G. Charache - Academia.edu (original) (raw)
Papers by G. Charache
This paper reports recent progress in the development of quaternary III-V thermophotovoltaic (TPV... more This paper reports recent progress in the development of quaternary III-V thermophotovoltaic (TPV) devices based on MBE grown GaâIn{sub 1-x}As{sub y}Sb{sub 1-y}. TPV is of great interest for a variety of applications. The objective of this work is to develop a TPV cell which is tunable to the emission spectrum of a heated blackbody, at temperatures in the range of
This paper reports the growth, materials characterization, and device performance of lattice-matc... more This paper reports the growth, materials characterization, and device performance of lattice-matched GaInAsSb/GaSb thermophotovoltaic (TPV) devices with cutoff wavelength as long as 2.5 μm. GaInAsSb epilayers were grown lattice matched to GaSb substrates by organometallic vapor phase epitaxy (OMVPE) using all organometallic precursors including triethylgallium, trimethylindium, tertiarybutylarsine, and trimethylantimony with diethyltellurium and dimethylzinc as the n- and p-type dopants, respectively. The growth temperature was 525 C. Although these alloys are metastable, a mirror-like surface morphology and room temperature photoluminescence (PL) are obtained for alloys with PL peak emission at room temperature as long as 2.5 μm. In general, however, a trend of decreasing material quality is observed as the wavelength increases. Both the surface roughness and PL full width at half-maximum increase with wavelength. In spite of the dependence of material quality on PL peak emission wavelength, the internal quantum efficiency of TPV devices with cutoff wavelengths of 2.3 to 2.5 μm is as high as 86%.
Thermophotovoltaic (TPV) devices have been fabricated using epitaxial ternary and quaternary laye... more Thermophotovoltaic (TPV) devices have been fabricated using epitaxial ternary and quaternary layers grown on GaSb substrates. The GaInSb layers were grown by organometallic vapor phase epitaxy (OMVPE) and the InGaAsSb lattice-matched layers were grown by liquid phase epitaxy (LPE). Device fabrication steps include unannealed p-type ohmic contacts, annealed Sn/Au n-type ohmic contacts, and a thick Ag top-surface contact using a lift-off process. Devices are characterized primarily by dark I-V, photo I-V, and quantum efficiency measurements, which are correlated to microscopic and macroscopic material properties. Particular emphasis has been on material enhancements to increase quantum efficiency and decrease dark saturation current density. TPV device performance is presently limited by the base diffusion length, typically 1 to 2 microns.
Journal of Electronic Materials
A conversion from commercial titania (TiO2) nanoparticles to nanotubes was achieved by a hydrothe... more A conversion from commercial titania (TiO2) nanoparticles to nanotubes was achieved by a hydrothermal method. The titanate nanotube (titanate) film was then deposited on a Si (001) substrate using an electrophoretic deposition (EPD) technique. The post hydrothermal treatment was then carried out by annealing the films at 300-1000 °C for 30 min in the static air. A major amount of intercalated sodium (Na) in as-synthesized titanate nanotubes was removed during the electrodeposition process. The collapse of the tubular structure can be seen clearly when annealed above 500 °C. X-ray diffraction data indicate a significant increase in the anatase phase peak intensity with annealing temperature. O 1s peak is found to be built up of subpeaks of H2O, -OH, and Ti-O. Annealing results in an increase of the Ti-O peak intensity while other peaks disappear. Clear changes in the O 1s peak positions, symmetry, and shift towards lower energy (0.8 eV) are evident with the increasing annealing temperature. The doublet spectral lines of Ti 2p were found separated by an energy of 5.6 eV. Photoelectron spectroscopy show that during the annealing treatment, the chemical bonds such as H2O and -OH are removed from the titanate films as well as converting the bonded states of titanate to that of titania.
Third NREL Conference on thermophotovoltaic generation of electricity, 1997
ABSTRACT
The optical constants ε(E)[=εâ(E) + iεâ(E)] of single crystal GaSb at 300K have been measured... more The optical constants ε(E)[=εâ(E) + iεâ(E)] of single crystal GaSb at 300K have been measured using spectral ellipsometry in the range of 0.3--5.3 eV. The ε(E) spectra displayed distinct structures associated with critical points (CPs) at Eâ(direct gap), spin-orbit split Eâ + Îâ component, spin-orbit split (Eâ), Eâ + Îâ and (Eââ²), Eââ² + Îââ² doublets, as well as Eâ. The experimental data over the entire measured spectral range (after oxide removal) has been fit using the Holden model dielectric function [Phys.Rev.B 56, 4037 (1997)] based on the electronic energy-band structure near these CPs plus excitonic and band-to-band Coulomb enhancement effects at Eâ, Eâ + Îâand the Eâ, Eâ + Îâ doublet. In addition to evaluating the energies of these various band-to-band CPs, information about the binding energy (Râ) of the two-dimensional exciton related to the Eâ, Eâ + Îâ CPS was obtained. The value of Râ was in good agreement with effective mass/{rvec k} · {rvec p} theory. The ability to evaluate Râ has important ramifications for recent first-principles band structure calculations which include exciton effects at Eâ, Eâ, and Eâ.
The optical constants ε(E)[=εâ(E)+iεâ(E)] of two epitaxial layers of GaInAsSb/GaSb have been ... more The optical constants ε(E)[=εâ(E)+iεâ(E)] of two epitaxial layers of GaInAsSb/GaSb have been measured at 300 K using spectral ellipsometry in the range of 0.35--5.3 eV. The ε(E) spectra displayed distinct structures associated with critical points (CPs) at Eâ (direct gap), spin-orbit split Eâ+Îâ component, spin-orbit split (Eâ, Eâ+Îâ) and (Eââ², Eââ²+Îââ²) doublets, as well as Eâ. The experimental data over the entire measured spectral range (after oxide removal) has been fit using the Holder model dielectric function [Phys.Rev.B 56, 4037 (1997)] based on the electronic energy-band structure near these CPs plus excitonic and band-to-band Coulomb enhancement effects at E{sub 0.}, Eâ+Îâ and the Eâ, Eâ+Îâ doublet. In addition to evaluating the energies of these various band-to-band CPs, information about the binding energy (Râ) of the two-dimensional exciton related to the Eâ, Eâ+Îâ CPs was obtained. The value of Râ was in good agreement with effective mass/k{sup â}·p{...
Theoretical predictions for the plasma frequency and Moss-Burstein shift (optical band gap) of de... more Theoretical predictions for the plasma frequency and Moss-Burstein shift (optical band gap) of degenerately doped (n[gt]10[sup 19] hthinsp;cm[sup [minus]3]) InAs, In[sub x]Ga[sub 1[minus]x]As, and InP[sub 1[minus]y]As[sub y] are presented. These systems are of interest because they possess desirable optical properties for thermophotovoltaic (TPV) applications. The studies presented are based on electronic band structures calculated using the Full Potential Linearized Augmented Plane Wave (FLAPW) method which includes non-local screened exchange (sX-LDA) and spin-orbit effects. The plasma frequency and Moss-Burstein shift are calculated vs. doping assuming a [open quotes]rigid band[close quotes] approximation (i.e. conduction band filling of the [open quotes]undoped[close quotes] bands). The doping dependence of the effective mass (band non-parabolicity) plays an important role at the high dopings considered here. This effect leads to a maximum in the plasma frequency vs. doping (2[endash]3[times]10[sup 14]/s) and a significant departure from the [open quotes]constant effective mass[close quotes] prediction for the optical band gap vs. doping. These calculations are in good agreement with measurements. [copyright] [ital 1999 American Institute of Physics.]
The complex optical constants (real and imaginary components of the dielectric function and index... more The complex optical constants (real and imaginary components of the dielectric function and index of refraction) and absorption coefficient in the range 0.3--5.5 eV have been evaluated at 300K using spectral ellipsometry for a series of n-(5.7 à 10¹ⷠcmâ»Â³ < n < 5.5 à 10¹⹠cmâ»Â³) and p-(6.5 à 10¹ⷠcmâ»Â³ < p < 5.0 à 10¹° cmâ»Â³) doped
The internal quantum efficiency is a very sensitive characteristic of the diffusion length and su... more The internal quantum efficiency is a very sensitive characteristic of the diffusion length and surface/interface recombination velocity in the neutral regions of a TPV cell. In this paper we show how we have modeled the absorption constant of InxGa1−xAs alloys as to be able to calculate the internal quantum efficiency using a closed form computer model of the cell. Comparison between the experimentally measured and the modeled quantum efficiency allows extraction of the diffusion length and surface/interface recombination velocity.
In this paper we present the measurement techniques that we have developed in our laboratory for ... more In this paper we present the measurement techniques that we have developed in our laboratory for characterizing the performance of larger area InGaAs or similar TPV cells. Characteristics obtained in our measurement procedures include: I-V characteristics under dark and illuminated conditions, lifetime of excess carriers using open circuit voltage decay, doping concentration, contact potential, external quantum efficiency and surface reflectivity. The I-V characteristics under dark and illuminated conditions are done on a pulsed basis to avoid using an elaborate heat sink. The measurement results obtained in InGaAs cells are presented.
In this paper we present two methods for determining the conversion efficiency of TPV devices. In... more In this paper we present two methods for determining the conversion efficiency of TPV devices. In the first, the conversion efficiency is calculated from measurements of the external quantum efficiency and reflection as a function of wavelength, and from the I-V characteristics under high-level illumination. This is an indirect method based on separate differential measurements. In the second method, a novel heat transfer technique is utilized to combine both the voltaic diode and spectral control efficiency into a single measurement.
This paper reports recent progress in the development of quaternary III-V thermophotovoltaic (TPV... more This paper reports recent progress in the development of quaternary III-V thermophotovoltaic (TPV) devices based on MBE grown GaâIn{sub 1-x}As{sub y}Sb{sub 1-y}. TPV is of great interest for a variety of applications. The objective of this work is to develop a TPV cell which is tunable to the emission spectrum of a heated blackbody, at temperatures in the range of
This paper reports the growth, materials characterization, and device performance of lattice-matc... more This paper reports the growth, materials characterization, and device performance of lattice-matched GaInAsSb/GaSb thermophotovoltaic (TPV) devices with cutoff wavelength as long as 2.5 μm. GaInAsSb epilayers were grown lattice matched to GaSb substrates by organometallic vapor phase epitaxy (OMVPE) using all organometallic precursors including triethylgallium, trimethylindium, tertiarybutylarsine, and trimethylantimony with diethyltellurium and dimethylzinc as the n- and p-type dopants, respectively. The growth temperature was 525 C. Although these alloys are metastable, a mirror-like surface morphology and room temperature photoluminescence (PL) are obtained for alloys with PL peak emission at room temperature as long as 2.5 μm. In general, however, a trend of decreasing material quality is observed as the wavelength increases. Both the surface roughness and PL full width at half-maximum increase with wavelength. In spite of the dependence of material quality on PL peak emission wavelength, the internal quantum efficiency of TPV devices with cutoff wavelengths of 2.3 to 2.5 μm is as high as 86%.
Thermophotovoltaic (TPV) devices have been fabricated using epitaxial ternary and quaternary laye... more Thermophotovoltaic (TPV) devices have been fabricated using epitaxial ternary and quaternary layers grown on GaSb substrates. The GaInSb layers were grown by organometallic vapor phase epitaxy (OMVPE) and the InGaAsSb lattice-matched layers were grown by liquid phase epitaxy (LPE). Device fabrication steps include unannealed p-type ohmic contacts, annealed Sn/Au n-type ohmic contacts, and a thick Ag top-surface contact using a lift-off process. Devices are characterized primarily by dark I-V, photo I-V, and quantum efficiency measurements, which are correlated to microscopic and macroscopic material properties. Particular emphasis has been on material enhancements to increase quantum efficiency and decrease dark saturation current density. TPV device performance is presently limited by the base diffusion length, typically 1 to 2 microns.
Journal of Electronic Materials
A conversion from commercial titania (TiO2) nanoparticles to nanotubes was achieved by a hydrothe... more A conversion from commercial titania (TiO2) nanoparticles to nanotubes was achieved by a hydrothermal method. The titanate nanotube (titanate) film was then deposited on a Si (001) substrate using an electrophoretic deposition (EPD) technique. The post hydrothermal treatment was then carried out by annealing the films at 300-1000 °C for 30 min in the static air. A major amount of intercalated sodium (Na) in as-synthesized titanate nanotubes was removed during the electrodeposition process. The collapse of the tubular structure can be seen clearly when annealed above 500 °C. X-ray diffraction data indicate a significant increase in the anatase phase peak intensity with annealing temperature. O 1s peak is found to be built up of subpeaks of H2O, -OH, and Ti-O. Annealing results in an increase of the Ti-O peak intensity while other peaks disappear. Clear changes in the O 1s peak positions, symmetry, and shift towards lower energy (0.8 eV) are evident with the increasing annealing temperature. The doublet spectral lines of Ti 2p were found separated by an energy of 5.6 eV. Photoelectron spectroscopy show that during the annealing treatment, the chemical bonds such as H2O and -OH are removed from the titanate films as well as converting the bonded states of titanate to that of titania.
Third NREL Conference on thermophotovoltaic generation of electricity, 1997
ABSTRACT
The optical constants ε(E)[=εâ(E) + iεâ(E)] of single crystal GaSb at 300K have been measured... more The optical constants ε(E)[=εâ(E) + iεâ(E)] of single crystal GaSb at 300K have been measured using spectral ellipsometry in the range of 0.3--5.3 eV. The ε(E) spectra displayed distinct structures associated with critical points (CPs) at Eâ(direct gap), spin-orbit split Eâ + Îâ component, spin-orbit split (Eâ), Eâ + Îâ and (Eââ²), Eââ² + Îââ² doublets, as well as Eâ. The experimental data over the entire measured spectral range (after oxide removal) has been fit using the Holden model dielectric function [Phys.Rev.B 56, 4037 (1997)] based on the electronic energy-band structure near these CPs plus excitonic and band-to-band Coulomb enhancement effects at Eâ, Eâ + Îâand the Eâ, Eâ + Îâ doublet. In addition to evaluating the energies of these various band-to-band CPs, information about the binding energy (Râ) of the two-dimensional exciton related to the Eâ, Eâ + Îâ CPS was obtained. The value of Râ was in good agreement with effective mass/{rvec k} · {rvec p} theory. The ability to evaluate Râ has important ramifications for recent first-principles band structure calculations which include exciton effects at Eâ, Eâ, and Eâ.
The optical constants ε(E)[=εâ(E)+iεâ(E)] of two epitaxial layers of GaInAsSb/GaSb have been ... more The optical constants ε(E)[=εâ(E)+iεâ(E)] of two epitaxial layers of GaInAsSb/GaSb have been measured at 300 K using spectral ellipsometry in the range of 0.35--5.3 eV. The ε(E) spectra displayed distinct structures associated with critical points (CPs) at Eâ (direct gap), spin-orbit split Eâ+Îâ component, spin-orbit split (Eâ, Eâ+Îâ) and (Eââ², Eââ²+Îââ²) doublets, as well as Eâ. The experimental data over the entire measured spectral range (after oxide removal) has been fit using the Holder model dielectric function [Phys.Rev.B 56, 4037 (1997)] based on the electronic energy-band structure near these CPs plus excitonic and band-to-band Coulomb enhancement effects at E{sub 0.}, Eâ+Îâ and the Eâ, Eâ+Îâ doublet. In addition to evaluating the energies of these various band-to-band CPs, information about the binding energy (Râ) of the two-dimensional exciton related to the Eâ, Eâ+Îâ CPs was obtained. The value of Râ was in good agreement with effective mass/k{sup â}·p{...
Theoretical predictions for the plasma frequency and Moss-Burstein shift (optical band gap) of de... more Theoretical predictions for the plasma frequency and Moss-Burstein shift (optical band gap) of degenerately doped (n[gt]10[sup 19] hthinsp;cm[sup [minus]3]) InAs, In[sub x]Ga[sub 1[minus]x]As, and InP[sub 1[minus]y]As[sub y] are presented. These systems are of interest because they possess desirable optical properties for thermophotovoltaic (TPV) applications. The studies presented are based on electronic band structures calculated using the Full Potential Linearized Augmented Plane Wave (FLAPW) method which includes non-local screened exchange (sX-LDA) and spin-orbit effects. The plasma frequency and Moss-Burstein shift are calculated vs. doping assuming a [open quotes]rigid band[close quotes] approximation (i.e. conduction band filling of the [open quotes]undoped[close quotes] bands). The doping dependence of the effective mass (band non-parabolicity) plays an important role at the high dopings considered here. This effect leads to a maximum in the plasma frequency vs. doping (2[endash]3[times]10[sup 14]/s) and a significant departure from the [open quotes]constant effective mass[close quotes] prediction for the optical band gap vs. doping. These calculations are in good agreement with measurements. [copyright] [ital 1999 American Institute of Physics.]
The complex optical constants (real and imaginary components of the dielectric function and index... more The complex optical constants (real and imaginary components of the dielectric function and index of refraction) and absorption coefficient in the range 0.3--5.5 eV have been evaluated at 300K using spectral ellipsometry for a series of n-(5.7 à 10¹ⷠcmâ»Â³ < n < 5.5 à 10¹⹠cmâ»Â³) and p-(6.5 à 10¹ⷠcmâ»Â³ < p < 5.0 à 10¹° cmâ»Â³) doped
The internal quantum efficiency is a very sensitive characteristic of the diffusion length and su... more The internal quantum efficiency is a very sensitive characteristic of the diffusion length and surface/interface recombination velocity in the neutral regions of a TPV cell. In this paper we show how we have modeled the absorption constant of InxGa1−xAs alloys as to be able to calculate the internal quantum efficiency using a closed form computer model of the cell. Comparison between the experimentally measured and the modeled quantum efficiency allows extraction of the diffusion length and surface/interface recombination velocity.
In this paper we present the measurement techniques that we have developed in our laboratory for ... more In this paper we present the measurement techniques that we have developed in our laboratory for characterizing the performance of larger area InGaAs or similar TPV cells. Characteristics obtained in our measurement procedures include: I-V characteristics under dark and illuminated conditions, lifetime of excess carriers using open circuit voltage decay, doping concentration, contact potential, external quantum efficiency and surface reflectivity. The I-V characteristics under dark and illuminated conditions are done on a pulsed basis to avoid using an elaborate heat sink. The measurement results obtained in InGaAs cells are presented.
In this paper we present two methods for determining the conversion efficiency of TPV devices. In... more In this paper we present two methods for determining the conversion efficiency of TPV devices. In the first, the conversion efficiency is calculated from measurements of the external quantum efficiency and reflection as a function of wavelength, and from the I-V characteristics under high-level illumination. This is an indirect method based on separate differential measurements. In the second method, a novel heat transfer technique is utilized to combine both the voltaic diode and spectral control efficiency into a single measurement.