T. Comyn, 'The Reactive Ion Etching of Transition Metals via the Formation of Metal-Organic Species', MSc Thesis, 1994.

The work presented here explores the use of organic process gases in the reactive ion etching of Au, Cr, Cu, NiFe and Pt via the creation of organometallic species, for uses in commercial applications. Presented also is a section of work using an existing computer simulation model, TRIM, which predicts the outcome of physical sputtering.

The practical work on chemical etching had a limited degree of success. Cr was etched using a Cl2/O2 mixture at a rate in excess of 700Å/min, and Cr, Pt and Au were all etched to some degree by a Cl2/CO mixture.

CO/H2, NO/H2, CO/NO/H2, HFAC/H2, ACAC/H2, DMHFOD/H2 and NO/Cl/He were all used, but were observed to etch material only via physical sputtering.

It was observed that HFAC caused Cr to blister heavily, as NO did to NiFe. This effect was most marked at low concentrations of HFAC and NO, when diluted in H2.

The model TRIM predicted that the metals tested were physically sputtered with ease in the order Au = Cu > Cr > Pt, and that the heavier ions (such as Cl+) remove surface ions far easier than lighter ions (such as H+). This was in agreement with the experimental work.

Optical emission spectroscopy was used in an effort to determine the identity of the molecular fragments in the plasma mixtures used. OH and CH+ were detected in a CO/H2 plasma, HNO, OH and NH were detected in a NO/H2 plasma, C2, CH2O and CF in a HFAC/H2 plasma. ClO was observed in a Cl2/O2 plasma. Optical emission spectroscopy also allowed a sensitive indication of N2 contamination.