Total Leakage Current Research Articles

The effect of copper contamination on field overlap edges and perimeter junction leakage current

This work demonstrates that copper contamination present on pre gate-oxidation silicon surfaces results in yield and reliability problems particularly at field overlap edges. Similarly, the junction leakage current associated with the junction perimeter dominates the total leakage current. These detrimental device effects are shown to be caused by copper that is present close to the silicon surface even after thermal processing. Because field overlap and junction perimeter defects become relatively more important when the critical dimensions of circuits are scaled to smaller sizes, they dominate yield loss of high-density circuits, magnifying their importance for future technology generations.

Heavily irradiated double-sided wedge silicon microstrip detector

One double-sided trapezoidal microstrip n-type detector was irradiated with 1MeV neutrons up to a dose of 7.9×1013n/cm2 at PSAIF (CERN) and was characterized with laboratory measurements. Biasing resistance and interstrip impedance measurements show that the silicon bulk is inverted to p-type, as expected. After irradiation, the full depletion voltage is about 90V, at which value the total leakage current is 56μA/cm2. The full depletion voltage value obtained from the I–V characteristics agrees with the value calculated with a bulk damage model (M. Bruzzi et al., Nucl. Instr. and Meth. A 388 (1997) 345; J.A.J. Mathews et al., Atlas internal note, INDET-NO-118, 1995).

Passivation and surface leakage in CdZnTe spectrometers

The subject of surface leakage and passivation in spectrometers fabricated in high pressure Bridgman CdZnTe crystals with evaporated gold contacts has been examined. The surface leakage has been found to be the dominant mechanism of dark current leakage in metal–semiconductor–metal detectors with noninjecting contacts. Several passivation techniques have been examined with various degrees of success in terms of reducing the mean value and the power spectral density of dark noise characteristics. An insulating polymer was applied on the bare surface reducing the total leakage current by the factor of 6–7, while favorably affecting the dark noise spectral density by decreasing the corner frequency of the low frequency noise. However, a considerable surface leakage has been observed even in these apparently passivated spectrometers. A three terminal device with iso-potential high voltage contacts has been fabricated and tested, exhibiting a reduction of the total dark current leakage by approximately two orders of magnitude. With this three terminal design, the bulk resistivity of typical high pressure Bridgman crystals is correctly determined and found to be ∼2⋅1011 Ω⋅cm.

Conduction mechanism of oxide-nitride-oxide film formed on the rough polycrystalline silicon surface

The conduction mechanism of oxide-nitride-oxide film formed on the rough polycrystalline silicon surface, which has non-uniform grain sizes, is studied. At the positive bias, Poole-Frenkel (PF) conduction of holes dominates in the total leakage current. At the negative bias, although the PF conduction of holes is dominant in the nitride, Fowler-Nordheim tunneling conduction of electrons also occurs at a higher electric field. This phenomenon is discussed with a model assuming an electric field concentration at the convex edge of the plate electrode, which is formed between neighboring grains.

Numerical analysis for conduction mechanism of thin oxide-nitride-oxide films formed on rough poly-Si

The conduction mechanism of thin oxide-nitride-oxide films formed on rough poly-Si, in which grain sizes are not uniform, is studied for low and negative applied voltage. By assuming an electric field concentration at the convex edge of the plate electrode, the numerical analysis for direct tunneling (D.T.) is carried out. From the results, it is thought that the D.T. currents to the convex edge of the plate electrode dominate the total leakage currents.

Carrier leakage in blue-green II–VI semiconductor lasers

Carrier confinement in blue-green II–VI semiconductor lasers was investigated. For devices longer than 300 μm an energy barrier of 260–280 meV was found to confine the electrons, the carrier being mainly responsible for leakage, within the active region. Shorter devices show more leakage due to an increased importance of mirror losses which require higher threshold gain. Due to the low conductivity of the p-type cladding layer there is a sizable contribution of drift to the total leakage current.

Investigation of leakage current behaviour of Schottky gates on InAlAs/InGaAs/InP HFET structures by a 1D model

The influence of the layer structure on the leakage current through Schottky contacts on the heterosystem InAlAs/InGaAs is investigated with the help of a fast one-dimensional model. In the presented model the total leakage current is determined by the calculation of a tunneling component and a current caused by thermionic emission. The model includes the calculation of the sheet carrier concentration in the two-dimensional electron gas (2DEG) and the barrier lowering caused by the Schottky effect. The good agreement of simulated and measured currents in dependence on the applied voltage and the device temperature demonstrates the validity of the model. A detailed analysis of leakage current behaviour in dependence on the layer structure with regard to heterostructure FET (HFET) applications is presented in this paper.

The magnetic field dependence of R0A products in n-on-p homojunctions and p-on-n heterojunctions from Hg0.78Cd0.22Te liquid phase epitaxy films

The analysis of R0A products as a function of magnetic field in n-on-p diodes using a simple diffusion current model has previously been shown to yield both Jep/Jtotal ratio (the relative contribution of the p-side diffusion current) and μep (the minority carrier, electron mobility). In this paper, we report the good agreement between the experimental and theoretical dependence of μep on the hole concentration over a wide range between 1 x 1016 and 4 x 1017 cm−3 in n-on-p homojunction diodes fabricated on undoped p-type Hg0.78Cd0.22Te liquid phase epitaxial (LPE) films. The averaged Jep/Jtotal ratio varied between 68 and 90% with the hole concentration. These Jep/Jtotal ratios indicate that other leakage current mechanisms than the p-side diffusion current were not negligible. Also, for the first time, comparative measurements were made on p+/n heterojunction diodes consisting of As-doped Hg0.07Cd0.30Te and In-doped Hg0.78Cd0.22Te LPE layers. Unlike a typical change in R0A products by a factor of 2–3 in n-on-p homojunction diodes, the R0A products in p+/n heterojunction diodes at 7 kG were typically only 2–3% higher than that at the zero field. The typical Jep/Jtotal ratio in p+/n heterojunction diodes was about 3–4 %, which confirms the general belief that the p+ cap layer, due to the high doping and a larger bandgap, contributes very little to the total leakage current.

Electrical characteristics of oxide-nitride-oxide films formed on tunnel-structured stacked capacitors

Current conduction mechanism of oxide-nitride-oxide films formed on tunnel-structured stacked capacitor (TSSC) was studied. At positive and negative bias, the Poole-Frenkel (P-F) conduction of holes in the nitride (SiN/sub x/) film dominates the total leakage current. From the P-F plot, the relative dielectric constant of the SiN/sub x/ was calculated. The electric field inside the structure was also calculated by assuming the model that characterizes the electrode shape inside the tunnel. From the results, the reason why the reliability of the TSSC is not lowered in comparison with the conventional stacked capacitor is discussed.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Guard ring design for high voltage operation of silicon detectors

Abstract The termination of the depletion zone towards the non-depleted part of silicon affects the total device leakage current, the long term stability, the noise level and the radiation hardness of silicon detectors. This paper describes computer simulations and experiments to develop guard ring structures for use in silicon detectors requiring thick depletion layers, high operating voltages and biasing beyond depletion without increase in the leakage current and the noise. Computer simulation of a simplified structure is used to understand the influence from the oxide charges and the substrate doping concentration for a segmented guard structure with several floating diffusion strips. Results from the simulations are compared with measurements on devices. The numerical results are found to be in agreement with experimental data. It is found that segmented guard structures with floating diffusion strips have high breakdown voltages and low leakage currents. The effects of floating metal field plates over the oxide between the floating diffusion strips are studied on two different guard structures by measuring the potential on the diffusion strips and the leakage currents in the guard and active diode. The results show that floating intermediate field plates reduces the influence from oxide charges and stabilises the device against environmental influence.

Degradation of silicon AC-coupled microstrip detectors induced by radiation

Results are presented showing the radiation response of FOXFET (field-oxide field effect transistor) biased AC-coupled microstrip detectors and related test patterns to be used in the CDF microvertex detector. The radiation tolerance of detectors to gamma and proton irradiation has been tested, and the radiation induced variations of the DC electrical parameters have been analyzed. The long-term post-irradiation behavior of detector characteristics has been studied, and the relevant room-temperature annealing phenomena have been considered. The main radiation damage effects after gamma or proton irradiation of FOXFET biased microstrip detectors consist of an increase of the total leakage current, while both the detector dynamic resistance and FOXFET switching voltage decrease. No radiation-induced variation of the strip self bias Vso is measured up to 1 Mrad. The above variations are due to the generation of Si/SiO/sub 2/ interface traps and positive charge accumulation in the oxide film, as confirmed by C-V measurements on MOS capacitors. >

Radiation effects on ac-coupled microstrip silicon detectors

The radiation hardness of FOXFET biased ac-coupled microstrip detectors, foreseen for the upgraded version of the vertex detectors (SVX') of the CDF experiment, has been studied. The detectors have been irradiated with 60Co γ rays with doses up to 1 Mrad. Two basic effects have been investigated: a variation of the switching voltage up to 100% due to oxide charge accumulation and Si/SiO 2 interface trap formation, and a 12-fold increase of the total leakage current, likely to come from surface damage. Moreover, a 10-fold decrease of the FOXFET dynamic resistance is observed at low gate bias. Post-irradiation annealing effects at room temperature of the electrical detector parameters are currently under investigation, in particular on a single shot 1 Mrad irradiated device. Recovery phenomena toward the pre-irradiation values are observed only for the switching voltage due to the annealing of trapped oxide charge.

High-intensity electrostatic-field exposure system for cultured biological cells.

We describe a new system for exposing cultured biological cells that have been plated on coverslips to strong electrostatic fields at magnitudes greater than 10(3) V/cm. Techniques are described that make use of mineral oil to render insignificant electrical conduction currents (total leakage current is less than 1.0 nA or less than 0.1 nA/coverslip), joule heating (less than 10(-6) W), or current-induced magnetic fields (less than 10(-13) T) in regions inhabited by cells. The mineral oil also eliminates a reduction in the strength of the applied field, which otherwise can occur from increased electrode-to-medium impedance at the site of application. Thus the applied field is reliably specified in the vicinity of a cell membrane. Control and electrostatic field chambers are housed in a grounded metal incubator. Cylindrical mu-metal shields can be used to reduce background magnetic fields in each chamber from 40 microT static and approximately 1 microT ac to, respectively, less than 3 microT static and approximately 100 nT ac. Contamination of cells by impurity atoms that may leach from electrodes was measured by atomic-absorption spectrophotometry and found to be negligible. Stray magnetic- and electric-field components within the incubator were measured, as were background fields around the laboratory.

Thermally stimulated and leakage current analysis of neutron irriadated silicon detectors

Thermally stimulated current measurement of neutron irradiated silicon detectors have been performed. Several energy levels have been observed, revealing in particular the presence of the vacancy-oxygen complex (A center) in its double configuration, the divacancy in single (V 2 −) and double (V 2 −−) minus charge state and the vacancy-phosphorous complex (E center). The increase of the leakage current of the irradiated diodes seems to be caused mainly by the presence of the E and V 2 − defects. The E centers give the main contribution to this effect, nearly 25 times higher than V 2 − one, while the current terms related to the other defects are almost irrelevant in the generation of the total leakage current of the device. A general agreement with leakage current constant measurements and annealing behaviours has been observed.

Effect of annealing on the ac leakage components of the ZnO varistor. II. Capacitive current

Following an earlier study on the temperature dependence of the resistive component of the leakage current of a ZnO varistor, a detailed study has been undertaken in this paper to compare the temperature dependence of the resistive (IR), capacitive (IC), and total (IT) leakage currents between unannealed and 600 °C annealed samples. It is shown that, as before, an Arrhenius plot of IC and IR versus temperature can best be represented by a break in the plot comprising a high-temperature regime (125–165°C) with a high activation energy and a low-temperature regime (30–100 °C with a low activation energy. The activation energies of IR-T curves are generally higher than those of IC-T curves in both temperatures regimes. Upon annealing at the critical temperature of 600 °C, the activation energy changes for both current components, accompanied by increased stability of the varistor. For IR, the activation energy becomes higher in the low-temperature regime in agreement with previous observation, and for IC, it becomes lower in the higher-temperature regime. The relative contributions of these two current components on total current are also different in the two temperature regimes. In the low-temperature regime, the total current is mostly capacitive, and in the high-temperature regime, it is mostly resistive. The varistor is thus prone to more Joule heating in the high-temperature regime than in the low-temperature regime.

A study of the role of electrical parameters in the Electrochemical Etching (ECE) process

ABSTRACT The effect of frequency on the density and size of fission tracks and fast neutron induced recoil atom tracks was studied. It was observed that while the choice of frequency is wide (from less than 1 kHz to 4 kHz) for highly ionizing particles like fission fragments, low frequencies (0.25 – 0.5 kHz) are better suited for recoil atom tracks. The differences observed between normally incident fission tracks and recoil atom tracks are probably due to the fact that the latter tracks have a spectrum of energies (extending to almost zero energy at the lower end) and secondly that they lie at all angles in 2π geometry. Alcoholic solutions (50% v/v) produce higher densities of recoil atom tracks and give rise to higher background too. Higher voltages at any given frequency give improved results and hence are preferred. When the above data on the effect of frequency was compared with data on Leakage Current (LC) versus frequency, no apparent relationship appeared to exist. However, when two films of different thicknesses were studied under identical conditions of ECE (for normally incident fission fragments) i.e. at constant frequency of 1 kHz and one of the following conditions: same total voltage or LC or Vp/μm or it was observed that for the same total LC, tracks of nearly the same sizes were obtained. This suggests that the total LC should be an important parameter in any further investigations.

An automated dose mapping system for the TRIUMF biomedical pion beam

The dosimeter used in a transmission ionisation chamber and as the measurement is by charge integration, the sensitivity is limited only by the total leakage current of the detector, signal cable and electrometer. A minicomputer is used for on-line control of and data acquisition from the dosimetry system. Computer control also makes possible changes in the beam tune during dose mapping as is required in dynamic control of the momentum. Continuous monitoring of all the beam line elements by the same minicomputer can result in terminated or rejected dosimetry in the case of component failure. Direct computer storage of the dose maps facilitates software manipulation of the data.

Mutual relation of leakage and emission currents in the Al-Al2O3-Al system

In the present paper the analysis of VA characteristics of the leakage current determined experimentally for the Al-Al2O3-Al system is discussed. It is shown that the total leakage current is given by a sum of several local currents from the single places of the dielectric given by various mechanisms acting simultaneously. The tunnel component of the leakage current passes through a part of the dielectric area only. This current component obeys the same regularities holding for the case of emission current of electrons into vacuum. Comparison of the values of applied field intensity obtained for this tunnel component leakage current with the values corresponding to the Fowler-Nordheim equation suggests the assumption of a nonuniform potential distribution in the dielectric, i.e. the existence of a region where the potential drop agrees with the theoretical values and the region of a lower potential drop. Further, the influence of the position of potential barrier and the transmission probability over the top electrode on the emission VA characteristics is shown.

Electrical Properties of Silicon Diode Array Camera Targets Made by Boron Ion Implantation

Silicon diode targets used as the image-sensing element of video camera tubes have been successfully fabricated with one photolithographic step using boron ion implantation to form the array of ∼ 800 000 p+n diodes. Targets have been made reproducibly in this way with total leakage currents ≲ 10 nA at 12-V bias and 0–2 low-intensity white video defects per target. The results, which are insensitive to changes in dose and energy, show that heavily boron-implanted oxides can be annealed to give low-surface recombination velocity (S0 ∼ 7 cm/sec) and fixed charge (Qss/q ≃ 2.0 × 1011 cm−2). In addition, the bulk effective lifetime in the space-charge region of the implanted diodes is measured to be τ &amp;gt; 200 μsec.

Precision Continuous Current Integrator

An all-electronic precision current integrator for use with a charged particle accelerator has been developed which eliminates the substantial deadtime normally associated with multicycle current integrators utilizing relays as their basic switching elements. Integration is accomplished in the conventional fashion of charging a precision condenser with the current to be integrated, electronically noting the completion of a predetermined voltage excursion and then returning the condenser to its initial charge state for repetition of the cycle. A unique feature of this integrator is the return of the integrating condenser to its initial charge state through the transfer of a precise amount of negative charge to the condenser by an all-electronic charge transfer circuit. Integration continues during the charge transfer with no interruption and with no effect on the precision of integration. An electronic counter circuit interrupts the integration at the end of a preset number of cycles and gates off any associated equipment. Current range is approximately 1–500 μa. Accuracy at low currents is limited primarily by a total circuit leakage current of the order of 1.5×10−9 amp. At higher currents, electrometer and other circuit instabilities limit accuracy to slightly better than 0.01%.