Transition-edge sensor

Optical image of four tungsten transition edge sensors for near-infrared single-photon detection. Image credit: NIST.

A transition edge sensor or TES is a type of cryogenic particle detector that exploits the strongly temperature-dependent resistance of the superconducting phase transition.

History

The first demonstrations of the superconducting transition's measurement potential appeared in the 1940's, thirty years after Onnes's discovery of superconductivity. D.H. Andrews demonstrated first a transition-edge bolometer, a current-biased tantalum wire which he used to measure an infrared signal, and thereafter a transition edge calorimeter, made of niobium nitride and used to measure alpha-particles. However, the TES detector did not gain popularity for about 50 years, due primarily to the difficulty of signal readout from such a low-impedance system. A second obstacle to the adoption of TES detectors was in optimizing stable operation in the narrow transition edge. Joule heating in a current-biased TES can lead to thermal runaway that drives the detector normal-conducting. A solution to the readout problem has been found in superconducting quantum interference devices (SQUIDs) which are now designed to pair effectively with the TES detectors; the additional development of voltage-biased operation for TESs has facilitated wide-spread adoption of TES detectors since the late 1990's.

Setup, Operation, and Readout

A TES single-photon detector is voltage-biased by driving a current source I_bias through a load resistor R_L. The voltage is chosen to put the TES in its so-called "self-biased region" where the power dissipated in the device is constant with the applied voltage. When a photon is absorbed by the TES, this extra power is removed by negative electrothermal feedback: the TES resistance increases, causing a drop in TES current, the dissipated Joule power in turn drops, cooling the whole device back to its equilibrium power dissipation in the self-bias region. In a common SQUID readout system, the TES is operated in series with the input coil L which is inductively coupled to a SQUID series-array. Thus a change in TES current manifests as a change in the input flux to the SQUID, whose output is further amplified and read by room-temperature electronics.

Functionality

Any bolometric sensor employs three basic components: an absorber of incident energy, a thermometer for measuring this energy, and a (weak) thermal link to base temperature to dissipate the absorbed energy and re-cool the detector.

Absorber

The simplest absorption scheme can be applied to TESs operating in the near-IR, optical, and UV regimes. These devices generally utilize a tungsten TES as its own absorber, thereby absorbing up to 20% of the input radiation. If high-efficiency detection is desired, the TES may be fabricated in a multi-layer optical cavity tuned to the desired operating wavelength and employing a backside mirror and frontside anti-reflection coating. Such tricks can decrease the transmission and reflection from the detectors to negligibly low values, resulting in a device with a device whose efficiency in principle exceeds 99%. (In fact, 95% detection efficiency has been observed.) At higher energies, the primary obstacle to absorption is transmission, not reflection, and thus an absorber with high photon stopping-power and low heat capacity is desirable--a Bismuth film is often employed. Any absorber should have low heat capacity with respect to the TES. Higher heat capacity in the absorber will contribute to noise and decrease the sensitivity of the detector (since a given absorbed energy will not produce as large of a change in TES resistance). For far-IR radiation into the millimeter range, the absorption schemes are varied and complicated, including resonant cavities as described above, feedhorns, and more.

Thermometer

The TES operates as a thermometer in the following manner: absorbed incident energy automatically increases the resistance of the voltage-biased sensor within its transition edge, and the integral of the resulting drop in current is proportional to the energy absorbed by the detector . The output signal is proportional to the temperature change of the absorber, thus for maximal sensitivity, a TES should have low heat capacity and a narrow transition edge. Important TES properties, including not only heat capacity but also thermal conductance and thermal noise, are strongly temperature dependent, so the choice of transition temperature T_c is critical to the device design. Furthermore, T_c should be chosen to accommodate the available cryogenic system. Tungsten has been a popular choice for elemental TESs as thin-film tungsten displays two phases, one with T_c ~15 mK and the other with T_c ~1-4 K, which can be combined to finely tune the overall device T_c. Bilayer and multilayer TESs are another popular fabrication approach, where thin films of different materials are combined to achieve the desired T_c.

Thermal conductance

Finally, it is necessary to tune the thermal coupling between the TES and the bath; a low thermal conductance is necessary to ensure that incident energy is seen by the TES rather than being lost directly to the bath, however the thermal link must not be too weak, as it is necessary to cool the TES back to bath temperature after the energy has been absorbed. The two primary approaches to control the thermal link are by electron-phonon decoupling, and by mechanical machining. In a superconductor, heat is carried not by conduction electrons but by phonons, and in some cases the electron and phonon systems of the same material may differ in temperature. Tungsten is an example of such a system, and tungsten TESs exploit the inherent weak phonon-electron coupling as their thermal link to bath. Other devices use mechanical means of controlling the thermal conductance such as building the TES on a sub-micron membrane over a hole in the substrate or in the middle of a sparse "spiderweb" structure.

Advantages and Disadvantages

TES detectors are attractive to the scientific community for a variety of reasons. Among their most striking attributes are an unprecedented high detection efficiency customizable to wavelengths from the millimeter regime to gamma rays , and a theoretical negligible background dark count level (less than 1 event in 1000 s from intrinsic thermal fluctuations of the device ). (In practice, although only a real energy signal will create a current pulse, a nonzero background level may be registered by the counting algorithm or the presence of background light in the experimental setup, such as blackbody radiation which is nonetheless seen by a TES optimized for use in the visible regime.)

TES single-photon detectors suffer nonetheless from a few disadvantages as compared to their avalanche photodiode (APD) counterparts. APDs are manufactured in small modules which count photons out-of-the-box with a dead time of a few nanoseconds and output a TTL pulse corresponding to each photon with a jitter of tens of picoseconds. In contrast, TES detectors must be operated in a cryogenic environment, output a signal which must be further analyzed to identify photons, and have a jitter of approximately 100 ns . Furthermore, a single photon spike on a TES detector lasts on the order of microseconds.

Applications

TES arrays are becoming increasingly common in physics and astronomy experiments such as SCUBA-2, the Atacama Cosmology Telescope, the Cryogenic Dark Matter Search, the E and B Experiment, the South Pole Telescope, and the Spider polarimeter.

See also

• Bolometer
• Cryogenic particle detectors