A. Rothwarf

Time-dependent open-circuit voltage in CuInSe2/CdS solar cells: theory and experiment

The increase with illumination time of the open‐circuit voltage of the CuInSe2/CdS solar cell has been studied as a function of illumination intensity, temperature, and preillumination bias voltage. The results indicate time constants on the order of minutes (40 to 500 s at room temperature), with an activation energy of ∼0.20 eV. The magnitude of the effect (ΔVoc ≤30 mV) and the time constants both decrease with increasing preillumination voltage bias, indicating that the effect is due to voltage rather than illumination. A model involving the tunneling of electrons trapped in deep states in the CdS, near the junction with the CuInSe2 layer to holes in the CuInSe2 valence band, seems to account for the magnitude, as well as the time and voltage dependence of the effect. The increased positive charge near the interface shifts the division of the diffusion voltage between the two sides of the junction, thereby increasing Voc. The fill factor and efficiency of the cell are also increased, while the short‐ci...

Effects of a voltage‐dependent light‐generated current on solar cell measurements: CuInSe2/Cd(Zn)S

Chopped beam measurements on CuInSe2/Cd(Zn)S solar cells have revealed a failure of the superposition principle, with the light‐generated current showing a strong voltage dependence. The light I‐V curve can be fully explained with no change in the diode current. The voltage‐dependent light‐generated current dominates the light I‐V curve below Voc, requiring that the analysis of all the traditional solar cell characterization measurements be modified. We have derived the general expressions needed to obtain the diode factor from the light I‐V and Voc versus intensity curves. We have also shown how the temperature variation of Voc is affected.

A p-i-n heterojunction model for the thin-film CuInSe 2 /CdS solar cell

By treating the high-resistivity CdS and CuInSe2layers in high efficiency cells as insulating, a simple p-i-n model results that predicts the behavior seen in these cells. The relatively low open-circuit voltage and the diode factor are both directly related to the width of the insulating CdS layer. Substantial improvements in both Vocand fill factor can be expected if the width of this CdS layer can be reduced.

Computer simulation and modeling of graded bandgap CuInSe/sub 2//CdS based solar cells

This paper proposes the graded bandgap absorber material, Cu/sub 1-x/Ag/sub x/In/sub 1-y-z/Ga/sub y/Al/sub z/Se/sub 2(1-u

CuInSe2/Cd(Zn)S solar cell modeling and analysis

/ -/sub w/)S/sub 2u/Te/sub 2w/ (CIS*) multinary system, to improve the low open-circuit voltage (V/sub OC/) seen in CuInSe/sub 2//CdS solar cells, without sacrificing the short-circuit current density (J/sub sc/). It also proposes a p-i-n model for the CuInSe/sub 2//CdS solar cell, where the intrinsic region is the graded bandgap CIS*. Reflecting surfaces are provided at the p-i and n-i interfaces to trap the light in the narrow intrinsic region for maximum generation of electron and hole pairs (EHPs). This optical confinement results in a 25-40% increase in the number of photons absorbed. An extensive numerical simulator was developed, which provides a 1-D self-consistent solution for Poissons equation and the two continuity equations for electrons and holes. This simulator was used to generate J-V curves to delineate the effect of different grading profiles on cell performance. The effects of a uniform bandgap, normal grading, reverse grading, and a low bandgap notch have been considered. Having established the inherent advantages to these grading profiles an optimal doubly graded structure is proposed with grading between 1.5 eV and 1.3 eV regions which has V/sub OC/=0.86 V, /spl eta/=17.9%, FF=0.79 and J/sub sc/=26.3 mA/cm/sup 2/ compared to 0.84 V, 14.9%, 0.76, and 23.3 mA/cm/sup 2/, respectively, for the highest efficiency 1.4-eV uniform bandgap cell. Replacing the thick CdS(2.42 ev) layer assumed in our simulations with a wide gap semiconductor such as ZnO(3.35 ev) increases all current densities by about 5 mA/cm/sup 2/, and increases the optimal calculated efficiency from 17.9% to roughly 21% for a doubly graded structure with a thickness of 1 /spl mu/m and bandgaps ranging from 1.3 eV to 1.5 eV.

A new quantum mechanical channel mobility model for Si MOSFET's

Abstract The CuInSe 2 /CdS solar cell can be modeled as a heterojunction with charge on its interface. Many of the time-dependent and intensity-dependent phenomena seen in current-voltage and capacitance-voltage measurements can be treated with this model. A comparison of the electron-beam-induced current (EBIC) results and the spectral response of the cells shows that only EBIC profiles which show strong response at the heterojunction are consistent with the flat (from 0.5 to 1.2 μm) spectral response of the cell.

Hot carrier effects in the collection efficiency of solar cells: a‐Si:H

A new model is presented for the transverse field-dependent mobility in MOSFETs. It is based on the diffuse scattering of electrons at the Si-SiO2interface in the narrow inversion layer, due to the uncertainty principle related momentum, of the electrons in induced quantum levels. Our calculations for unstressed material yield µ ≈ F-αwith α ≈ 0.67 at high fields. For stressed material α falls to less than 0.2. Combining this channel mobility with bulk phonon scattering gives an excellent fit to a large body of data. Process-induced stress may account for the wide range in µ and α reported in the literature.

Noise characteristics and detectivity of YBa2Cu3O7 superconducting bolometers: Bias current, frequency, and temperature dependence

Recent models for the falloff in spectral response at short wavelengths in low mobility solar cell materials have been based upon the diffusion of thermalized carriers against the electric field to high recombination contacts or interfaces. These models neglect the excess energy imparted to the carriers by short wavelength light, that enable them to easily move against the field while thermalizing. Calculations of the distance the carriers diffuse before losing their excess energy by phonon emission indicate that in the high absorption coefficient region of a‐Si:H up to 50% of the carriers can reach a contact or interface before thermalizing.

YBa2Cu3O7−δ infrared bolometers: Temperature‐dependent responsivity and deviations from the dR/dT curve

Meander line patterned infrared detectors, with values of resistance Ronset at the onset temperature Tc onset of 3–5 kΩ, were fabricated from YBa2Cu3O7−x superconducting material on MgO and SrTiO3 crystalline substrates. Noise voltages from the samples were measured versus bias current, radiation modulation frequency, and temperature, in both the normal and superconducting states. Four major types of voltage noise were identified according to where they occurred in temperature relative to Tconset and the zero resistance temperature Tc zero, and their dependence on frequency and bias current. They were also associated with the granularity of the superconducting film, which is related to the substrate material used. From these observations a specific cause for each type of noise is suggested. The results are as follows. (i) In the normal state with temperature T≳Tc onset, noise with a magnitude that is consistent with thermal (Johnson) noise is seen, but it depends linearly on bias current above a threshold...

An Aether Model of the Universe

The phase and amplitude of the response to infrared radiation of superconducting bolometers was investigated. The detectors were fabricated with 120–550 nm Y1Ba2Cu3O7−δ films on MgO, SrTiO3, and LaAlO3 substrates. A model for the response is developed and compared to the experimental results. The response versus frequency of the samples shows bolometric behavior, in agreement with the measured time dependence of the signals at low frequencies. Different techniques were developed to measure the key parameter in the response of the bolometers, G, the thermal conductance. Several anomalies were observed in the study that provide insight into heat conduction in these devices. The dc or low modulation frequency thermal conductance, G(0), of the samples is found to be limited by the substrate/cold‐head thermal boundary resistance. It was also found to decrease with increasing substrate thickness, but still limited by the substrate/cold‐head interface. This thickness dependence of G(0) is attributed to scattering of phonons within the substrate, that changes their transmission rate through the substrate/cold‐head interface. The results from simultaneous measurements of the IR response and dR/dT (using R versus T) at low modulation frequencies (20 Hz) show that the magnitude of the response differs by up to 30% from the dR/dT curve. The discrepancy is found to be frequency dependent, increasing with decreasing modulation frequency.This can be treated by use of temperature‐dependent thermal constants (G and the heat capacity, C) in the model for the bolometric response. The discrepancy is also observed to be dependent on the superconducting transition, which suggests a possible correlation between the heat conduction through the substrate (and its interfaces) and the absorption mechanism in the superconductor. The phase of the response versus temperature shows an abrupt change at the transition. This is evidence for a change in thermal constants in the bolometer as it goes through the superconducting transition, affecting both the phase and magnitude of the response. Joule heating at even high bias currents has little effect on the response (and its deviation from the dR/dT curve) in our samples, and the effect of noise on the response is significant only at very low bias currents.