MINERALOGICAL MAGAZINE, MARCH 198o, VOL. 43, PP. 633-7 The effects of surface damage and oxide films on the optical properties of cuprite P. RASTALL, K. Mc. CLARKE, AND E. F. I. ROBERTS Department of Metallurgy and Materials, City of London Polytechnic, Whitechapel High Streetl London El 7PF SUMMARY. Spectral ellipsometry in the range 240 nm to 540 nm has been used to investigate the influence of surface damage induced by mechanical polishing and surface films of CuO on the optical properties of pure synthetic cuprite. Comparison is made between bulk CuO and thin film CuO produced by low- temperature oxidation of cuprite. The effects of strain and disorder are discussed in relation to the suppression of excitonic transitions. Recently developed techniques have been used to simplifythe analysis and take full advantage of the spectral data. IT has long been known that structural defects in crystal lattices, whether point, line, or plane, have a marked effect on absorption phenomena and elec- tronic transitions. Semi-conductors are particu- larly sensitive as their electrical properties are strongly affected by those defects. Cuprous oxide is a typical 'deficit' semi-conductor with a band gap of approximately 2.4 eV (Roberts and Rastall, I978) and much investigation has been carried out into the excitonic transitions in its absorption spectrum. Excitons are electron hole pairs in orbit pro- duced by photon absorption in various parts of the spectrum, and their nature and properties have been discussed at length (e.g. Gross, I962). Various workers have studied the dependance of the excitonic structure on lattice perfection. Agekyan et al. (I975) have shown that the method of pro- duction of Cu20 crystals strongly affects the sharp- ness of the exciton peaks, hydrothermally grown crystals exhibiting sharper transitions than less perfect material prepared from Cu metal by oxida- tion. The defect structure and stoichiometry of Cu20 is markedly affected by annealing, and Spyrideles et al. (1977) have indicated that vacancy aggregation in bulk samples, controlled by oxygen partial pressure in the annealing process, is respon- sible for suppression or enhancement of the exciton peaks. Work carried out by Hayfield (i96 I) on thin- film Cu20 obtained by low-temperature oxidation of copper again suggests a relationship between lattice defects and optical absorption. ~) Copyright the Mineralogical Society Experiments involving reflected light such as the ellipsometric measurements in this study are not only sensitive to bulk structural defects but are also affected by the state of the crystal surface. Surface conditions which affect the ellipsometric para- meters are: 0) Surface roughness: degree of polish, presence of etch pits, grain boundaries, and scratches, etc. (2) Polish damage: disordering, in- creased density of the surface layers. (3) Surface film presence: incidental atmospheric contamina- tion or controlled oxidation. The purpose of this present study is to evaluate the effects of (2) and (3), above, on the optical properties of synthetic cuprite, and to investigate anomalously high values of the absorption index obtained in our previous study (Roberts and Rastall, I978), which we suggested might be due to the presence of a thin absorbing surface film of CuO. Surface condition and etching procedure. Speci- mens of pure, polycrystalline cuprite were prepared by a similar method to that described previously. After oxidation, however, the samples were not cooled slowly in a partial vacuum, but were given a separate annealing treatment of 70o ~ and IO-~ tort for 24 hours. Spectral scans were carried out on an automatic following ellipsometer, the details of which are described by Roberts and Meadows (I964). The system has since been modified by the replacement of silica and glass Faraday-cell cores by water cores to reduce systematic errors and facilitate experi- mental procedure (Hunt, 1979). To remove the polish-damaged layer before further measurement, samples were etched by im- mersing them in a mixture of concentrated nitric (2o ~), phosphoric (4~ %), and acetic (40 %) acids (v/v) for 2o seconds, at room temperature. All the specimens were washed in soap solution, ultra- sonically cleaned in distilled water, and dried in pure ethanol and warm air. Oxidation of Cu20. Oxidation studies, both batch and in situ, were carried out on polished and etched