This paper describes the device fabrication process and characteristics of AlGaN/GaN heterostructure field-effect transistors (HFETs) aimed for millimeter-wave applications. We developed three novel techniques to suppress short-channel effects and thereby enhance high-frequency device characteristics: high-Al-composition and thin AlGaN barrier layers, SiN passivation by catalytic chemical vapor deposition, and sub-100-nm Ti-based gates. The Al_0.4Ga_0.6N/GaN HFETs with a gate length of 30nm had a maximum drain current density of 1.6A/mm and a maximum transconductance of 402mS/mm. The use of these techniques led to a current-gain cutoff frequency of 181GHz and a maximum oscillation frequency of 186GHz.

View full abstract

We have fabricated AlGaN/GaN HEMTs with a thin InGaN cap layer to implement normally-off HEMTs with a small extrinsic source resistance. The key idea is to employ the polarization-induced field in the InGaN cap layer, by which the conduction band is raised leading to the normally-off operation. Fabricated HEMT with an In_0.2Ga_0.8N cap layer with a thickness of 5nm showed normally-off operation with a threshold voltage of 0.4V and a maximum transconductance of 85mS/mm for the device with a 1.9-μm-long gate. By etching-off the In_0.2Ga_0.8N cap layer at the region except under the gate using gate and ohmic electrodes as etching masks, the sheet resistance has decreased from 2.7 to 0.75kΩ/_??_, and the maximum transconductance has increased from 85 to 130mS/mm due to a reduction of the extrinsic source resistance. The transconductance was increased from 130 to 145mS/mm by annealing the devices at 250°C for 20 minutes in a N₂ atmosphere.

View full abstract

AlN/GaN Metal Insulator Semiconductor Field Effect Transistors (MISFETs) were designed, simulated and fabricated. DC, S-parameter and power measurements were also performed. Drift-diffusion simulations using DESSIS compared AlN/GaN MISFETs and Al₃₂Ga₆₈N/GaN Heterostructure FETs (HFETs) with the same geometries. The simulation results show the advantages of AlN/GaN MISFETs in terms of higher saturation current, lower gate leakage and higher transconductance than AlGaN/GaN HFETs. First results from fabricated AlN/GaN devices with 1μm gate length and 200μm gate width showed a maximum drain current density of ∼380mA/mm and a peak extrinsic transconductance of 85mS/mm. S-parameter measurements showed that currentgain cutoff frequency (f_T) and maximum oscillation frequency (f_max) were 5.85GHz and 10.57GHz, respectively. Power characteristics were measured at 2GHz and showed output power density of 850mW/mm with 23.8% PAE at V_DS=15V. To the authors knowledge this is the first report of a systematic study of AlN/GaN MISFETs addressing their physical modeling and experimental high-frequency characteristics including the power performance.

View full abstract

We have fabricated enhancement-mode n-channel GaN MOSFETs with overlap gate structure on a p-GaN using thick HfO₂ as a gate insulator. The maximum transconductance of 23mS/mm which is 4 times larger, to our knowledge, than the best-reported value of the normally-off GaN MOSFETs with SiO₂ gate oxide has been obtained.

View full abstract

The mechanism of current collapse of AlGaN/GaN heterojunction field-effect transistors (HFETs) was investigated by gate bias stress with and without illumination. It is clarified that there are two positions where negative charges accumulate, at the gate edge and in the bulk epi-layer. In the gate-edge mode, the charge comes either through the passivation film or the AlGaN layer, depending on the resistance of the films. Reduction of leakage current in the passivation film will be important to suppress the surface-related collapse.

View full abstract

In AlGaN/GaN high electron mobility transistors (HEMTs), drain current reduction by current collapse phenomenon is a big obstacle for a high efficient operation of power amplifier application. In this study, we investigated the effects of SiN passivation film quality on the electrical characteristics of AlGaN/GaN HEMTs. First, we conducted some experiments to investigate the relationship between electrical characteristics of AlGaN/GaN HEMTs and various conditions of SiN passivation film by plasma enhanced chemical vapor deposition (PE-CVD). We found that both gate current leakage and current collapse were improved simultaneously by SiN passivation film deposited by optimized condition of NH₃ and SiH₄ gas flow. It is found that the critical parameter in the optimization is a I_N-H/I_Si-H ratio measured by Fourier transforms infrared spectroscopy (FT-IR) spectra. Next, a thermal CVD SiN was applied to the passivation film to be investigated from the same point of view, because a thermal CVD SiN is well known to have good quality with low hydrogen content and high I_N-H/I_Si-H ratio. We confirmed that the thermal CVD SiN passivation could improve much further both of the gate leakage current and the current collapse in AlGaN/GaN-HEMTs. Furthermore, we tried to apply the thermal CVD SiN to the gate insulator in MIS (Metal Insulator Semiconductor) structure of AlGaN/GaN HEMTs. The thermal CVD SiN passivation was more suitable for the gate insulator than PE-CVD SiN passivation in a view of reducing current collapse phenomena. It could be believed that the thermal CVD SiN film is superior to the PE-CVD SiN film to achieve good passivation and gate insulator film for AlGaN/GaN HEMTs due to the low hydrogen content and the high I_N-H/I_Si-H ratio.

View full abstract

To reduce the gate leakage current of AlGaN/GaN HEMTs, we selected ITO/Ni/Au for Schottky electrodes and Schottky characteristics were compared with those of Ni/Au electrodes. ITO/Ni/Au and Ni/Au electrodes were deposited by vacuum evaporation and annealed at 350°C in nitrogen atmosphere. From the I-V evaluation results of ITO/Ni/Au electrodes, forward and reverse leakage currents were reduced. Schottky characteristics of ITO/Ni/Au electrodes were also improved compared to these of Ni/Au electrodes. In addition, substantial decrease of leakage currents was confirmed after the annealing of HEMTs with ITO/Ni/Au electrodes. This may be explained that ITO/AlGaN interface state became lower by the annealing. By the temperature dependence of I-V curves, clear dependence was confirmed for the gates with ITO/Ni/Au electrodes. On the other hand, small dependence was observed for those with Ni/Au electrodes. From these results, tunnel leakage currents were dominant for the gates with Ni/Au electrode. Thermal emission current was dominant for the gates with ITO/Ni/Au electrode. The larger temperature dependence was caused that ITO/AlGaN interface states were smaller than those for Ni/Au electrode. It was suggested that suppressed AlGaN Schottky barrier thinning was caused by the surface defect donors, then tunneling leakage currents were decreased. We evaluated HEMT characteristics with ITO/Ni/Au electrode and Ni/Au electrode. I_{d max} and G_{m max} were similar characteristics, but V_th, with ITO/Ni/Au electrode was shifted +0.4V than that with Ni/Au electrode due to the higher Schottky barrier. It was confirmed to have a good pinch-off currents and low gate leakage currents by ITO/Ni/Au electrodes.

View full abstract

Low-power dry-etching process has been adopted to study the influence of dry-etching on Ohmic contact to p-GaN. When the surface layer of as-grown p-GaN was removed by low-power SiCl₄/Cl₂-etching, no Ohmic contact can be formed on the low-power dry-etched p-GaN. The same dry-etching process was also applied on n-GaN to understand the influence of the low-power dry-etching process. By capacitance-voltage (C-V) measurement, the Schottky barrier heights (SBHs) of p-GaN and n-GaN were measured. By comparing the change of measured SBHs on p-GaN and n-GaN, it was suggested that etching damage is not the only reason responsible for the degraded Ohmic contacts to dry-etched p-GaN and for Ohmic contact formatin, the original surface layer of as-grown p-GaN have some special properties, which were removed by dry-etching process. To partially recover the original surface of as-grown p-GaN, high temperature annealing (1000°C 30s) was tried on the SiCl₄/Cl₂-etched p-GaN and Ohmic contact was obtained.

View full abstract

The heterogeneous integration of GaAs HEMTs on a polyimide-covered AlN ceramic substrate was demonstrated using a fluidic self-assembly (FSA) technique. We used thin device blocks for the FSA, which have various advantages. In particular, they can reduce the drain-source capacitance C_ds of the assembled HEMTs if the substrate has a low dielectric constant. This is a novel kind of semiconductor-on-insulator (SOI) technology. The dc and RF properties of the GaAs HEMTs on the polyimide/AlN substrate were studied and the reduction of C_ds was confirmed. This technique was successfully applied to the SPDT switch, where a low C_ds is essential for good isolation.

View full abstract

This paper will review the development of SiC power devices especially SiC power junction field-effect transistors (JFETs). Rationale and different approaches to the development of SiC power JFETs will be presented, focusing on normally-OFF power JFETs that can provide the highly desired fail-save feature for reliable power switching applications. New results for the first demonstration of SiC Power ICs will be presented and the potential for distributed DC-DC power converters at frequencies higher than 35MHz will be discussed.

View full abstract

On the basis of the RF characteristics of p-type diamond field-effect transistors (FETs) with hydrogen surface termination, we establish an equivalent circuit (EQC) model. From comparisons of three cases we reveal that to represent the device performance in the EQC, the source, gate, and drain resistance should be considered but that the gate-source and gate-drain resistance can be ignored. The features of diamond FETs are (1) a plateau of the gate capacitance in a certain gate voltage range. (2) maximum f_T and f_MAX cut-off frequencies near the threshold gate voltage, and (3) a high f_MAX/f_T ratio∼3.8. We discuss these features in terms of the energy barrier between the gate metal and the two-dimensional hole channel and drift region below the gate.

View full abstract

Heterostructure field-effect transistors (HFETs) composed of antimonide-based compound semiconductor (ABCS) materials have intrinsic performance advantages due to the attractive electron and hole transport properties, narrow bandgaps, low ohmic contact resistances, and unique band-lineup design flexibility within this material system. These advantages can be particularly exploited in applications where high-speed operation and low-power consumption are essential. In this paper, we report on recent advances in the design, material growth, device characteristics, oxidation stability, and MMIC performance of Sb-based HEMTs with an InAlSb upper barrier layer. The high electron mobility transistors (HEMTs) exhibit a transconductance of 1.3S/mm at V_DS=0.2V and an f_TL_g product of 33GHz-μm for a 0.2μm gate length. The design, fabrication and improved performance of InAlSb/InGaSb p-channel HFETs are also presented. The HFETs exhibit a mobility of 1500cm²/V-sec, an f_max of 34GHz for a 0.2μm gate length, a threshold voltage of 90mV, and a subthreshold slope of 106mV/dec at V_DS=-1.0V.

View full abstract

We report on continuous-wave optoelectronic terahertz (THz) measurements using low-temperature grown (LTG) GaAsSb as photomixer material. A broadband log-periodic antenna and a six interdigital finger photomixer with 1μm gap is fabricated on LTG-GaAsSb for THz generation and detection. At 0.37THz, the resonance frequency of the inner most antenna tooth, we obtained a power of 6.3nW. A Golay cell was used as detector. The photocarrier lifetime of the material was determined to be 700fs by pump-probe experiments with an optical wavelength close to the band gap of LTG-GaAsSb. The band gap was 1.0eV, measured by wavelength dependent pump-probe measurements.

View full abstract

For realization of hexagonal BDD-based digital systems, active and sequential circuits including inverters, flip flops and ring oscillators are designed and fabricated on GaAs-based hexagonal nanowire networks controlled by Schottky wrap gates (WPGs), and their operations are characterized. Fabricated inverters show comparatively high transfer gain of more than 10. Clear and correct operation of hexagonal set-reset flip flops (SR-FFs) is obtained at room temperature. Fabricated hexagonal D-type flip flop (D-FF) circuits integrating twelve WPG field effect transistors (FETs) show capturing input signal by triggering although the output swing is small. Oscillatory output is successfully obtained in a fabricated 7-stage hexagonal ring oscillator. Obtained results confirm that a good possibility to realize practical digital systems can be implemented by the present circuit approach.

View full abstract

Monolithic gyrators are proposed on the basis of integrating resonant tunneling diodes (RTDs) and HEMT toward realization of broadband and high-Q passives. Feasibility of millimeter-wave active inductors using the gyrator are described with equivalent circuit analysis and numerical calculations assuming InP based RTDs and a HEMT to be integrated.

View full abstract

The GaAs industry has been growing immensely during recent years. This is mainly driven by the tremendous growth of the wireless communication market, which is still continuously growing. Additionally, an emerging mmW market with applications in automotive, defense and optoelectronics is further driving the demand for GaAs components. The two largest European GaAs fabrication companies, UMS and Filtronic are very well positioned to address the complete frequency range from 1GHz up to 100GHz for commercial, high volume low cost markets, as well as individual niche applications. An overview of the companies' structures, their processes and design capabilities and also their new product developments will be presented in this paper.

View full abstract

We have developed 0.5-μm-emitter InP-based HBTs with high reliability. The HBTs incorporate a passivation ledge structure and tungsten-based emitter metal. A fabricated HBT exhibits high collector current density and a current gain of 58 at a collector current density of 4mA/μm². The results of dc measurements indicate that the ledge layer sufficiently suppresses the recombination current at the emitter-base periphery. The HBT also exhibits an f_t of 321GHz and an f_max of 301GHz at a collector current density of 4mA/μm². The f_t does not degrade even though the emitter size is reduced to as small as 0.5μm×2μm. The results of an accelerated life test show that the degradation of dc current gain is due to thermal degradation of the interfacial quality of semiconductors at the passivation ledge. The activation energy is expected to be around 1.5eV, and the extrapolated mean time to failure is expected to be over 10⁸ hours at a junction temperature of 125°C. These results indicate that this InP HBT technology is promising for making over-100-Gbit/s ICs with high reliability.

View full abstract

The performance of recently developed Large Signal (LS) HBT model was evaluated with extensive LS measurements like Power spectrum, Load pull and Inter-modulation investigations. Proposed model has adopted temperature dependent leakage resistance and a simplified capacitance models. The model was implemented in ADS as SDD. Important feature of the model is that the main model parameters are taken directly from measurements in rather simple and understandable way. Results show good accuracy despite the simplicity of the model. To our knowledge the HBT model is one of a few HBT models which can handle high current & Power HBT devices, with significantly less model parameters with good accuracy.

View full abstract

A 76-GHz Gunn voltage-controlled oscillator (VCO) with a high output power and a wide tuning-frequency range was fabricated by optimizing VCO circuits and using laser micromachining. The tuning-frequency range of the fabricated Gunn VCO was more than two times higher than that attained in our previous experiments by optimizing VCO circuits. The VCO attained a tuning-frequency range of 493MHz, out-put power variation of 1.0dB, and tuning-frequency linearity of 6.1% over a tuning-voltage range from 0 to 10V. Its power consumption was 2.0W at operation voltage of 3.6V. And it measured out-put power was 13.3dBm with DC-RF conversion efficiency of 1.0% at 76.5GHz. Moreover, under fundamental-mode operation, it achieved low phase noise of -107.8dBc/Hz at an offset frequency of 1MHz. Since laser micromachining was used in fabricating the Gunn VCO, the reproducibility of its RF performance was improved.

View full abstract

We have developed a power amplifier IC for Bluetooth Class 1 operating at single low voltage of 1.8V for both control and drain voltages. We can realize it due to fully enhancement-mode hetero-junction FETs utilizing a re-grown p⁺-GaAs gate technology. The power amplifier is a highly compact design as a small package of 1.5mm×1.5mm×0.4mm with fully integrated gain control and shutdown functions. An impressive power added efficiency of 52% at an output power of 20dBm is achieved with an associated gain of 22dB. Also, sufficiently low leakage current of 0.25μA at 27°C is exhibited, which is comparable to conventional HBT power amplifiers.

View full abstract

We have successfully demonstrated a GaInAs/InP multiple quantum well (MQW)-based wavelength switch composed of the straight arrayed waveguide with linearly varying refractive index distribution by changing the refractive index using thermo-optic effect. Since optical path length differences between waveguides in the array were achieved through refractive index differences that were controlled by SiO₂ mask design in selective metal-organic vapor phase epitaxy (MOVPE), wavelength demultiplexing, and the output port switching in each wavelength of light by the refractive index change in the array waveguides through the thermooptic effect were achieved. We have obtained the wavelength switching and the change of transmission spectra in each output ports.

View full abstract

We propose a polymeric waveguide optical switch using a novel drive mechanism. The switch uses a flexible polymeric waveguide film where trenches are formed at cross-points of the waveguides. Light passes through the trench while it is closed. When the trench opens, light path changes by total internal reflection between the air gap and the polymeric waveguide. Therefore, we can control light paths by changing the trench state between closed and open one. In order to realize this, a rotating arm is inserted near the trench. As rotational force transfers to the trench through the arm and the film, the trench switches from closed to open state and vice versa. We investigated this rotary drive mechanism by three-dimensional (3D) structural analysis, designed the optical switch, and experimentally demonstrated the switching operation.

View full abstract

An all-optical phase multiplexing scheme for phase-modulated signals is proposed and experimentally demonstrated using fourwave mixing (FWM) in a highly-nonlinear fiber (HNLF). Two 10-Gb/s π/2-shifted differential phase-shift keying (DPSK) wavelength-division multiplexing (WDM) signals are experimentally demonstrated to be converted and phase-multiplexed into a 20-Gb/s differential quadrature phase-shift keying (DQPSK) signal with non-return-to-zero (NRZ) and return-to-zero (RZ) formats, respectively. Experimental results show that, due to phase-modulation-depth doubling effect and phase multiplexing effect in the FWM process, a DQPSK signal is successfully generated through the proposed all-optical phase multiplexing with improved receiver sensitivity and spectral efficiency.

View full abstract

The dispersion and the splice characteristics of optical fibers with trench-index profile are investigated. The normalized distance between core and trench is preferably larger than 3.0 to realize complete compatibility with the standard G.652 fiber in terms of chromatic dispersion. The optical fiber realizes compatibility with ITU-T Recommendation G.652 fiber and bend-insensitivity simultaneously. Fabricated fibers with the trench-index profiles can be spliced to standard single-mode fiber with low losses, which have similar values with simulation results.

View full abstract

For the construction of photonic crystal fiber (PCF) systems using their unique properties, a PCF coupler (PCFC) is one of the key components of the systems. The characteristics of the PCFC depend on the state of air holes in the tapered region of the PCFC because the state of air holes in the tapered region affects light propagation in the PCFC taper. In this paper, coupling characteristics of PCFCs were theoretically investigated. In PCFCs with air hole remaining tapers, we found that a smaller elongation ratio i. e. a stronger elongation is required to obtain optical coupling as an air hole pitch or a ratio of air hole diameter to pitch is larger. In PCFCs with air hole collapsed tapers, it was clarified that a dependence of extinction ratio on air hole collapsed elongation ratio is higher for smaller elongation ratio. It was also clarified that an air hole remaining PCFC has slow wavelength characteristics in extinction ratio compared to an air hole collapsed PCFC. Air hole remaining PCFCs and air hole collapsed PCFCs were fabricated using a CO₂ laser irradiation technique. We could successfully control whether air holes in the PCFC taper were remaining or collapsed by adjusting the irradiated laser power in the elongation process of the PCFC fabrication. It was experimentally clarified that the air hole remaining PCFC has slow wavelength characteristics in extinction ratio compared to the air hole collapsed PCFC. The tendencies of the measured wavelength characteristics of PCFCs agree with those of numerical results.

View full abstract

A compact model is proposed for accurately incorporating effects of STI (shallow trench isolation) stress into post-layout simulation by making layout-dependent corrections to SPICE model parameters. The model takes in-plane (longitudinal and transverse) and normal components of the layout-dependent stress into account, and model formulas are devised from physical considerations. Not only can the model handle the shape of the active-area of any MOSFET conforming to design rules, but also considers distances to neighboring active-areas. Extraction of geometrical parameters from the layout can be performed by standard LVS (layout versus schematic) tools, and the corrections can subsequently be back-annotated into the netlist. The paper spells out the complete formulation by presenting expressions for the mobility and the threshold voltage explicitly by way of example. The model is amply validated by comparisons with experimental data from 90nm-and 65nm-CMOS technologies having the channel orientations of, respectively, ‹110› and ‹100›, both on a (100) surface. The worst-case standard errors turn out to be as small as 1.7% for the saturation current and 8mV for the threshold voltage, as opposed to ∼20% and ∼50mV without the model. Since device characteristics variations due to STI stress constitute a significant part of what have conventionally been treated as random variations, use of the proposed model could enable one to greatly narrow the guardbands required to guarantee a desired yield, thereby facilitating design closure.

View full abstract

Configurable clock is necessary for many applications such as digital communication systems, however, using the conventional direct digital frequency synthesizer (DDS) as a pulse or clock generator may cause jitter problems. People usually employ phase-interpolation approaches to generate a pulse or clock with correct time intervals. This work proposes a new phase-interpolation DDS scheme, which uses the output of the phase accumulator to provide an initial voltage on an integration capacitor by pre-charging in the first phase, and then performs integration operation on the same integration capacitor in the second phase. By using single capacitor integration, the instability of the delay generator existed in the phase-interpolation DDS can be avoided, and the impact caused by capacitance error in the circuit implementation also can be reduced. Furthermore, without ROM tables, the proposed DDS using pre-charging integration not only reduces the spurious level of the clock output, but also has a low hardware complexity.

View full abstract

We formulated the excitation rate of VUV and emitted visible light from rare gas on PDP by using the Boltzmann equation with electron-atom collision integral term and obtained the excitation rate as the function of Temperature and Mass. This form of excitation rate was firstly derived in PDP area. In addition we showed the Pressure dependence of intensity ratio of Ne/VUV as the application of our excitation rate formulae.

View full abstract

Radar Cross Section (RCS) can be obtained from near-field data by using near-field to far-field RCS transformation methods. Phase errors in near-field data cause the degradation of the prediction accuracy. In order to overcome the difficulty, we propose the far-field RCS prediction method from one-dimensional intensity data in near-field. The proposed method is derived by extending the phase retrieval method based on the Gerchberg-Saxton algorithm with the use of the relational expression between near-fields and scattering coefficients. The far-field RCS can be predicted from the intensity data of scattered fields measured at two different ranges. The far-field RCS predicted by the proposed method approximately coincides with the computed one. The proposed method also has significant advantages of simple and efficient algorithm. The proposed method is valuable from a practical point of view.

View full abstract

In this letter, we propose a partial access mechanism to be used on a register file for low-cost embedded multimedia processor architecture. In the embedded system, supporting the SIMD operations is a burden because of the wide SIMD register file and its execution unit. So a new architecture is proposed to increase the performance of SIMD operations with minimal additional hardware overhead. To evaluate the performance and hardware overhead, this architecture is adopted to an embedded multimedia processor and simulated with five DSP benchmarks. The simulation results indicate that the performance is increased by 38% and the total area is increased by 13.4%. The proposed partial access mechanism may be useful for low-cost embedded multimedia ASIP.

View full abstract