Search

This tutorial paper focuses on the physical origin of thermal droop, i.e., the decrease in the luminescence of light-emitting diodes (LEDs) induced by increasing temperature. III-nitride-based LEDs are becoming a pervasive technology, covering several fields from lighting to displays, from automotive to portable electronics, and from horticulture to sensing. In all these environments, high efficiency is a fundamental requirement, for reducing power consumption and system cost. Over the last decade, a great deal of effort has been put in the analysis of the efficiency droop, the decrease in LED internal quantum efficiency (IQE) induced by high current density. On the other hand, an IQE decrease is observed also for increasing temperature, a phenomenon usually referred to as thermal droop. For commercial LEDs, the IQE decrease related to thermal droop can be comparable to that of efficiency droop: for this reason, understanding thermal droop is a fundamental step for making LEDs capable of operating at high temperature levels. In several fields (including street lighting, automotive, photochemical treatments, projection, entertainment lighting, etc.), compact and high-flux light sources are required: typically, to reduce the size, weight, and cost of the systems, LEDs are mounted in compact arrays, and heat sinks are reduced to a minimum. As a consequence, LEDs can easily reach junction temperatures above 85–100 °C and are rated for junction temperatures up to 150–175 °C (figures from commercially available LED datasheets: Cree XHP70, Osram LUW HWQP, Nichia NVSL219CT, Samsung LH351B, and LedEngin LZP-00CW0R) and this motivates a careful analysis of thermal droop. This paper discusses the possible physical causes of thermal droop. After an introduction on the loss mechanisms in junctions, we will individually focus on the following processes: (i) Shockley–Read–Hall (SRH) recombination and properties of the related defects; (ii) Auger recombination and its temperature dependence, including the discussion of trap-assisted Auger recombination; (iii) impact of carrier transport on the thermal droop, including a discussion on carrier delocalization, escape, and freeze out; (iv) non-SRH defect-related droop mechanisms. In addition, (v) we discuss the processes that contribute to light emission at extremely low current levels and (vi) the thermal droop in deep ultraviolet LEDs, also with reference to the main parasitic emission bands. The results presented within this paper give a tutorial perspective on thermal droop; in addition, they suggest a pathway for the mitigation of this process and for the development of LEDs with stable optical output over a broad temperature range. [ABSTRACT FROM AUTHOR]

We present a detailed simulation study on plasmonic-organic hybrid electro-optic modulators based on coupled symmetric or asymmetric plasmonic slots. An electro-optic polymer is exploited as an active material, and the device is compatible with a silicon photonics platform. The proposed device operates at 1550 nm wavelength, typical of data center or long-haul telecommunication systems. The device performance in terms of area, plasmonic losses, optical bandwidth, intrinsic modulation bandwidth and energy dissipation are comparable to already proposed Mach-Zehnder solutions, but with potentially better extinction ratio, coupling losses due to photonic-plasmonic transitions, and flexibility in exploiting, without any performance penalty, asymmetric slots to shift the ON and OFF states bias. Finally, the bias dependence of the modulation chirp is investigated, comparing through and cross-coupling configurations. [ABSTRACT FROM AUTHOR]

Herein, an overview on the physics‐based and system‐oriented modeling of organic electro‐optic Mach–Zehnder modulators for optical communication systems is presented. State‐of‐the‐art solutions in electro‐optic organic materials and modulator designs are reviewed, with particular stress on silicon organic hybrid (SOH) and plasmonic organic hybrid (POH) solutions. Then, the physics‐based simulation of concentrated and traveling‐wave modulators is discussed both through 3D optical simulations and a combination of 2D and 3D models, where the modulator is partitioned into convenient subdomains. Neural network behavioral models are finally discussed and case studies are proposed on the physics‐based and system‐level simulation of SOH and POH Mach–Zehnder modulators. [ABSTRACT FROM AUTHOR]

Defects can significantly modify the electro-optical characteristics of InGaN light-emitting diodes (LEDs); however, modeling the impact of defects on the electrical characteristics of LEDs is not straightforward. In this paper, we present an extensive investigation and modeling of the impact of defects on the electrical characteristics of InGaN-based LEDs, as a function of the thickness of the quantum well (QW). First, we demonstrate that the density of defects in the active region of III-N LEDs scales with increasing thickness of the InGaN QW. Since device layers with high indium content tend to incorporate more defects, we ascribed this experimental evidence to the increased volume of defects-rich InGaN associated to thicker InGaN layers. Second, we demonstrate that the current–voltage characteristics of the devices are significantly influenced by the presence of defects, especially in the sub turn-on region. Specifically, we show that the electrical characteristics can be effectively modeled in a wide current range (from pA to mA), by considering the existence of trap-assisted tunneling processes. A good correspondence is obtained between the experimental and simulated electrical characteristics (I–V), by using—in the simulation—the actual defect concentrations/activation energies extracted from steady-state photocapacitance, instead of generic fitting parameters. [ABSTRACT FROM AUTHOR]

The ICT scene is dominated by short-range intra-datacenter interconnects and networking, requiring high speed and stable operations at high temperatures. GaAs/AlGaAs vertical-cavity surface-emitting lasers (VCSELs) emitting at 850–980 nm have arisen as the main actors in this framework. Starting from our in-house 3D fully comprehensive VCSEL solver VENUS, in this work we present the possibility of downscaling the dimensionality of the simulation, ending up with a multiphysics 1D solver (D1ANA), which is shown to be capable of reproducing the experimental data very well. D1ANA is then extensively applied to optimize high-temperature operation, by modifying cavity detuning and distributed Bragg's reflector lengths. [ABSTRACT FROM AUTHOR]

We present a multiband envelope-function model for wurtzite nanostructures based on a rigorous numerical procedure to determine operator ordering and band parameters from nonlocal empirical pseudopotential calculations. The proposed approach, implemented within a finite-element scheme, leads to well-posed, numerically stable envelope equations that accurately reproduce full-Brillouinzone subband dispersions of quantum systems. Although demonstrated here for III-nitride nonlocal empirical pseudopotentials, the model provides a general theoretical framework applicable to ab initio electronic structures of wurtzite semiconductors. [ABSTRACT FROM AUTHOR]

In state-of-the-art, large format, HgCdTe-based infrared focal plane arrays the typical pixel size (3–10 μ m ) is of the order of the operating wavelength and much smaller than the carrier diffusion length. This makes inter-pixel crosstalk a limiting factor, especially in planar structures. Employing three-dimensional electromagnetic and electrical simulations we show that, besides reducing the dark current through Auger suppression, majority carrier depletion of the detector absorber is also effective in curtailing the inter-pixel crosstalk due to carrier diffusion. In the case of a 5 μ m -pitch pixel, a proper design of the absorber composition and doping profile allows to reduce inter-pixel crosstalk by more than a factor of two when increasing the reverse bias from - 0.1 to - 0.5 V , keeping the contribution to crosstalk coming from carrier diffusion between 2 and 12% in the mid- and long-wavelength infrared spectrum. [ABSTRACT FROM AUTHOR]

Probing the thermal effects that limit the performance of vertical-cavity surface-emitting lasers (VCSELs) reveals an intricate interplay between carrier transport, recombination mechanisms, thermal conduction, and optical features. An understanding of this interplay requires an accurate yet computationally-efficient multiphysical approach at the microscopic level. We address this task by characterizing oxide-confined multimode 850 nm VCSELs over a wide temperature range, and by simulating them with VENUS, our in-house multiphysics simulator. The agreement with the experimental results in the whole explored temperature range provides physical insight into the mechanisms that determine VCSEL rollover and turn off, and demonstrates the predictive capabilities of the model. A complete set of model parameters determined through this combined experiment-simulation approach is presented. [ABSTRACT FROM AUTHOR]

The properties of vertical-cavity surface-emitting lasers (VCSELs) are investigated by means of a multiphysical Vcsel Electro-opto-thermal NUmerical Simulator (VENUS). VENUS includes a three-dimensional vectorial electromagnetic code, a description of the quantum well optical response, a heat equation solver, and a quantum-corrected drift-diffusion simulator. The proposed suite includes coupling mechanisms often overlooked in VCSEL simulation and allows to reassess the impact of parameters, which can be critically dependent on implementation details inaccessible in commercial codes. The agreement with experimental results holds the promise of the application of this framework to the computer-aided design of innovative VCSEL concepts. [ABSTRACT FROM AUTHOR]

We present a k·p model for wurtzite semiconductors that allows the accurate approximation of the electronic structure over the entire Brillouin zone. The inclusion of an additional expansion point besides Γ allows significant improvements over standard full-Brillouin-zone approaches while keeping a manageable number of model parameters. We provide complete information about the Hamiltonian matrices of both expansion points and discuss the details of the optimization process used to determine the matrix parameters. As a demonstration of our scheme, we propose an approximation of the electronic structure of wurtzite ZnO, optimized for application to full-band Monte Carlo electron transport simulation. (A MATLAB implementation of the k·p model for ZnO is available from the authors.) [ABSTRACT FROM AUTHOR]

A nonlocal semiempirical pseudopotential calculation of the electronic structure of wurtzite ZnO is proposed. The local and nonlocal components of the atomic effective potentials have been sequentially optimized and an excellent quantitative agreement has been achieved with a wide range of band features (energy gaps at high symmetry points, valence band ordering, in-plane and perpendicular components of the effective masses for electrons and holes at Γ), selected not only from available experimental and ab initio results, but also from new calculations performed with the code developed by the ABINIT project. The valence band description has been further improved through the inclusion of spin-orbit corrections. The complex dielectric function along the main crystallographic directions corresponding to the optimized electronic structure is also presented, along with extensive comparisons of all the computed quantities with the literature data. [ABSTRACT FROM AUTHOR]

Wave-vector-dependent rates of disorder-induced alloy scattering have been computed for wurtzite AlGaN and InGaN to determine the transport properties of III-nitride alloys through full band Monte Carlo simulation. Contrary to previous studies, the empirical selection of a constant alloy scattering potential has been replaced by a more fundamental approach based on detailed information about the electronic structure and the corresponding screened atomic potentials. Band structures and atomic potentials have been determined in the framework of the nonlocal empirical pseudopotential method; good agreement of the fundamental energy gap with available experimental information has been achieved over the entire composition range of the alloys with the inclusion of a disorder contribution in the pseudopotential. The calculated alloy scattering potential is in reasonable agreement with the few indirect measurements available for AlGaN. Calculations of electron steady-state velocity-field curves confirm that alloy scattering has significant effects on the transport properties of AlGaN and InGaN. [ABSTRACT FROM AUTHOR]

In this article we discuss the effect of temperature on the impact ionization coefficients in wide band-gap semiconductors and compare it to that of bulk GaAs. The impact ionization coefficients as a function of temperature are examined for three semiconductors: gallium arsenide, cubic phase silicon carbide, and zinc-blende phase gallium nitride. It is found that the magnitude of the phonon energy is principally responsible for changes in the impact ionization coefficients as a result of temperature change. While the energy band gap of a material does have a temperature dependence that directly affects the impact ionization transition rate, that change is much smaller than the relative change in the magnitude of the phonon-scattering rates in all of the materials studied here. The phonon energies are found to play a vital role in the magnitude of the change in the scattering rates as a function of temperature. Materials with relatively small phonon energies have phonon scattering rates that change considerably with temperature, and therefore have impact ionization coefficients that also change considerably with changes in temperature. Conversely, the phonon scattering rate in materials with a large phonon energy is less affected and thus these materials have impact ionization coefficients that are relatively insensitive to changes in the temperature. [ABSTRACT FROM AUTHOR]

This work presents nonlocal pseudopotential calculations based on realistic, effective atomic potentials of the wurtzite phase of GaN, InN, and AlN. A formulation formulation for the model effective atomic potentials has been introduced. For each of the constitutive atoms in these materials, the form of the effective potentials is optimized through an iterative scheme in which the band structures are recursively calculated and selected features are compared to experimental and/or ab initio results. The optimized forms of the effective atomic potentials are used to calculate the band structures of the binary compounds, GaN, InN, and AlN. The calculated band structures are in excellent overall agreement with the experimental/ab initio values, i.e., the energy gaps at high-symmetry points, valence-band ordering, and effective masses for electrons match to within 3%, with a few values within 5%. The values of the energy separation, effective masses, and nonparabolicity coefficients for several secondary valleys are tabulated as well in order to facilitate analytical Monte Carlo transport simulations. © 2000 American Institute of Physics. [ABSTRACT FROM AUTHOR]

This work presents detailed information on the band structures of the III-nitride wurtzite ternary alloys, computed through the virtual crystal approximation approach. The key ingredient of this study is the set of realistic atomic effective potentials described in Part I of the present work, dedicated to the constituent binary compounds. The model relies on the linear interpolation of the structural parameters and of the local and nonlocal effective potentials: no further empirical corrections are included. The dependence on the mole fraction is computed for the energy gaps at all the high-symmetry points, the valence-band width, and the electron effective masses in the valleys relevant for carrier-transport simulation. © 2000 American Institute of Physics. [ABSTRACT FROM AUTHOR]

The electron capture time in single quantum wells is calculated by considering capture and escape as scattering events in Monte Carlo simulation. The calculation is performed for an AlGaAs/GaAs quantum well as a function of the well width at 300 K. The overall capture time of carriers is found to be controlled by the transition from the free state to the uppermost confined levels. Subsequent interband transitions cause rapid decay into lower levels. [ABSTRACT FROM AUTHOR]

Combined optical and electrical simulations of infrared HgCdTe-based focal plane arrays under broadband, non-monochromatic illumination are critically relevant to the design of small-volume detectors with sub-wavelength pixel pitches. We present an efficient technique, based on a single finite-difference time-domain electromagnetic simulation, that provides the photogeneration rate profile due to realistic, broadband optical sources, avoiding multiple monochromatic simulations. This technique is applied to assess the effects of the temperature of blackbody optical sources on quantum efficiency and inter-pixel crosstalk of planar LWIR arrays. [ABSTRACT FROM AUTHOR]

Recent studies indicate as an important technological step the development of infrared HgCdTe-based focal plane arrays (FPAs) with sub-wavelength pixel pitch, with the advantage of smaller volume, lower weight, and potentially lower cost. In order to assess the limits of pixel pitch scaling, we present combined three-dimensional optical and electrical simulations of long-wavelength infrared HgCdTe FPAs, with 3 μm, 5 μm, and 10 μm pitch. Numerical simulations predict significant cavity effects, brought by the array periodicity. The optical and electrical contributions to spectral inter-pixel crosstalk are investigated as functions of pixel pitch, by illuminating the FPAs with Gaussian beams focused on the central pixel. Despite the FPAs being planar with 100% pixel duty cycle, our calculations suggest that the total crosstalk with nearest-neighbor pixels could be kept acceptably small also with pixels only $$3\,\mu \hbox {m}$$ wide and a diffraction-limited optical system. [ABSTRACT FROM AUTHOR]

We present a simulation study of HgCdTe-based long-wavelength infrared detectors, focusing on methodological comparisons between the finite-difference time-domain (FDTD) and ray-tracing optical models. We performed three-dimensional simulations to determine the absorbed photon density distributions and the corresponding photocurrent and quantum efficiency spectra of isolated n-on- p uniform-composition pixels, systematically comparing the results obtained with FDTD and ray tracing. Since ray tracing is a classical optics approach, unable to describe interference effects, its applicability has been found to be strongly wavelength dependent, especially when reflections from metallic layers are relevant. Interesting cavity effects around the material cutoff wavelength are described, and the cases where ray tracing can be considered a viable approximation are discussed. [ABSTRACT FROM AUTHOR]

The potential applications of germanium and its alloys in infrared silicon-based photonics have led to a renewed interest in their optical properties. In this letter, we report on the numerical determination of Auger coefficients at T=300K for relaxed and biaxially strained germanium. We use a Green's function based model that takes into account all relevant direct and phononassisted processes and perform calculations up to a strain level corresponding to the transition from indirect to direct energy gap. We have considered excess carrier concentrations ranging from 1016 cm-3 to 5×1019 cm-3. For use in device level simulations, we also provide fitting formulas for the calculated electron and hole Auger coefficients as functions of carrier density. [ABSTRACT FROM AUTHOR]

A conformal-mapping approach to model the skin-effect conductor losses of thick multiconductor coplanar waveguides and multiconductor coplanar strips is proposed. The model allows for arbitrary strip or slot number and widths and is accurate for conductor thicknesses up to around 40% of the minimum strip and slot width. Examples are presented to demonstrate the accuracy of the approach when compared to the results from a finite-element method numerical code. [ABSTRACT FROM PUBLISHER]

A combined experimental and numerical simulation study is presented on two sets of nominally identical $$\hbox {Hg}_{1-x}\hbox {Cd}_{x}\hbox {Te}$$ single-color back-illuminated midwave-infrared n-on- p photodetectors grown by liquid-phase epitaxy, p-doped with Hg vacancies and with Au, respectively. The present numerical model includes a novel formulation for band-to-band tunneling, which overcomes the intrinsic limitations of the classical Kane description without introducing numerical issues typical of other approaches. Our study confirms that adopting n-on- p architectures, avoiding metal vacancy doping, and reducing the acceptor density in the absorber region are prerequisites for obtaining high-operating-temperature photodetectors. A significant contribution to the dark current in both sets of devices is attributed to impact ionization, crucial to obtain a satisfactory explanation for the measured characteristics also at low to intermediate bias. [ABSTRACT FROM AUTHOR]

We present a combined theoretical, numerical and experimental investigation on trap-assisted tunneling (TAT) in the subthreshold regime of III-nitride-based light-emitting diodes (LEDs). Starting from the basic formulation of the TAT models provided by Hurkx and Schenk, we discuss the derivation of a detailed approach based on both multiphonon and elastic nonlocal processes. A sensitivity study conducted over the main trap- and phonon-related physical parameters of this nonlocal TAT model confirms the importance of tunneling assisted by lattice defects on the LED electrical behavior in the low-medium forward bias range. Comparisons with measured temperature-dependent electrical characteristics $$I(V;T)$$ of a single quantum well LED grown on a highly conductive SiC substrate demonstrate that $$I(V;T)$$ can be accurately reproduced in the range between 200 and 400 K by implementing the nonlocal model for TAT processes via traps in the electron-blocking and spacer layers. [ABSTRACT FROM AUTHOR]

In a combined experimental and numerical investigation, we present the effects of trap-assisted tunneling on the sub-threshold forward bias characteristics of a blue InGaN/GaN single-quantum-well LED test structure grown on a SiC substrate. The different role of donor- and acceptor-like traps has been studied, for the information it can provide on the role played by point defects. Using the energy Et and trap density Nt as the only tunneling-related fitting parameters, the behavior of the measured I( V) curves is well reproduced by our model over a wide current and temperature range. The very good agreement between simulations and experiments suggests that trap-assisted forward tunneling is one of the most relevant contributions to the current flow below the optical turn-on of the diode. [ABSTRACT FROM AUTHOR]

A self-energy-based model for the calculation of the electron capture time through phonon emission in presence of carrier population is applied to nitride-based quantum wells (QWs). The role of plasma dynamical screening on the Frölich interaction is investigated, highlighting its central role in the dependence of capture time on carrier density. The proposed approach yields a closed-form expression for capture time as a function of the carrier densities in QW and barrier states. Its application to simple two-population rate equations allows to reproduce available experimental data with excellent agreement, offering an accurate yet practical alternative to the usual approximation of a constant capture time in modeling light-emitting diodes and lasers. [ABSTRACT FROM AUTHOR]

Recent experiments of electron emission spectroscopy (EES) on III-nitride light-emitting diodes (LEDs) have shown a correlation between droop onset and hot electron emission at the cesiated surface of the LED p-cap. The observed hot electrons have been interpreted as a direct signature of Auger recombination in the LED active region, as highly energetic Auger-excited electrons would be collected in long-lived satellite valleys of the conduction band so that they would not decay on their journey to the surface across the highly doped p-contact layer. We discuss this interpretation by using a full-band Monte Carlo model based on first-principles electronic structure and lattice dynamics calculations. The results of our analysis suggest that Auger-excited electrons cannot be unambiguously detected in the LED structures used in the EES experiments. Additional experimental and simulative work are necessary to unravel the complex physics of GaN cesiated surfaces. [ABSTRACT FROM AUTHOR]

What has turned into highly complex and somewhat misunderstood efficiency loss mechanisms occurring in light-emitting diodes (LEDs) based on the InN-GaN material system at high injection levels are discussed. Suggestions are made as to the dominant mechanism (s) in an open forum format as well as pointing out some of the shortcomings of the methodologies used and premises forwarded. It is unequivocally known that increased junction temperature would cause a reduction in radiative power due to mainly the reduction in radiative recombination efficiency. Another obvious mechanism is the asymmetry in doping in wide bandgap semiconductors, such as GaN, wherein the hole concentration lags well behind that of electrons in the active region. Because an electron and a hole are required for radiative recombination, the radiative efficiency cannot keep up with increasing carrier injection due to progressively lagging hole population. This results in either electron escape without radiative recombination or electron accumulation, which in turn changes the internal bias of the device, manifested as reduced internal forward bias, which reduces the rate of increase in light intensity. Some of the reports ascribe the efficiency loss at high injection levels to Auger recombination (mainly through indirect and recently reportedly direct deductions) as the main and or the only source of efficiency loss by in many cases simply relying on the temperature and injection independent (not well taken) A, B, C coefficients to fit a third order polynomial to the efficiency vs. injection current. As for the direct deduction, the spectroscopic analysis of Auger kicked hot electrons as they traverse through the G and L bands before being emitted into the vacuum by means of cesiated surface challenges the existing theories and some experiments regarding carrier scattering and G-L separation. Despite just a few reports to the contrary, the bulk of the resonant optical emission experiments do not support the Auger argument as being the main cause. In parallel, there exists a body of theoretical and experimental reports for electron overflow of ballistic/quasi-ballistic electrons traversing the active region to p-GaN, escaping recombination altogether in the active region. In fact and from the get go, the LED industry ubiquitously employed, and continues to do so, an (Al,In)GaN electron-blocking layer (EBL) to prevent electron escape for improved light output that in and of itself would more than suggest that the electron escape (overflow) does indeed occur. The only adverse effect of EBL is that it impedes hole injection due to the valence band offset between the p-type (Al,In)GaN EBL and p-GaN and also generates piezoelectric (if not lattice matched) and differential spontaneous polarization induced fields that pull down the conduction band edge at the interface reducing EBL's effectiveness. To at least reduce the aforementioned aggravating factors to some extent, the electron overflow and the associated efficiency loss can be reduced substantially (particulars of which depend on the active layer design) by inserting a stair-case electron injector (SEI) with a step-like indium composition to act as an "electron cooler" or by linearly graded cooler in some form or another prior to the active region. Use of multiple layered heterostructures for the active layers also plays the role of electron cooler, albeit not necessarily in the most optimum fashion. As if oblivious to the raging issues alluded to above, the LED industry has been moving along very successfully with a 2-prong approach. In one, dubbed the "high voltage LED", a set of LEDs (most likely configured in the form of a full-wave bridge rectifier) operating at low currents, where the efficiency is at its maximum, is used, which also has the added benefit of much reduced power supply complexity and weight. The other is the continual improvement of layer quality and optimum active layer design, taking technological parameters into consideration, which at the time of writing sported 63% wall-plug efficiency at an injection current of 350 mA for ~1.1 mm x 1.1 mm LEDs, which translates to approximately 75% efficiency once the voltage (about 3 V) and phosphor conversion efficiencies are taken into account. Assuming the same extraction and external quantum efficiencies, one obtains about 86% for each. It would not be an exaggeration to conclude that mid 90% internal quantum efficiencies are very likely in play here, which means that the purported inherent problems dominating the discussion are practically reduced to an academic exercise. Despite aggravating factors involving holes and hetero-barriers, commercial LEDs almost invariably use multilayer heterostructure designs for the active layers, loosely termed as multiple quantum wells, presumably to circumvent technological challenges and the ramifications of the polarization induced field even though double heterojunction varieties could be preferable from the point of view of hole transport. Eventually, the real limiting factors are the extent of hole supply (can be mitigated by increased hole concentration, which is well known and continually explored) and the proverbial thermal wall (can be mitigated by increased efficiency and efficient heat removal) of course. [ABSTRACT FROM AUTHOR]

By systematically comparing one-, two-, and three-dimensional (1D, 2D, 3D) simulations of a single-color HgCdTe n-on- p photodiode, we discuss nontrivial differences between 2D and 3D results and observe the need for at least a 2D description in order to get qualitatively correct photocurrent spectra. We show that the photoresponse of a typical dual-band mid-wavelength/long-wavelength infrared pixel can be investigated also in 1D under certain conditions, and we assess the limitations of 1D and 2D modeling against reference 3D simulations. [ABSTRACT FROM AUTHOR]

Electroluminescence (EL) characterization of InGaN/GaN light-emitting diodes (LEDs), coupled with numerical device models of different sophistication, is routinely adopted not only to establish correlations between device efficiency and structural features, but also to make inferences about the loss mechanisms responsible for LED efficiency droop at high driving currents. The limits of this investigative approach are discussed here in a case study based on a comprehensive set of currentand temperature-dependent EL data from blue LEDs with low and high densities of threading dislocations (TDs). First, the effects limiting the applicability of simpler (closed-form and/or one-dimensional) classes of models are addressed, like lateral current crowding, vertical carrier distribution nonuniformity, and interband transition broadening. Then, the major sources of uncertainty affecting state-ofthe- art numerical device simulation are reviewed and discussed, including (i) the approximations in the transport description through the multi-quantum-well active region, (ii) the alternative valence band parametrizations proposed to calculate the spontaneous emission rate, (iii) the difficulties in defining the Auger coefficients due to inadequacies in the microscopic quantum well description and the possible presence of extra, non-Auger high-current-density recombination mechanisms and/or Auger-induced leakage. In the case of the present LED structures, the application of three-dimensional numerical-simulation-based analysis to the EL data leads to an explanation of efficiency droop in terms of TD-related and Auger-like nonradiative losses, with a C coefficient in the 10-30 cm6/s range at room temperature, close to the larger theoretical calculations reported so far. However, a study of the combined effects of structural and model uncertainties suggests that the C values thus determined could be overestimated by about an order of magnitude. This preliminary attempt at uncertainty quantification confirms, beyond the present case, the need for an improved description of carrier transport and microscopic radiative and nonradiative recombination mechanisms in device-level LED numerical models. [ABSTRACT FROM AUTHOR]

In this paper, we describe the state of the art in the numerical simulation of the carrier transport properties of GaN and its ternary alloys. We outline the characteristics of our state of the art full-band Monte Carlo model that includes a fitting parameter free approach to compute the carrier-phonon interaction and a full quantum mechanical model for multiband transport which is critical to understand the high-field transport properties of these materials. Finally, we provide several examples of applications of the model to the calculation of the low-field electron mobility of GaN and In0.18Al0.82N, drift velocity in GaN and the impact ionization coefficients in AlxGa1-xN alloys. [ABSTRACT FROM AUTHOR]

A microscopic model, based on a full-Brillouin-zone description of the electronic structure, is used to investigate Auger transitions in InGaN/GaN quantum wells. The lack of momentum conservation along the confining direction enhances direct (i.e., phononless) Auger transitions, leading to Auger coefficients in the range of those predicted for phonon-dressed processes in bulk InGaN. The dependence of Auger coefficients on temperature and quantum well thickness is analyzed. The limitations of conventional multiband models, based on zone-center approximations of the band structure, are discussed. [ABSTRACT FROM AUTHOR]

Band-to-band Auger recombination mechanisms in HgCdTe are investigated as functions of temperature in the small modulation limit by using realistic electronic structures obtained by empirical pseudopotential calculations and their corresponding wavevector-dependent dielectric functions. The calculated Auger lifetimes are compared with the semiempirical Beattie-Landsberg-Blakemore (BLB) model, which has been extensively employed to reproduce experimental data. The Auger-1 lifetime can be fitted well to the BLB model with a constant overlap integral | F F| = 0.16, near the lower limit of the range reported in the literature. The role of the Auger-7 process in p-type HgCdTe is also investigated and the ratio γ between the intrinsic Auger-7 and Auger-1 lifetimes is found to be about 10. [ABSTRACT FROM AUTHOR]

2-D numerical simulation is employed to assess a number of possible design approaches aimed at optimizing the internal quantum efficiency (IQE) of ZnO-based light-emitting diodes (LEDs) grown along the c-axis. First, the relative performance of similar ZnO-based and GaN-based LED structures is compared and discussed. Second, the effects on IQE of thickness, doping, and alloy composition of the MgZnO electron blocking layer (EBL) is studied in order to maximize the carrier confinement in the active region. The optimum number of quantum wells is also addressed, and different strategies for barrier doping are considered, showing that, if the EBL is doped p-type, a similar doping in the barriers is not required to compensate for the spontaneous and piezoelectric interface charges and to enhance hole transport. Different choices of the geometrical and doping parameters of the n-type access region are considered, and the impact of different values of the electron mobility is determined. Finally, the analysis of a ZnO/BeZnO LED structure suggests that the incorporation of BeZnO layers does not provide significant advantages. [ABSTRACT FROM PUBLISHER]

As a preparatory step toward establishing reliable numerical design tools for ZnO-based optoelectronic devices, we have reassessed the available information on material parameters relevant for the simulation of light-emitting diodes (LEDs) with active regions including ZnO, MgZnO, and BeZnO layers. The impact of different approximations for the electronic structure and the interface polarization charge on the optical properties of bulk ZnO and ZnO/MgZnO quantum wells has been evaluated, and a consistent set of parameters has been used not only for systematic comparison of ZnO/MgZnO and ZnO/BeZnO single quantum well structures but also for the first simulation of a realistic ZnO/BeZnO multiple quantum well LED. [ABSTRACT FROM AUTHOR]

A full-band Monte Carlo model has been developed for understanding the carrier multiplication process in HgCdTe infrared avalanche photodiodes. The proposed model is based on a realistic electronic structure obtained by pseudopotential calculations and a phonon dispersion relation determined by ab initio techniques. The calculated carrier–phonon scattering rates are consistent with the electronic structure and the phonon dispersion relation, thus removing adjustable parameters such as deformation potential coefficients. The computation of the impact ionization transition rate is based on the calculated electronic structure and the corresponding wavevector-dependent dielectric function. The Monte Carlo model is applied to investigate key performance figures of long-wavelength infrared (LWIR) and mid-wavelength infrared (MWIR) HgCdTe avalanche photodetectors such as carrier multiplication and noise properties. Good agreement is achieved between simulations and experimental results. The multiplication process in LWIR ( λc = 9.0 μm at 80 K) and MWIR ( λc = 5.1 μm at 80 K) devices is found to be initiated only by electrons, as expected from excess noise measurements. This single-carrier multiplication behavior can be traced back to the details of the computed valence-band structure and phonon scattering rates. [ABSTRACT FROM AUTHOR]

Two alternative approximations of the electronic structure of CdTe and HgTe are proposed, both suited to the needs of accuracy and numerical efficiency of full-band carrier transport simulation: a local empirical pseudopotential (EPM) parametrization including relativistic corrections, and an original full-Brillouin-zone (FBZ) $${{\user2 {k}}\cdot {\user2 {p}}}$$ model using two expansion points (Γ and W). The EPM and $${{\user2 {k}}\cdot {\user2 {p}}}$$ band structures closely match the available experimental and ab initio information, complemented with the results of new density functional theory (DFT)-local density approximation (LDA) calculations, for the conduction and valence bands relevant in transport phenomena. The EPM description of the binary compounds, featuring transferable Te pseudopotentials, is the basis for a computation of the electronic structure of the ternary alloy Hg1− xCd xTe in the framework of disorder-corrected virtual crystal approximation. The composition dependence of energy gaps, effective masses, and high-frequency dielectric constants are discussed and compared with available experimental data, and the novel FBZ approach is applied to the case of x = 0.7. [ABSTRACT FROM AUTHOR]

The low-field electron mobility in bulk ZnO has been the object of extensive experimental studies, mainly through measurement of the temperature-dependent Hall effect. In this work, we reassess the experimental results through direct simulations of Hall measurements, performed with a Monte Carlo transport model and taking into account all the major scattering mechanisms. The deformation potentials required to compute acoustic and optical phonon scattering are derived from first-principles computations, and an original theory for charged-dislocation-line scattering is proposed and implemented. Monte Carlo results for the electron mobility and the Hall factor are compared with analytical expressions derived with the relaxation-time approximation, which is found to be adequate at low temperatures where inelastic scattering effects due to optical phonons are negligible. [ABSTRACT FROM AUTHOR]

Ab initio computations of the structural and electronic properties of wurtzite BeO have been performed with the codes of the abinit project, and the resulting band structure has been approximated with the nonlocal empirical pseudopotential method and a new full-zone $${k \cdot p}$$ model using two expansion points (Γ and L). The very good overall quality demonstrated by both empirical approaches should allow their application to the accurate yet numerically efficient description of the electronic structure of the BeZnO materials system. [ABSTRACT FROM AUTHOR]

Electron and hole transport in wurtzite phase ZnO is studied using an ensemble full band Monte Carlo method. The model includes an accurate description of the electronic structure obtained with the nonlocal pseudopotential method and numerically calculated impact ionization transition rates based on a wavevector-dependent dielectric function. Results of transport simulations at both low and high electric fields are presented. It is found that the low field electron mobility is close to 300 cm² V-1 s-1 at room temperature, and the peak electron drift velocity is 2.2 x 107 cm/s at a field of 275 kV/cm. The determination of the ionization coefficients is affected by some uncertainties due to the incomplete knowledge of the high energy phonon scattering rates. Nevertheless, the present calculations of the ionization coefficients provide a reasonably accurate estimate of the impact ionization process. [ABSTRACT FROM AUTHOR]

An efficient finite-element reduced-order quasi-TEM model for the frequency-dependent characteristics of lossy multiconductor transmission lines is presented. Conductor losses are evaluated as functions of frequency through a magneto-quasi-static model. Numerically generated problem-matched basis functions reduce the problem size and, therefore, the CPU time required by frequency sweeps without appreciable loss of accuracy. The proposed approach is applied to complex coplanar waveguides and to multiconductor interconnects; its results are compared with quasianalytical techniques and with the full-wave finite-element method. [ABSTRACT FROM AUTHOR]

A theoretical justification is presented of the partial-capacitance (PC) approach, widely exploited in the modeling of coplanar waveguides on finite-thickness and multilayered substrates. The analysis is based on the static spectral-domain approach to the computation of the capacitance of a set of planar conductors embedded into a multilayered substrate. It is shown that the PC method can be derived from approximating the static Green's function of a multilayered dielectric as the sum of partial contributions; such a decomposition can be applied to the Green's function either in parallel (admittance) or series (impedance) form. The resulting (parallel or series) PC approaches are shown to be accurate with substrates having layers of decreasing or increasing permittivity, respectively. Lines backed by magnetic or electric walls are introduced as limiting cases of multilayered structures. [ABSTRACT FROM AUTHOR]

Presents a class of coaxial broad-band directional couplers for high-power radio frequency testing. Drawbacks of classical stripline and coaxial couplers overcome by the devices; Features of the design technique for the longitudinally nonuniform transversally nonsymmetrical couplers.

Proposes a conformal-mapping design tools for coaxial couplers with complex cross section. Accurate evaluation of per-unit-length capacitances, impedances and coupling factors; Even- and odd-mode characteristic impedance of a dielectric-filled circular coaxial coupler.