The etch rates for 317 combinations of 16 materials (single-crystal silicon, doped, and undoped polysilicon, several types of silicon dioxide, stoichiometric and silicon-rich silicon nitride, aluminum, tungsten, titanium, Ti/W alloy, and two brands of positive photoresist) used in the fabrication of microelectromechanical systems and integrated circuits in 28 wet, plasma, and plasmaless-gas-phase etches (several HF solutions, H/sub 3/PO/sub 4/, HNO/sub 3/+H/sub 2/O+NH/sub 4/F, KOH, Type A aluminum etchant, H/sub 2/O+H/sub 2/O/sub 2/+HF, H/sub 2/O/sub 2/, piranha, acetone, HF vapor, XeF/sub 2/, and various combinations of SF/sub 6/, CF/sub 4/, CHF/sub 3/, Cl/sub 2/, O/sub 2/, N/sub 2/, and He in plasmas) were measured and are tabulated. Etch preparation, use, and chemical reactions (from the technical literature) are given. Sample preparation and MEMS applications are described for the materials.

/pdf/etch-rates-for-micromachining-processing-29z80ltijs.pdf

Etch rates for micromachining processing

Deep-leve transient spectroscopy (DLTS) of bulk traps and interface states in Si MOS diodes are theoretically studied and energy levels, capture cross-sections and spatial and energy density distributions of majority-carrier traps are measured. In P+-implanted unannealed MOS diodes, four bulk traps are measured at Ec-0.18 eV, Ec-0.20 eV, Ec-0.31 eV and Ec-0.45 eV. Their spatial distributions are found to be the same among them within experimental error and thought to be corresponding to the distribution of implanted ions qualitatively. Bulk traps are distinguished from interface states experimentally. The capture cross-section of interface states in non-implanted MOS diodes are measured to be of the order of 10-16 cm2 in the energy range of Ec-0.15 eV to Ec-0.30 eV. The interface state density measured with DLTS is found to be in a reasonable agreement with those detetmined by other methods.

https://iopscience.iop.org/article/10.1143/JJAP.18.113/pdf

Deep Level Transient Spectroscopy of Bulk Traps and Interface States in Si MOS Diodes

Optical interference fringe measurements of the thickness of transparent layers can be rapid, accurate, and nondestructive. If the refractive index n of the layer being measured is known, it may be combined directly with interference fringe information to yield the layer thickness t. If, however, n is unknown, the measurement procedure necessarily becomes more complicated. In this paper, a new and simpler optical interference method is presented for the approximate determination of both n and t of a transparent layer on a transparent substrate. The required experimental information is obtained from a single spectrophotometric recording of either the reflectance or transmittance of the layer and its substrate. The theory of the method is presented, and an application of the method to measurements of layers of SIPOS (Semi-Insulating POlycrystalline Silicon) is described. The method requires that the layer being measured must be uniformly deposited on a flat substrate, and it must neither absorb nor scatter the light passing through it. The major approximation inherent in the method is that both the layer and the substrate are assumed to be nondispersive over the wavelength region of interest.

Optical interference method for the approximate determination of refractive index and thickness of a transparent layer

Silicon nitride (Si(3)N(4)) thin film optical waveguides with propagation losses of less than 0.1 dB/cm for the TE(0) mode at lambda(0) = 6328 A have been successfully grown by low-pressure chemical vapor deposition. Silicon wafers 5 cm in diameter were used as substrates, and the Si(3)N(4) was separated from the substrate by a steamoxide SiO(2) buffer layer. Propagation losses are examined for the various waveguide modes, and their dependence on waveguide parameters and wavelength are discussed and compared with exact calculations. Leakage into the silicon substrate is shown to be a major loss mechanism, especially at longer wavelengths and for higher mode numbers.

Silicon nitride films on silicon for optical waveguides

Single- and multiple-nanopore membranes are both highly interesting for biosensing and separation processes, as well as their ability to mimic biological membranes. The density of pores, their shape, and their surface chemistry are the key factors that determine membrane transport and separation capabilities. Here, we report silicon nitride (SiN) membranes with fully controlled porosity, pore geometry, and pore surface chemistry. An ultrathin freestanding SiN platform is described with conical or double-conical nanopores of diameters as small as several nanometers, prepared by the track-etching technique. This technique allows the membrane porosity to be tuned from one to billions of pores per square centimeter. We demonstrate the separation capabilities of these membranes by discrimination of dye and protein molecules based on their charge and size. This separation process is based on an electrostatic mechanism and operates in physiological electrolyte conditions. As we have also shown, the separation capabilities can be tuned by chemically modifying the pore walls. Compared with typical membranes with cylindrical pores, the conical and double-conical pores reported here allow for higher fluxes, a critical advantage in separation applications. In addition, the conical pore shape results in a shorter effective length, which gives advantages for single biomolecule detection applications such as nanopore-based DNA analysis.

https://www.pnas.org/content/pnas/106/50/21039.full.pdf

Versatile ultrathin nanoporous silicon nitride membranes

The water content of phosphoric acid in etching silicon nitride and silicon dioxide plays an important role. An increase in water content increases the etch rate of silicon nitride and decreases the etch rate of silicon dioxide. The highest possible temperature for a fixed water content at atmospheric pressure in the system is realized by boiling the liquid and refluxing the vapor phase. Refluxed boiling phosphoric acid at 180°C was found to be a useful etchant for silicon nitride films. The etch rate is 100 Aa/min. Under the same conditions deposited silicon dioxide had an etch rate of 0–25 Aa/min depending on the method of preparation, and elemental silicon 3 Aa/min. Etch rates of silicon nitride, silicon dioxide, and silicon in refluxed boiling phosphoric acid were measured as a function of temperature (and concentration) in the range of 140°–200°C. All etch rates increased with temperature. The "apparent" activation energies are 12.7, 27.6, and 26.4 kcal/mole, respectively. The etch rate of silicon nitride in phosphoric acid of constant concentration was measured as a function of temperature only. In this case the "real" activation energy was 22.8 kcal/mole. The difference in etch rate between silicon nitride, deposited silicon dioxide, and silicon offers a technique for etching contact holes in silicon nitride using deposited silicon dioxide as a mask. Such a technique was used successfully in making transistors with silicon nitride over as a junction seal.

https://iopscience.iop.org/article/10.1149/1.2426757/pdf

The Etching of Silicon Nitride in Phosphoric Acid with Silicon Dioxide as a Mask

https://iopscience.iop.org/article/10.1149/1.2407927/pdf

Silicon Impurity Distribution as Revealed by Pulsed MOS C‐V Measurements

The thickness of SiO 2 and Si 3 N 4 films grown or deposited on silicon substrates is determined by an interference technique, using a spectrophotometer in the u.v. and visible wavelength range. The accuracy is estimated at ±50 A . The precision is ±0.3 percent. Empirically calibrated fringe charts are presented for SiO 2 and Si 3 N 4 avoiding errors due to dispersion and ‘net’ phase shift. The refractive index and the net phase shift for both dielectric materials is measured as a function of wavelength using the empirical calibration. The refractive index of the SiO 2 films was found to be about 2 percent higher than published values for fused silica. The refractive index of the Si 3 N 4 was 1.97 at λ = 546 mμ . It is shown that the thickness of double layers of nitride on oxide or oxide on nitride can be determined if the thickness of one layer is known. A color chart for Si 3 N 4 is presented.

Optical thickness measurement of SiO2Si3N4 films on silicon

A Laser Interferometer System to Monitor Dry Etching of Patterned Silicon

Most semiconductor devices today have a costly vacuum-tight encapsulation that provides a microenvironment for high reliability and electrical connections to the circuit in which it is used. A junction seal consisting of a metal-insulator-silicon (MIS) system of materials has been developed to replace the vacuum-tight encapsulation. The MIS junction seal, consisting of platinum silicide-titanium-platinum-gold contacts and a Silicon nitride overcoat, provides the necessary encapsulation for high reliability. Electrical and mechanical connections are provided by gold beam-deads. During fabrication, the contact windows are opened in the deposited silicon nitride layer either by etching with boiling phosphoric acid using SiO 2 as a mask or by anodically converting the silicon nitride in the windows to a soluble oxide. The multilayer contact is then applied to complete the junction seal. The initial characteristics of sealed-junction transistors fabricated by the above methods were similar to those of the unsealed transistors. The reliability of the sealed-junction transistors determined by accelerated aging after an intentional sodium contaminafion of 1017atoms/cm2surpassed that of the standard silicon planar transistors sealed in a vacuum-tight enclosure.

W. van Gelder

Papers

The Etching of Silicon Nitride in Phosphoric Acid with Silicon Dioxide as a Mask

Silicon Impurity Distribution as Revealed by Pulsed MOS C‐V Measurements

Optical thickness measurement of SiO2Si3N4 films on silicon

A Laser Interferometer System to Monitor Dry Etching of Patterned Silicon

A metal-insulator-silicon junction seal