International Journal of High Speed Electronics and Systems

This paper presents high performance device results using an ultra-thin AlN/GaN structure on sapphire substrate with a 100-nm T-gate. Excellent dc and RF characteristics are reported, including an extrinsic transconductance of 500 mS/mm and an extrinsic f_t/f_max(U) ratio of 78/111-GHz which is among the highest reported for AlN/GaN HFETs. Low gate leakage results are also presented despite the small barrier thickness and absence of a gate dielectric. Modeling of the small signal parameters is also discussed to gain an understanding of the limiting and contributing performance factors.

We report the design, fabrication and characterization of metamorphic high electron-mobility transistors (MHEMTs) with self-aligned ohmic electrodes. In this work, asymmetrically recessed 50-nm Γ-gates have been successfully used as the shadow mask for ohmic metal deposition. Extremely high extrinsic transconductance over a wide drain bias from 0.1 to 1.25 V can be made possible by fabricating devices with small gate-source spacing, small source-drain spacing, and the non-alloyed ohmic. Measured maximum extrinsic transconductance of 3 S/mm is a new record for all HEMT devices on GaAs and equals the best results from InP-based HEMTs. The same devices also show a voltage gain of 22, maximum stable gain of 10.8 dB at 110 GHz, and breakdown voltage of 4.3 V, which all are the highest among any self-aligned HEMTs based on InGaAs channel. The outstanding performance is the result of the seamless integration of the asymmetric gate recess and Γ-gate-based self-aligned ohmic process.

Pivotal in the design of circuits is the ability to efficiently translate available transistor gain to high gain per stage. Remarkably, for 35-nm InP HEMT transistors, the efficiency of this translation remains high even up to ~0.5 THz. The ever shrinking wavelength correlated with higher frequencies necessitates a scaling of not only the device layout, but also of the passive elements and wafer thickness. Furthermore, to avoid distributed effects, the length of transistor gate fingers must be reduced.

In this paper, we will review the design, construction and performance of the two-way Ka-band waveguide branch-line and asymmetric magic-T based unequal power combiners. The manufactured combiners were designed to combine input signals that are equal in phase and with an amplitude ratio of two. Next, the design, construction and performance of a three-way branch-line unequal power combiner, achieved by serially interconnecting two 2-way branch-line hybrids and optimizing the dimensions using software tools, is presented. The application of the two-way and three-way combiners for combining the output from two or three MMIC PAs was demonstrated. The observed efficiencies for all three power combining configurations are 90 percent or greater at Ka-band (31.8 to 32.3 GHz).

We have found that Scandium metal is near ohmic as deposited on GaN, but when it is annealed at high temperatures a large barrier height Schottky forms. In this study we used Sc-Au contacts to form Schottky barrier diodes on AlGaN/GaN HEMT material. We have found that the morphology remains unchanged even after an 800 degrees centigrade anneal. This investigation has revealed that the reverse leakage current of this metal system is an order of magnitude lower than a conventional Ni-Au contact and supports a reverse breakdown that is 1/3rd larger. The similarity of the anneal temperatures to ohmic contacts enable gates and contacts to be annealed at the same time thus simplifying processing. The lack of morphology change supports the use of Sc-Au for E-beam alignment marks as well. Diode contacts on AlGaN/GaN with Schottky-ohmic separation of 10 microns demonstrated reverse breakdown in excess of 100V when the contacts were annealed at 800C. These results suggest this metallization may have applications as a new HEMT gate metal, and Schottky diodes.

A high-frequency a high-power GaN HEMT was analyzed using our full band Cellular Monte Carlo (CMC) simulator, in order to extract small signal parameters and figures of merit, and to correlate them to carrier dynamics and distribution inside the device. A complete RF and DC characterization of the device was performed using experimental data to calibrate the few adjustable parameters of the simulator. Then, gate-related capacitances, such as C_g, C_gd, and C_gs, were directly and indirectly extracted combining small-signal analysis and DC characterization.

Bi-directional solid-state circuit breakers (BDSSCBs) can provide performance benefits over mechanical fault protection devices. A common-source configuration of normally ON, junction field effect transistors (JFETs) is favorable for BDSSCB implementations. SiC 0.1-cm² 1200-V JFETs designed for normally-ON operation at a zero-volt gate bias, and having low leakage currents, were used in the fabrication of a 30-A BDSSCB switch module. Operation of the module under continuous current and during turn-OFF transitions was evaluated to verify the parallel scalability of the common-source configuration. A bi-directional snubber connected across the switch module mitigated inductive voltage overshoot during BDSSCB turn-OFF transitions. At turn-OFF, under maximum power tests in both directions, the load current was reduced from 30 A to 0 A in approximately 10 μs, with a supply voltage of 600 V, and a BDSSCB peak voltage of 680 V. These results demonstrate the functionality and current scalability of this BDSSCB topology.

Army applications require power components that are utilized in pulsed power systems to be reliable and durable, compact, have high power and energy density, and lastly, to be easily integrated into combat vehicles. The U.S. Army Research Laboratory (ARL) is currently investigating silicon carbide Super-gate turn-off thyristors (SiC SGTOs) to meet the future pulsed power system requirements for the Army. ARL's methodology is to evaluate the device performance in pulsed power circuits that emulate the electrical stresses that the power devices would encounter in a pulsed power system. Each module being evaluated utilized four 0.6 cm²SiC SGTOs. The packaging design for the module utilizes ThinPak technology. The module has reliably been pulsed over 2000 times at peak currents greater than 8 kA with a pulse width of 170 μs. The module has also been pulsed greater than 3.6 kA at a pulse width of 1 ms, corresponding to an action rate of 7000 A²s and a current density of 2.5 kA/cm² over the cathode mesa area. This work evaluates SiC SGTO modules under various pulsed conditions. This paper includes details on device structure, module package design, and module pulsed switching and reliability.

We report for the first time operation of GaN Schottky rectifiers under cryogenic temperatures. A 600V, 4A GaN Schottky rectifiers from Velox Semiconductors has been used for the characterization. Forward conduction and reverse blocking performance was measured down to 77K. Two Schottky barrier heights have been noticed at low temperatures and a tunneling limited reverse leakage current was observed for the rectifier.

The safe operating area of E-pHEMTs have been explored using device simulations. It is proposed that the second breakdown is triggered by the debiasing of the channel potential due to the gate reverse current at high drain voltages. A mathematical description of this behavior and experimental results are presented.

In recent years, increasing demand for hybrid electric vehicles (HEVs) has generated the need for reliable and low-cost high-temperature electronics which can operate at the high temperatures under the hood of these vehicles. A high-voltage and high temperature gate-driver integrated circuit for SiC FET switches with short circuit protection has been designed and implemented in a 0.8-micron silicon-on-insulator (SOI) high-voltage process. The prototype chip has been successfully tested up to 200°C ambient temperature without any heat sink or cooling mechanism. This gate-driver chip can drive SiC power FETs of the DC-DC converters in a HEV, and future chip modifications will allow it to drive the SiC power FETs of the traction drive inverter. The converter modules along with the gate-driver chip will be placed very close to the engine where the temperature can reach up to 175ΰC. Successful operation of the chip at this temperature with or without minimal heat sink and without liquid cooling will help achieve greater power-to-volume as well as power-to-weight ratios for the power electronics module.

A III-Nitride/SiC separate absorption and multiplication avalanche photodiode (SAM-APD) offers a novel approach for fabricating high gain photodetectors with tunable absorption over a wide spectrum from the visible to deep ulltraviolet. However, unlike conventional heterojunction SAM APDs, the formation of polarization-induced charge at the hetero-interface arising from spontaneous and piezoelectric polarization can dramatically affect the performance of this detector. In this paper we report on the role of this interface charge on the performance of GaN/SiC SAM APDs. Simulations of the electric field profile within this device structure while biased near avalanche breakdown indicate that the density of positive interface charge may be sufficient to confine the electric field within the SiC multiplication region with negligible punch-through into the GaN absorption region, a distribution that is likely undesirable for efficient collection of photo-generated holes due to the presence of defects at the hetero-interface. Simulations further show that the incorporation of a p-type doped interface charge control layer at the hetero-interface can modify the total density of charge at the interface and allow for the tailoring of the electric field profile within this device. Experimental results are provided that correlate well with the simulation results.

Low dislocation density pseudomorphic epitaxial layers of Al_xGa_1-xN have been grown on c-face AlN substrates prepared from high quality bulk crystals. As reported previously, pseudomorphic growth yields very low dislocation density layers with atomically smooth surfaces throughout the active region of a full LED device structure. An advantage of the low dislocation density is the ability to n-type dope the high aluminum content Al_xGa_1-xN (x ~ 70%) epitaxial layers required for UVLED devices to obtain sheet resistances less than 350 Ohm/square for 0.5 μm thick layers. Here, we report on the characterization of our pseudomorphic epitaxial AlGaN layers via cathodoluminescence (CL) and on-wafer and initial packaged level characterization of fully fabricated pseudomorphic ultraviolet LEDs (PUVLEDs) with an emission wavelength between 250 - 265 nm. An additional benefit of PUVLED devices is the ability to run these devices at high input powers and current densities. Further, the aforementioned low dislocation density of the epitaxial structure results in improved device performance over previously published data. Mean output powers of greater than 4 mW were obtained on-wafer prior to thinning and roughening while output powers as high as 45 mW were achieved for packaged devices.

Bio-sensing sensitivity of a spectrally selective nanoparticle based ultraviolet (UV) photodetector is characterized in comparison to a silicon photodiode and a photomultiplier tube (PMT). The nanoparticle based photodetector is comprised of poly-vinyl alcohol (PVA) coated zinc-oxide (ZnO) nanoparticles deposited on an aluminum-gallium-nitride (AlGaN) epitaxially grown substrate. The sensitivity was determined by measuring the fluorescence intensity of the native fluorophore, tryptophan, in Escherichia coli (E-coli, ATCC-25922) cells. Tryptophan intrinsically fluoresces with a peak at 340 nm under 280 nm UV light illumination. It is shown that this detector can sense the concentration of E-coli to 2.5 × 10⁸ cfu/mL while the silicon photodiode cannot detect the intrinsic fluorescence at all. Nevertheless, the PMT outperformed the ZnO nanoparticle-AlGaN substrate based photodetector with the ability to sense E-coli concentrations to 3.91 × 10⁶ cfu/mL. However, because PMT based systems are commonly limited by high dark current, susceptible to environmental changes, sensitive to ambient light, are not spectrally selective and have high power consumption, biological detection systems comprised of these ZnO nanoparticle-AlGaN substrate based photodetectors can be more effective for near real time characterization of potential bacterial contamination.

A step-graded Al_xGa_1-xN electron blocking layer (EBL) is introduced to the InGaN-based edge-emitting blue-violet laser diode (LD) structure to suppress the undesired built-in interface polarization charges. When compared to a conventional abrupt Al_0.18Ga_0.82N EBL design, the step-graded Al_xGa_1-xN EBL design may help reduce the electron accumulation at the edge of the active region and hence improve the quantum efficiency in LD operation. The effects of the step-graded Al_xGa_1-xN EBL on the fabricated device performance are also investigated. LDs with the step-graded Al_xGa_1-xN EBL demonstrated significantly reduced threshold current density and increased slope efficiency under the continuous-wave operation.

Low-resistance Ohmic contacts are essential for the fabrication of electrical devices. While low contact resistance has been achieved to p-type layers or n-type layers separately, contacts are likely to degrade when both types need to be integrated into a single fabrication process, in particular when prior mesa etching is required. We present a solution to the problem, resulting in low-resistance Ohmic contacts on n-type GaN layers without post-deposition thermal anneal, while maintaining the quality of typical p-type contacts. We implement an integrated process for both, n- and p-contacts, involving an oxygen pretreatment to fabricate light emitting diodes with lower series resistance in the contacts and lower voltage drop at high current when compared to separately optimized contacts.

The background photon shot noise, induced by the signal itself and all other in-band surroundings, provides the fundamental infrared imaging sensitivity limit (background limited performance or BLIP). Optical diffraction constrains resolution angle to values greater than the ratio of the detection wavelength to the optics diameter. Imaging technology tries to attain BLIP at spatial frequencies approaching the diffraction limit while minimizing system size, weight, power, and life-cycle cost.

To minimize costs, detector array technology must raise the detector operating temperature while keeping dark currents and other noise mechanisms below background shot noise. This requires reducing material and process defects. Recently HgCdTe has seen the MWIR-LWIR device dark currents to be limited only by fundamental mechanisms over a wide range of temperatures and wavelengths. Novel materials systems (e.g. Type-II superlattices) in theory have even lower fundamental limits than HgCdTe, but as yet lag in demonstrated performance. At least for now HgCdTe provides a convenient well-characterized benchmark and gives insight into further improvement possibilities

Infrared imaging technology must also exploit advanced processing and packaging techniques to reduce detector pitch to a few wavelengths of detected light. Mid- and long-wave infrared imaging arrays with ~15μm detector dimensions are now approaching this goal.

We report time-resolved photoluminescence measurements on a set of long-wave infrared InAs/GaSb type II superlattice absorber samples with various widths as a function of temperature and excitation density. Careful analysis of the photoluminescence data determines the minority carrier lifetime and background carrier density as a function of temperature, and provides information on the acceptor energy and density in each sample. Results indicate that carrier lifetime is dominated by Shockley-Read-Hall recombination with a lifetime of ~30 ns at 77 K for all samples. Below 40 K, background carriers are observed to freeze-out in conjunction with increased contributions from radiative recombination. An acceptor energy level of ~20 meV above the valance band is also determined for all samples. Variations of carrier lifetime between each sample do not strongly correlate with absorber width, indicating that barrier recombination is not the dominant factor limiting the carrier lifetime in our samples.

Reduction in the dark current and improvement in signal to noise ratio in the quantum dots in a well infrared photodetectors using resonant tunneling barriers have been demonstrated. Ultra-low dark current levels and high detectivity of 3×10¹⁰ cm.Hz^1/2/W at 77K for f/2 optics has been obtained for longwave infrared detection. In another experiment, the ability to control the excited state in the DWELL has been demonstrated by systematically varying the quantum well thickness. These detectors demonstrate high operating temperature with high detectivity values, even for high operating temperatures.

Tunneling currents and surface leakage currents are both contributors to the overall dark current which limits many semiconductor devices. Surface leakage current is generally controlled by applying a post-epitaxial passivation layer; however, surface passivation is often expensive and ineffective. Band-to-band and trap assisted tunneling currents cannot be controlled through surface passivants, thus an alternative means of control is necessary. Unipolar barriers, when appropriately applied to standard electronic device structures, can reduce the effects of both surface leakage and tunneling currents more easily and cost effectively than other methods, including surface passivation. Unipolar barriers are applied to the p-type region of a conventional, MBE grown, InAs based pn junction structures resulting in a reduction of surface leakage current. Placing the unipolar barrier in the n-type region of the device, has the added benefit of reducing trap assisted tunneling current as well as surface leakage currents. Conventional, InAspn junctions are shown to exhibit surface leakage current while unipolar barrier photodiodes show no detectable surface currents.

The conduction channel of a semiconductor Field Effect Transistor (FET) or a heterostructure High Electron Mobility Transistor (HEMT) can act as a plasma wave resonator for density oscillations in electron gas, at frequencies significantly higher than the transistor cut-off frequency in a short channel device. The hydrodynamic model predicts a resonance response to electromagnetic radiation at the plasma oscillation frequency. In particular, the hydrodynamic nonlinearities produce a constant source-to-drain voltage when gate-to-channel voltage has a time-harmonic component. In the Dyakonov-Shur detector a short channel HEMT is used for the resonant tunable detection of terahertz radiation. Starting with the quasi-classical Boltzmann equation for a semiconductor and graphene channels, we derived the viscous hydrodynamic model with temperature dependent transport coefficients in both cases. We evaluated the detector response function and in the case of semiconductor channel we also obtained the temperature dependence of the quality factor of the plasma resonance. The present treatment extends the theory of Dyakonov-Shur plasma resonator and detector to account for the temperature dependence of viscosity. In the case of semiconductor channels the treatment here also includes the energy balance equation into the analysis. The numerical results are given in cases of GaAs and GaN channels. We showed that in high mobility semiconductor channels at low temperature the quality of the resonance is strongly limited by the viscosity of the electron fluid. In the case of graphene channel the hydrodynamic model derived here accounts both for electrons and holes, and includes the related diffusion currents. When the gate voltage is a few volts, the Fermi temperature of the electron-hole liquid is considerably higher than the room temperature. In such cases the diffusion currents can be ignored, and from the simplified hydrodynamic equations we evaluated the non-linear response of the plasma in graphene channel to the external perturbation. The results are of interest in potential application to graphene based detectors due to potential of obtaining a channel with the room temperature mobility considerably higher than the mobility in semiconductor channels.

We have created a low profile, high efficiency half waveplate for operation at terahertz (THz) frequencies. The waveplate is a periodic gold/polyimide composite with a physical thickness of λ/10. Our reflection based waveplate has an intensity throughput of 80% at the design frequency of 350 GHz. This is quite high in comparison to transmissive THz components which typically suffer a large insertion loss due to Fresnel reflections. Simulations suggest a halfwave rotation of over 99% of the reflected THz radiation from 320 – 372 GHz. Experiments at 350 GHz confirm the basic operation of our electromagnetic composite as a functional half waveplate.

InAs/AlSb/GaSb heterostructure backward tunnel diodes are sensitive millimeter-wave detectors that achieve their high sensitivity by leveraging interband tunneling current in a broken-gap heterostructure. A new physics-based tunneling model that accurately reproduces experimental results is proposed in this work. This model, which uses Kane's eight-band k·p band structures and tunneling coefficients calculated using the transfer matrix method, predicts IV curves that closely match measured data. It also predicts critical device parameters for detector performance (including curvature coefficient and junction resistance) as a function of device structure, and good agreement with experiment results has been obtained. It accurately predicts the junction resistance's exponential dependence on tunnel barrier thickness, as well as more subtle effects such as the impact of cathode doping profile adjustment and anode ternary composition. The predictive capabilities of this model are promising to provide guidance for future detector heterostructure optimization and may be useful in modeling the characteristics of other heterostructure designs and devices using 6.1 Å lattice constant materials.

In this paper, we report our studies on field effect transistor (FET) THz detectors operating in the non-resonant mode based on the Dyakonov-Shur plasma wave detection theory, where the quantum capacitance dominates. The influence of quantum capacitance in detector response is theoretically developed and numerically simulated at low and high frequencies. Fundamental constraints in the upper frequency limit are also analyzed for FET THz detectors based on various materials, showing advantages of GaN for 8 - 20 THz applications. Experiments at microwave and THz frequencies have been carried out for GaN based devices showing agreement with the theory.

A subwavelength aperture-microbolometer sensor was simulated using a finite-difference time domain (FDTD) numerical method. The system consisted of a focusing bullseye grating with a central subwavelength aperture. A thin vanadium oxide sensing layer was suspended behind the aperture for sensing purposes, backed by a thin layer of silicon dioxide. The maximum electric field across the vanadium oxide was measured as a function of incident field. Total noise equivalent power (NEP) was estimated to be 180 pW/√Hz at a frequency of 0.665 THz, comparable to similar current micro sensors. The formation of surface plasmonics greatly increases its sensitivity to particular incident wavelengths. This fact, along with its compact size, make this system a promising candidate for numerous low-cost, small-scale terahertz sensing applications.

GaN-based pseudomorphic heterostructures with their demonstrated superior thermal performance suggest an alternative to the standard GaAs-based technology to realize high power lasing at THz frequencies. A larger electron effective mass in GaN based system results in the energy levels lying deeper within the quantum well compared to its GaAs counterpart resulting in longer carrier lifetimes assisting transitions required for THz radiation while reducing tunneling current. However, the presence of spontaneous and piezoelectric polarization and dependence of bandgap and band offsets on structural and bias induced strain reduces carrier lifetime.

We report a room temperature 570-630 GHz frequency domain terahertz (THz) spectroscopy system developed on the basis of a broadband quasi-optical zero bias Schottky diode detector. The detector is designed to cover the frequency range of 100 GHz to nearly 900 GHz. A responsivity of 300-1000 V/W has been measured, and the noise equivalent power (NEP) is estimated to be 5-20 pW/√Hz based on the measurements of similar detectors. For a prototype demonstration, the frequency domain THz spectroscopy system was operated within the region of 570-630 GHz using a VDI (Virginia Diodes, Inc.) frequency extension module (FEM) to provide the THz radiation. Mylar thin films with different thicknesses and THz metal mesh filters have been measured using this system, demonstrating a measurement accuracy of ~2%. This system has been applied to measure biomolecules in liquid-phase, and nano-material samples in solid-phase. Initial results and discussion are presented.

Spatial Wavefunction-Switched (SWS) Field-Effect Transistors (FETs) consist of inversion layers comprising two or more coupled quantum wells (QWs). Carriers can be localized in any of the wells and vertically transferred between them by changing the gate voltage. In addition, carriers can also be laterally transferred between adjacent SWSFET devices by the manipulation of the gate voltages (Vg). This enables processing of two more bits simultaneously by changing the spatial location of the carrier ensemble wavefunction, which in turn determines the state of the device [e.g., electrons in well W2 (01), in W1 (10), in both (11), in neither (00)]. Experimentally, the capacitance-voltage data, having a distinct peak, has been presented in InGaAs-AlInAs two-quantum well structures. The peak(s) are attributed to the appearance of carriers, first in the lower well and subsequently their transfer to the upper well. Use of multiple channels allows for CMOS-like configuration with both transistors having n-channel mobilities. Simulation of an InGaAs SWS inverter computes a gate delay of 0.24ps. A cut-off frequency in excess of 8THz is computed for 12nm channel length InGaAs SWSFETs. Examples, including logic gates and a 3-bit full-adder, are presented to show the reduction of device count when SWS-FETs are employed.

This paper describes design of analog-to-digital converters (ADCs) and digital-to-analog onverters (DACs) using field-effect transistors that exhibit three states in their transfer characteristics. An intermediate state "i" has been observed in the transfer characteristics (drain current-gate voltage) of FETs when two layers of cladded quantum dots (e.g. SiO_x-Si and GeO_x-Ge) are introduced in the gate region above the tunnel insulator between the source and drain regions. Three states in such a transistor, defined as quantum dot gate field-effect transistor (QDG-FET) include two stable states (ON and OFF) and a low-current saturation state "i" in its transfer characteristics. QDG-FETs are quite different in construction than nanodot based nonvolatile memories, reported in the literature, where the quantum dots are sandwiched between a tunnel gate insulator and a relatively thick control gate dielectric.

In this paper we present analog-to-digital converters (ADCs) using comparators based on QDG-FETs. A comparator is designed with fewer three-state QDG-FETs. Designs of 3-bit ADC, using 25 nm QDG-FETs, are simulated showing a signal-to-noise ratio (SNR) of ~18. In addition, the R-2R ladder problem, encountered in conventional analog-to digital converters (ADCs), is absent in QDG-FET based architecture.

Graphene transistors using large area chemical-vapor-deposited (CVD) monolayer graphene and advanced dielectric stacks are constructed and examined. Top-gated devices with a SiO₂/Al₂O₃ gate-dielectric have a Dirac Point (DP) located at less than 5 V and asymmetric electron/hole mobility. In contrast, devices based on an advanced AlN interfacial layer have a DP located near 0V and a near symmetric carrier mobility- characteristics that could be more suitable for applications that require ambipolar behavior and low-power operation. For the first time, a measured RF cut-off frequency range of 1GHz is measured for top-gated transistors using CVD graphene. The results are of importance for the realization of graphene based, wafer-scale, high frequency electronics.

The effect of external optical feedback on a packaged monolithic quantum dot passively mode-locked laser is investigated. The radio-frequency (RF) linewidth narrows from 8 KHz in the free-running situation to a value as low as 170 Hz under relatively low feedback and temperature control. The RF linewidth characterization under resonant feedback at a multiple of the laser cavity length agrees well with the published theory. The narrow RF linewidth with external optical feedback makes the chip-scale quantum dot mode-locked laser an attractive source for ultra low noise applications such as optical clocking and optoelectronic oscillators.

A photovoltaic photodetector harnessing near infrared band gap absorption by thin films of post-synthetically sorted semiconducting single walled carbon nanotubes (s-SWCNTs) is described. Peak specific detectivity of 6×10¹¹ Jones at -0.1 V bias at 1210 nm is achieved using a heterojunction device architecture: indium tin oxide/ ca. 5 nm s-SWCNT / 120 nm C₆₀ / 10 nm 2,9-dimethyl-4,7-diphenyl-1,10-phenanthroline (BCP) / Ag. The photodiodes are characterized by a series resistance of 2.9 Ω cm² and a rectification ratio of 10⁴ at ±1V. These results are expected to guide the exploration of new classes of solution-processable, mechanically flexible, integrable, thin film photovoltaic photodetectors with tunable sensitivity in the visible and infrared spectra based on semiconducting carbon nanotubes.

We have recently reported on the synthesis and characterization of a new form of nanostructured graphene that we call "nanoperforated graphene". Nanoperforated graphene is fabricated by etching a periodic array of nanoscale holes into atomic membranes of graphene to create an ultrathin superlattice-like structure. Nanoperforated graphene demonstrates semiconductor-like behavior and we have realized room-temperature field-induced conductance modulation as high as 450 (compared with < 10 for unpatterned graphene) with field-effect mobilities of ~ 1 cm²V^-1s^-1. Here, we discuss the conduction mechanisms in nanoperforated graphene and the relevance of this new material for field-effect transistor devices. In nanoperforated graphene with 15 nm nanoconstrictions, we observe that the low-bias mobility is independent of temperature, consistent with elastic scattering-limited conduction. At low temperatures, a transport gap limits conduction in the sub-threshold regime and affects the threshold voltage for band conduction. We show that the high-bias electrical characteristics of nanoperforated graphene are similar to "artificial solids," a class of materials made of 2D arrays of Coulomb islands, consistent with observed Coulomb Blockade features in the sub-threshold regime. Currently, the device characteristics of the nanopatterned graphene material are found to be suitable for large-area, thin-film transistor applications. Future higher-performance applications are expected.

Energy efficient hetero-junction tunneling transistors in a simple "cross-point" configuration that utilizes top gates are analyzed for use at extremely scaled sub-10 nm gate-lengths. The active tunneling region comprises of a vertical p++/n+ heterojunction (for example formed at the cross point of p++SiGe/n+Si), where modulation of the energy-bands in the gated n+Si region with a top-gate is used to control the degree of band overlap and tunneling distance and hence current. The sub-threshold swing characteristic of these devices is shown to be potentially highly immune to extreme downscaling to 6 nm gate length allowing for an intact and efficient switching behavior to be retained. The extreme scalability and ultra-low voltage operation could make such cross-point devices useful for alternative applications and architectures that require ultimate energy efficiency.

In this paper, a technology comparison between monolithic passively mode-locked lasers (MLLs) fabricated from 1.24 μm InAs dots-in-a-Well (DWELL) and 1.25 μm InGaAs single quantum well (SQW) structures grown using elemental source molecular beam epitaxy (MBE) is presented. 5 GHz optical pulses with sub-picosecond RMS jitter, high pulse peak power (1W) and narrow pulse width (< 10 ps) are typical of these monolithic two-section InAs DWELL passive MLLs. An InGaAs single quantum well MLL with the 42% indium is shown to exhibit a superior high-temperature performance. Compared with the typical operating range of the InAs DWELL devices (<60°C), the operation is in excess of 100 °C and is particularly attractive for clocking applications in next generation microprocessors. Based on an 8-band k.p analysis the reduction in the band edge density of states for such a quantum well is discussed.