Plasma etching surfaces exposed

A wide spectrum of heterogeneous processes can occur at solid surfaces exposed to a glow discharge (5,25,27). The primary processes of interest in plasma etching (and in plasma deposition) are summarized in Table II i20). These interactions result from the bombardment of surfaces by particles. Although vacuum UV photons and soft x-rays present in the plasma are sufficiently energetic to break chemical bonds, electron and particularly ion bombardment has been found to be the most effective method of promoting surface reactions (25). 

In the ion induced damage mechanism energetic ions break crystal bonds on the film surface thereby making the film more accessible and more reactive to the active chemical etchant. However, the side walls remain relatively unperturbed, and etching proceeds at the nominal chemical etch rate. Consequently, material removal proceeds far more rapidly in the ion flux direction, resulting in anisotropy. In actuality, the surfaces exposed to the plasma are likely to be composed of a chemisorbed coating of etchant... 

The mechanism we believe is responsible for the large SiOj-to-Si etch-rate ratios which have been obtained in fluorine-deficient discharges is based on several experimental observations. First of all, it has been shown that there are several ways in which carbon can be deposited on surfaces exposed to CF, plasmas. One way is to subject the surface to bombardment with CF ions which are the dominant positive ionic species in a CF plasma. The extent to which this can occur is shown by the Auger spectra in Fig. 3.3. Curve (a) is the Auger spectrum of a clean silicon surface and curve (b) is the Auger spectrum of the same surface after bombardment with 500 eV CFj" ions. Note that the silicon peak at 92 eV is no longer visible after the CFj bombardment indicating the presence of at least two or three monolayers of carbon. Another way in which carbon can be deposited on surfaces is by dissociative chemisorption of CFj or other fluorocarbon radicals. 

The core of the nucleation model proposed in Refs. [362, 363] is an assumption based on the experimental data that epitaxially oriented nucleation sites are formed in the SiC layer of about 10-nm thickness during the bias treatment. These sites are exposed at the SiC surface, while plasma etching of SiC is occurring during both the BEN treatment and the successive diamond growth process. The model of nucleation process is schematically depicted in Figure 11.57 ... 

Figure 21. AES survey spectra from Al surfaces exposed to plasma etching (a) and not exposed to plasma etching (b). (Reproduced with permission. Copyright Perkin-Elmer Corporation.)...

---