David Redfield
Stanford University
Network
Latest external collaboration on country level. Dive into details by clicking on the dots.
Publication
Featured researches published by David Redfield.
Applied Physics Letters | 1989
David Redfield; Richard H. Bube
Generation of light‐induced metastable defects in amorphous Si:H(a‐Si:H) is shown to follow the same stretched exponential (SE) that describes relaxation of thermally induced metastability at room temperature for a simple case. Apparent power laws derived from the central part of the SE are (time)0.3 and (intensity)0.6, agreeing well with the (time)1/3 and (intensity)2/3 dependences often reported in the mid range of defect density, thus providing an alternative description of defect generation. The SE link between light‐induced and thermally induced instabilities suggests that the thermal effects are also due to defect processes, and offers an alternative defect‐based explanation to a macroscopic ‘‘structural relaxation’’ or ‘‘glass transition.’’
Applied Physics Letters | 1988
David Redfield
A unified explanation is derived relating ‘‘stable’’ and metastable deep‐level defects in amorphous Si:H, the existence of a universal saturated density of metastable defects, the temperature dependence of that saturation density, the increases with doping of both the density of ‘‘stable’’ defects and sensitivity to light, and the slow relaxation of dark properties. This follows from the defect energetics, and a completed analysis of the kinetics of formation and anneal of defects that is fully symmetric. It is concluded that the dominant defects are extrinsic in origin, probably involving a complex with a dopant atom.
Journal of Applied Physics | 1989
Richard H. Bube; David Redfield
Nine recombination‐driven mechanisms are possible in principle for the kinetics and steady‐state effects of illumination on defects in hydrogenated amorphous silicon. By comparing the different mechanisms, and making choices based on the experimental observation that the carrier lifetime varies inversely with the metastable defect (dangling bond) density, these nine are reduced to three. They correspond to recombination that generates metastable defects (1) taking place at metastable defects, (2) being associated with direct electron‐hole recombination, or (3) taking place at other specific defect sites in the material. Each of these three models is considered under the assumptions that (a) the rate constants involved are not functions of time, and (b) the rate constants involved are functions of time in a way similar to that observed for the decay of thermally induced defects, described by a stretched exponential formulation. Comparison with experimental data on kinetics as a function of time and light intensity indicates that both models 2a and 1b are capable of describing the reported results. If, however, it is maintained that all defect‐related kinetics in amorphous silicon should exhibit the dispersive behavior leading to the stretched exponential description, then only model 1b is acceptable. Experiments for distinguishing between models 2a and 1b are suggested.
Journal of Applied Physics | 1992
Richard H. Bube; Lisa E. Benatar; M. Grimbergen; David Redfield
The constant photoconductivity measurement (CPM) technique has been widely used to determine the density of defects in thin films of materials where only a small absorption coefficient exists for these defects, particularly in determining the dangling‐bond density in hydrogenated amorphous silicon. Interpretation and modeling in the literature assume that CPM gives an accurate value for the density of defects present. The limitations of this method are examined as a function of several defect parameters, and particularly as a function of the Fermi energy, in models consisting of one level, a Gaussian distribution of levels, two independent levels, and the levels typical of a multivalent defect, applying the results of the latter to actual data on optical degradation kinetics in amorphous silicon. It is concluded that a prerequisite for reliably relating CPM‐determined densities to actual defect densities is a knowledge of the relative location of the equilibrium Fermi level and the energy level of the defect.
Applied Physics Letters | 1990
Richard H. Bube; Lisa Echeverria; David Redfield
It is shown that a growing number of experimental results, including some that are very recent and detailed, provide a strong support for a stretched‐exponential description of optically induced degradation in hydrogenated amorphous silicon. This support consists of the confirmation of predictions of intensity‐independent saturation of the density of light‐induced defects, as well as several effects involved in the time dependence of degradation. These rate effects include variations with (a) light intensity, (b) temperature during degradation, (c) the range of density variation, and (d) the optical properties of the material.
Journal of Applied Physics | 1989
Richard H. Bube; David Redfield
A simple model for photoconductivity in a‐Si:H is shown to be capable of describing the variations of photoconductivity observed both with doping and with optical degradation. The model consists of a trivalent recombination center with consideration only of transitions connecting the centers and extended states, and of exponential conduction and valence‐edge tail states with occupancy described simply by the locations of the dark or quasi‐Fermi levels. The model describes the changes in the magnitude of photoconductivity with doping or optical degradation as arising primarily from changes in the density of effective recombination centers caused by shifts in the location of the dark Fermi level, and only secondarily from changes in the total density of dangling bonds. The model also describes the change in the exponent for the variation of photoconductivity as a power of the excitation rate; this exponent changes from a value of 0.50, when the density of electrons trapped in conduction tail states is propo...
Journal of Applied Physics | 1993
Lisa E. Benatar; David Redfield; Richard H. Bube
At least three quantities have been referred to as ‘‘activation energies’’ in association with fits to metastable defect kinetics in hydrogenated amorphous silicon. Most commonly cited is Eτ, the activation energy determined from stretched‐exponential fits to kinetics data measured over a range of temperatures. The stretched exponential can also be written in terms of a rate constant, K, which has been reported as being thermally activated. In addition, a stretched‐exponential model describing defect kinetics includes an energy, E2, in its rate equation. In this article, we clarify the interpretations of Eτ, EK, and E2, and discuss the possible physical significance of Eτ.
Applied Physics Letters | 1986
David Redfield
Recombination‐induced defect reactions are reversible in principle; i.e., any recombination (or trapping) process that induces defect formation can also induce recovery of that defect. This reversibility principle applies to crystalline and amorphous semiconductors. Although the rates in the two directions are generally quite different, evidence is cited indicating that in hydrogenated amorphous silicon they may be comparable, and offer new explanations for several observations. Illustrative rate equations are presented along with a series of testable experimental predictions.
Journal of Non-crystalline Solids | 1991
David Redfield; Richard H. Bube
Competing types of models of dangling-bond defects in hydrogenated amorphous silicon are analyzed in the light of recent advances in descriptions of light-induced defects such as saturation of their density and stretched-exponential kinetics. The familiar intrinsic models (Si-Si bond breaking) are found to have several serious problems—more than generally recognized—while extrinsic models have few. In particular, for the rehybridized two-site model based on foreign atoms, the one important need is to explain defects in undoped material.
Journal of Non-crystalline Solids | 1989
David Redfield; Richard H. Bube
Abstract The time dependence of generation of light-induced metastable defects in a-Si:H is fully describable by the same stretched exponential (SE) as the relaxation of a quench-induced state. The time constant of this SE is shown to contain information about the dependence of defect generation on light intensity and optical properties of the material. This explains reported differences in generation rates in different materials.