physics Related Articles

Overview: The article explores the limitations of silicon-based power devices and highlights gallium nitride as a promising alternative. It discusses the importance of accurate modeling for optimizing circuit designs in power electronics. Power electronics and semiconductor device technologies are vital components for electrical energy conversion to be reliable and efficient. Due to several key advantages, silicon (Si) has been the dominant material for power electronic devices. They are still widely used in power electronics, and their level of development is high. What are the challenges faced by Si-based power devices?Power devices that are primarily Si-based dissipate the largest power losses in a power converter. The challenges faced by Si-based power devices include maximum attainable switching frequency, maximum permissible junction temperature, and maximum reverse voltage blocking capabilities. Si IGBTs, having a high breakdown voltage capability of up to 6.5 kV, are restricted to lower switching rates. Because of their slower switching speeds, large-sized inductive and capacitive elements are employed to manage power. These factors have caused silicon power devices to approach their operational limit regarding performance and efficiency for current power electronic converters. Developing power electronics toward more effective, compact, high-voltage, high-frequency operation requires new developments in power semiconductor devices. Advantages of Gallium NitrideGallium Nitride (GaN), a wide bandgap semiconductor, has gained attention as a possible material for the upcoming generation of power semiconductor devices. GaN has a bandgap of approximately 3.4 eV compared to silicon's 1.2 eV. GaN offers higher carrier density and lower resistance, enabling speeds up to 100 times faster than silicon. GaN has superior thermal conductivity, allowing for better heat dissipation in high-power applications. Importance of Modeling of GaN DevicesA prototype that can simulate the intended performance characteristics of the circuit must be used to develop power electronics circuits. This enables more efficient development of high-performance power electronics systems. Accurate models allow engineers to predict and optimize the performance of GaN devices without extensive physical prototyping. Models help verify designs before fabrication, reducing development time and costs associated with physical prototyping and testing. The semiconductor device model is a crucial component of circuit simulations to predict design performance, offers a structure for design centering and tolerancing, and facilitates troubleshooting. There is a need for a precise compact model of the GaN power devices in the market. Fig. 1 depicts the evolution of various GaN HEMT models. Fig. 1 Evolution of GaN HEMT large-signal models. Source: ScienceDirect Classification of GaN ModelingThe modeling of GaN semiconductor devices can be classified into several main categoriesEmpirical ModelsPhysics-Based ModelsCompact ModelsBehavioral ModelsLarge-Signal ModelsThermal ModelsTrapping ModelsMulti-Physics ModelsThis article provides an overview of both empirical and physics-based models. Empirical ModelsEmpirical models were the typical models to emerge first, mostly because they characterize HEMT performances using relatively simple expressions. Examples includeAngelov modelDynaFET modelEEHEMT modelStatz and Crutice models These models are based on hyperbolic tangent functions. The accuracy, user-friendliness, and high degree of adjustable nature of the empirical models make them extremely suitable for a wide range of industrial applications. Angelov ModelIn recent years, a number of GaN-based device models have become available. For RF HEMT devices, as well as other Si and SiC devices, one of the common models is the Angelov model, which has been employed extensively. Numerous updated Angelov models have emerged in recent years to address the evolving requirements of various new application scenarios. The Angelov model is a widely used large-signal model for modeling GaN HEMTs and other field-effect transistors. Uses exponential and hyperbolic tangent functions to model I-V characteristics. Designed to accurately capture the nonlinear behavior of FETs, especially for RF/microwave applications. However, because the model is entirely empirical and has been in widespread use for a long time, it may be applied to any semiconductor device. DynaFET ModelArtificial neural network (ANN)--based GaN HEMT models are commonly employed in machine learning (ML)-based GaN HEMT models. Naturally, the ANN-based models can effectively describe a device's nonlinearity with a low computing cost because of their great fitting ability and reasonable deployment cost. A common example of an ANN-based model is Keysight's DynaFET model. Physics-Based ModelsEfforts were directed toward developing GaN HEMT models based on the underlying physical characteristics. These models are grounded in the fundamental physics of the device. These models show a far higher scalability than empirical models. Physics-based models includeMIT Virtual Source (MVS) HEMT modelAdvanced SPICE Model (ASM) HEMT modelHiroshima-University Starc Igfet Model for GaN HEMT (HiSIM GaN HEMT model)École Polytechnique Fédérale de Lausanne (EPFL) HEMT model Rigid physical equations serve as the foundation for physical models. Physical models are more scalable, compact, and robust than empirical models. MIT GaN ModelMIT GaN model, also known as the MIT Virtual Source GaNFET-High Voltage (MVSG-HV) model, is a physics-based compact model developed for GaN HEMTs. This model has been validated for an EPC GaN device and for depletion-mode RF devices. Additionally, the model is able to simulate charge trapping and radiofrequency noise under high-frequency RF power amplifier operation. However, the model has proven to be reliable and capable of good convergence on depletion-mode RF GaN devices. ASM-HEMT ModelThe ASM-HEMT (Advanced SPICE Model for High Electron Mobility Transistors) is a compact model developed specifically for GaN HEMTs. It is a physics-based compact model for GaN HEMTs designed for accurate circuit simulation of GaN devices. It has a surface potential-based approach for improved scalability and physical accuracy. It is based on using Schrodinger's and Poisson's equations to formulate the surface potential in the channel. This model has been proven to have both small- and large-signal characteristics appropriate for RF devices. It is also applicable to power electronics, and it is highly suitable for RF GaN applications. In conclusion, the GaN HEMT large-signal models face numerous obstacles in the 2020s and beyond. Advances in semiconductor technologies are the source of some of them. The lack of physical meaning in empirical models results in poor scalability. Also, physical models may be less widely applicable due to their complexity and lack of tunability. Artificial intelligence and measurement technology are changing quickly, so it will be interesting to see if they can lead to new models. Summarizing the Key PointsGallium nitride offers significant advantages over silicon, including higher efficiency, faster switching speeds, and superior thermal conductivity, making it ideal for modern power electronics applications.Accurate modeling of GaN HEMTs is crucial for optimizing circuit designs, reducing development costs, and enhancing performance, enabling engineers to predict device behavior without extensive prototyping.Various modeling approaches exist, including empirical and physics-based models, each with unique benefits,understanding these can help select the proper model for specific applications. ReferenceKotecha, Ramchandra M., Md Maksudul Hossain, Arman Ur Rashid, Asif Imran Emon, Yuzhi Zhang, and H. Alan Mantooth. “Compact Modeling of High-Voltage Gallium Nitride Power Semiconductor Devices for Advanced Power Electronics Design.” IEEE Open Journal of Power Electronics 2 (January 1, 2021): 75–87. https://doi.org/10.1109/ojpel.2021.3055531.Luo, Haorui, Wenrui Hu, and Yongxin Guo. “On Large-Signal Modeling of GaN HEMTs: Past, Development and Future.” Chip 2, no. 3 (September 1, 2023): 100052. https://doi.org/10.1016/j.chip.2023.100052.

SummaryIron-air batteries promise a considerably higher energy density than present-day lithium-ion batteries. In addition, their main constituent -- iron -- is an abundant and therefore cheap material. Scientists from Forschungszentrum Jülich are among the driving forces in the renewed research into this concept, which was discovered in the 1970s. Together with American Oak Ridge National Laboratory (ORNL), they successfully observed with nanometre precision how deposits form at the iron electrode during operation. A deeper understanding of the charging and discharging reactions is viewed as the key for the further development of this type of battery rechargeable to market maturity. The results were published in the journal Nano Energy--Charging and discharging reactions during operation shown with nanometer precision.  BodyFor reasons including insurmoutable technical difficulties,research into metal-air batteries was abandoned in the 1980s for a long time.The past few years, however, have seen a rapid increase in research interest. Iron-air batteries draw their energy from a reaction of iron with oxygen. In this process, the iron oxidizes almost exactly as it would during the rusting process. The oxygen required for the reaction can be drawn from the surrounding air so that it does not need to be stored in the battery. These material savings are the reason for the high energy densities achieved by metal-air batteries. Iron-air batteries are predicted to have theoretical energy densities of more than 1,200 Wh/kg. By comparison, present-day lithium-ion batteries come in at about 600 Wh/kg, and even less (350 Wh/kg) if the weight of the cell casing is taken into account. Lithium-air batteries, which are technically considerably more difficult and complicated to realize, can have energy densities of up to 11,400 Wh/kg. When it comes to volumetric energy density, iron-air batteries perform even better: at 9,700 Wh/l, it is almost five times as high as that of today's lithium-ion batteries (2,000 Wh/l). Even lithium-air batteries have "only" 6,000 Wh/l. Iron-air batteries are thus particularly interesting for a multitude of mobile applications in which space requirements play a large role. "We consciously concentrate on research into battery types made of materials that are abundant in the Earth's crust and produced in large quantities," explains institute head Prof. Rüdiger-A. Eichel. "Supply shortages are thus not to be expected. The concept is also associated with a cost advantage, which can be directly applied to the battery, particularly for large-scale applications such as stationary devices for the stabilization of the electricity grid or electromobility." What Cause These Difficulties?The insights obtained by the Jülich researchers create a new basis for improving the properties of the battery in a targeted manner. Using in situ electrochemical atomic force microscopes at the Center for Nanophase Materials Sciences at Oak Ridge National Laboratory, they were able to observe how deposits of iron hydroxide particles (Fe(OH)2) form at the iron electrode under conditions similar to those prevalent during charging and discharging. "The high pH of 13.7 alone represents a borderline condition for the instrument," explains Henning Weinrich from Jülich's Institute of Energy and Climate Research (IEK-9). "We were the first at Oak Ridge to successfully conduct such an experiment under realistic conditions," says Weinrich, who stayed in the USA for three months especially for the measurements. Capacity IncreasingWe should notice that deposits do not decrease the power of the battery.On the contrary, since the nanoporous layer increases the active surface area of the electrode, it contributes to a small increase in capacity after each charging and discharging cycle. Thanks to the investigations, the researchers have for the first time obtained a complete picture of this layer growth. "It was previously assumed that the deposition is reversed during charging. But this is obviously not the case," explains Dr. Hermann Tempel from Jülich's Institute of Energy and Climate Research (IEK-9). Furthermore, a direct link was verified for the first time between the layer formation at the electrode surface and the electrochemical reactions. There is, however, still a long way to go until market maturity. Although isolated electrodes made of iron can be operated without major power losses for several thousand cycles in laboratory experiments, complete iron-air batteries, which use an air electrode as the opposite pole, have only lasted 20 to 30 cycles so far.