Table of contents

High-J_c YBa₂Cu₃O_7-x (YBCO) films on metal tapes have been successfully fabricated by the metalorganic deposition method using trifluoroacetates (TFA-MOD method) with optimal firing conditions. In the firing process, the key factors by which to obtain high-J_c YBCO films on metal tapes are to prevent the films from deterioration by water vapour and to avoid BaCeO₃ creation between YBCO and the CeO₂ buffer layer. With the optimal conditions, we have successfully formed a 0.24 µm thick YBCO film on a metal tape, having a J_c of 2.5 MA cm^-2 (77 K, 0 T).

We propose a new method for fabricating resistively shunted intrinsic Josephson junctions (RS-IJJs). Two mesas, measuring 6 × 6μm² and 100 × 100 μm² respectively in the a-b plane, have been fabricated in a Bi₂Sr₂CaCu₂O_8+x single crystal with a SiO layer evaporated in-between as an insulator. A gold stripe was sputtered onto the insulator to connect the mesas, thus effectively shunting the smaller one. The current–voltage (I–V) curves of such RS-IJJs have shown the typical behaviour of a resistively shunted junction with greatly reduced hysteresis and voltage jump. Under microwave irradiation at a 3 mm waveband, we have registered three types of Shapiro steps: conventional, sub-harmonic and zero-crossing. In this rapid communication, we discuss the fabrication details, as well as the observations of the Shapiro steps.

Mathematically speaking, the bound states (stationary Josephson vortices) of the magnetic flux in Josephson junctions are presented as solutions of some nonlinear differential equation, whose form and respective boundary conditions depend especially on the geometry of the junction. The generated boundary value problem contains a number of parameters (Filippov et al 1987 Phys. Lett. A 120 47) such as the bias current, the boundary magnetic field, the sizes of junction and inhomogeneities in the inhomogeneous case as well, etc. Depending on the values of both physical and geometrical parameters, the bound states can be stable or unstable with respect to small space–time perturbations. The transition of some concrete vortex from one state to another is defined as a bifurcation of this solution. The bifurcations of the possible bound states in Josephson junctions at change in the main physical parameters—the bias current and the boundary magnetic field, in the one-dimensional case, are well discussed. At the same time, the influence of the geometrical parameters (sizes) of the Josephson junctions on the bifurcations of the vortices has been studied insufficiently so far. The necessity of the above study is conditioned by at least two reasons. First, the sizes of the physical devices utilizing Josephson junctions depend themselves on their sizes and, in particular, on the length of Josephson junctions in the 'one-dimensional' case. Hence, from a 'technological' point of view it is important to minimize these sizes. In this paper we show numerically that for every nontrivial stable or unstable vortex there exists a minimal length of Josephson junction, for which this state remains stable/unstable. Second, a similar problem arises from a theoretical point of view as well. In the one-dimensional case generally adopted is the division of Josephson junctions to 'long' or 'short' depending on the inequality λ_J >> 1 (held or not) for their lengths, related to the Josephson penetration depth λ_J. Our numerical results enable us to make that definition. More strictly, we consider every Josephson junction as 'long', in which there exists at least one nontrivial stable distribution of the magnetic flux for fixed values of all the physical and other geometrical parameters. In this way, whether a sample of Josephson junction is 'long' or 'short' is determined by the specific operating conditions.

We numerically integrate the time-dependent Schrödinger equation in a single-degree-of-freedom model of SQUID with a variable potential barrier between the basis flux states. We find that linear superpositions of the basis states, with relatively little residual excitation, can be formed by pulsed modulations of the barrier, provided the pulse duration exceeds the period of small oscillations of the flux. Two pulses applied in sequence exhibit strong interference effects, which we propose to use for an experimental determination of the decoherence time in SQUIDs.

Dense MgB₂/Cu wires with Ta as a buffer layer were successfully fabricated by the powder-in-tube (PIT) method. The microstructure was investigated by optical microscopy. Magnetization measurements were carried out by using a superconducting quantum interference device (SQUID) magnetometer at magnetic fields up to 7 T from 5 K to 35 K. The transition temperature of the MgB₂ wire is around 38.4 K and the irreversibility field is 6.6 T at 5 K. The critical current density as high as 10⁵ A cm⁻² (5 K, self-field) and 10⁴ A cm⁻² (20 K, 1 T) has been obtained. The results suggest that the powder-in-tube (PIT) process is promising in preparing high-quality MgB₂ wires.

The influence of the seed distance (d_S) on the properties of grain boundaries (GBs) has been studied in order to obtain a precipitate-free GB. On the top of one monolith, two distances between two seeds were chosen, i.e. 4 and 11 mm. For d_S≃11 mm, some copper-rich phases and large pores can be observed throughout the GB. In contrast, for d_S≃4 mm, neither residual copper-rich phases nor pores are observed within the GB. The growth direction for the two samples seems to be parallel or nearly parallel to the GB. It reveals that the absence of misorientation between the two domains cannot prevent the formation of a residual copper phase layer throughout the GB; only a shorter seed distance will do this. However, element-mapping measurements revealed the occurrence of macrosegregation of the Y-211 particles on top of the monolith and centred on the GB. This tends to vanish towards the bottom of the sample and only a thin layer remains. Magneto-optical measurements performed on the sample with d_S≃4 mm, revealed that the critical current density through the GB differs according to the growth directions.

Using the liquid phase epitaxy process, a series of REBa₂Cu₃O_z/NdBa₂(Cu_1−xME_x)₃O_z double-layers were grown (REBCO/ME-NdBCO, RE = Y, Sm and Nd; ME = Mg, Ni). The high critical superconducting transition temperature (T_c) value of 93 K was achieved from the YBCO/Mg-NdBCO sample. The peak effect at a higher applied field on the critical current density (J_c) curve was unpredictably displayed by detecting samples of YBCO/ME-NdBCO, implying a pinning centre. The T_c reduction due to the ME contamination was investigated.

Preparation of extremely low-resistivity contacts (10^-9 Ω cm²) to the surface of bi-axially oriented high-temperature superconductor thin films is achieved. The low-resistivity contacts were made at the surfaces of Cu_1-xTl_xBaSrCa₂Cu₂O_8-y (Cu_1-xTl_x-1212), Cu_1-xTl_xBaSrCa₂Cu₃O_10-y (Cu_1-xTl_x-1223) and Cu_1-xTl_xBa₂Ca₂Cu₃O_12-y (Cu_1-xTl_x-1234) superconductor thin films (Khan et al 1999 Physica C 320 39). The results of contact formation at the surface of Cu_1-xTl_x-1223 thin films are presented in this paper. The parameters investigated in order to produced low-resistivity contacts were oxygen ion etching, sputtering conditions, the post-deposition annealing temperature and the ambient (air or O₂) during annealing. It was found that post-annealing in air at 500 °C after sputter deposition were the optimal conditions for low-resistivity contacts (10^-9 Ω cm²).

The reactive-field-assisted-sintering technique (RFAST) has been applied to freeze-dried decomposed precursor powders with a starting cation composition of Bi:Pb:Sr:Ca:Cu =1.7:0.3:2.0:2.5:3.5 Bi:Pb:Sr:Ca:Cu = 1.7:0.3:2.0:2.5:3.5. A short time (<15 min) RFAST treatment at 800-900 °C did not lead to the formation of Bi₂Sr₂Ca₂Cu₃O_10+x (2223-phase). But RFAST was favourable to Bi₂Sr₂CaCu₂O_8+y (2212-phase) formation particularly in samples with a low content of the 2212-phase in the precursor powder. Heat treatment (HT) after RFAST showed that the 2223 phase could form more rapidly than in freeze-dried pressed and sintered pellets produced by the traditional technique. The resultant ceramics after 70 h of annealing contain 75% 2223-phase and have T_{c(R = 0)} = 100.7 K, T_c = 109.5 K and ΔT_c = 22.4 K.

Powdered and sintered MgB₂ samples, both pure and substituted or doped, have been prepared and characterized through magnetic measurements performed from T = 5 K up to a few degrees above the transition temperature of about 39 K. For all the samples, the irreversibility line (IL) appears much lower than the H_c2–T line and very far from that of both high-T_c (YBCO) and low-T_c (Nb₃Sn) materials indicating the need to increase the pinning in this material to make it attractive for technological applications. Moreover, we have verified through different procedures that the sintered samples behave as well-connected bodies, showing no trace of granularity; therefore, the critical current density values J_c may be obtained by applying the 'critical state model' in a straightforward way. The hysteresis loop measurements allowed estimation of J_c both for powders and sintered samples and confirmed the strong field decrease of J_c, implicit in IL behaviour. We attempt to introduce defects in the MgB₂ structure by different chemical treatments like substitution of lithium on the magnesium site and doping of the precursor boron powders with aluminium and silicon. J_c always increased in doped or substituted samples (up to a factor 3) and this fact is meaningful, in particular in the light of the small level of substitution or doping we performed. The best result in terms of J_c is achieved by silicon doping that, moreover, does not significantly decrease the transition temperature.

Multi-domain YBa₂Cu₃O_y (123) superconducting fabrics with controlled domain orientations have been prepared from Y₂O₃ precursor cloths by multiple seeding during an infiltration and growth process. Long-length samples containing up to four domains with controlled orientations yielding the (110)/(110) and (100)/(100) junctions respectively have been produced by placing the seed crystals accordingly. The normal state resistivity and the transport currents along the domain boundaries are discussed and correlated with their microstructure. The (110)/(110) boundaries are observed to exhibit superior transport currents over the (100)/(100) boundaries due to the formation of relatively clean boundaries. The presence of boundary junctions is found to be effective in increasing the normal state resistivity of the samples. The coupling of domains rather than their crystallographic alignment is observed to be the dominant factor in deciding the final J_c of the multi-seeded samples.

Single crystals of the (Hg, Pb)(Ba, Sr)₂Ca₂Cu₃O_8+δ or (Hg, Pb)-1^(Ba, Sr)223 phase of 100-200 µm in size have been grown by a liquid-assisted solid-state recrystallization method employing an improved encapsulation technique developed for preventing the possible explosion of the fused-quartz ampoule. Strontium-rich crystals with up to 75% Sr-for-Ba substitution were successfully obtained. Based on the results of crystal morphology investigation and the identification of the polycrystalline secondary phases, a possible growth mechanism involving the formation of small amounts of a liquid phase and a layer-by-layer growth of the (Hg, Pb)-1^(Ba, Sr)223 crystals is proposed. The size of the obtained crystals was found to depend on both the density of the starting mixture pellet and the amount of oxygen in the ampoule. In order to establish the actual chemical composition of the crystals ICP analysis was performed. The as-grown crystals showed a sharp superconductivity transition at T_c = 115-128 K depending on the Sr content.

A natural mixture of heavy rare earth oxides has been used to prepare high critical current bulk superconducting ceramics which may be used in electrotechnical devices. Melt processing of a starting mixture of REBa₂Cu₃O₇ and Y₂BaCuO₅ powders (RE denotes natural mixture of rare earth elements), leads to a final ceramic composite REBa₂Cu₃O₇/RE₂'BaCuO₅, where RE and RE' are different mixtures of heavy rare earths and yttrium. It is found that the superconducting matrix and the core of the large RE₂'BaCuO₅ particles are yttrium rich while the small RE₂'BaCuO₅ particles are rich in heavy RE ions. The differential solubility of the rare earths in the high-temperature semisolid state leads to a strong inversion of the rare earth composition between the superconducting matrix and the insulating precipitates which is originated during the melt growth process. The superconducting properties of the new compound having natural RE mixtures are found to be very similar to those of pure YBa₂Cu₃O₇/Y₂BaCuO₅ melt textured composites.

Two types of Ag-clamped Bi₂Sr₂CaCu₂O_x (Bi-2212) thick films have been prepared using the screen-printing method and the melt-solidification process conducted with and without Bi₂Al₄O₉. The melt-solidification process with Bi₂Al₄O₉ leads to the critical-current density (J_C) having a sharper dependence on the maximum sintering temperature (T_M) and to the rise of the optimum T_M value. We compare the microstructure and J_C properties of both samples sintered at each optimum T_M to investigate the influence of the bismuth vapour dispersed from Bi₂Al₄O₉ during the melt-solidification. The results of inductively coupled plasma analysis, x-ray diffraction analysis and scanning electron microscopy studies show that the composition of the Bi-2212 layer and the precipitation of the impurities for both samples are almost the same. The intergranular coupling of the Ag-clamped sample is independent of the existence of the bismuth vapour during the melt-solidification, which is the exact opposite of the previous result obtained for a thick film with a single Ag substrate. On the other hand, it is found that the intragranular pinning of the Ag-clamped sample is damaged by the melt-solidification with Bi₂Al₄O₉. We speculate that the degradation of the intragranular pinning is caused by the trapped oxygen between the Ag substrates at high temperatures during the melt-solidification process.

Temperature-dependent electrical resistivity in a set of YBa₂Cu₃O₇/Ag thick films prepared by reaction diffusion technique is analysed to estimate the extent of granularity in them. Percolative nature of charge transfer in the films was examined by studying the excess conductivity in the superconducting order parameter fluctuation (SCOPF) region (T ≥ T_c), the approach to zero-resistance state in the tail region, the current-controlled tailing of the resistivity transition near T_c0(T_c0 < T < T_c0) and the current–voltage characteristics in the paracoherent regime (T < T_c0) with their associated exponents. The percolative current model was thus utilized to quantify granularity in the films of such parameters like the weak link resistivity across grain boundaries, current path lengthening factor arising due to grain misalignment, and due to voids and cracks. Evolution of these parameters with Ag content in the films indicates reduced structural defects (voids and cracks) at higher Ag concentrations. The result points to a mechanism of grain growth in the composites, which accounts for the improved grain growth with narrower grain size distribution in these systems as compared to that in Ag-free YBCO sintered material. These microstructural modifications cause lowering of percolation threshold from the value expected in a granular system where the grain size approaches to one of the system dimensions (the film thickness) making the conduction through grain boundary essentially two-dimensional.

The dynamics of magnetic flux distributions across a YBa₂Cu₃O_7−δ strip carrying transport current is measured using magneto-optical imaging at 20 K. The current is applied in pulses of 40–5000 ms duration and of magnitude close to the critical one, 5.5 A. During the pulse some extra flux usually penetrates the strip, so the local field increases in magnitude. When the strip is initially penetrated by flux, the local field either increases or decreases depending on both the spatial coordinate and the current magnitude. Meanwhile, the current density always tends to redistribute more uniformly. Despite the relaxation, all distributions remain qualitatively similar to the Bean-model predictions.

In this paper, we report on the recent progress in fabrication and high-frequency experiments of Bi₂Sr₂CaCu₂O_8+x (BSCCO) intrinsic Josephson junctions (IJJs). Using a newly developed double-sided fabrication process, we single out a well-defined number of IJJs in a well-defined geometry from inside a slice of a BSCCO single crystal. We discuss in detail the observations of both Shapiro steps and harmonic mixings in the terahertz region, and we indicate the possible applications in spectroscopy.

Effects of bend strains on monocore Bi-2223 tapes were studied by transport measurements, magneto-optical imaging and electron microscopy. The strain tolerance of tapes that were sintered for 100 (uniform and well-textured microstructure with high I_c) and 50 h (nonuniform microstructure with low I_c) was evaluated. The strain tolerance of the 100-h sintered tape was superior to that of the 50-h sintered tape. We have shown that the observed strain tolerance is largely determined by the interplay of nonuniform strains and microstructural variation across the tape cross section, leading to nonuniform I_c values. An explanation that accounts for an observed concentration of the critical-current density at the Ag/BSCCO interface was presented for the measured strain dependence of the retained I_c.

Biaxially textured films of YBa₂Cu₃O_7−∂ have been deposited in situ on (100) single crystal SrTiO₃ substrates by a one step spray pyrolysis process. Good out-of-plane (FWHM = 0.4°) and in-plane (FWHM = 8°) textures have been achieved for films grown at temperatures between 800 °C and 900 °C, at atmospheric pressure. The ac susceptibility measurements have shown that the films have T_c onsets of 90 K without the need of a subsequent oxygenation procedure. Magneto-optic imaging studies have indicated that the films are homogeneous, well connected and crack free. Calculations from magnetization hysteresis loops have shown that the films can have J_c values as high as 1.9 × 10⁵ A cm⁻² at 77 K in zero field.

Large grain Y–Ba–Cu–O (YBCO) superconductor doped with various amounts of depleted UO₂ and containing excess Y₂BaCuO₅ (Y-211) and Y₂O₃ have been fabricated by top seeded melt growth (TSMG). The effect of depleted UO₂ on the large grain microstructure has been studied systematically in samples with and without added Pt. It is found that UO₂ refines the size of the second phase particles in the superconducting YBa₂Cu₃O_7−δ (Y-123) matrix to dimensions of a few hundred nanometres with an approximately spherical morphology. Addition of Y₂O₃ to the uranium-doped precursor powder, rather than Y-211, yields a significantly finer distribution of second phase particles and an associated higher critical current density J_c at increased magnetic field.

LTC Josephson junction arrays (JJA) have recently gained interest as a model for the study of magnetization in superconductor materials. The presence of paramagnetic Meissner effect in such devices when field-cooled makes them an excellent candidate to discriminate between the different explanations given for such phenomenon. On the other hand, the understanding of the JJA physics is not simple due to the interplay between the non-linear behaviour of Josephson devices and the mutual induction effects in large JJA. The use of a SQUID microscope (SSM) to catch the magnetic image of the array is a distinctive feature of the experiment. So a study of SQUID read-out based on the actual JJA magnetic behaviour is needed in order to have a correct picture of what is going on. In this paper, using a simple way to reconstruct the complex array far-field viewed by the SQUID, we study how the flux read-out is dependent on array resolution, distance, tilting and noise for the typical situations in which arrays are found.

This paper was presented at the 8th International Superconductive Electronics Conference, Osaka, Japan, 19–22 June 2001

We present a new type of microwave circuits for Josephson voltage standards. Coplanar strips (CPS) are used instead of microstriplines or coplanar waveguides (CPW) to integrate the Josephson junctions in a microwave transmission line. Compared to microstripline and CPW circuits the CPS design doubles the scale of integration and simplifies the fabrication technology. Well-operating 10 V circuits in CPS design with 17,000 SIS junctions and 1 V SINIS circuits were tested in a new cryoprobe. In this cryoprobe a low loss dielectric waveguide consisting of a teflon strip is used for the microwave transmission in the 70 GHz frequency range. The dielectric waveguide together with a fibreglass-reinforced plastic supporting tube reduces the contribution of the cryoprobe to the liquid helium boil-off rate drastically.

This paper was presented at the 8th International Superconductive Electronics Conference, Osaka, Japan, 19–22 June 2001.

An HTS SQUID gradiometer consisting of a series gradiometer pickup loop in flip-chip configuration with a small washer SQUID is presented. The series gradiometer configuration is advantageous for applications without shielding because of the avoidance of circulating shielding currents in the pickup loop. With a 20 mm × 10 mm substrate for the pickup loop a large effective area of 1.25 mm² and a gradiometer baseline of 9.6 mm are achieved. The balance of the gradiometer impaired by the washer read-out SQUID can significantly be enhanced by the weighted subtraction of a reference SQUID's signal. With this a balance of 5 × 10⁴ is reached. The noise-limited magnetic field gradient resolution of the balanced SQUID gradiometer is 38 fT cm⁻¹ Hz^−1/2.

This paper was presented at the 8th International Superconductive Electronics Conference, Osaka, Japan, 19–22 June 2001.

In this paper we discuss the application of rapid single flux quantum (RSFQ) superconducting digital technology for multiuser detection in wideband code division multiple access base stations. The extraordinary speed of the RSFQ circuits allows us to implement one of the most reliable types of multiuser detector, the successive interference canceller (SIC). The RSFQ SIC consists of the partial cross-correlation unit, responsible for the regeneration of the interference at the receiver and linear combinations of the partial cross-correlations using the path gains, and the iterative linear system solver, which removes the interference from the received signal. We expect that the receiver can yield a capacity gain a factor of two greater than that of a conventional receiver.

This article was presented at the 8th International Superconductive Electronics Conference, Osaka, Japan, 19-22 June 2001.

A high-speed superconducting x-ray calorimeter that uses a transition edge sensor (TES) was developed and tested for industrial applications such as microanalysis used in the semiconductor industry. The calorimeter consisted of a Au/Ti bilayer TES formed on a bridge-type SiN_x membrane. A count rate of at least 10 000 cps was achieved by designing the calorimeter to have a decay time constant of less than 16 μs. Preliminary experiments involved using the calorimeter (cooled by using a ³He refrigerator with a base temperature below 350 mK) to detect heat pulses and x-rays. To evaluate the response of the calorimeter, we measured the dependence of the decay time constant on the bias current, the pulse height of a heat pulse and the current noise. At the optimal bias point, the calorimeter showed a measured decay time constant of 15 μs and a pulse height of 4 μA when detecting x-rays.

This paper was presented at the 8th International Superconductive Electronics Conference, Osaka, Japan, 19–22 June 2001

To design practical planar high-T_c SQUID gradiometers for magnetocardiographs, a signal-to-noise (S/N) chart that will enable the gradiometers to detect P-waves was developed. The signal was estimated as the gradient field generated by the mean current dipole of P-waves of 20 healthy volunteers. The current dipoles were obtained by using a conventional magnetocardiograph comprising low-T_c SQUID gradiometers. On the other hand, the noise was estimated from studies on previously reported gradiometers based on the limit of intrinsic gradient field noise. The obtained S/N chart indicates that a gradiometer with a baseline longer than 12 mm is necessary for detecting P-waves without averaging. The reliability of the S/N chart was confirmed by detecting actual P-waves with 100-beat averaging by using a fabricated planar high-T_c SQUID gradiometer with a baseline of 6.75 mm.

This paper was presented at the 8th International Superconductive Electronics Conference, Osaka, Japan, 19–22 June 2001.

The performance of a hot-electron bolometric (HEB) mixer based on NbN from 0.9 to 2.5 THz was investigated using a quasi-optical receiver configuration. An HEB mixer is an ultra-thin superconducting NbN strip located at the feed point of a thick normal conducting Au spiral antenna on a high-resistivity Si substrate. The active area of the mixer was 3 nm thick, 0.4 μm long and 4 μm wide. The quasi-optics consisted of an MgO hyperhemisphere with anti-reflection caps made of Kapton-JP polyimide film and an offset parabola to reduce input losses. The frequency dependence of the double-sideband receiver noise temperature was investigated at several frequencies by using a backward wave oscillator or an optically pumped far-infrared laser as the local oscillator. Results demonstrated low-noise and wide-band characteristics, below 1 K GHz⁻¹ over the measured frequency range. At 917 GHz, the measured receiver noise temperature was 550 K across a 500 MHz intermediate-frequency bandwidth centred at 1.5 GHz, which is slightly better than that of other HEB mixers at around this frequency.

This paper was presented at the 8th International Superconductive Electronics Conference, Osaka, Japan, 19–22 June 2001.

We have developed a new type of superconducting interference device (SQUID) microscope. A direct-coupled SQUID magnetometer with a high-μ metal needle was used and the substrate was machined to create a dimple for the needle at the centre of the pick-up loop. One end of the needle penetrated through the superconducting pick-up loop in a vacuum; the needle was fixed in the vacuum window with the other end at room temperature in the outside atmosphere. Several kinds of simulation using a Maxwell simulator were performed and the results were applied to our system. As a demonstration, a laser-printed output was scanned by the microscope. Line bars with a line width of 100 μm and a spacing between lines of 200 μm were clearly imaged.

This paper was presented at the 8th International Superconductive Electronics Conference, Osaka, Japan, 19–22 June 2001.

The sensitivity of galvanically coupled single-layer gradiometers on bicrystal substrates for homogeneous magnetic fields can be minimized by an increase of the balance, which is the ratio of the effective area of the antennas and the effective area of the SQUID. In this paper, we describe possible realizations of new gradiometric SQUIDs (G-SQUIDs) on 10 × 10 mm² STO-bicrystal substrates, which minimize the effective area of the SQUID, and therefore the parasitic area of the gradiometer, to 34 μm². We integrate these gradiometric SQUIDs in well-defined gradiometer antennas. The G-SQUID layouts with coupling inductances between 20 pH and 100 pH do not show any dependence of the parasitic area from the SQUID inductance. With a gradient field resolution of 1.6 pT(cm√Hz)⁻¹ (white, unshielded) at 24 pH and 1.8 pT(cm√Hz)⁻¹ at 55 pH, these gradiometers possess outstanding noise characteristics. So the balance of comparable conventional gradiometers, which show values of about 100, can be increased by more than an order of magnitude to 1070 with the new layouts.

This paper was presented at the 8th International Superconductive Electronics Conference, Osaka, Japan, 19–22 June 2001.

We have proposed a cell-based design approach based on a binary decision diagram (BDD) for the design of rapid single flux quantum (RSFQ) logic circuits. In this design approach, any logic function can be implemented by simply embedding the limited number of basic cells. We have constructed a BDD RSFQ cell library and prepared a top-down CAD environment. In this study, we investigated the tolerance of the BDD RSFQ basic cells to the circuit parameter variations. It was found that theoretical and measured dc bias margins of the basic cells agree well if we assume appropriate parameter variations due to the fabrication process. The dependence of the dc bias margin of the circuits on the circuit size was also examined, where we have implemented a 2-bit multiplexer, a 4-bit data-driven self-timed shift register and a 1-bit ALU. The low-speed test results reveal that dc bias margin of the circuits containing several hundreds of Josephson junctions is about ±15%, whereas theoretical dc bias margin is about ±33%.

This paper was presented at the 8th International Superconductive Electronics Conference, Osaka, Japan, 19–22 June 2001.

We have been developing a third-order sinc decimation filter in order to realize an oversampling sigma delta A/D converter with a second-order modulator. The sinc filter is formed from a multistage decimation structure, which is suitable for a high-speed operation. A 4-bit destructive readout (DRO) counter, which was an elementary circuit of the sinc filter and was also utilized as a first-order sinc filter with a decimation factor of 8, was designed and its proper operation was confirmed with a dc bias current margin of ±33% from a low-frequency measurement. The 4-bit DRO counter also operated up to the data rate of 4.8 Gbps with a dc bias current margin of ±29% from a high-frequency measurement. The third-order sinc filter with a decimation factor of 2, which was an elementary circuit block of our multistage decimation third-order sinc filter, was designed and its proper operation was confirmed from a low-frequency measurement. The third-order sinc filter with a decimation factor of 2 consisted of a 3-bit shift register, three 2-bit DRO counters and a 4-bit DRO counter, containing 441 junctions, and its power consumption was estimated to be 59 μW.

This paper was presented at the 8th International Superconductive Electronics Conference, Osaka, Japan, 19–22 June 2001.

We have applied an Ar ion beam etching technique to trim the critical current, I_c, and the normal resistance, R_n, of step-edge Josephson junctions which provides a means to optimize SQUID performance. By using this technique, I_c can be reduced by more than one order of magnitude and R_n can be increased to over 10 Ω. Significant improvements in SQUID performance have been demonstrated with this technique. White noise of 80 f T Hz^−1/2 was achieved from a trimmed 5 mm dc SQUID that had poor initial performance.

This paper was presented at the 8th International Superconductive Electronics Conference, Osaka, Japan, 19–22 June 2001.

The incorporation of magnetic material buffer layers for superconducting microwave devices is examined in order to allow Al₂O₃ single crystal to be used as a substrate for YBa₂Cu₃O_7−δ (YBCO) growth and to provide functionality for microwave devices. Two magnetic materials are examined; Ni_0.35Zn_0.65Fe₂O₄ (NZFO), which is a spinel-type ferrite, and Pr_0.8Ca_0.2MnO₃ (PCMO), a perovskite-type ferromagnetic material. The use of magnetic buffer layers with YBCO is applicable to magnetically controllable superconducting microwave devices. The (0001) surface of the Al₂O₃ and [113]-oriented YBCO are used in this study due to the similarity of ion configuration at the interface. YBCO grown on Al₂O₃(0001) buffered with NZFO does not exhibit superconducting behaviour, whereas that grown on the PCMO buffer superconducts at a critical temperature of 87.5 K. This result indicates that PCMO is a suitable magnetic buffer layer for YBCO growth on Al₂O₃(0001).

This paper was presented at the 8th International Superconductive Electronics Conference, Osaka, Japan, 19–22 June 2001.

In order to operate high-T_c SQUID magnetometers in an environment with no or weak magnetic shielding, we incorporated slots along the wide (3 mm) grain boundary junction in the pickup coil and the flux dam of direct-coupled SQUID magnetometers. It was confirmed that the slots formed along the wide grain boundary junction suppressed vortex motion and prevented a large increase in low-frequency noise. When we applied external fields above a threshold value of the flux dam after zero-field cooling, the flux-locked-loop output fluctuated, and the low-frequency field noise B_n increased but became stable after a time. The relaxation time after field application was longer for higher fields. A slight increase in the low-frequency noise remained, which was attributed to the fluctuation of the flux trapped in the slots having grain boundaries.

This paper was presented at the 8th International Superconductive Electronics Conference, Osaka, Japan, 19–22 June 2001.

The subgap structure in current–voltage (I–V) characteristics of a stack of intrinsic Josephson junctions in high-T_c superconductors is studied. An analytical formula for the I–V characteristics is obtained which had taken into account the influence of the dynamically breaking of charge neutrality (DBCN) in CuO₂ layers on the subgap structure. It is shown that DBCN does not affect the positions and the amplitudes of the subgap peaks, but changes the curvature of the branches in the I–V characteristics. As a possible manifestation of the non-equivalence of the junction, the experimental I–V characteristics of intrinsic Josephson junctions are presented.

This paper was presented at the 8th International Superconductive Electronics Conference, Osaka, Japan, 19–22 June 2001.