Fluorocarbon plasmas

Figure C2.13.7. Change between polymerizing and etching conditions in a fluorocarbon plasma as detennined by tire fluorine-to-carbon ratio of chemically reactive species and tire bias voltage applied to tire substrate surface [36].

Figure 10, Schematic of the influence of fluorine to carbon ratio and electrode bias on etching versus polymerization processes in fluorocarbon plasmas, (Reproduced with permission from Ref, 40 J...

A brief description of the key plasma diagnostic techniques, which have been especially useful in delineating the gas phase processes in fluorocarbon plasmas, will be given followed by an extensive discussion of plasma etching and polymerization mechanisms. 

The etch resistance of poly (butene-1 sulfone) in fluorocarbon-based plasmas can be enhanced by prior treatment of the surface in an oxygen plasma. This pretreatment inhibits or retards the depolymerization reaction that characterizes normal etching in fluorocarbon plasmas, thereby permitting formation of a surface-modified layer which exhibits a substantially reduced etch rate. Pretreating PBS in an oxygen plasma enables it to be used subsequently in selective reactive-ion etch processes involving fluorocarbon plasmas to delineate submicron, anisotropically etched patterns. 

Figure 1. PBS etch rate versus inverse temperature in fluorocarbon plasma. Curve a, Ref. 3 curve b, Ref. 7.

Figure 2. Auger spectra of PBS surface post (a) no plasma exposure (b) 30 second and (c) 90 second fluorocarbon plasma exposure.

PBS/SUBSTRATE ETCHING IN FLUOROCARBON PLASMAS WITH PRIOR OXYGEN EXPOSURE... 

Figure 8 shows a plot of thickness of PBS removed versus time in both the CF4/O2 and CF4/He/02 plasmas for samples priorly exposed to an oxygen plasma (lOOW, 0.5 Torr, 3 minutes 16X). The etching curves in the fluorocarbon plasma are characterized by two distinct regions. Initially, the etch rate of PBS is quite high being comparable to that of samples not subjected to pretreatment in O2 plasma (cf. Figure 1). The etch rate then quickly diminishes to a low constant value of 12 2A/min (for CF4/He/02 and 29 5A/min in CF4/O2. When the linear removal rate, obtained from a least-squares plot of the thickness removed versus plasma exposure time, is plotted as an Arrhenius expression at different temperatures (Figure 9), an activation energy of zero is obtained.

Figure 8. PBS thickness removed vs time in fluorocarbon plasma post 3 minute oxygen plasma exposure.

Table 1 summarizes the plasma etching conditions which enable PBS to be used as an etch mask in a fluorocarbon plasma environment. It should be noted that the conditions cited result in isotropic etching of the substrate. In order to maintain good resolution (<1.0Mm) without excessive linewidth loss, the substrate must be relatively thin. 

Figure 10. PBS features on tantalum substrate post 3 minute oxygen plasma and 90 second fluorocarbon plasma exposure.

Under normal conditions, etching of PBS in fluorocarbon plasmas is dominated by depolymerization which results from scission of the main polymer chain caused by the activated species in the plasma. 

Depolymerization of PBS in fluorocarbon plasmas results in system-dependent etching rates which are difficult to reproduce. This is due to plasma-initiated polymerization of the gaseous degradation products which competes with the degradation reaction. 

Pretreatment of PBS films in an oxygen plasma allows the creation of a chemically altered surface during subsequent fluorocarbon plasma treatment which provides enhanced etching resistance. 

Fluorocarbon plasmas produce two different kinds of reactive species in the glow, namely, CF and F atoms. The density of CF ... 

---