Residual free reactive ion etching of the Bell contact Ti/Pt/Au

Abstract

The etching of the complete Bell contact consisting of a layer of Ti/Pt/Au was performed in highly reactive plasmas containing Cl 2 for Ti, PF 3 =NF 3 for Pt, and Cl 2 and/or BCl 3 for Au. All the constituents of the Bell contact form volatile compounds in either capacitively-coupled low-density plasmas or high-density plasmas generated by electron cyclotron resonance. This is conditio sine qua non for surfaces and sidewalls which have to remain free of any residues. r

Gold etching for microfabrication

“…Another consequence of the low volatility of AuCl x species is the tendency for etched gold to redeposit elsewhere on the substrate and in the process chamber [17,18]. This is highly undesirable as it leads to contamination of the wafer and necessitates frequent chamber cleaning [18].…”

“…This is highly undesirable as it leads to contamination of the wafer and necessitates frequent chamber cleaning [18]. A related phenomenon is that of low volatility etch products redepositing on the sidewalls of the etch mask during the etch process [16,17]. This can arise from ion-induced sputtering of the nonvolatile etch product that forms on the surface and also by direct condensation of etch by-products.…”

“…This can arise from ion-induced sputtering of the nonvolatile etch product that forms on the surface and also by direct condensation of etch by-products. The resulting deposits are known as 'fences' or 'veils' and are often difficult to remove in subsequent processing steps [17,19]. This problem is particularly severe when a thick photoresist mask is used as a mask.…”

“…Franz et al [17] reported the reactive ion etching of a Ti/Pt/ Au ohmic contact using a high-density plasma etcher and a variety of etch chemistries. The top Au layer was etched using a thin Al 2 O 3 hard mask or a thick photoresist.…”

Transport studies in quasi one-dimensional III-V wires, tubes, and fin structures

“…Although this step is identical to the gate fabrication in the conventional CMOS process framework with a lithography and etch sequence, many challenges still remains because of Figure 6.5a, no visible etching can be clearly observed. However, residues forming dark droplets on the SEM pictures are accumulated around the patterned structure and were suspected to be the by-products formed during the reaction between BCl 3 and Ti 285 or InAs at this low temperature. Indeed, it has been shown that the dry etching of In-based material with a Chlorine based plasma require the heating of the sample at high temperature (~200°C depending on the plasma conditions) to make InCl x products volatile at the pressure level.…”