Langmuir Probe Measurements on  CHF 3 and  CF 4 Plasmas: The Role of Ions in the Reactive Sputter Etching of SiO2 and Si

Abstract

Langmuir probe measurements on CHFs, CFt, CF4-H~, and CF4-O2 plasmas with concurrent etch rate measurements on SiO2 and Si have been made in a reactive sputter etching system= At an rf power of 300W and a pressure of 15 mTorr, the ion density is typically 1.5 • 101~ cm -~, the electron temperature 2 eV, the plasma potential 30 eV, and the ion energy (i.e., the sheath potential) of the order of 500 eV. The dependence of these parameters on rf power, pressure, reactor geometry, and plasma gas composition is investigated, as well as the spatial uniformity of the plasma. For SiO~, etch rates vary essentially in proportion to ion densities with all the different plasmas. For Si in CF4-O2 plasmas, the anisotropic part of the etch rate also parallels the ion density. From the measured ion densities the ion flux to the substrate is estimated. Thus, etch rates per ion can be determined, which turn out to be similar to those in reactive ion beam etching, although in reactive sputter etching ion energies are much smaller.ABSTRACT A laser interferometer system to monitor plasma etch rate, and to control etched depth of isolation areas in silicon for oxide isolated bipolar devices, is reported. The etching process is stopped as soon as the desired etched depth is achieved. The accuracy of the system is 4-6%, depending on the masking material. It is found that this method of laser interferometry can be particularly useful in process development for observations in real time of changes in etch rate.Laser interferometers have been used in integrated circuit processing to monitor the etch rates of dielec-* Electrochemical Society Active Member.

Reactive ion etching (CF4/Ar) and ion beam etching of various glasses for diffractive optical element fabrication

“…As previously discussed, reactive ion etching of a wide range of materials occurs by a process of either physical sputtering or ion‐enhanced chemical etching. In both cases, the etch rate is proportional to the square root of the ion energy, which was shown as directly proportional to the DC bias voltage − V b . Thus, etch rates proportional to the square root of − V b are expected, as long as sufficient densities of reactive gas molecules are available (Figure ).…”

“…In both cases, the etch rate is proportional to the square root of the ion energy, which was shown as directly proportional to the DC bias voltage −V b . 24 Thus, etch rates proportional to the square root of −V b are expected, as long as sufficient densities of reactive gas molecules are available (Figure 4). Figure 5 represents the etch rates in a CF 4 plasma (8 Pa, 50 sccm) for different glass samples as function of the plasma power.…”

Plasma Textured Glass Surface Morphologies for Amorphous Silicon Thin Film Solar Cells-A review

“…Konagai et al mostly employed CF 4 gas for the random texturing of glass surface morphologies. Various metal masks like Al, SiN3, and PR are used for etching purposes according to the applications [48][49][50]. Table 1 illustrates the short historical survey about the textured glass surface morphologies used for solar cells.…”

“…The textured soda-lime glass surface morphology changed from a flat surface to one with large craters and lateral feature sizes between 4,000 nm and 6,000 nm, as well as having rms roughness increase from 0.6 nm to 360 nm with the increase of gas pressure from 0 Pa to 13 Pa. The different surface textures at various gas pressures may be caused by the variation of gas radical concentrations like ion-radical and neutral etchant species in the dry plasma [7,50,51]. Figure 4 shows the textured glass surface morphologies for various RIE etching times.…”

Dormant Cathode Erosion in a Multiple-Cathode Gridded Ion Thruster

“…In this equation, is a dimensionless current correction developed by Laframboise [24] that depends on probe size, plasma number density, and temperature. For the temperatures and number densities obtained in most ion thruster plasma, Steinbrüchel [28] suggested that is given to within 3% error by Eq. (3):…”