AVS 53rd International Symposium
    Thin Film Wednesday Sessions
       Session TF-WeA

Paper TF-WeA3
Properties of Grain Boundaries in Cu(In,Ga)Se2

Wednesday, November 15, 2006, 2:40 pm, Room 2022

Session: Thin Films for Energy Applications in Photovoltaics, Fuel Cells, Hydrogen Storage & Batteries
Presenter: D. Hebert, University of Illinois, Urbana-Champaign
Authors: D. Hebert, University of Illinois, Urbana-Champaign
A. Hall, University of Illinois, Urbana-Champaign
A. Aquino, University of Illinois, Urbana-Champaign
C.H. Lei, University of Illinois, Urbana-Champaign
I.M. Robertson, University of Illinois, Urbana-Champaign
A. Rockett, University of Illinois, Urbana-Champaign
Correspondent: Click to Email
CuInSe@sub 2@ (CIS) and related materials yield the highest performance thin film solar cells and show excellent promise for very high efficiency multijunction devices if adequate single junction devices can be produced. One of the primary questions concerning the nature of these devices is the optoelectronic behavior of grain boundaries in the material. We have grown epitaxial bicrystals of these materials on polished GaAs bicrystalline substrates and have studied the grain boundaries by a variety of methods including atomic force microscopy, Kelvin probe force microscopy, position-dependent photoluminescence, cathodoluminescence, transmission electron microscopy, Hall effect and other methods. The bicrystals show distinct differences in the two grains, one of which has a nearly (100) surface orientation while the other has a nearly (110) orientation. The latter is highly faceted. The grain boundary is fully dense. The (110) grains show a lower energy luminescence characteristic of subgap emission from deep defect states, while the (100) grain show higher energy luminescence. No distinct behavior directly connected with the grain boundary is evident in any of the measurements. Furthermore, there is no measurable chemical change in the CIS chemistry at the grain boundary that might account for surface state passivation. Although this negative result is surprising, it is consistent with the observation that grain boundaries are inactive in the devices. This confirms the evidence that grain boundaries are surprisingly innocuous in this material.