CHAPTER 2 HEMT DEVICES AND BACKGROUND

CHAPTER 2 HEMT DEVICES AND BACKGROUND 2.1 Overview While the most widespread application of GaN-based devices is in the fabrication of blue and UV LEDs, the fabrication of microwave power devices has attracted much attention because of large potential markets. The world market for transistors is divided between bipolar and field effect transistors, depending on application. For GaN applications there has been research into both Heterojunction Bipolar Transistors (HBTs) and High Electron Mobility Transistors (HEMTs). State-of-the-art AlGaN/GaN HBTs generally still offer poor performance and typically have current gains of ~10. Makimoto et al., have achieved promising results with GaN/InGaN Double Heterojunction Bipolar Transistors (DHBT) having current gains greater than 2000 and breakdown voltages above 50V. However, the progress made in HBTs has been very slow owing mostly to the poor acceptor activation in the base layer. As a result, most research has focussed on field effect transistors with both Metal Semiconductor Field Effect Transistors (MESFETs) and HEMTs receiving considerable attention. In this dissertation, the work has focussed on AlGaN/GaN HEMTs and this chapter will discuss the theory behind the operation of HEMTs, 2 Dimensional Electron Gas (2DEG) formation, various effects in HEMTs and different techniques involved in HEMT breakdown voltage improvement. 2.2 HEMT Structure The HEMT is a heterojunction device with superior performance to its homojunction counterpart, the MESFET. The principle of operation in a HEMT is very similar to Metal Insulator Semiconductor Field Effect Transistor 10

(MISFET). However, instead of carrying current in a thick channel, a HEMT relies on the formation of a two dimensional electron gas at the heterojunction interface. A typical cross-sectional schematic of AlGaN/GaN HEMT device is shown in Figure 2.1. The device is usually grown on a semi-insulating substrate which has a high thermal stability and close lattice matching with GaN. A buffer layer is grown on top of the substrate to act as an isolation layer between the substrate and channel. Any lattice mismatching or crystal defects from the substrate are minimised using this GaN buffer layer. The device usually uses a schottky gate contact and ohmic source and drain contact. The channel in a HEMT is formed at the heterojunction interface of the AlGaN barrier and GaN channel layer. The following section describes more about the operation of HEMT. Source - Ohmic Gate - Schottky Drain - Ohmic AlxGa(1-x)N Barrier UID GaN Channel 2DEG Buffer Layer Semi Insulating Substrate Figure 2.1. Layer structure of a typical AlxGa(1-x)N/GaN HEMT. 11

2.3 HEMT Operation In the most common HEMT structures, the wide bandgap barrier is doped n-type while the narrow bandgap channel remains undoped. As a result, electrons diffuse from the wide bandgap material into the narrow bandgap material to minimise their energy. This process continues until a balanced Fermi level is formed in the two materials and equilibrium is established. Because of the resulting electrostatics, a new triangular well forms on the narrow bandgap side of the heterojunction. Which we call it as two dimensional quantum well and the electrons confined inside the well is called Two Dimensional Electron Gas (2DEG). The n-doped barrier in the device supplies electrons to the undoped channel, thus spatially separating the channel charge carriers from their ionised donors. In this manner, the heterostructure channel is capable of delivering high carrier concentration with high mobility as impurity scattering is minimised in the undoped channel. As an added advantage, surface scattering is also reduced by moving the current-carrying region below the barrier. To understand the principle of operation and techniques, the formation of 2DEG and different effects involved in the HEMT are explained in the following sections. 2.3.1 Two Dimensional Electron Gas (2DEG) in HEMT A schematic band diagram of a modulation doped heterostructure is shown in Figure 2.2. It consists of a wide gap semiconductor (AlGaN) and a semiconductor with narrower gap (GaN). At the interface a triangular quantum well is formed in the undoped narrow gap material. Electrons from the wider band gap material fall into this potential well and are confined within the well. Because of such quantum mechanical confinement in a very narrow dimension, they form a high density of electron gas in two dimensions. Electrons can move freely within the plane of the heterointerface, while the motion in the direction perpendicular to the heterointerface is restricted to a well-defined space region by energy, momentum, and wave function quantization, thus forms the so-called 12

2DEG. As the narrow gap material is undoped and these electrons are away from the interface, the electron mobility can be simultaneously increased with high concentration of carriers. Figure 2.3 shows the polarization vectors in the AlGaN and underlying GaN. Within the AlGaN layer there are two polarization vectors Ppe, AlGaN and Psp, AlGaN for the piezoelectric and spontaneous polarizations respectively. The polarization in the AlGaN causes dipole charges to form at the borders of the material with a negative sheet charge at the surface and an equal positive sheet charge at the AlGaN/GaN junction. The polarization in the GaN layer causes a negative sheet charge at the AlGaN/GaN junction and an equal positive sheet charge on the bottom surface. Since the total polarization in the AlGaN is larger, the overall result is a net positive sheet charge at the AlGaN/GaN interface. Figure 2.2 An idealised AlGaN/GaN heterostructure 2DEG formation Ppe, AlGaN + Psp, AlGaN Psp, GaN Figure 2.3 Spontaneous and piezoelectric polarization vectors in AlGaN and GaN. 13

Notice that the bottom GaN and AlGaN/GaN interfaces are both positive while the charge density at the top AlGaN interface is negative. The critical point to remember is that the interface sheet charges here are not free carriers. They are induced charges as in a typically polarised dielectric. However, it is the presence of these charges and the polarization induced electric field in the AlGaN which allows for 2DEG formation without barrier doping. 2.3.2 Different types of Effects in HEMT There are different effects present in the operation of HEMT devices such as Polarization effects and Trapping Effects. The following section describes more details about various effects in the HEMT devices. 2.3.2.1 Polarization in AlGaN/GaN Heterostructures It is now widely recognized that built in electric fields due to polarization induced charges play an important role in the electrical and optical properties of Nitride heterostructures grown in [0001] orientation (Parvesh Gangwani et al., 2007). These fields also provide a source of a 2-dimensional electron gas in AlGaN/GaN heterostructures. In HFETs, the understanding and controlling of the source of electrons is important for the optimization of their performance. The polarization effect in HFETs are well explained by Jia L., et al (2001). Spontaneous Polarization: Al-N and Ga-N bonds are highly ionic and each carries a strong dipole. For example, because the electronegativity of N is much higher than that of Ga, the electron wave function around the Ga-N pair is offset to the nitrogen side. The effect is even more exaggerated in the Al-N pair. This is a special feature of the III-nitrides as the degree of spontaneous polarization is more than five times greater than in most III-V semiconductors. Figure 2.4 shows GaN grown in the Ga-face and N-face orientation which is the norm for high performance AlGaN/GaN heterostructures. The c-axis polarization vector points from the nitrogen to the Gallium, as indicated, and creates an internal electric field pointing in the opposite direction, this is referred 14

to as spontaneous polarization. In an AlGaN/GaN HEMT both the GaN and AlGaN layers have spontaneous polarization vectors which point in the same direction, from the N to the Ga(Al) towards the substrate (Figure 2.3). Figure 2.4 Ga-face and N-face Wurtzite GaN structure Piezoelectric Polarization: The lattice constants a0 and c0 for GaN are slightly larger than those for AIN. As a result, thin AlGaN layers grown on GaN are tensile strained (the GaN is relaxed due to the thick buffer grown on the chosen substrate). In the nitride system the piezoelectric constants are more than ten times greater than those typical in most III-V semiconductors and this creates very large polarizations. In Ga-faced material under tensile strain the piezoelectric polarization due to the deformation of the AlGaN layer points parallel to the spontaneous polarization vectors, e.g. towards the substrate in Figure 2.4 A thorough treatment calculating the size of the spontaneous and piezoelectric polarizations in the AlGaN and GaN systems is given by Ambacher et al. 2.3.2.2 Trapping in Semiconductor Materials AlGaN/GaN HEMTs show strong current slump which is widely considered to be caused by electron trapping. Traps are very often surface related so it is not unreasonable to suppose they may contribute to current slump effects 15

in AlGaN/GaN HEMTs. Aside from surface effects, traps may also be formed by dislocations, point defects or impurities. 2.4 Contacts in a HEMT Besides the electronic properties of the layer structure such as carrier mobility or conductivity of 2DEG, the metal contacts also determine the DC and RF properties of the final device. The quality of the contact is crucial to stable operation of transistor. Ohmic contact have to carry signal with minimal resistance and without rectification. On the other side Schottky contact have to dispone with high barrier height to AlGaN/GaN structure. 2.4.1 Ohmic Contacts The next type of contact present in the device is ohmic contact. In general, an ohmic contact is referred to a non-injecting contact in which the currentvoltage relationship under both reverse- and forward-bias condition is linear and symmetrical. However, in reality, a contact is considered ohmic if the voltage drop across the metal semiconductor interface is negligible compared to the voltage drop across the bulk semiconductor. It is difficult to make ohmic contact to wide-gap semiconductor (e.g. III-group nitrides), because it does not generally exist in metal with low-enough work function to yield a low barrier. Therefore the practical way to obtain a low resistance ohmic contact is to increase the doping level near the metal-semiconductor interface to very high level. So in some cases a highly doped GaN layer is placed at the top of AlGaN/GaN heterostructure in effort to lower the barrier. Ohmic contacts allow current to pass into and out of the underlying semiconductor with ease. Formation of ohmic contacts with low resistivity is critical for optimal device performance. The cut-off frequency of a FET is strongly determined by the transconductance, gm, of the device. Ft= gm/2πcg (2.1) 16

The transconductance in equation (2.1) is the extrinsic transconductance, that is, the overall transconductance offered by the device, including the parasitic ohmic contacts. The transconductance, and thus ft, in equation (2.1) can be increased by reducing the resistivity of the ohmic contacts. gm= gmo/(1+rsgmo) (2.2) Equation (2.2) shows how the extrinsic transconductance is related to the intrinsic transconductance gmo. The value of gmo is determined by the layer quality and basic gate geometry and represents the maximum transconductance value that is attainable for the device. Rs, the source resistance, is the resistance between the top of the source ohmic contact, through the semiconductor, to the gate edge. Equation (2.2) expresses how the source resistance is a parasitic circuit element and the reduction of these parasitic elements results in improved device performance. Formation of a good ohmic contact ideally requires choosing a metal with a very small work function. 2.4.2 Schottky Contacts Schottky contacts is one of the important building block for HEMT devices. The device formed using schottky contact may have great impact in the device performance. The critical elements at play for the Schottky contacts are diode idealities close to unity, low gate leakage current to allow for effective channel modulation, and small gate length to improve cut-off frequency. Ideally, fabrication of quality Schottky contacts on n-type material requires the use of metals with large work functions. Besides, the metalsemiconductor junction must be thermally stable and the gate metal must also be highly conductive to minimise the gate resistance. The energy barrier height ɸB is a key parameter of the junction, controlling both the width of depletion region in the semiconductor and the electron current across the interface. Barrier height is defined as the energy difference between 17

the semiconductor conduction-band edge of the interface and the Fermi level in the metal. For undoped semiconductor the electrons cross through the barrier mainly by passing over the barrier, by thermionic emission. In the case of doped heterostructure, electrons also tunnel through the barrier at some elevated energies, by thermionic field emission. In the case of very high doping the depletion region is very thin and electrons tunnel through the barrier. The IV characteristic gets linear, when the resistance is low and the contact becomes ohmic. The value of barrier height depends on the difference between the electron affinities for the metal and the semiconductor. This is more or less only a theoretical case. In reality the deposition of metal to semiconductor gives a large number of interface states at the metal-semiconductor interface. High interface state density causes that Fermi level is pinned at certain level in the energy gap. Then the calculation of Schottky barrier requires detailed information on the distribution of interface states. The Schottky- barrier height is normally determined from experimental current voltage characteristics. Schottky Contact in HEMT Devices: High frequency RF device requires a perfect contact to work in close synchronisation with the high mobility 2DEG channel. A Schottky Source/Drain (SSD) HEMT reduces the leakage current and increases the on current of the device, the reduced leakage current simultaneously increases the breakdown voltage of the device. The Schottky Contact technology can be adopted to achieve a high improvement in the Offstate breakdown voltage (BV) of the lattice-matched In0.17Al0.83N/GaN HEMT. The Schottky-Source/Drain and Schottky-source (SS) AlGaN/GaN HEMTs are used. A Schottky-barrier normally off InAlN-based HEMT is another kind of HEMT device, which finds more application where high frequency is required. High Off-state breakdown, low gate leakage current and high frequency are the main advantage in these kind of device. For the fabrication of the device 18

structure, Metal Organic Chemical Vapor Deposition (MOCVD) is used. To create a negative polarization in the junction 1-nm InAlN/1-nm AlN barrier stack is capped with a 2-nm-thick undoped GaN. Negative polarization charge at the GaN/InAlN junction depletes the channel below the gate and reduces the gate leakage. After removing the GaN cap at access regions, electrons populate the channel. The relationship between Schottky gate leakage current and the breakdown voltage of AlGaN/GaN HEMT is based on the surface defect charge model. The leakage current caused by the positive charge in the surface portion of AlGaN layer induced by process damage such as nitrogen vacancies are represented in this model. To effectively suppress this surface charge influence, a field plated structure is effective. The gate leakage current increases with the defect charge due to thinning of the Schottky barrier and the field plated device structure reduces the electric field concentration at the gate edge. This will effectively increase the device breakdown voltage. To suppress the surface damage at the AlGaN layer, a low-damage fabrication process is most important point regarding development of the AlGaN/GaN-HEMT for the power electronics application due to low leakage current and high breakdown voltage. To further enhance the power performance of the device, the drain region design should be optimized. 2.5 Proposed Work The main objective of this research is to improve the breakdown voltage characteristics of the GaN based HEMT device. In order to achieve this we have proposed 3 different breakdown voltage improvement techniques. Figure 2.5 shows the research flow diagram followed in this thesis. The three different technique is designed and simulated using Sentauraus TCAD simulator and comparison has been made to analyse the performance and characteristics of each technique. The following section explains more details about the current high power HEMT techniques. 19

Gallium Nitride (GaN) Based High Power High Electron Mobility Transistor (HEMT) Breakdown voltage improvement techniques in AlGaN/GaN HEMT Schottky Source drain contact AlGaN/GaN HEMT High-K Passivation Layer Technique Gate Field Plated Engineering Simulation and Analysis of HEMT Device Comparative analysis of all the three technique to identify the technique relevant application Figure 2.5 Research flow diagram 2.5.1 Gallium Nitride (GaN) Based High Power HEMT HEMT was originally developed for high frequency RF application due to its outstanding performance in high frequency operation. The development of high quality GaN epitaxial films gave an extra mileage for the HEMT device in high power application. Nowadays it finds application where high frequency and high power are demanded, such as satellite communication and space applications. 20

2.5.2 Breakdown voltage improvement techniques in AlGaN/GaN HEMT Breakdown is an important problem for high power HEMT devices. In order to tackle breakdown issue of HEMT many investigations are carried out to understand its physical origin. Some of the relevant breakdown mechanism in HEMT devices are, Source-drain breakdown due to short channel effect Hetero structure fabrication defects and trap related problem. Impact ionisation mechanism, sudden increase in drain current due to carrier collapse. High gate electric field, due to the schottky contact leakage current or through surface related conduction. There are different techniques available to improve the Breakdown voltage of a transistor, some of the common techniques and its review is given. Schottky source/drain contact HEMT Field plated HEMT Passivation layer in HEMT 2.5.3 Schottky Source/Drain Contact technique in HEMT In this type of power HEMT, schottky contact is utilized on both source and drain contact to suppress the ON-Resistance (Ron) and leakage current. Usually Ni/Au Schottky contacts are used for contact in the source and drain contact. To achieve a high breakdown voltage it is necessary to have a reduced leakage current and Ron. The Schottky source/ drain (SSD) HEMT is dominated and induced by the gate leakage and gives a reduced leakage current and Ron, this makes the SSD HEMT a good candidate for high breakdown voltage application. In Schottky contact method, a metal layer is grown over the AlGaN/GaN using various techniques. The Schottky contact at the source region can 21

effectively suppress the source carrier injection since the smooth Schottky metallization produces a more uniform electric-field distribution in GaN buffer. Figure 2.6 shows a GaN/AlGaN schottky source/drain contact HEMT. Source (Ni/Au) Gate Passivation Passivation Al0.15Ga0.85N Undoped Drain (Ni/Au) S.I GaN Buffer Layer Figure 2.6 Source/Drain schottky contact GaN/AlGaN HEMT. 2.5.4 Passivation Layer technique in HEMT Passivation is becoming increasingly important for GaN high electronmobility transistors (HEMTs) as the demand for high power devices and circuits for applications at higher frequencies is getting ever stronger. This is because the field plate, which has been widely adopted in GaN HEMTs for improving power and reliability, would degrade the device gain due to the additional parasitic capacitance that it would introduce. Furthermore, the vertical scaling for thinner gate barrier, needed for keeping up the aspect ratio of devices with reduced gate lengths to be operational at higher frequencies, would make the channel of the GaN HEMT more sensitive to the traps originating from surface states. The resulting trapping effects could lead to severe deterioration in power performance and poor uniformity. These trapping effects are assumed to be associated with surface states created by dangling bonds, threading dislocations accessible at the surface and ions absorbed from ambient environment. In order to reduce this degradation, the HEMT surface is passivated with suitable dielectric materials which is shown in Figure 2.7. The permittivity and 22

thickness of the dielectric material are the key factors for the improvement in the breakdown voltage. It has been found that by increasing the permittivity or by increasing the thickness of the passivation layer, the breakdown voltage also increases because of the weakening of the electric field at the drain edge of the gate. Thus by using high-k material and thick passivation layer, breakdown voltage can be increased. Figure 2.7 Schottky with High-k Passivation in the GaN/AlGaN HEMT. 2.5.5 Field Plate engineering technique in HEMT A noteworthy improvement in the device performance can be obtained by adopting the field plate technique which is shown in Figure 2.8. The main purpose of the field plate is to reshape the electric field distribution in the channel and to reduce its peak value on the drain side of the gate edge. Introducing a field plate in AlGaN/GaN HEMTs reduces the current collapse which in turn increases the power performance. The benefit is an increase of the breakdown voltage and a reduced high-field trapping effect. Overall the power density is increased. Various field plate methods like gate field plate, a floating field plate, multiple field plate, etc have been reported so far. It has been found that by increasing the gate to drain distance and by reducing the field plate length, breakdown voltage increases for a particular structure. 23

Figure 2.8 Schottky with High-k Passivation in the GaN/AlGaN HEMT. In The field plate technique, an additional lithography is done in order to place a metal plate covering the gate and extending to the access region on the drain side. This metal plate is electrically connected to the gate. It tracks the potential of the gate electrode, acting as a quasi-electric field plate.the field distribution at the drain of the gate region splits and reshape the distribution of the electric field, hence the peak electric field reduces and increases the device breakdown. The field-plate length dictates the size of the field-reshaping region. This not only increases device breakdown voltage but also reduces the high-field trapping effect, hence enhancing current-voltage swings at high frequencies. The trade-off of the field plate structure includes addition of the gate-drain capacitance at low voltages and extension of the gate-drain depletion length at high voltages, which reduces gain as well as frequency of the device. 2.5.6 Simulation and Analysis of HEMT Device using TCAD Technology Computer-Aided Design (TCAD) describes the process of developing and optimizing the semiconductor processing technologies of the devices by using computer simulations. The software solves fundamental, 24

physical partial differential equations, such as diffusion and transport equations, to model the structural properties and electrical behavior of a self-designed device structure. Sentaurus TCAD is one such software provided by Synopsys for design and analysis of the devices. Technology CAD (TCAD) tools are widely used today in the semiconductor industry and academic research. TCAD simulation and modelling can be used to predict the device performance and expedite the device development / optimization process for new technology. They can also help us to understand device physics and operation mechanisms. The TCAD tool used in this work is Sentaurus, advanced 1D, 2D, and 3D device simulator capable of simulating the electrical, thermal, and optical characteristics of silicon and compound semiconductor devices from Synopsis. The TCAD simulation work consists of the following steps: Use the Sentaurus structure editor or Sentaurus process tool to virtually fabricate the device Create an appropriate mesh for device simulation; Define simulation parameters and models Perform numerical simulations and get device performance; Generate output plots. The software package provides a selection of tools to perform above calculations which provides structured application of the software with visualization aids to observe the physical properties inside the device structure. The tools which have been used in this thesis include. 2.5.6.1 TCAD Carrier Transport Models Sentaurus device supports several carrier transport models for semiconductors. They can be written in the form of continuity equations, which describe charge conservation. The transport models differ in the expressions used to compute the electron current density and the hole current density. Depending 25

on the device under investigation and the level of modelling accuracy required, you can select four different transport models: Drift-diffusion Thermodynamic Hydrodynamic Monte Carlo The main feature of the HEMT is its high carrier mobility and polarization effect. So in order to simulate a high mobility device with polarization effect sentaurus device uses a physics model for piezoelectric polarization. Semi classical transport models such as the Drift-Diffusion (DD) and the Hydro-Dynamic (HD) model cannot be applied directly to the simulation of quantum devices. The Hydrodynamic Density Gradient model (HDG) including the Poisson equation, the electron continuity equation, the quantum potential equation, hot carrier effects, scattering and energy balance equation is used along with polarization model to simulate GaN/AlGaN HEMT devices. Drift-Diffusion Model: The drift-diffusion model is the default carrier transport model in Sentaurus Device. It uses Isothermal simulation, suitable for low-power density devices with long active regions. It is not possible to use the driftdiffusion model for a particular carrier type and to solve the carrier temperature for the same carrier type. Thermodynamic Model: The model differs from drift-diffusion when the lattice temperature equation is solved. It accounts for self-heating. Suitable for devices with low thermal exchange, particularly, high-power density devices with long active regions. Hydrodynamic Model: Accounts for energy transport of the carriers. Suitable for devices with small active regions. This model takes into account the contribution due to the spatial variations of electrostatic potential, electron affinity, and the band gap etc.. 26

Monte Carlo Model: Solves the Boltzmann equation for a full band structure. Piezoelectric Polarization: There are two type of polarization piezoelectric and spontaneous polarization. Sentaurus Device provides two models (strain and stress) to compute polarization effects in GaN devices. They can be activated in the Physics section of the command file as follows: Physics { Piezoelectric_Polarization (strain) Piezoelectric_Polarization (stress) } The piezoelectric polarization vector and the piezoelectric charge can be plotted by: Plot {PE_Polarization/vector PE_Charge } 2.5.6.2 RF Parameter Extraction using TCAD AlGaN/GaN HEMTs receive considerable attention for high-power and high-frequency applications because of the advantages of high breakdown voltage and high electron velocity. High frequency analysis of the HEMT devices are based on the measurement and evaluation of scattering parameters (S-parameters). S-parameters are related to travelling waves that are scattered or reflected when an n-port network is inserted into a transmission line. HEMT is measured as a two-port network (Figure 2.9) where the input port corresponds to gate-source and the output port to drain-source. The various RF parameters of GaN/AlGaN HEMT such as Cut-off Frequency, Maximum Frequency of oscillation, Mason s Unilateral gain, Region of Stability, Polar plot and Smith Chart can be analyzed with the help of this two port network method. In this two-port network, the voltage sources are attached to the gate (port 1) and drain (port 2) terminals and all other terminals are grounded. 27

PORT 1 PORT 2 HEMT Figure 2.9 HEMT in a two-port network S-parameters of two-port network at different conditions are defined in equations (2.3) to (2.6). by, (2.3) (2.4) (2.5) (2.6) where and are incident and reflected waves at input and output of the transistor. S-parameters are measured as the function of the frequency. The is called the input reflection coefficient, the reverse transmission coefficient, the forward transmission coefficient, and the output reflection coefficient. The basic quantities which are important for RF characterization of device can be calculated directly from measured S-parameters. They are: Maximum Available Power Gain MAG: Maximum available power gain expresses maximal power gain by conjugate matching at input and output. 28

Maximum Stable Gain MSG: MSG gives power gain of transistor at the stability border, at K=1. K is called stability factor, speaks about stability of the transistor in frequency range Maximum Stable Gain MSG: MSG gives power gain of transistor at the stability border, at K=1. Unilateral Power Gain GU: Every transistor exhibits a retroactive capacitance from the output toward input. GU considers only the power gain that is obtained by compensation of these retroaction parameter by losses neutralization circuit. For the practical application of transistor the figures of merit are unity current gain frequency ft and unity power gain fmax. Current Gain Cutoff Frequency ft: ft is the frequency at which magnitude of the transistor incremental short-circuit current gain h21 drops to the unity. It is a key estimator of transistor high-speed performance. Mathematically ft is defined in equation (2.7) ft :h21(ft) = 1 (2.7) The Cut-off frequency of the device is plotted as a function of voltage bias for three different extraction methods such as unit-gain-point, extract-at-db point and extract-at-frequency of the two-port network. The Cut-off frequency of the GaN/AlGaN HEMT is given by the equation (2.8), Cut-off frequency, ft = gm/(2π(cgs+cgd)) (2.8) Where, gm =Transconductance of the device Cgs = Gate to Source capacitance Cgd=Gate to Drain Capacitance Maximum Frequency of Oscillation fmax: fmax is defined as the frequency at which GU of transistor drops to the unity. fmax is typically different from ft, because of addition to current gain, fmax takes into account the possibility of 29

voltage gain. Practically, to estimate ft, S-parameters are converted into the h- parameters and current gain is plotted versus frequency. fmax is then estimated by extrapolation of the measured unilateral power gain versus frequency. 2.6 Conclusion In this chapter the HEMT devices and its background are discussed. The device simulation tools and its physical model are also explained in detail. In the coming chapters the design and analysis of various techniques in AlGaN/GaN HEMT devices is discussed. The different techniques used in the design and analysis are schottky source/drain contact technique, different high-k passivation technique and field plate engineering Technique. The next chapter discusses the schottky source/drain contact technique in AlGaN/GaN HEMT device for high power applications. 30