Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
RADIATIVE PROCESSES AND DIAGNOSTICS 27 Information Resources The database for surface reactions relevant to plasma processing of silicon is also highly dispersed. Useful information is included in several well-known texts,33 refereed journals, symposia proceedings, trade journals, and course notes. The complete characterization of a modem process is often proprietary to the manufacturer or to the equipment vendor. Many surface reactions have been studied using clean surfaces under ultrahigh-vacuum (UHV) conditions, and the different types of reactions and possibilities are reasonably well characterized. The difficulty arises when the surface is not that of a freshly cleaned crystal. The coefficients of the present database are not expected to be applicable to the surfaces found in low-pressure plasma reactors. Further, these surfaces are usually exposed simultaneously to fluxes of radicals, ions, and photons. Research is therefore required to bridge the gap between the clean surface science techniques that lead to fundamental understanding, and the etching/deposition/plasma- induced synergisms and the effective etching/deposition rates and selectivities that are important to the process engineer. It is by now well documented34 that infrared spectroscopy is one of the most powerful and broadly applicable of surface-sensitive diagnostic techniques. A key factor in its emergence has been the development of generations of capable, reliable, and relatively inexpensive FTIR spectrometers. Experimental configurations include reflection-absorption at grazing or oblique angles, transmission (through a thin film at the Brewster angle), emission, and multiple internal reflection using specially prepared substrates. Some of these techniques are suitable not only for laboratory experiments but also for monitoring and process control applications. In addition to identifying chemical species and their environments, infrared spectroscopy can be the method of choice for measuring surface temperature. Although many features in surface spectra are readily identifiable based on gas phase band positions, a great deal of the information contained in these spectra is in the properties of these bands (such as details of their shape), or in new bands, which result from environments specific to the surface. Unfortunately, there is a lack of controlled laboratory experiments to conclude that a band assignment is correct beyond all doubt. However, most of the experimental and theoretical techniques exist to create the needed database for infrared surface spectroscopy. For example, isotopic substitution can often supply key evidence for both molecular identities and orientations. Coupling of infrared studies with other surface spectroscopies that do not have in situ potential but that can be applied in UHV systems has already formed the basis for some very fruitful studies, but much remains to be done. A particularly important avenue of investigation, now that atomic force microscopy is becoming a widely available technique, is the correlation of surface roughness with the results of optical techniques such as reflection-absorption. New Diagnostic Techniques Other new methods that are assisting the understanding of surface processes include the following: 1. Imaging of radicals interacting with surfaces (IRIS) combines a molecular beam source of radicals in a vacuum chamber with laser-induced fluorescence detection of both incident and reflected species.35 This technique is described in Chapter 4, "Heterogeneous Processes." 2. Photoluminescence can be used to monitor process-induced damage. Calibration is needed to allow for the effects of temperature, surface condition, and other variables. 3. There is widespread agreement that future ultralarge-scale integration (ULSI) and other applications such as diamond films will be strongly influenced by initial surface properties and nucleation phenomena on surfaces. Recent studies have used not only transmission electron microscopy (TEM), scanning electron microscopy (SEM), and atomic force microscopy (AFM), but also synchrotron radiation x-ray photoelectron spectroscopy to determine actual surface bonding and oxidation mechanisms. Calibration of the signals from these surfaces and from the much more complex photoresist surface is a formidable
