Nitrogen related absorption bands of the IR spectra are used to determine the content of nitrogen and N—H bonds in a—Si56 Ge44 films. A preferential attachment of nitrogen to silicon is observed. Nitrogen incorporation enhances the normalized photoconductivity to 5 · 107 cm2/V at maximum. This is related to N—H complexes acting as donors. A correlation between optical gap and single oscillator energy is reported.

We have studied wet-chemical and dry etching properties of doped and undoped hydrogenated amorphous silicon (a-Si:H) films with bonded hydrogen content varying from 0 to 20 at.%. Etching processes studied include (1) wet-chemical etching using solutions of KOH, isopropyl alcohol (IPA), and H2O, (2) hydrogen plasma etching, and (3) XeF2 vapor etching.

The slope etching technology of hydrogenated amorphous silicon film (a-Si:H) has been very useful for its devices; thin film transistor, contact image sensor and other large area electronic components. To obtain a good step coverage slope etching was performed by RIE using CF4 gas and optimum amount of O2 gas after modification of photoresist(PR) patterns by thermal treatment. In this experiment, primary factors were

1) thermal treatment temperature and time of PR.

2) CF4 gas flow rate, etch rate of a-Si:H film and ash rate of PR in proportion to O2 gas flow rate.

As a result, the slope angle of 11° was obtained when CF4 gas flow rate was 46 SCCM, O2 gas flow rate was 7 SCCM and thermal treatment temperature and time was 140°C and 30 minutes respectively.

Capacitance vs. temperature, deep-level transient spectroscopy (DLTS), and transient photocapacitance spectroscopy have been used to investigate the amorphous-crystalline silicon interface region of a device made of hydrogenated amorphous silicon deposited on a lightly doped n-type crystalline silicon.

By comparing our results between substrates with and without oxide contamination with those in a earlier study, we have been able to correlate the effects of substrate preparation on the density of interface states.

The bulk and interface charge trapping phenomena of fluorinated oxides have been studied by various electronic measurements. Fluorine is introduced into dry oxides by low energy (25kev) implantation followed by a 1000°C N2 ambient anneal to remove physical damage. Both the flat band and the mid gap voltage shifts of such MOS devices are measured during avalanche electron injection. We have developed techniques to separate effects due to interface state generation from bulk trapping effects. The bulk electron traps in the fluorinated oxides have a different cross section from the known water-related traps in conventional oxides. The generation of fast and slow interface states for different dosages of fluorine implantation is discussed based on Q-V and C-V measurements. The fast interface donor states, generated during avalanche injection, are charged at flat band but discharged at mid gap and beyond. An optimum dosage of fluorine implantation is found to suppress the so called turn-around effect during avalanche injection due to the formation of slow donor states. Finally, injection under high temperature (120°C-150°C) anneals out most of these donor states.

In the optical reflectance spectrum of the random multilayers of a-Si:H/a-Si3N4+x:H, it is observed an anomalous peak which is explained by the classical localization of light propagation. The following two subjects are discussed in this report; (a). The Kramers-Kronig transformation is done including the anomalous disorder-related reflectance peak. Extra absorption coefficients Δα are obtained by this analysis. Energy dependence of the localization length 1(hv) of light propagation are obtained by 1(hv)=1/Δα(hv). (b). The other is the experiment on scaling where the ratio of the disorder and the average layer thickness is kept constant but the size of each layer is changed in each experiment

We discuss the absorption-length and film thickness dependence of photoconductivity in amorphous silicon using measurements of the ambipolar diffusion length and the Debye screening length.

Photoluminescence absorption spectroscopy or PLAS has been used to measure the below gap absorption in a-Si:H at 77 K. As the absorption probed by this technique is composed of contributions from the bulk and the surfaces of the a-Si:H, we have developed an improved sample structure to separate these two contributions. Two kinds of interface have been investigated: a-Si1−xNx:H/a-Si:H and a-Si1−xOx:H/a-si:H. For the studies ofa-Si1−xNx:H/a-Si:H.x, two samples have been employed. The first sample consisted of an a-Si1−xNx:H/a-Si:H/a-Si:H/a-Si1−xNx:H/NiCr layered structure; the second one had a similar structure except that the a-Si:H layer was interrupted periodically by two thin (100 A°) a-Si1−xNx:H layers to increase the contribution of the interface absorption. A shoulder of the absorption around 1.2 eV for the second sample, which was not found in the first one, is probably due to the interfaces between a-Si:H and a-Si1−xNx:H. All samples were light-soaked using an Ar+ laser (5145 A°). The increases in the absorption measured by PLAS at 77 K are essentially the same as the results of PDS at 300 K, which suggests that the interfaces do not contribute to the light-induced absorption.

Carrier transport in amorphous silicon pin-diodes has been analyzed by steady state photocarrier charge collection applying strongly absorbed light. For low intensities at room temperature electron charge collection is limited by recombination in the generation region. For increasing intensity ø and/or decreasing temperature charge collection becomes nonlinear in ø and shows S-like characteristics versus voltage.

We present a model for this behaviour, including space charge limitation which e.g. for holes in a-Si:H limits charge collection even at room temperature due to low extended state mobility of holes and dark Fermi level position above midgap.

Charge transport in a-Si:H films is studied by measuring the intensity and the rate of decay of the transient photocurrent with optical bias illumination. The optical bias enhancement of the transient photocurrent due to the probe beam decreases at first as the temperature is decreased, reaches a minimum at about 21OK, and then increases as the temperature is further decreased. This observation, together with the decay time measurements, indicates a change in the mechanism of charge propagation from conduction band transport at high temperature to hopping conduction in the low temperature. A theory of hopping conduction in tail states is developed to explain the experimental results. A quantitative comparison of the theoretical and experimental results is made to determine the localization length of the tail states.

The determination of the ambipolar diffusion length, L*, and the effective lifetime, τ*, in p/i and a-Si:H Schottky barriers (ITO/p/a-Si:H/Al-Si; Cr/a-Si:H/Cr/Ag) have been determined by Flying Spot Technique, FST. This technique consists in the transient analysis of the photocurrent/photopotential induced by a laser beam that moves perpendicularly to the structure with a constant motion ratio, at different velocities. Taking into account the competition between the diffusion/drift velocities of the excess carriers and the velocity of the flying spot, it is possible to solve the transport equations and to compute separately L* and τ*, from the asymmetrical distribution responses.

Basic consideration of parameters including electron de Broglie wavelength, resonant tunneling time (τr) and inelastic scattering tunneling (τi) were studied to identify the intrinsic tunneling properties of amorphous double—barrier (DB) structure. The order of magnitude of τr/τi. of amorphous DB structure is about 5 to 12 which is far beyond that of crystalline case which is 1 to 2. The transmissivity versus applied voltage or versus electron energy were calculated to identify the influence of barrier height, barrier thickness and electron effective mass. The influence of carrier distribution will also be discussed.