New dielectric material system of Nd(Mg1/2Ti1/2)O3–CaTiO3 at microwave frequency

The microwave dielectric properties of (1 − x)CaTiO₃–xNd(Mg_1/2Ti_1/2)O₃ (0.1  x  1.0) ceramics prepared by the conventional solid state method have been investigated. The system forms a solid solution throughout the entire compositional range. The dielectric constant decreases from 152 to 27 as x varies from 0.1 to 1.0. In the (1 − x)CaTiO₃–xNd(Mg_1/2Ti_1/2)O₃ system, the microwave dielectric properties can be effectively controlled by varying the x value. At 1400 °C, 0.1CaTiO₃–0.9Nd(Mg_1/2Ti_1/2)O₃ has a dielectric constant (ε_r) of 42, a Q × f value of 35 000 GHz and a temperature coefficient of resonant frequency (τ_f) of −10 ppm/°C. As the content of Nd(Mg_1/2Ti_1/2)O₃ increases, the highest Q × f value of 43 000 GHz for x = 0.9 is achieved at the sintering temperature 1500 °C.     

Extraction of interface states at emitter–base heterojunctions in AlGaAs/GaAs heterostructure bipolar transistors using sub-bandgap photonic excitation

Distribution of interface states at the emitter–base heterojunctions in heterostructure bipolar transistors (HBTs) is characterized by using current–voltage characteristics using sub-bandgap photonic excitation. Sub-bandgap photonic source with a photon energy E_ph which is less than the energy bandgap E_g (E_g,GaAs = 1.42, E_g,AlGaAs = 1.76 eV) of emitter, base, and collector of HBTs, is employed for exclusive excitation of carriers only from the interface states in the photo-responsive energy range at emitter–base heterointerface. The proposed method is applied to an Al_0.3Ga_0.7As/GaAs HBT (A_E = W_E × L_E = 250 × 100 μm²) with E_ph = 0.943 eV and P_opt = 3 mW. Extracted interface trap density D_it was observed to be D_it,max  4.2 × 10¹² eV⁻¹ cm⁻² at emitter–base heterointerface.     

A high performance pMOSFET with two-step recessed SiGe-S/D structure for 32 nm node and beyond

A novel SiGe-S/D structure for high performance pMOSFET called two-step recessed SiGe-source/drain (S/D) is developed with careful optimization of recessed SiGe-S/D structure. With this method, hole mobility, short channel effect and S/D resistance in pMOSFET are improved compared with conventional recessed SiGe-S/D structure. To enhance device performance such as drain current drivability, SiGe region has to be closer to channel region. Then, conventional deep SiGe-S/D region with carefully optimized shallow SiGe SDE region showed additional device performance improvement without SCE degradation. As a result, high performance 24 nm gate length pMOSFET was demonstrated with drive current of 451 μA/μm at V_dd of 0.9 V and I_off of 100 nA/μm (552 μA/μm at V_dd of 1.0 V). Furthermore, by combining with V_dd scaling, we indicate the extendability of two-step recessed SiGe-S/D structure down to 15 nm node generation.     

Low-frequency low-noise circuits design using an E n-In model

In view of the limitations of a R _n-G_n model in the low frequency range and the defects of an E _n-I_n model in common use now, this paper builds a complete E _n-I_n model according to the theory of random harmonic. The parameters for the low-noise design such as the equivalent input noisy voltage E _ns, the optimum source impedance Z _sopt and the minimum noise figure F _min can be calculated accurately by using this E _n-I_n model because it considers the coherence between the noise sources fully. Moreover, this paper points out that it will cause the maximum 30% miscalculation when neglecting the effects of the correlation coefficient γ. Using the series-series circuits as an example, this paper discusses the methods for the E _n-I_n noise analysis of electronic circuits preliminarily and demonstrates its correctness through the comparison between the simulated and measured results of the minimum noise figure F _min of a single current series negative feedback circuit.     

Interaction of α radiation with iron-doped n-type silicon

Deep-level transient spectroscopy (DLTS) measurements were carried out on low-doped n-silicon before and after irradiation with 5.48 MeV α particles at room temperature with a fluence of 10¹⁰ α particles/cm². The DLTS measurements on the samples identified three electron levels E₁, E₂ and E₃ before irradiation. The deep-levels characteristic studies include emission rate signatures, activation energies, defect concentrations and capture cross sections. It was found that all pre-existing defects decreased their amplitudes during irradiation. The decrease in activation energy of level E₃ and noticeable suppression of level E₁ was also observed after irradiation. It was clearly seen that the composite peak E₃ (combination of E₂ and E₃) was successfully resolved after irradiating with α particles. α-irradiation is seen to lead a significant suppression of the iron interstitial defect, and without causing any change in its room temperature annealing characteristics.     

Thermal stress concentration factors for defects in plated-through-vias

Plated-through-vias (PTVs) are subject to thermal stress during soldering and in service. Plating and manufacturing defects in the PTV walls (barrels) can create stress concentrations that frequently become the sites of crack initiation during thermal fatigue. This is of growing concern as processing temperatures increase due to the use of lead-free solders, and boards become thicker generating increased stress as PTV aspect ratios rise, making it more difficult to achieve uniform plating thickness throughout the barrel. Finite element analysis (FEA) has been used to develop correlations that can be used to calculate the stress concentration factors (SCFs) for five typical defects as a function of various geometric parameters. In terms of their severity, the defects were ranked as: (1) rapid thickness reduction (SCF  13), (2) occasional waviness (SCF  4.6), (3) gradual thickness reduction (SCF  4), (4) wicking (SCF  4) and (5) waviness (SCF  3.1). The SCF due to an internal pad and the influence of an external pad-barrel corner crack were also investigated. The maximum stress at the defect is found by multiplying the SCF by the von Mises stress at the mid-plane of the corresponding idealized PTV. The latter can be found using either an analytical model or correlations as a function of the aspect ratio and board-to-barrel thickness ratio that were developed using FEA.     

Numerical solutions for orthogonal wavelet filters by Newton method

The wavelet transform has recently generated much interest in applied mathematics, signal processing and image coding. Mallat (1989) used the concept of the function space as a bridge to link the wavelet transform and multiresolution analysis. Daubechies (1990) added regularity conditions to find 2N, 2N10, tap coefficients for orthogonal wavelet filters. Owing to the difficulty of finding their closed solutions for large N a numerical method called the Newton method is proposed. We constructed the orthogonal wavelet filter with 2N-tap coefficients by N linear equations and N nonlinear equations. The 2N-tap, 2N10, coefficients we found are very consistent with those of Daubechies. Also, the method can be used to find the orthogonal wavelet filter with N-tap coefficients for N>10.     

An alternative method to fabricate the planar-type resonant-cavity light-emitting diodes by using silicon oxide for short-reach communications

In this article, the silicon oxide (SiO_x) planarization technique is presented to fabricate the 650-nm resonant-cavity light-emitting diodes (RCLEDs). The performances of RCLEDs are characterized by forward voltage, light output power, external quantum efficiency, emission spectrum, and dynamic response. As a result, the device with the SiO_x-planarized layer exhibits a low operating voltage of 2.3 V at 20 mA, a maximum light output power of 304 μW at 15 mA, and the best external quantum efficiency of 3% at 1.2 mA. In addition, the SiO_x-planarized device exhibits temperature insensitivity as compared to the device without it. The RCLED with a 30-μm diameter shows the maximum 3 dB frequency bandwidth of 275 MHz at a driving current of 40 mA. Finally, the RCLED with a SiO_x-planarized layer shows a clear eye-opening feature as operating at 100 Mbit/s at 20 mA. These results indicate that such LEDs are excellent candidates for use in high-speed short-reach plastic optical fiber communications.     

Delay-dependent H ∞ control for 2-D discrete state delay systems in the second FM model

This paper is concerned with the problem of delay-dependent H _∞ control for two-dimensional (2-D) discrete state delay systems described by the second Fornasini and Marchesini (FM) state-space model. Based on a summation inequality, a sufficient condition to have a delay-dependent H _∞ noise attenuation for this 2-D system is given in terms of linear matrix inequalities (LMIs). A delay-dependent optimal state feedback H _∞ controller is obtained by solving an LMI optimization problem. Finally, a simulation example of thermal processes is given to illustrate the effectiveness of the proposed result.

Enhancing optical properties of nanocrystalline silicon films with air exposure

We report the effect of air exposure and deposition temperatures, T_d, on the optical property of nanocrystalline silicon (nc-Si). The nc-Si thin films were investigated by photoluminescence (PL), optical absorption, X-ray diffraction (XRD), Fourier-transform infrared (FTIR) absorption and Raman scattering. Experimental results show the structural change from an amorphous to a nanocrystalline phase at T_d=80 °C. In addition, it suggests that T_d low condition leads to the increase in the density of SiH-related bonds and a decrease in the average grain size, δ. The oxygen absorption peak increases with the air-exposure time. The PL exhibited two peaks at around 1.75–1.78 and 2.1–2.3 eV. The PL increases and blue shifts consistently with the decrease of δ and increase of oxygen content. The first peak may be related to nanocrystallites in nc-Si films and the origin of another one may be due to defect-related oxygen. Thus, by the plasma-enhanced chemical vapor deposition (PECVD) technique at low T_d, we can produce the nc-Si films with different grain sizes, causing the corresponding luminescent properties. The new method processes advantages of low deposition temperature and effective oxidation of nc-Si on inexpensive substrates, thus making it more suitable for developing low-cost array or flexible nc-Si optoelectronic devices.     

Lateral base design rules for optimized low-frequency noise of differentially grown SiGe heterojunction bipolar transistors

The low-frequency noise dependence on lateral design parameters was investigated for SiGe heterojunction bipolar transistors fabricated by differential epitaxy. The low-frequency noise was found to vary substantially as a function of the extrinsic base design. The dominant noise sources were located either at the interface between the polycrystalline and epitaxial Si/SiGe base, in the epitaxial Si/SiGe base link region, in the base–emitter depletion region, or at the thin SiO₂ interface layer between the polysilicon and monosilicon emitter. Boron was found to passivate interfacial traps, acting as low-frequency noise sources. Generation–recombination noise with a strong dependence on the lateral electrical field was observed for some of the designs.     

Chemical compositions and optical properties of HfOxNy thin films at different substrate temperatures

High-k HfO_xN_y thin films have been grown by radio frequency (rf) reactive sputtering of metal Hf target in N₂/Ar/O₂ ambient at different substrate temperatures. The chemical compositions of the films have been investigated as a function of substrate temperature by X-ray photoelectron spectroscopy (XPS). XPS measurements showed that nitrogen concentration increases with an increase in substrate temperature. Room-temperature spectroscopic ellipsometry (SE) with photon energy 0.75–6.5 eV was used to investigate the optical properties of the films. SE results demonstrated that refractive index n increases with an increase in substrate temperature. Based on TL parameters which were obtained from the best fit results used in a simulation of the measured spectra, meanwhile, we conclude that the energy band gap (E_g) decreases with an increase in substrate temperature.     

Optical characterization of as-prepared and rapid thermal oxidized partially strain compensated Si1−x−y GexCy films

The optical properties of as-prepared and rapid thermal oxidized (RTO) heteroepitaxial Si_1−x−yGe_xC_y alloys grown on Si substrate have been characterized using spectroscopic ellipsometry. The critical points E₁, E₀′, E₂ band gaps were determined by line shape fitting in the second derivative spectra of the pseudo-dielectric functions. For as-prepared films, the E₁ gap increases with C concentration and a linear dependence on C content was observed. However, the E₂ gap decreases as the C concentration increases. For the RTO samples, the amplitude of E₂ transition reduces rapidly and the E₁ transition shifts to a lower energy. The reduction in the amplitude of E₂ transitions is due to the presence of oxide layer. A high Ge content layer and the low C content in the RTO films account for the E₁ shift to lower energy and the increase of the refractive indices.     

Interfacial layer quality effects on low-frequency noise (1/f) in p-MOSFETs with advanced gate stacks

The effects of interfacial layer quality on the low-frequency noise behavior of p-channel MOSFETs with high-κ gate dielectric and metal gate are investigated. Devices with chemically grown SiO₂ interfacial layers (0.8 nm) are compared with N₂O (0.8 nm) interfacial oxides. A 0.4 nm SiO₂ interfacial layer device is used for comparison purposes. A cross-over kind of behavior has been observed in N₂O devices, which occur at lower gate voltages (1.2–1.3 V) when normalized spectral densities and input referred noise are investigated. This behavior is found to be closely related to the observed transconductance variation in these devices. The dominant mechanism of 1/f noise is found to be Hooge’s mobility fluctuations. Hooge’s parameter, as a figure of merit, shows an increase for 0.4 nm devices when compared to 0.8 nm devices, while 0.8 nm N₂O devices confirm their cross-over nature.     

Low power, GHz class ADC for broadband applications

The design of ultra-low power (<100 mW), high-speed analogue to digital converter (ADC) is an essential element for the next generation radio telescope, the square kilometre array (SKA). CMOS technology is limited in high precision applications, such as ADCs due to the stringent requirement of device matching. Also to achieve high-speed (f_t>100 GHz) CMOS requires deep sub-micron gates (90 nm or less) where expensive phase shift masks are required. This paper describes the design and simulation of a low-power high-speed (4 GS/s) analogue to digital converter based on an InP/InGaAs heterojunction bipolar transistor (HBT). The technology used was developed at the University of Manchester using MBE growth which relied upon two novel developments. Firstly stoichiometric conditions permitted growth at a fairly low temperature of 420 °C while conserving extremely high-quality materials. Secondly dimeric phosphorus was generated from a gallium phosphide (GaP) decomposition source leading to excellent RF device properties. The DC and RF performance of the fabricated HBTs showed characteristics ideally suited to low-power IC designs, with current gain 70, low offset voltages and achieving an f_t=91 GHz and f_max=83 GHz on a 1.5×5 μm² emitter area device using fairly relaxed optical lithography.     

Noise Attributes of LWIR HDVIP HgCdTe Detectors

Time series and Fourier-transform data on high-density vertically integrated photodiodes (HDVIP) at 0 and 50 mV reverse bias in the dark have been studied. The detectors have a cutoff wavelength λ _c (60 K) of 10.5 μm. Examination of the detector current time series and Fourier-transform curves of these devices reveals a variety of interesting characteristics: (i) time series displaying switching between four states characteristic of random telegraph signal (RTS) noise, the noise current power spectrum having Lorentzian or double Lorentzian type characteristics, (ii) time series data exhibiting wave-like characteristics with the noise current power spectrum being 1/f ²-like at low frequencies, (iii) time series having a mean value independent of time with the noise current power spectrum being white, and (iv) time series nearly independent of time with the noise current power spectrum having 1/f characteristics. Although from a single array, the excess noise characteristics at low (mHz) frequencies were varied, most of the detectors measured fell into one of these four categories. The predominance of detectors examined had minimal excess low-frequency noise down to ~ 10 mHz. Detectors that displayed RTS noise in reverse bias were repeatable under subsequent measurement. However, when measured at zero bias, the same detectors exhibited no RTS noise, the noise current power spectrum being white in nature.     

On-line disjoint path routing for network capacity maximization in energy-constrained ad hoc networks

In this paper we consider on-line disjoint path routing in energy-constrained ad hoc networks. The objective is to maximize the network capacity, i.e. maximize the number of messages routed successfully by the network without any knowledge of future disjoint path connection request arrivals and generation rates. Specifically, in this paper we first present two centralized on-line algorithms for the problem. One is based on maximizing local network lifetime, which aims to minimize the transmission energy consumption, under the constraint that the local network lifetime is no less than γ times of the optimum after the realization of each disjoint path connection request, where γ is constant with 0 < γ  1. Another is based on the exponential function of energy utilization at nodes, and the competitive ratio of this latter algorithm is also analyzed if admission control mechanism is employed. We then conduct extensive experiments by simulations to analyze the performance of the proposed algorithms, in terms of network capacity, network lifetime, and the transmission energy consumption for each disjoint path connection request. The experimental results show that the proposed algorithms outperform those existing algorithms that do not take into account the power load balancing at nodes in terms of maximizing the network capacity.     

H ∞ filtering of 2-D discrete state-delayed systems

In this paper, we deal with delay-independent and delay-dependent H _∞ filtering problems for a class of two-dimensional (2-D) discrete time-invariant systems with state delays. The 2-D systems are described by local state-space (LSS) Fornasini–Marchesini (FM) second model. First, delay-dependent bounded real lemma is proposed through introducing free weighting matrices. Then the delay-independent and delay-dependent H _∞ filtering designs are developed to assure the stability and H _∞ performance γ of filtering error systems via LMIs’ feasibility. Furthermore, the minimum H _∞ norm bound γ can be obtained by solving linear convex optimization problems. Numerical examples demonstrate the effectiveness and advantages of our results.

Optimization of electric properties of high-k zirconium dioxide by varying deposition and annealing conditions

ZrO₂ thin films with a smooth surface were synthesized on silicon by atomic vapor deposition™ using Zr[OC(CH₃)₃]₄ as precursor. The maximum growth rate (7 nm min⁻¹) and strongest crystalline phase were obtained at 400 °C. The increase of the deposition temperature reduced the deposition rate to 0.5 nm min⁻¹ and changed the crystalline ZrO₂ phase from cubic/tetragonal to monoclinic. These films showed no enhancement of the dominating monoclinic phase by annealing. The values of the dielectric constant (up to 32) and leakage current density (down to 1.2×10⁻⁶ A cm⁻² at 1×10⁶ V cm⁻¹) varied depending on the deposition temperature and film thickness. The midgap density of interface states was N_it=5×10¹¹ eV⁻¹ cm⁻². The leakage current and the density of interface states were lowered by the annealing to 10⁻⁷ A cm⁻² at 1×10⁶ V cm⁻¹ and to 10¹⁰ eV⁻¹ cm⁻², respectively. However, this also led to a decrease of the dielectric constant.     

S/N Optimized, Programmable 1–10 MHz Bandwidth and 0–20 dB Gain, Precision I-Q Communication Channel for Digital Video Applications

An analog part of a digital-video quadrature demodulation scheme is built using a 7 GHz, 0.8 m biCMOS process. The scheme provides for 1–10 MHz cutoff frequency and 0–20 dB gain controls and dissipates 250 mW from a power supply of 5 V. The channel filtering is realized by two identical 4th Order Butterworth lowpass filters built with the g _m -C technique. They are equipped with cutoff programming and in-package trim tuning for cutoff adjustment. A programmable gain amplifier is placed in front of each filter for better joint noise and intermodulation performance. Such an arrangement allows to operate the filter at a maximum signal level improving the worst-case channel S/N by 6.5 dB. For the in-band components the worst case S/N is better than 41 dB, whereas THD and IMD are less than –48 dB. This single-ended channel achieves PSRR of 42 dB.