussed. It is found that grid line densities of more than 20 lines cm−1 will be necessary to achieve a reasonable fraction of the ultimate efficiency under the best conditions. However, good device performance can be expected under a much wider variety of conditions with grid line densities of 40 lines cm−1 and above. For the higher grid line densities it is shown that the device performance drops rapidly as the insulator charge density decreases below a critical value. Finally, modelling calculations of IL solar cells carried out by others are discussed and compared with those presented here.

CIGSS solar cells based on CVD ZnO buffer layers

This paper describes investigations of cells based on Siemens Solar material, CIGSS, and highly resistive ZnO (i-ZnO) buffer layers grown by MOCVD. Resistive i-ZnO buffer layers are grown on CIGSS at 100/spl deg/C, after heating the substrate to 250/spl deg/C and using nitrogen as a carrier gas. The use of a KCN etch on the CIGSS surface prior to growth of i-ZnO buffer layers was determined to be beneficial to cell performance. XPS studies show that the etching step removes oxygen from the substrate surface that had complexed with Se to form SeO/sub 2/. Cells that do not receive a KCN etch typically have shunted I-V curves leading to relatively low open circuit voltages, fill factors and efficiencies. A tentative model is proposed for the effect of KCN. Finally, one cell (using a KCN etch) exhibited a total area efficiency of 12.7 % with an open circuit voltage of 0.577 Volts.

High efficiency CIGS and CIS cells with CVD ZnO buffer layers

This paper describes investigations of CIS and CIGS solar cells with ZnO buffer layers. These studies are a result of a team effort between investigators at Washington State University (WSU), the Institute Of Energy Conversion (IEC) and the National Renewable Energy Laboratory (NREL). Cells with ZnO buffer layers were fabricated with both Siemens CIS and NREL CIGS substrates. An active area efficiency of 13.95% was achieved for a ZnO/CIGS cell. ZnO buffer layers are grown by reacting a zinc adduct with tetrahydrofuran using a two-step approach: growth of approximately 100 /spl Aring/ of ZnO at 250/spl deg/C; and then growth of 500 to 700 /spl Aring/ of ZnO at 100/spl deg/C. The high temperature step is necessary to achieve good cell performance. It appears that exposure of CIGS to hydrogen at 250/spl deg/C may remove contaminants and/or passivate recombination centers on the surface and subsurface regions.

PC-1D modeling of depletion layer recombination in GaAs solar cells

Theoretical studies of current-loss mechanisms in GaAs solar cells using PC-1D are presented. The contribution to current-voltage characteristics of depletion-layer recombination via traps at various locations in the forbidden gap has been modeled using PC-1D. The results of these studies are used to interpret experimental data for high-efficiency GaAs solar cells. It is found that the double-mechanism characteristic observed for high-efficiency GaAs cells can usually be interpreted in terms of two mechanisms acting in parallel, one due to recombination via a midgap trap with n=2 and another at higher voltages due to emitter/base recombination with n=1, or a component characterized by n on the order of 1.1 to 1.5 due to recombination via a trap located between midgap and the conduction or valence band edge.<<ETX>>

High voltage ZnSe/CuInSe2 solar cells

This paper describes investigations of CIS solar cells based on ZnSe window layers deposited by MOCVD. Investigations of ZnSe/CIS solar cells are being carried out to determine if ZnSe is a viable alternative to CdS as a window material. MOCVD growth of ZnSe is accomplished in a SPIRE 500XT reactor and conductive n‐type ZnSe is grown by using iodine as a dopant. ZnSe films have been grown on CIS substrates at 200 °C to 250 °C. ZnO is also being deposited by MOCVD by reacting tetrahydrofuran (THF) with a zinc adduct. ZnSe/CIS heterojunctions have been studied by growing n‐ZnSe films onto 2 cm × 2 cm CIS substrates diced from materials supplied by Siemens. Test cells are fabricated by depositing thin, transparent Al contacts 2.8 mm in diameter on top of the ZnSe to serve as contacts. These test devices typically exhibit open circuit voltages ≳500 mV and estimated active area efficiencies ≳13%. Efforts to deposit ZnO onto ZnSe/CIS structures for top contact layers have usually resulted in photocurrent suppre...

Challenge of replacing CdS in CuInSe2-based solar cells

This paper discusses some key issues concerning the replacement of CdS buffer layers in CIS solar cell structures, and describes investigations of alternative buffer layers deposited by MOCVD. One apparently unique property of CdS buffer layers grown by CBD is that a ZnO TCO can be deposited on top of a CdS/CIS structure without significantly degrading the photovoltaic properties of the CdS-CIS junction. Investigation of alternative buffer materials such as high resistance ZnO (i-ZnO), ZnSe and InSe have first identified MOCVD growth procedures that yield Al/X/CIS test structures (X=i-ZnO, ZnSe and InSe) with good properties, and then addressed the challenge of fabricating efficient, complete cells with conductive ZnO top contact layers. These studies have been conducted with Siemens CIS and CIGSS substrates, and with NREL CIGS substrates. A total area efficiency of 12.7% and estimated active area efficiency of 13.4% is reported for a CIGS cell with an i-ZnO buffer layer grown by MOCVD.

Efficient direct ZnO/CIS solar cells

This paper describes investigations of CIS solar cells with ZnO window layers deposited by MOCVD. These studies have been conducted with graded absorber CIS substrates obtained from Siemens Solar. Cell fabrication involves surface preparation of the Siemens substrate, growth of 200 to 400 A of undoped ZnO by MOCVD, deposition of a highly conducting ZnO top contact layer and deposition of a Ni/Ag collector grid. MOCVD growth of ZnO is accomplished in a SPIRE 500XT reactor by reacting a zinc adduct and tetrahydrofuran. Processing development has been conducted by forming test cells on ZnO/CIS structures by depositing thin, transparent Al contacts 2.8 mm in diameter on top of the ZnO window layer to serve as contacts. Several cells have been completed with a total area efficiency ≥11.0%, with the best result being 11.3%. The best active area efficiency is approximately 12%. Other topics discussed include current‐voltage characteristics of direct ZnO/CIS cells.

Alternative buffer layers for CuIn(Ga)Se2 solar cells

Although 12% to 14% cells have been fabricated with highly resistive ZnO buffer layers grown by MOCVD, an improved understanding of the required processing to achieve high efficiency is still required. For example, it is found that it is beneficial to “age” i-ZnO/CIS cell structures in air for several weeks before completing the cell with a TCO and collector grid. SIMS depth concentration profiles have been acquired for i-ZnO/CIS film structures grown on polycrystalline CIS and also for epitaxial CIS grown on GaAs. These profiles clearly establish that Zn and oxygen diffuse along grain boundaries during ZnO growth, and probably in such a manner that the Zn concentration exceeds that of oxygen. It is proposed that the beneficial effect of air exposure in the aging process is due to additional oxygen diffusing along grain surfaces and combining with excess zinc such that recombination centers are passivated, and values of FF and Voc are increased.