Separate Gate Research Articles

The separated shorted-anode insulated gate bipolar transistor with the suppressed negative differential resistance regime

The separated shorted-anode LIGBT (SSA-LIGBT), which suppresses effectively the negative differential resistance regime, is investigated by performing 2-dimensional numerical simulation. In order to suppress the negative differential resistance regime, the SSA-LIGBT increases the pinch resistance by employing the highly resistive n-drift region as an electron conduction path instead of the lowly resistive n buffer region of the conventional SA-LIGBT. The SSA-LIGBT shows the remarkably decreased forward voltage drop when compared with the conventional SA-LIGBT and shows the one-order faster turn-off time than that of the LIGBT.

Formalised method for effecting multiple modes in single MOS gated power devices

A method is proposed which allows the safe operation of devices with many modes and eliminates the need for separate gates. The method relies on the adjustment of the MOS channel gate threshold voltages in various regions of the structure. Then a single MOS gate controls the operating modes, which are selected automatically as the gate voltage is changed. An example structure is described and its operation discussed. It is concluded that investing such control in the familiar single MOS gate will make this approach applicable to a large range of new and emerging power semiconductor devices and other electronic structures.

Nonlocality in Quantum Computing

The operation of a quantum computer is inherently nonlocal due to the entanglement of separated quantum gates. This paper is concerned instead with the use of nonlocal effects to construct the quantum gates themselves. A nonlocal interaction in which pairs of atoms exchange two photons is expected to produce a large nonlinear phase shift. An interesting feature of this effect is that energy is not conserved at the location of an individual atom but it is conserved overall. The preliminary results from a series of experiments in a sodium vapor cell are in apparent agreement with the predictions of the theory. These nonlinear phase shifts can be used to construct quantum logic gates, such as a Controlled-NOT.

Flow analysis of MHC molecules and other membrane proteins in isolated phagosomes

A method was developed to apply flow cytometry analysis to the characterization of individual phagosomes. Macrophages were incubated with latex beads and homogenized to release the phagosomes. Intact cells and nuclei were removed by low speed centrifugation, and a crude phagosome preparation was fixed with paraformaldehyde. Distinct optical properties of latex bead phagosomes allowed their analytic isolation from other organelles and cell fragments by flow analysis using a narrow gate based on scatter parameters. Furthermore, separate gates were established for phagosomes containing one, two and even three beads, which were sorted and examined by electron microscopy (EM). EM showed that the phagosomal membrane was closely apposed to the latex bead in most phagosomes, but some more spacious phagosomes were also observed. Phagosomes were immunolabeled and subjected to flow analysis for MHC-I and MHC-II molecules and lysosomal membrane markers (LAMPs). The proportion of LAMP-positive phagosomes increased with incubation time, reflecting maturation of phagolysosomes. Significant staining for MHC-I and MHC-II was demonstrated and remained relatively constant with time. Flow analysis of phagosomes allows the characterization and comparison of individual phagosomes, and the identification of subpopulations of phagosomes with differing membrane compositions. It also provides the advantage of analytically isolating phagosomes from other components of the cell without the need for extensive prior physical purification. Thus, it can be used to rapidly assess changes in phagosomal membrane composition as a function of phagosome maturation.

Three-color flow cytometry in the diagnosis of malignant lymphoma based on the comparative cell morphology of lymphoma cells and reactive lymphocytes.

This study examines the identification of unusual cell populations highly associated with lymphoma cells (UCP-L) in diagnostic biopsy specimens using three-color flow cytometry (3-FCM). Patterns of surface antigen expression were used to compare the morphology of distinct lymphoid cell populations present in biopsy specimens and determine the presence or absence of UCP-L. UCP-L were identified by their larger size as compared to admixed reactive lymphocytes, and the method is based on the concept that neoplastic lymphoma cells are larger than reactive lymphocytes. The comparison of relative cell sizes was determined by overlaying forward scatter histograms by multicolor gating using PAINT-A-GATE software. In order for separate gates to be set on UCP-L and reactive cell populations, UCP-L had to fulfill one or more immunophenotypic criteria. These included: (1) belonging to a subset of B cell antigen-positive cells showing restricted expression of kappa or lambda light chains; (2) belonging to a subset of CD4-positive cells having dim or absent expression of CD45RA; (3) showing alterations in antigen expression (loss, dimmer or brighter); or (4) expressing an immunophenotype that is present on only rare cell populations or is absent from reactive lymph nodes. The immunophenotypic profiles of the respective cell populations were demonstrated by cubic representations to assess more easily the co-expression of three antigens. The common morphology of UCP-L as defined by forward and side scatter grams was consistent with a 'lymphoid appearance' except in several cases of HTLV-I-positive T cell lymphoma and gammadelta T cell lymphoma. The immunophenotypic profiles of UCP-L were confirmed to correspond to the presumptive lymphoma cell population by use of a live gating procedure on the large cells, which eliminated interference by reactive cells or necrotic tissue fragments. Using this method, we identified UCP-L in 208 of 293 (71%) consecutive cases of non-Hodgkin's lymphomas, while no UCP-L were seen in 72 cases of non-specific hyperplasia of lymph nodes. Twenty-seven cases could not properly be examined about the existence of UCP-L because of massive necrosis, extensive fibrosis or strong non-specific staining reactions of unknown cause. When those cases were eliminated from the analysis, 80% of non-Hodgkin's lymphoma were found to contain UCP-L. In B cell lymphoma, the incidence of UCP-L in nodal lymphomas (80%) was much higher than in extranodal lymphomas (47%). Only one of 21 cases of Hodgkin's lymphoma was found to have UCP-L. The 3-FCM procedure was validated by the combined use of immunohistochemistry, morphologic examination, cytogenetic and antigen receptor gene rearrangement analysis by Southern blot hybridization. Our findings indicate that detection of UCP-L by 3-FCM is a reliable method to distinguish non-Hodgkin lymphomas from reactive hyperplasias in the majority of cases, even when the reactive cell population predominates over the malignant cell population.

Species-specific voltage-gating properties of connexin-45 junctions expressed in Xenopus oocytes

Gap junctions composed of connexin-45 (Cx45) homologs from four species, zebrafish, chicken, mouse, and human, were expressed in pairs of Xenopus oocytes. The macroscopic conductance (gj) of all Cx45 junctions was modulated by transjunctional voltage (Vj) and by the inside-outside voltage (Vm), and the modulation was species specific. Although their gating characteristics varied in voltage sensitivity and kinetics, the four Cx45 junctions shared 1) maximum conductance at Vj = 0 and symmetrical gj reduction in response to positive and negative Vj of low amplitude, with little residual conductance; and 2) gj increases in response to simultaneous depolarization of the paired cells. The formation of hybrid channels, comprising Cx45 hemichannels from different species, allowed us to infer that two separate gates exist, one in each hemichannel, and that each Cx45 hemichannel is closed by the negativity of Vj on its cytoplasmic side. Interestingly, the Vm dependence of hybrid channels also suggests the presence of two gates in series, one Vm gate in each hemichannel. Thus the Vj and Vm dependence provides evidence that two independent voltage gates in each Cx45 hemichannel exist, reacting through specific voltage sensors and operating by different mechanisms, properties that have evolved divergently among species.

Vertical Single Electron Transistors With Separate Gates

We demonstrate independent gate action in a vertical sub-micron single electron transistor with four separate gates fabricated by a new technology from a double barrier structure starting material. When the gate voltage is swept over a given range, the number of conductance oscillations increases systematically as the number of gates “squeezing” the quantum dot is varied from one to four. Alternatively, the average period of the conductance oscillations is essentially independent of the number of “squeezing” gates if the gate voltage is appropriately scaled to an “effective gate voltage”. We show that subtle changes occur in the appearance of the conductance oscillations when the lateral geometry of the quantum dot is controllably deformed.

Multiple-gated submicron vertical tunnelling structures

We propose and demonstrate a new hybrid technology for fabricating submicron vertical resonant tunnelling structures with separate gates which has the potential to allow single-electron charging of, and single-electron tunnelling through, quantum dot structures containing just a `few' electrons to be investigated in new ways.

Effect of interhemispheric sections of the hypothalamus on milk-ejection bursts of supraoptic oxytocin neurones during bilateral and unilateral suckling in the rat

The burst activity of oxytocin neurones was recorded from the supraoptic nucleus of lactating rats with or without interhemispheric sectioning of the hypothalamus during bilateral or unilateral suckling. The results showed that extensive interhemispheric sectioning did not abolish the burst of oxytocin neurones during bilateral and contralateral suckling (100% of 25 neurones and 95% of 20 neurones, respectively), but significantly reduced the number of oxytocin neurones that showed the burst during ipsilateral suckling (5.3% of 19 neurones). To orientate the crossing site of the signals for bilateral synchronization of the bursts of oxytocin neurones, interhemispheric sectioning of the rostral or caudal hypothalamus was attempted, but either sectioning partially blocked the occurrence of the bursts during ipsilateral suckling. These results suggest that there are two separate gates located on different sides of the hypothalamus and that the neural connections between the gates seem distributed diffusely.

IVBT-documented platelet function correlates with flow cytometric data

Thrombocytopenic patients with identical platelet counts often show different bleeding tendencies owing to significant differences in the platelet function. This could be demonstrated by the in vitro bleeding test (IVBT). Using flow cytometry, we tried to find characteristics of platelet antigen expression in order to explain these differences in function. Thirty patients with bone marrow hypoplasia receiving 65 platelet transfusions (mainly from a cell separator) were observed for 3 to 29 days. Size, granulation and fluorescence of platelet-rich plasma ( n = 522 samples) were evaluated using monoclonal antibodies against GP IIIb (collagen receptor), GP IIb/IIIa (fibrinogen receptor) and GP Ib (thrombin receptor). We defined separate gates for each antibody using the results from 50 normals and by laying an orthograde cross over the gate to divide the gate into four equal quadrants. The platelet populations were divided into four different groups according to the occlusion time (OT) of the IVBT and the Simplate® time (ST). The thrombocytes with the most impaired function (OT ≥ 485 s/ST > 30 min) had significantly less platelet fluorescence when marked with antibodies against GP IIIb and GP Ib than those with short OT and ST (OT < 100 s/ST < 15 min). Similar results were obtained when evaluating the data relative to the bone marrow status: patients with < 1000 WBC/μl showed significantly less platelet fluorescence when marked with anti-GP IIIb and anti-GP Ib than thrombocytopenic patients, who had a spontaneous platelet rise beyond 30,000 platelets/μl a few days later. One day after platelet transfusion, significantly more platelets with high GP IIIb and Ib expression could be found. We were also able to document better transfusion efficacy of platelet concentrates with high GP IIIb and Ib expression. Finally, patients with high bleeding scores showed less GP Ib expression on the platelets than patients with low bleeding scores. In summary, the IVBT-documented platelet function clearly corresponded to an increased expression of the collagen receptor and the thrombin receptor of platelets.

Distinct voltage-dependent gating behaviours of a swelling-activated chloride current in human epithelial cells.

1. The swelling-activated chloride current is critical in the homeostatic regulatory volume decrease (RVD) of both excitable and non-excitable cells. Although not activated by voltage, it displays kinetic behaviour similar to voltage-gated Shaker-type potassium currents. We have studied the voltage-dependent properties of this current in single T84 human cell line epithelial cells using whole-cell patch clamp methodology. 2. An external anion permeability sequence of I- > Cl- > methanesulphonate (MeSO3-) was observed for the swelling-activated current. Extracellular application of the chloride channel blocker DIDS (100 microM) resulted in approximately 50% block of the current in a voltage-dependent manner. 3. At positive membrane potentials, the swelling-activated chloride current undergoes time-dependent inactivation. Following such inactivation, recovery of both the inward and outward components of the macroscopic current was found to be voltage dependent. The time constants describing these two individual recovery processes were identical over a range of membrane potentials. In addition, the magnitude of current recovery was directly dependent upon the degree prior inactivation of current at positive voltage. 4. We further observed that the swelling-activated current undergoes a form of steady-state, voltage-dependent inactivation that appears to differ from the inactivation observed at positive potentials. This steady-state inactivation occurred over the physiological voltage range, with a membrane potential at half-maximal inactivation (V1/2) of -72 mV, and differed from the time-dependent inactivation observed at positive membrane potentials, which occurred with a V1/2 of 40 mV. These observations demonstrate two distinct forms of voltage-dependent inactivation, probably reflecting two separate gating processes at the level of the channel. 5. These latter properties are thus anticipated to regulate voltage-dependent chloride efflux under cell swelling conditions and further influence RVD and membrane excitability in cells generating action potentials.

MMIC KA band oscillator

The design of two MMIC KA band oscillators compatible with the Philips DO2AH process is outlined. Both designs use 0.2 µm PHEMT devices: the first employs separate gate and drain bias supplies, the other includes bias resistors on chip. Design procedures and results obtained from simulations and measurements are presented.

Simplified ohmic and Schottky contact formation for field effect transistors using the single layer integrated metal field effect transistor (SLIMFET) process

A new III-V semiconductor device fabrication process for GaAs-based field effect transistors (FET) is presented which uses a single lithographic process and metal deposition step to form both the ohmic drain/source contacts and the Schottky gate contact concurrently. This single layer integrated metal FET (SLIMFET) process simplifies the fabrication process by eliminating an additional lithographic step for gate definition, a separate gate metallization step, and thermal annealing for ohmic contact formation. The SLIMFET process requires a FET structure which incorporates a compositionally graded In/sub x/Ga/sub 1-x/As cap layer to form low resistance, nonalloyed ohmic contacts using standard Schottky metals. The SLIMFET process also uses a Si/sub 3/N/sub 4/ mask to provide selective removal of the InGaAs ohmic layers from the gate region prior to metallization without requiring an additional lithographic step. GaAs MESFET devices were fabricated using this new SLIMFET process which achieved DC and RF performance comparable to GaAs MESFET's fabricated by conventional methods.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Accurate Single-Section Model of Mosfet Including Substrate Resistivity

control devices as a 4 terminal, 3-port active network. The purpose of this work is to present a new model in the general case of the MOSFET structure including substrate control effects by suitable approximation of the MOSFET terminal admittances. Device admittances are calculated in terms of the ’Uniform RC double transmission line’. The combined gate-substrate excitation mode can be treated in terms of an RC transmission line having a single, non-uniform capacitance distribution and a uniform resistance similar to the case of negligible substrate effects. The separate gate and substrate excitation modes, however, require for their representation an RC transmission line analog having two nonuniform capacitance distributions. We have the admittance parameters as:

Cmos Polysilicon Thin Film Transistors with Simultaneously Deposited Layers for Source-Drain and Gate

AbstractA new process architecture for fabricating CMOS thin film transistors (TFTs) using in-situ-doped polysilicon source-drain layers is proposed. In the new architecture, a top-gate n-channel TFT and a bottom-gate p-channel TFT, or vice versa, form a CMOS pair. This allows an n+ - doped polysilicon bottom (or top) layer to serve simultaneously as the source-drain layer of the n-channel TFTs and the gate layer of the p-channel TFTs; while a p+ - doped polysilicon top (or bottom) layer serves as the source-drain layer of the p-channel TFT and the gate layer of the n-channel TFT. It thus eliminates the deposition of a separate doped gate layer normally required in the conventional process flow. In addition, a thin tri-layer stack, consisting of undoped-poly / gate dielectric / undoped-poly, separates the two doped polysilicon layers, thus allowing the use of a single island mask to define the channel regions for both the n- and p-channel TFTs. As a result, the photolithographic steps are also reduced by one mask. Working n- and p-channel TFTs with both top- and bottomgate structures, obtained by reversing the dopant types of the top and the bottom layers, have been successfully demonstrated using low-temperature (&lt; 600 °C) polysilicon technology.

Three-dimensional CMOS IC's Fabricated by using beam recrystallization

A three-dimensional (3-D) CMOS integrated circuit with a structure, in which one type of transistor is fabricated directly above a transistor of the opposite type with separate gates and an insulator in between, has successfully been fabricated by using laser beam recrystallization. Seven-stage ring oscillators fabricated in the 3-D structure have a propagation delay of 8.2 ns. In the present experiment, a double-layer of silicon-nitride and phospho-silicate-glass (PSG) film has been used as an intermediate insulating layer between the top and the bottom devices. This CMOS structure and the process technology we have developed here can be the basis for realizing a multilayered 3-D device composed of vertically stacked transistors with separate gates and an insulating layer in between.

A novel p-n junction polycrystalline silicon gate MOSFET†

A segmented multiple gate MOSFET utilizing a single level of polysilicon gate layer was fabricated and characterized. Data presented for both p- and n-channel devices in which the polysilicon gate layer is segmented into three separate lateral gates by alternate p- and n-typo doping. It was found that current conduction takes place from the source to the drain by applying appropriate potentials to the end gates oven though the central gate was left floating. The harrier potential of the diodes formed within the polysilicon gate layer was measured together with their threshold voltage which was found to be substantially different under the p- and n-type gates..It is concluded that this difference is only partially duo to the difference in work function of the two types of gates ; the rest is due to a difference in the effective surface state density Qss which was, as expected, always found to be positive. It was also found that for both p- and n-channel devices, the Qss was larger for n+ gate in comparison with the pH gate. Transfer characteristics of the device were also measured and modelled satisfactorily by applying the standard MOSFET theory suitably modified to account for the different threshold voltage of the p+ and n+ polygate. Based on this DC characterization, it is proposed that the device would be suitable for CCD or logic gates in which the single polysilicon gate layer could afford higher packing density and yield.

Stacked MOSFET's in a single film of laser-recrystallized PolySilicon

The feasibility of building stacked MOSFET's in a single laser recrystallized polysilicon film is demonstrated. Devices fabricated in this initial study use separate gates to obtain independent enhancement mode behavior on each surface of the recrystallized film. The transistors worked simultaneously and inverter action was demonstrated using a CMOS configuration.

Submicron gate MOSFET's with channel-doped separate gate structures (SG-MOSFET's)

Submicron gate MOSFET's with a new device structure are presented. The device features gate separation between the source and gate and between the gate and drain. The minimum gate length limited by V TH lowering is extended into the submicron range. Experimental results showed pentode-like current-voltage characteristics without punchthtough, even in the submicron range. Experimental results of inverter circuits and theoretical analysis predict high-speed operation in the subnanosecond region.

Theoretical study of a channel-doped separate gate Si MOSFET (SG-MOSFET) by two-dimensional computer simulation

A new device structure is proposed for Si MOSFET, featuring an insulated gate structure, channel doping, and finite spacing between gate and source and between gate and drain. Two-dimensional numerical analysis shows that punchthrough is suppressed and that minimum gate length, limited bypunchthrough or V T shift, is extended into the submicrometer range.