ESA Science & Technology - Publication Archive

In "Ultraviolet-Optical Space Astronomy Beyond HST" ASP Conference Series 164, ed. Jon A. Morse, J. Michael Shull, & Anne L. Kinney. Astronomical detectors based on Superconducting Tunnel Junction (STJ) technology are the next logical step beyond CCD and MCP detectors for optical and ultraviolet astronomy. Rather than merely registering each photon event with high efficiency, these devices also measure the wavelength of the detected photons throughout the far-UV through near-IR. Large format STJ arrays promise to provide the ultimate '3D' astronomical detector capable of covering simultaneously more than a decade in wavelength. The STJ detector holds particular promise in space-based ultraviolet applications where its inherent spectral resolution is the highest and its naturally high quantum efficiency overcomes the sensitivity limit set by available UV photocathode materials. The capability to separate spectral orders on the detector also opens up exciting new possibilities for novel and highly efficient grating spectrometer designs.

A number of 6×6 element arrays of Ta-based superconducting tunnel junctions have been manufactured for photon counting applications with moderate energy resolution in ground-based optical astronomy. The individual array elements show low leakage, uniform responsivity across the array, good simultaneous Josephson current suppression and minor crosstalk between adjacent pixels. The same arrays have been characterized in the soft X-ray range (E=270-1500 eV). The base electrode response shows good energy resolving power (E/DeltaE = ~140). Unwanted spectral features originating from other parts of the detector can be largely eliminated by rise-time filtering. Modifications in the layering are necessary in order to improve the soft X-ray detection efficiency.

Despite considerable progress over the past years, the detection of medium-energy X-ray photons (E>1 keV) with STJs near the energy-resolution limit, set by the Fano and tunnel noise, remains an elusive goal. There is presently little doubt that the spatially inhomogeneous response of the STJ is responsible for the degradation of the energy resolution. We review several proposed mechanisms against experimental data for Nb- and Ta-based STJs, of various sizes and in single or array-format. We argue against a single mechanism behind the resolution degradation. The experimental results presented here support a model in which quasi-particles are lost at the edges of the STJ, but also indicate that losses into the leads seriously degrade the energy resolution. Finally, an example is given of how fabrication details may play a role as well.

We summarise the results of a number of X-ray experiments on an epitaxial GaAs device carried out in both our laboratory and at the PTB Radiometry Laboratory at the BESSY Synchrotron Radiation Source. The detector has a diameter of 1.5 mm and is fully depleted to a depth of 40 micron. It has been characterized as a function of energy, bias and temperature. At -35 °C we determine the charge collection efficiency to be 97% and find that energy resolutions ranging from 730 to 930 eV fwhm can be readily achieved using conventional pre-amplifiers over the energy range 6-60 keV. By considering the various contributions to the fwhm, we show that leakage current and charge trapping noise dominate the resolution function. From detector modelling we determine the effective electron and hole density/cross section products to be 7 and 3 cm^-1, respectively.

The spectral response of a 100 x 100 micron² tantalum based superconducting tunnel junction to 5.9 keV x-ray photons from a ⁵⁵Fe source has been studied. In full illumination the energy resolution for the Mn K_alpha line complex is 56 eV, dominated by spatial nonuniformity in the response of the detector. When illuminating selectively a 5-10 micron diam spot in the center of the detector, the energy resolution improves to 22 eV, corresponding to 15.7 eV for the individual Mn K_{alpha 1} and Mn K_{alpha 2} lines. This exceeds the predicted theoretical energy resolution of 7.3 eV for this type of device by only a factor of ~ 2.

Photon counting experiments at wavelengths ranging from near infrared to x-ray with niobium based superconducting tunnel junctions with aluminium trapping layers are presented. Single photons can be detected up to a wavelength of 1 micron. The response in the ultraviolet to near-infrared region is characterized by a good energy linearity (< 2.5%), a capability to handle event rates up to ~ 3 kHz, and moderate energy resolving power (E/DeltaE = ~7 for E = 4 eV). The x-ray response at 6 keV is characterized by anomalously high signals compared to the low energy response, a severe energy nonlinearity and a relatively poor energy resolution of ~ 140 eV, full width at half maximum.

Asymmetric NbN-Nb-Al-AlO_x-Al-Nb superconducting tunnel junctions have been investigated as photon counting detectors at x-ray and ultraviolet (UV)-visible wavelengths. The inclusion of a thin NbN passivation layer on the top electrode of the devices in place of the natural niobium oxides has reduced the quasiparticle loss rates, thereby enhancing the probability of multiple tunnel processes. As a consequence, the detector responsivity has increased from 900e^-/eV, up to values in excess of 2000e^-/eV in the temperature range 0.30-0.8 K. Such a responsivity level has allowed single photon counting performance at wavelengths as long as 700 nm and at operating temperatures as high as 830 mK. The devices show a linear response in the UV-visible range, while at 6 keV the expected nonlinearities in the energy response and moderate energy resolution similar to that found in Nb-Al junctions are observed.

Some recent results associated with the development of tantalum based photon counting superconducting tunnel junctions (STJ) suitable for use as broad-band low resolution spectrometers for optical and ultraviolet astronomy are presented. A 20x20 micron square tantalum based STJ, operated at a temperature of 0.3 K, has demonstrated a limiting resolution of ~ 8 nm at 200 nm and ~ 80 nm at 1000 nm. The device is extremely linear in response with photon energy, and covers the waveband from 200 nm to 2 micron while measuring the individual photon wavelength and arrival time. The short wavelength limit is currently constrained by the current experimental configuration (a fibre optic) as well as to some extent the sapphire substrate. The estimated quantum efficiency for single photons is over ~ 50% between 200 and 700 nm with a maximum of ~ 75% at 550 nm. Such an STJ when packaged into an array could contribute significantly to many fields of near infrared, optical and ultraviolet astronomy being able to provide efficiently and simultaneously the broad band spectrum and photon arrival time history of every single object in the field over a very wide dynamic range.

We discuss the capabilities of superconducting tunnel junctions as detectors for ultraviolet, optical, and near-infrared astronomy. Such junctions have recently been shown to allow the detection of individual optical and ultraviolet photons with an inherent spectral resolution related to the critical temperature of the absorbing superconductor. Limiting resolutions at 500 nm ranging from 5 - 40 nm (for materials with critical temperatures between 0.1 to 10 K) should be achievable. These detectors should have a high quantum efficiency (50 per cent) over a very broad wavelength range from the ultraviolet to the near infrared (100 - 2000 nm). The overall efficiency is limited by reflection from the superconducting film, and should be improved significantly by appropriate anti-reflection coatings. The devices function at very high incident photon rates - with count rates of order 10 kHz or higher being feasible, and photon arrival time datation possible to microsec-level accuracy. It is realistic in the future to envisage that these devices, of a size typically 20-50 micron², could be packaged into imaging arrays. These key characteristics imply that many areas of optical and ultraviolet astronomy could benefit significantly from their further development.

We report the detection of individual optical and ultraviolet photons using a different approach to photon detection based on a superconducting tunnel junction. A 20 × 20 micron² junction, employing a 100 nm niobium film and operated at a temperature of ~ 0.4 K, has been used to detect individual photons with inherently high quantum efficiency (> 45%) over a broad wavelength range (between 200 and 500 nm), yielding high temporal (sub-ms) resolution, spatial resolution determined by the junction size, under conditions of minimal dark current, and in the absence of read noise. The quantum efficiency is limited by surface reflection, and could be improved by the deposition of antireflection coatings. The theoretical wavelength response range continues into the far UV and soft x-ray region, and is presently limited beyond 500 nm largely by the available signal processing electronics. The device intrinsically functions at very high incident photon rates - with count rates of order ~ 10 kHz or higher being feasible and again currently limited primarily by the signal processing electronics - thus providing a correspondingly enhanced dynamic range by several orders of magnitude compared with previous panoramic photon counting detectors. The measured charge output from the device is highly linear with photon energy resulting in an optical photon detection system with intrinsic spectral resolution, related to the critical temperature of the junction material and, in the current device, providing a limiting spectral resolution of about 50 nm. It is realistic in the future to envisage that these devices could be packaged into arrays, with the resulting system characteristics offering advantages over detectors based on semiconductors

The Mini-STEP concept was conceived from a desire by NASA to reduce the cost of the satellite test of the equivalence principle (STEP) experiment below that of the already downsized Quick STEP concept. The goal was for the total cost, including payload, spacecraft, launch vehicle, reserves and operations, to be in the 50 million US dollar range. Stanford University and the Jet Propulsion Laboratory studied this simplified STEP concept between March and June 1995. A similar concept was developed in parallel by the European Space Agency (ESA) as an alternative to the M3 STEP mission.

The x-ray performance of Nb-Al-AlO_x-Al-Nb superconducting tunnel junctions deposited on sapphire has been studied for phonon mediated detection of x rays absorbed in the substrate in the energy range 750 to 6000 eV. Two separate channels of phonon propagation can be identified. One produces a discrete signal peak, due to high frequency phonons originating from the x-ray absorption sites in a shallow layer below the junction. The other contributes to a monotonic signal tail, due to low frequency phonons, reaching the junction after diffusive or multiple scattering at surfaces.

An investigation into the phonon contamination of x-ray sensitive superconducting tunnel junctions arising from the x-ray photoabsorption in various substrates has been conducted. Results are presented on the design of a superconducting tunnel junction (STJ) which substantially reduces or even eliminates phonon induced noise from the substrate. Such noise is the predominant feature in x-ray spectra from junctions due to the bulk of the photons being absorbed in the substrate rather than in the thin superconducting film. The design involves the choice of a suitable buffer sandwich between the substrate and the STJ. Such a buffer would appear not only to attenuate the phonons created in the x-ray photoabsorption in the substrate but also to scatter the phonons inelastically, introducing a frequency down-conversion. Such a process ensures that few phonons of energy sufficient to break Cooper pairs in the superconducting film of the STJ enter the junction.

The performance of photon detectors based on superconducting tunnel junctions are related to their current - voltage (I-V) curve characteristics and, ultimately, to the quality of the thin tunnel barriers (of order 1 nm) which separate the two superconducting thin films. Both the optimization of the spectroscopic performance of these detectors and the development of a reproducible and high yield fabrication route, require a better understanding of barrier quality and growth techniques. Transmission electron microscopy (TEM) and atomic force microscopy (AFM) provide valuable tools for the investigation of the barrier region and for the control of the quality of the different thin films and related interfaces. In this paper, the results of a TEM and AFM evaluation of Nb-Al-AlOx-Nb tunnel junctions are reported, together with their interpretation on the basis of the I-V curve performance at low temperature (T >= 0.3 K). Thickness disuniformities of the Al plus AlOx overlayer and evidence of barrier defects have been found, which may place constraints on the spectroscopic performance of such devices. Through the use of TEM it has also been possible to confirm the epitaxial nature of the Nb base electrode. The junction counter electrode however appears to be polycrystalline, with a columnar morphology and an average grain width of 40 nm. The overall structure of the various layers may well place constraints on the tunneling characteristics of the device.

Recent experimental results show a linear energy response in high quality Nb-Al-AlO_x-Nb superconducting tunnel junction detectors for photon energies between 1.5 and 6.4 keV. The experimental data are based on both direct x-ray illumination and on the escape and re-absorption of fluorescent photons created in the junction electrodes and in the silicon substrate. The observed linearity of the energy response raises questions on the validity of some theoretical models which describe the relaxation process occurring in a superconducting thin film after x-ray photoabsorption. Such models generally predict nonlinear effects due to large quasiparticle number densities and short recombination times.

Current research into x-ray detection using superconducting tunnel junctions indicates that the poor spectral resolution obtained so far, in comparison with theoretical expectations, is partly due to the excellent acoustic coupling of the junction and substrate. The substrate acts both as a source of noise and as a heat sink for the nonequilibrium junction, thus masking the intrinsic response of the superconducting electrodes to photoexcitation. A new design for a superconducting tunnel junction based on an x-ray detector is presented. The design effectively decouples the substrate and junction and should therefore eliminate many causes of spectral degradation, bringing resolution closer to that predicted theoretically, and thus allowing experimental investigation of the intrinsic superconducting film response to x-ray photoexcitation. An outline of the way in which the design can be optimized geometrically to achieve the decoupling is given. Further optimization of the intrinsic film response to x-ray photons is achieved through the introduction of specific absorbing and trapping regions to improve both the quantum efficiency and charge output of the new design. The use of "pairing potential barriers" within the electrode leads will also improve the intrinsic resolution of this device.

Preliminary results on the X-ray performance of Nb-based superconducting tunnel junctions (STJs) with a highly transmissive barrier are reported. The results show that the energy resolution of these detectors can be improved by collimating the X-ray photons onto the junction barrier area, thus reducing illumination of the surrounding substrate and leads. A charge output of about 50% of the theoretical maximum has been recorded for these STJs, with full width at half maximum resolution of about 200 eV at 6 keV. Several mechanisms which are believed to degrade the energy resolution are also discussed. X-ray events are also detected by other junctions on the same chip which are not illuminated. This may indicate the presence of a marked phonon transmission along the sapphire substrate which acts as a phonon waveguide analogous to the light transmission mechanism in fiber optics

An insert based on a top-loading, fully welded stainless-steel design, with a special steel-copper weld is described. It has been developed explicitly for pumped helium cryostats, down to a temperature of about 1.2 K. The principal characteristics of this insert are a large high vacuum sample space, the absence of any cold seal, and the reliability and flexibility of use. Typical applications are optical/UV/x-ray photon counting experiments based on cryogenic detectors and the development of advanced heterodyne receivers using Superconductor-Insulator-Superconductor junctions as mixers in the millimeter and submillimeter regimes.

Refractory metal Nb/Al/Al-oxide/Al/Nb junctions are shown to be sensitive to 6-keV X-rays over the temperature range from 2.8 to 1.4 K. For such junctions, which have an observed minimum ionizing energy of 12 MeV, a limiting energy resolution of 8 eV is predicted. Currently an energy resolution of 250 eV is observed at 1.4 K which is primarily dominated by system electronic noise. The Nb-based junctions are shown to be very stable with respect to thermal cycling while the nonequilibrium physics can be simply scaled from the theory of Sn junctions. It is concluded that on-chip arrays having a broad band pass and good energy resolution should be feasible to construct.