Search

We describe a dual current source designed to provide static flux biases for a superconducting qubit and for the Superconducting QUantum Interference Device (SQUID) which measures the qubit state. The source combines digitally programmable potentiometers with a stabilized voltage source. Each channel has a maximum output of ±1 mA, and can be adjusted with an accuracy of about ±1 nA. Both current supplies are fully computer controlled and designed not to inject digital noise into the quantum bit and SQUID during manipulation and measurement of the flux. For a 275 μA setting, the measured noise current is 2.6 parts per million (ppm) rms, in a bandwidth of 0.0017–10 Hz, from which we estimate dephasing times of hundreds of nanoseconds in the particular case of our own qubit design. By resetting the current every 10 min, we are able to reduce the drift to no more than 5 ppm at a current of 750 μA over a period of 3 days. The current source has been implemented without thermal regulation inside a radiofrequenc...

DC Superconducting Quantum Interference Devices (SQUIDs) are widely used to amplify low-level, radio frequency (rf) electrical signals. SQUID amplifiers offer low noise, high gain, and low power dissipation. One method of implementing a SQUID rf amplifier for frequencies from a few hundred megahertz to several gigahertz is to operate the integrated input coil on top of the SQUID washer as a microstrip resonator. This is achieved by applying the input signal between one end of the coil and the SQUID washer, which acts as a groundplane; the other end of the coil is left open. Substantial levels of gain can be achieved from the microstrip SQUID amplifier for a signal frequency at the fundamental resonant frequency of the microstrip, at which the length of the microstrip is equal to one-half wavelength, λ/2. Since the length of the microstrip has to be made shorter for higher frequencies, however, the mutual inductance between a SQUID with a given geometry and the microstrip—and thus the gain—decreases with i...

Computer models are presented for a microstrip amplifier based on a dc superconducting quantum interference device (SQUID). In this device, the signal is applied between one end of the spiral input coil and the square washer on which it is deposited. The amplifier exhibits substantial power gain when the signal frequency is such that a half wavelength is approximately equal to the length of the microstrip formed by the coil and the groundplane. The resonant frequency is lowered significantly by the inductance of the square washer transformed into the input coil; this reduction is consistent with predictions of a simple model and with analog simulations. With the washer grounded, the gain of the amplifier peaks at a frequency that is lowered from the unloaded resonant frequency by the damping of the resistance associated with the source. The position and magnitude of the peak are in good agreement with both a lumped circuit model and with a model representing the microstrip as a transmission line. When the...

We describe a high transition temperature superconducting, first-order gradiometer intended for biomagnetic measurements in an unshielded environment. The gradiometer involves a single-layer, planar flux transformer with two loops of unequal size, the smaller of which is inductively coupled to the pickup loop of a directly coupled magnetometer. In this configuration, the presence of the flux transformer reduces the sensitivity of the magnetometer by only about 5%. The flux transformer is patterned in a thin film of YBa2Cu3O7−δ deposited on a 100 mm diam wafer, and has a baseline of 48 mm. The flux transformer and magnetometer substrates are permanently bonded together in a flipchip arrangement. The common mode rejection of uniform magnetic field fluctuations in any direction is better than 1 part per 100. The outputs of two such gradiometers are subtracted to form a second-order gradiometer, which rejects first-order gradient fluctuations to about 1 part in 100. With the aid of three orthogonally mounted ...

We describe the design and construction of a single-coil, two-channel probe for the detection of low-field magnetic resonance using dynamic nuclear polarization (DNP). The high-frequency channel of the probe, which is used to saturate the electron spins, is tuned to the electron Larmor frequency, 75 MHz at 2.7 mT, and matched to 50 Ω. Low-field, 1H nuclear magnetic resonance (NMR) is detected through the second, low-frequency channel at frequencies

A microscope has been constructed in which a high‐transition temperature dc superconducting quantum interference device (SQUID) in vacuum is brought to within 140 μm of samples at room temperature and pressure. The SQUID is mounted on the upper end of a sapphire rod, cooled to liquid nitrogen temperature, below a 75‐μm thick sapphire window. Samples can either be placed directly on the window or scanned over the SQUID to produce magnetic images. A square‐washer SQUID with inner and outer dimensions of 30 and 50 μm can resolve 130 μm magnetic features. The instrument is intended primarily for biological applications.

Measurements have been made of the linearity of a high transition temperature dc superconducting quantum interference device (SQUID) operated at 77 K with 130 kHz flux modulation in a flux‐locked loop. The degree of nonlinearity was determined from harmonic generation. A sinusoidal magnetic flux with harmonic content less than −130 dB was applied to the SQUID, which was cooled in a magnetic field below 10−7 T, and the harmonics at the output of the flux‐locked loop were measured with a spectrum analyzer. For input signals at frequencies up to 248 Hz and amplitudes up to 20Φ0 rms (Φ0 is the flux quantum), the second, third, and fourth harmonics were each at least 115 dB below the fundamental. At higher frequencies the harmonic content began to increase because of the reduction in the open‐loop gain of the flux‐locked loop. The magnitude of the harmonics was not measurably changed when the SQUID was cooled in a field of 100 μT. The amplitudes of the even harmonics depended critically on the amplitude of the...

A spectrometer based on a dc superconducting quantum interference device (SQUID) has been developed for the direct detection of nuclear magnetic resonance (NMR) or nuclear quadrupole resonance (NQR) at frequencies up to 5 MHz. The sample is coupled to the input coil of the niobium‐based SQUID via a nonresonant superconducting circuit. The flux locked loop involves the direct offset integration technique with additional positive feedback in which the output of the SQUID is coupled directly to a low‐noise preamplifier. Precession of the nuclear quadrupole spins is induced by a magnetic field pulse with the feedback circuit disabled; subsequently, flux locked operation is restored and the SQUID amplifies the signal produced by the nuclear free induction signal. The spectrometer has been used to detect 27Al NQR signals in ruby (Al2O3[Cr3+]) at 359 and 714 kHz.

A study was carried out to observe a heat-affected zone (HAZ) in a deformed Fe-Cr-Ni system containing α′ martensite using high-Tc superconducting quantum interference device (SQUID) microscopy. Microstructure and remanent magnetization images were studied by an optical microscope and a SQUID microscope, respectively. The HAZ, in which only the face-centered-cubic austenite phase exists, could be visualized by the SQUID microscope. It was also found that the SQUID images were consistent with the results from the microstructural analysis. It was concluded that a simple SQUID measurement may serve as an effective method for a nondestructive evaluation of ferromagnetic steel phases by correlating remanent magnetization images to microstructural characteristics.

A series of five dc superconducting quantum interference devices (SQUIDs) have been operated as microstrip amplifiers at frequencies ranging from 2.2 to 7.4 GHz. In these devices, the signal is connected between the SQUID washer and coil, which acts as a microstrip resonator. The gain measured at 4.2 K ranged from 12±1 to 6±1 dB. The noise temperature of three devices at 4.2 K in the frequency range 2.2–4 GHz was between 1 and 2 K, and the saturation temperature was between 150 and 250 K. Applications of these devices include readout for axion detectors, and intermediate-frequency amplifiers for superconductor–insulator–superconductor and hot-electron bolometer mixers.

A prototype three‐axis magnetometer has been developed, intended for geophysical applications, involving high‐transition temperature dc superconducting quantum interference devices (SQUIDs). Each SQUID was fabricated from a thin film of YBa2Cu3O7−x deposited on a SrTiO3 bicrystal to form two Josephson junctions, and was directly coupled to a single‐turn pickup loop patterned in the same film. The three orthogonally mounted sensors were immersed in liquid nitrogen in a fiberglass Dewar and operated in separate flux‐locked loops. In nominally zero ambient magnetic field, the best magnetic field resolution was 170 fT Hz−1/2 in the white noise, increasing to 225 fT Hz−1/2 at 1 Hz. The dynamic range was ±1.6×106 Hz1/2 in the white noise, and the highest slew rate achieved was 1.6 mT s−1, at 900 Hz. The 3 dB point in the frequency response was about 90 kHz. The three channels were mutually orthogonal to about 1%. The three‐axis magnetometer was also operated outside the laboratory in the presence of 60 Hz and r...

Springs for scanning force microscopy (SFM) are described which incorporate two separate advances: Their shape significantly reduces the effect of photon shot noise in the optical beam deflection method, and they are endowed with very sharp tips through a unique process based on controlled fracture of micromachined structures. With this readout method, the noise in the measured tip position is limited by photon shot noise, which is proportional to the ratio of the effective length of the spring to the length of the mirror. Cantilever springs optimized for this parameter are described, with a measured white noise level of 29 fm/√Hz at 210 μW of laser power; torsion springs for which the mirror length exceeds the effective length yielded a theoretical noise level of 7.2 fm/√Hz at this power. Cantilever springs optimized for this parameter are described, with a measured white noise level of 29 fm/√Hz at 210 μW of laser power; torsion springs for which the mirror length exceeds the effective length yielded a theoretical noise level of 7.2 fm/√Hz at this power. The latter is lower than the noise of an ideal fiber interferometer at the same power, demonstrating that the optical‐beam‐deflection method can be made competitive with interferometric methods for noise‐critical SFM applications. This novel tip making technique generates cube‐corner‐shaped tips, with a radius which is unresolvable in scanning electron microscope images, implying that it is less than 10 nm. These tips are intended for samples whose features are in the size range 0.5–20 nm, where the tip curvature is the resolution limiting factor. To demonstrate their performance in this regime, the tips were used to image single collagen molecules.

A detailed study has been made of the low‐frequency excess noise of rf superconducting quantum interference devices (SQUIDs), fabricated from thin niobium films and operated at 4.2 K, with rf bias frequencies of 0.15, 1.7, and 3 GHz. When the SQUIDs were operated in an open‐loop configuration in the absence of low‐frequency flux modulation, the demodulated rf voltage exhibited a substantial level 1/f noise, which was essentially independent of the rf bias frequency. As the rf bias frequency was increased, the crossover frequency at which the 1/f noise power was equal to the white noise power moved to higher frequencies, because of the reduction in white noise. On the other hand, when the SQUID was flux modulated at 50 kHz and operated in a flux locked loop, no 1/f noise was observed at frequencies above 0.5 Hz. A detailed description of how the combination of rf bias and flux modulation removes 1/f noise due to critical current fluctuations is given. Thus, the results demonstrate that the 1/f noise observ...

The construction of microelectronic circuits from high‐transition‐temperature (Tc) superconductors requires techniques for producing thin‐film wires, insulating crossovers, and vias (window contacts) between wires. Together, these three components form a superconducting interconnect technology. The challenges encountered in developing such a technology for high‐Tc superconductors involve factors associated with the materials, the circuits and the fabrication techniques. The use of pulsed laser deposition in conjunction with shadow mask patterning, photolithographic pattern definition, acid etching, ion‐beam etching, and surface cleaning to produce multilayer interconnects from YBa2Cu3O7−x (YBCO) is discussed. These processes have been used to construct a variety of passive high‐temperature superconducting components and circuits, including crossovers, window contacts, multiturn coils, and flux transformers. Integrated magnetometers incorporating superconducting quantum interference devices, multichip modules with semiconductor die bonded to YBCO interconnect structures, and analog‐to‐digital converters have also been successfully demonstrated.

We describe a gradiometer based on a high-transition temperature Superconducting Quantum Interference Device (SQUID) for improving the sensitivity of a SQUID-based biosensor. The first-derivative gradiometer, fabricated from a single layer of YBa{sub 2}Cu{sub 3}O{sub 7-x}, has a baseline of 480 {micro}m and a balance against uniform fields of 1 part in 150. Used in our SQUID ''microscope,'' it reduces parasitic magnetic fields generated by the measurement process to the level of the SQUID noise. The gradiometer-based microscope is two orders of magnitude more sensitive to super paramagnetic nanoparticles bound to biological targets than our earlier magnetometer-based microscope.

We present the multiplexed readout of two gamma-ray microcalorimeters made from transition-edge sensors. We use a frequency-domain multiplexing scheme in which each sensor is biased at an identifying frequency. We show that the energy resolution of the sensors is unaffected by multiplexing, and that crosstalk between the sensors is negligible. Our results indicate the feasibility of multiplexing 30 sensors or more to one readout line.

Point contact junctions made from two pieces of MgB2 can be adjusted to exhibit either superconductor-insulator-superconductor (SIS) or superconductor-normal metal-superconductor (SNS) current-voltage characteristics. The SIS characteristics are in good agreement with the standard tunneling model for s-wave superconductors, and yield an energy gap of (2.02 +/- 0.08) meV. The SNS characteristics are in good agreement with the predictions of the resistively-shunted junction model. DC Superconducting QUantum Interference Devices made from two SNS junctions yield magnetic field noise as low as 35 fT/Hz^{1/2} at 19 K., 4 pages, 4 figures

Each magnetometer consisted of a dc superconducting quantum interference device (SQUID), with the YBa2Cu3O7−x washer patterned into 4 μm lines, coupled to a multiturn, multilayer flux transformer containing flux dams. The noise at 1 Hz did not increase when bare SQUIDs or magnetometers were cooled and operated in fields up to values well above the magnetic field of the earth. When the magnetic field was changed, the noise in a bare SQUID was constant up to a threshold field of 12 μT. The addition of a flux transformer containing flux dams increased the magnetic field sensitivity by a factor of 43, while reducing the threshold field only moderately, to 5 μT.

A magnetic flux stabilization scheme involving a low-frequency flux-locked loop has been developed to regulate the gain of a radio-frequency amplifier based on a superconducting quantum inteference device (SQUID). The flux-locked loop largely eliminates drifts in the gain arising from drifts in the flux bias. Tests on a microstrip SQUID amplifier operating at 777 MHz inside a superconducting shield show that the peak-to-peak drift in the gain is no more than 0.3 dB/day.

The harmonic distortion and intermodulation products have been measured for a microstrip amplifier based on a dc superconducting quantum interference device. When the flux bias is adjusted to maximize the gain, in general both second harmonic and a series of odd harmonics are observed. The magnitude of the third harmonic is in good agreement with a simple model. The magnitude of the second harmonic, however, depends critically on the flux bias and is generally a minimum when the gain is slightly less than the maximum value. Similar results are found for the two lowest order intermodulation products.

A dc superconducting quantum interference device (SQUID) with a resonant microstrip input is operated as an amplifier at temperatures down to 20 mK. A second SQUID is used as a postamplifier. Below about 100 mK, the noise temperature is 52±20 mK at 538 MHz, estimated from measurements of signal-to-noise ratio, and 47±10 mK at 519 MHz, estimated from the noise generated by a resonant circuit coupled to the input. The quantum-limited noise temperatures are 26 and 25 mK, respectively. The measured noise temperature is limited by hot electrons generated by the bias current.

We present the design and experimental evaluation of a superconducting quantum interference device (SQUID) multiplexer for an array of low-temperature sensors. Each sensor is inductively coupled to a superconducting summing loop which, in turn, is inductively coupled to the readout SQUID. The flux-locked loop of the SQUID is used to null the current in the summing loop and thus cancel crosstalk. The sensors are biased with an alternating current, each with a separate frequency, and the individual sensor signals are separated by lock-in detection at the SQUID output. We have fabricated a prototype 8 channel multiplexer and discuss the application to a larger array.

A sensitive spectrometer, based on a dc superconducting quantum interference device, for the direct detection of low‐frequency pulsed nuclear magnetic resonance (NMR) and nuclear quadrupole resonance (NQR), is described. The frequency response extends from about 10 to 200 kHz, and the recovery time after the magnetic pulse is removed is typically 50 μs. As examples, NMR spectra are shown from Pt and Cu metal powders in a magnetic field of 6 mT, and NQR spectra are shown from 2D in a tunneling methyl group and 14N in NH4ClO4.

A spectrometer incorporating a high transition temperature dc superconducting quantum interference device (SQUID) is used to obtain nuclear magnetic resonance signals from protons in mineral oil at room temperature in fields up to 3 mT. The spatial separation between the SQUID magnetometer at 77 K and the sample at room temperature is less than 1 mm. At 2 mT, the signal is easily resolved in a single scan. Two-dimensional images of samples consisting of pieces of lucite or glass immersed in mineral oil are obtained at 2 mT.

We have fabricated planar first-derivative gradiometers in which an asymmetric flux transformer patterned in a single-layer YBa2Cu3O7−x film is permanently bonded to a directly coupled magnetometer based on a dc superconducting quantum interference device (SQUID). The gradiometer base line is 48 mm. The common mode rejection of uniform magnetic fields is typically one part in a few hundred. The outputs of two such gradiometers are subtracted digitally to form a second-derivative gradiometer. Additionally, three orthogonal SQUID magnetometers are used to reduce the residual response to uniform magnetic fields to better than 100 ppm. The system is able to detect magnetic signals from a human heart in an unshielded environment.

A dc superconducting quantum interference device (SQUID) with a resonant microstrip input and a cooled heterostructure field-effect transistor as a postamplifier is used as a radio-frequency amplifier in the frequency range 90–500 MHz. At liquid 3He temperatures, gains of 24 and 20 dB and intrinsic noise temperatures of 0.06±0.02 and 0.12±0.10 K were achieved at 89.6 and 438 MHz, respectively. The system noise temperature at 438 MHz was also estimated from the Nyquist noise produced by a resonant circuit coupled to the input of the microstrip SQUID.

We describe the concept of the spin‐polarized scanning tunneling microscope. It consists of a ferromagnetic tip with saturated magnetization from which spin‐polarized electrons tunnel into a ferromagnetic sample which has its axis of magnetization aligned with that of the tip. When the magnetization of the sample is alternated periodically from parallel to antiparallel, a portion of the tunnel current is predicted to oscillate at the same frequency, with an amplitude linearly proportional to the average tunnel current. The construction of a prototype microscope, operated in air, is described. When the sample magnetization is alternated, a modulation of the tunnel current is observed at the same frequency. This signal satisfies criteria developed to characterize spin‐polarized tunneling. Spurious signals are also observed and their probable origins identified. A number of improvements are suggested that should eliminate the spurious effects.

A planar, second-order gradiometer was fabricated from single-layer YBa2Cu3O7−x films. The gradiometer consists of a symmetric flux transformer with an overall length of 80 mm inductively coupled to a directly coupled magnetometer, and has a baseline of 31 mm. The mutual inductance between the flux transformer and the magnetometer is adjusted mechanically to reduce the response to a uniform magnetic field applied perpendicularly to the plane of the gradiometer to typically 50 ppm. From an independent measurement, the residual first-order gradient response was determined to be at most 1.4% relative to the second-order gradient response.

A dc superconducting quantum interference device (SQUID) was used to amplify radio-frequency signals that were coupled to one end of the microstrip formed by the input coil and the SQUID washer. For one device, the resonant frequency of the microstrip was increased from about 200 to 620 MHz by progressively shortening the length of the coil. At an operating temperature of 4.2 K, the gain was typically 18 dB, and the system noise temperature ranged from 0.5±0.3 K at 80 MHz to 3.0±0.7 K at 500 MHz.

Using a low transition temperature superconducting quantum interference device as a detector, we have obtained magnetic resonance images of laser-polarized 3He gas and solid 129Xe at 4.2 K in magnetic fields as low as 0.54 mT (3He) and 1 mT (129Xe), corresponding to Larmor frequencies of 17.6 and 11.8 kHz, respectively. The experimental resolution of the images is ∼500 μm for 3He in the gas phase and ∼950 μm for 129Xe in the solid state.

The planar gradiometer consists of a directly coupled magnetometer inductively coupled to the smaller loop of an asymmetric, two-loop flux transformer patterned in a single film of YBa2Cu3O7−x. The coupling is adjusted mechanically to balance the gradiometer to about 1 part in 3000 with respect to fields perpendicular to its plane. The baseline is 48 mm and the intrinsic balance with respect to in-plane fields about 1 part in 1400. When a magnetic field is applied solely to the magnetometer, its sensitivity is reduced by only 5% by the presence of the flux transformer.

High-transition temperature dc superconducting quantum interference devices (SQUIDs) have been fabricated with arrays of slots or holes patterned in the 186×204 μm2 washers. In two designs, with seven slots or 248 holes, the linewidth of the remaining YBa2Cu3O7−x film was 4 μm. These devices exhibited virtually no increase in their low-frequency noise when they were cooled in static magnetic fields up to a certain threshold, above 100 μT in the best case. Furthermore, the mutual inductance between these devices and a seven-turn, spiral coil was at least 85% of that for a solid SQUID with the same outer dimensions.

The spectral density SΦ(f ) of the low‐frequency 1/f noise of high transition temperature dc superconducting quantum interference devices (SQUIDs) with narrow linewidths was independent of B0, the magnetic field in which they were cooled, up to a threshold value, about 33 μT in the best case. Above this threshold, which is associated with the entry of flux vortices into the film, the noise increased rapidly. By contrast, for large square washer SQUIDs, SΦ(f ) scaled linearly with B0. Estimates indicate that the 1/f flux noise produced by the pickup loop of a directly coupled magnetometer is negligible.

Three magnetometers based on dc superconducting quantum interference devices (SQUIDs) fabricated from YBa2Cu3O7−x have been operated in a magnetically shielded room using a flux‐locked loop involving additional positive feedback with bias current reversal. Two of these devices, integrated multiloop dc SQUIDs with outer diameters of 7 mm, achieved white noise levels of 10 fT/√Hz for bicrystal junctions and 30 fT/√Hz for step‐edge junctions. The third magnetometer involved a flux transformer with a 10×10 mm2 pickup coil connected to a 16‐turn input coil which was inductively coupled to a bicrystal SQUID. This device achieved a white noise of 16.2 fT/√Hz. High quality magnetocardiograms were obtained without signal averaging.

Two integrated, thin‐film multiloop magnetometers (fractional turn SQUIDs) have been fabricated using a YBa2Cu3O7−x‐SrTiO3‐YBa2Cu3O7−x multilayer process. The magnetometers have 16 parallel loops, an outer diameter of 7 mm and effective areas of 1.84 and 1.89 mm2, respectively. The magnetic field noise of the better device, measured in a YBCO tube at 77 K with a bias reversal scheme, was 37 fT Hz−1/2 at 1 Hz and 18 fT Hz−1/2 at 1 kHz.

Using a liquid‐nitrogen‐cooled scanning SQUID magnetic microscope, we have developed a technique for broadband imaging of radio‐frequency (rf) and microwave fields with a spatial resolution of about 15 μm. We have produced images of the amplitude of 50 MHz fields with an rms noise of 2.6 nT and a 300 μm/s scan rate. Detection is accomplished by using the nonlinearity of the voltage‐flux characteristic of the SQUID to rectify the rf fields. Our present technique is limited by cavity mode resonances in the SrTiO3 substrate of our SQUID sensor. Using a small excitation probe, we have directly imaged these resonances at frequencies up to about 12.5 GHz.

We have fabricated asymmetric, two‐junction YBa2Cu3O7−x dc superconducting quantum interference devices (SQUIDs) on SrTiO3 bicrystal substrates and used them as three‐terminal amplifiers by injecting signal currents directly into part of the superconducting loop. Using a second, readout SQUID to monitor changes in the critical current, we have achieved current gains up to 2.3 at 77 K. By coupling two SQUIDs together, we have increased the gain to 8.5 at 77 K. The experimental results are compared with numerical simulations, which are used to discuss potential increases in the gain.

We have fabricated YBa2Cu3O7−x–SrTiO3–YBa2Cu3O7−x (YBCO–STO–YBCO) trilayers, in which each layer is patterned photolithographically, capping the first YBCO film with an in situ STO film. Atomic force microscopy demonstrates that the capping process dramatically improves the quality of the surface of the second layer, allowing the growth of an upper YBCO film with a substantially reduced level of low‐frequency flux noise. A magnetometer with a multiturn flux transformer coupled to a dc superconducting quantum interference device achieved a magnetic field noise of 74 fT Hz−1/2 at 1 Hz, improving to 31 fT Hz−1/2 at 1 kHz.

We have fabricated and tested picovoltmeters in which the voltage source is connected in series with a calibrated resistor and a coil inductively coupled to a high‐transition temperature superconducting quantum interference device operating at 77 K. The coil consists of either seven or ten turns of copper wire or of two turns patterned in a film of YBa2Cu3O7−x(YBCO). The lowest voltage noise achieved at 1 Hz was 30 pV Hz−1/2 and 2.3 pV Hz−1/2 for input coils made of copper and YBCO, respectively.

We present results on a near‐field scanning probe microwave microscope which is based on the nonlinear behavior of a YBa2Cu3O7 direct‐current superconducting quantum interference device (SQUID). Our system has a spatial resolution of about 30 μm and derives its probing field from circulating currents in the SQUID loop. The frequency is set by the Josephson relation and is continuously tunable up to about 200 GHz. By imaging small patterned normal metal thin‐film samples of Cu and Nb at 77 K, we have confirmed the operating bandwidth of the system and investigated the nature of the imaging process.

We have investigated the performance of YBa2Cu3O7−x dc superconducting quantum interference devices (SQUIDs) cooled in static magnetic fields, B0, of 0.01–1 mT. For fields less than the earth’s ambient field, about 0.05 mT, the white noise of the devices at 77 K is not materially affected. However, at a frequency f of 1 Hz the spectral density of the 1/f noise, SΦ (1 Hz), at 0.05 mT increases by an order of magnitude over that for zero field. Furthermore, SΦ (1 Hz) scales approximately linearly with B0, suggesting strongly that the noise originates in the motion of vortices in the YBCO film. This increase in noise is likely to be an issue for SQUIDs operated in the earth’s field.

We have used a scanning magnetic flux microscope based on a high transition temperature YBa2Cu3O7 superconducting quantum interference device (SQUID) to produce magnetic images of eddy currents in patterned Cu thin films and 11–30‐μm‐thick Cu on printed circuit boards. The fields produced by the eddy currents are imaged with a spatial resolution of about 80 μm over a 100‐mm2 sample area. With the sample and SQUID at 77 K, the microscope uses typical probing fields of 80 nT and can obtain simultaneously eddy current and static magnetic field images. At probing frequencies of 26–100 kHz, the system achieves a field sensitivity of about 7 pT Hz−1/2.

We have fabricated high‐transition temperature superconducting magnetometers involving a flux transformer patterned in a single film of YBa2Cu3O7−x (YBCO) on a 50‐mm‐diam substrate. This transformer is inductively coupled to a loop that in turn is directly coupled to a dc superconducting quantum interference device, patterned in a single layer of YBCO deposited on a SrTiO3 bicrystal. At 77 K, the lowest magnetic field noise achieved is 31 fT Hz−1/2 at frequencies above 5 Hz, increasing to 39 fT Hz−1/2 at 1 Hz.

We have fabricated magnetometers patterned in a single layer of YBa2Cu3O7−x involving dc superconducting quantum interference devices (SQUIDs) with bicrystalline grain boundary junctions. The magnetometers consist of either a SQUID with a large area square washer or a single turn pickup loop coupled directly to the body of a small area SQUID. We found that the transfer function falls off with increasing SQUID inductance much more rapidly than predicted; implications for magnetometer performance are discussed. When operated at 77 K with a bias reversal technique a directly coupled magnetometer had a noise of (105±10) fT Hz−1/2 at 1 kHz, increasing to (145±10) fT Hz−1/2 at 1 Hz.

We have developed a magnetic flux microscope which uses a liquid nitrogen cooled thin‐film YBa2Cu3O7 dc superconducting quantum interference device (SQUID) to produce two‐dimensional images of magnetic fields. The instrument operates as a scanning probe microscope with a spatial resolution of about 80 μm and a field of view of about 100 mm2. The equivalent magnetic field noise in the SQUID ranges from 80 pT Hz−1/2 at 1 Hz to 20 pT Hz−1/2 at 1 kHz, yielding a field resolution of about 200 pT in our images. We have used the microscope to image magnetic domains in ferromagnetic samples, small currents flowing in fine wires, and trapped flux and diamagnetic susceptibility in superconducting thin films.

At low frequencies f the 1/f noise power in single biepitaxial junctions of YBa2Cu3 O7−x peaks sharply for bias currents just above the noise reduced critical current and increases as I2 for high bias currents I. This behavior is explained by a model in which both critical current fluctuations δI0 and resistance fluctuations δR contribute to the measured voltage noise. The magnitude of the normalized critical‐current fluctuations ‖δI0/I0‖ is always much greater than that of the normalized resistance fluctuations ‖δR/R‖. Switching the bias current between positive and negative values at 2 kHz greatly reduces the magnitude of the 1/f noise from both sources, implying that the coherence of the noise generating process is not affected by the current reversal.

Our YBa2Cu3O7−x thin‐film magnetometer consists of a dc superconducting quantum interference device with bi‐epitaxial Josephson junctions, fabricated on one chip, and a flux transformer with a multiturn input coil, fabricated on a second chip. Photolithographic processing is used to pattern all layers. The magnetometer operates in a flux‐locked loop at temperatures up to 81 K, with the flux transformer improving the magnetic field sensitivity by a factor of 83±3. The low‐frequency rms magnetic field noise scales approximately as 1/f1/2, where f is the frequency, with a magnitude of 0.6 pT Hz−1/2 at 10 Hz and 0.09 pT Hz−1/2 at 1 kHz with the magnetometer immersed in liquid N2.

Using a dc superconducting quantum interference device (SQUID) at 4.2 K, we have measured the magnitude and temperature dependence of the flux noise generated by superconducting flux transformers made from YBa2Cu3O7−x. If the transformer loop is opened, the observed noise decreases by a factor of 10, demonstrating that moving vortices generate noise primarily by altering the screening current, rather than by coupling their magnetic flux directly into the SQUID. Model calculations are given for the various noise contributions, and shown to be reasonable predictions of the observed effects.

We report high‐resolution transmission electron microscopy and electron diffraction studies of the heteroepitaxial superconductor‐insulator‐superconductor system YBa2Cu3O7−x‐SrTiO3‐YBa2Cu3O7−x deposited on polished (001)MgO substrates by in situ laser ablation. The resulting films grow epitaxially and consistently preserve a parallel orientation between the close‐packed (001)YBa2Cu3O7−x planes and (001)SrTiO3 planes over the entire trilayer, even in the presence of ledges or steps along vicinal interfaces. Although the interface regions showed strain occasionally relieved by stacking faults, they were free of disorder and any evidence of impurity phases. The observed epitaxial growth is very likely responsible for the excellent electrical properties found in similarly constructed multilayer interconnects.