Exploiting Anisotropy for In Situ Measurement of Silicon Etch Rates in KOH Solution

Abstract

Anisotropic etching of V-grooves in a masked substrate provides the basis for two simple methods for in situ measurement of etch rates of Si͑100͒. The width of the ͑100͒ facet defining the base of the groove and thus its surface area decreases at a rate which is determined by the etch rate in the ͑100͒ direction. By measuring voltammograms at regular intervals during etching we were able to monitor the change in geometry of the groove. In a complementary approach, in situ optical microscopy was used for determining etch rates in real time. An application of the method is described. Anisotropic etching of silicon is an essential step in device fabrication.1,2 Microelectromechanical systems ͑MEMS͒ and optoelectronics rely on the characteristic forms that result from the differences in etch rate of different crystallographic orientations. [3][4][5] It is therefore essential to be able to measure etch rates accurately and to detect any changes that may occur in these rates during a fabrication process. In addition, it may be necessary to develop a new etchant or to adapt an existing one, e.g., to determine the influence of an additive in a wide concentration range or to investigate the effect of temperature or of a shift in potential of the sample on the dissolution rates.Methods based on the etching of wagon-wheel patterns 6-8 or of semiconductor ͑hemi͒spheres 9-11 have the advantage of yielding dissolution rates for a whole range of orientations in a single experiment. These methods have a number of disadvantages: they are not simple, analysis of the results is cumbersome and, most importantly, the etching conditions generally differ from those used for device fabrication, e.g., sample geometry, hydrodynamics, gas evolution and galvanic interaction between facets may be important in determining etch rate and morphology. 12,13 At the other extreme, one can measure etch rates simply by masking a wafer of a given orientation and, after etching, determining the etched depth with a profilometer. In addition, examination of the cleaved sample with optical microscopy or scanning electron microscopy ͑SEM͒ can give information about etch rates of other orientations ͑via the underetching 14 ͒. Because each etch rate requires a number of separate experiments this approach is tedious, especially if one wishes to screen a wide range of experimental conditions. Glembocki and coworkers adapted this approach by using strips of silicon masked with an oxide layer containing square windows.15 By retracting the sample stepwise from the solution they were able to vary the time that windows were exposed to the etchant and were therefore able to measure a range of etch rates ͑in their case as a function of potential͒ in a single experiment. A disadvantage of this approach is that etching close to the solution/air boundary is important. Conditions near this boundary are obviously not the same as in a typical etching experiment, in particular because gas is being evolved. In addition, quenching of the etching reaction above the solution a...

In-situ infrared study of silicon in KOH electrolyte: Surface hydrogenation and hydrogen penetration

“…When the potential is scanned positive of OCP, oxidation by surface holes gives rise to an anodic current peak followed by surface passivation [11][12][13]. In the range negative of OCP, the cathodic current is low because of the absence of electrons in the conduction band in the dark, and the chemical etch rate is close to that measured at OCP [13,14]. For the case of n-Si, at potentials negative of OCP, a cathodic current results from hydrogen evolution due to conduction-band electrons, and the etch rate is reduced [11][12][13].…”

“…There is evidence that electron injection also occurs to some extent with p-Si [12]. In both cases, chemical etching is suppressed as soon as anodic passivation occurs [4,12,14].…”

Electrochemical Growth of Micrometer-Thick Oxide on SiC in Acidic Fluoride Solution

“…A simple electrochemical method can be used to detect the point at which Si is completely removed. 44 This is necessary to prevent dissolution of the relatively reactive oxide layer. Alternatively, the work of Rotter et el [38][39][40] on GaN and AlGaN suggests that (photo)anodic oxidation might be an interesting approach to making dielectric oxide on SiC (possibly with a low density of interface states) for device applications.…”

Crystallographic Orientation Dependent Reactive Ion Etching in Single Crystal Diamond

“…Taking advantage of the anisotropic nature of monolithic materials, crystal direction dependent wet etching techniques have been developed and are most well-known for etching Si in KOH. In a KOH solution, the kinetics of chemical reactions vary on Si {100}, {110}, and {111} planes, leading to a crystallographic dependent etch [3][4][5][6][7][8][9] . This wet etch recipe can process large amount of samples in parallel and is one of the most important assets in modern MEMS technology [10] .…”

“…[3][4][5][6][7][8][9] This wet etch recipe can process large amounts of samples in parallel and is one of the most important assets in modern microelectromechanical systems (MEMS) technology. Taking advantage of the anisotropic nature of monolithic materials, crystal direction dependent wet etching techniques have been developed and are most well-known for etching Si in KOH.…”