Staff: Simon Doyle, Peter Ade and Enzo Pascale
Postdocs: Aurelien Bideaud, Dmitry Morozov and Edgar Castilo
PhD Students: Pete Barry, Sam Rowe and Tom Brien
Collaborators: Jochem Baselmans (SRON, Netherlands), Phil Mauskopf (Arizona State University, USA), Jonas Zmuidzinas and Matt Bradford (Caltech USA), Terry Whall and Evan Parker (Warwick University UK).
The Astronomy Instrumentation Group has several specific programs dedicated to detector development for both astronomical and commercial applications. The group focuses on development of cryogenic (low temperatures between 0.05 and 4K) detectors suitable for sensing light from millimetre wavelengths to the near infrared regions of the electromagnetic spectrum. This region of the electromagnetic spectrum contains a wealth of information for both astronomy and industry. In astronomical applications, our research focuses on both the development of detector technology for large scale imaging and spectroscopy for use on ground based telescopes as well developing ultra sensitive detectors for use in the next generation of space based observatories. Working with industrial partners, the group has used the experience gained in the development of cryogenic detectors and applied it to commercial applications developing arrays of detectors that image terrestrial objects. Some of the various research areas the detector group is currently pursuing are outlined below.
- Kinetic Inductance Detectors (KIDs)
- Transition-Edge Sensors (TES)
- Superconducting Single Photon Detectors
- Cold Electron Bolometers
- On chip spectrometers
- Terrestrial THz imaging
A 196 pixel 150 GHz LEKID array designed for the NIKA instrument. Each square patch on the silicon wafer is a detector. The entire array is cooled to 100mK (0.1K) and all 196 detectors read out via the two SMA connectors. This approach to creating large format arrays drastically simplifies complexity in both cryogenic electronics and array fabrication compared to competing ultra-sensitive mm and submm detector technology.
The Kinetic Inductance Detector (KID), is a superconducting detector that provides an elegant solution to producing the next generation of large format, ultra sensitive detector arrays. By multiplexing in the frequency domain, up to 1000 KIDs can be read out simultaneously on a single pair of coaxial cables. Furthermore, KIDs are naturally broad-band and can serve as continuum detectors in the submm – FIR or as energy resolving single photon detectors in the optical to X-ray. The Cardiff group has pioneered research in KID technology with the invention of the Lumped Element Kinetic Inductance Detector (LEKID) – a technology that has been adopted by many groups around the world working in this field. The LEKID architecture has been used in a dual band (150GHz and 220GHz) demonstrator camera on the IRAM telescope and in an optical photon counting demonstrator camera on the Palomar telescope. Proposed future instruments using LEKID technology include; NIKA (a large format 3 band millimeter camera for the IRAM 30 meter telescope, MAKO (a large format 350GHz camera for the CSO/CCAT telescope) and SuperSpec (a mm imaging spectrometer for the CCAT telescope).
The Cardiff KIDs group are currently optimising the LEKID architecture for a range of astronomical and commercial applications. Current research includes: Electromagnetic design of direct and lens-coupled LEKID absorbers; Materials and processing development for the next generation of cryogenic detector arrays and planar high frequency superconducting circuits; Developing KID arrays suitable for space based applications (astronomical and Earth observing); Development of a terrestrial sub-mm imaging system for a range of applications including airport security, medical imaging and manufacturing quality control; Superconducting on-chip spectrometers.
Transition-Edge Sensors (TES) are bolometric detectors with superconducting thermometers that can be fabricated into arrays and read out with low noise superconducting multiplexors. TES arrays are used in a number of ground-based and balloon-borne mm-wave to sub-mm instruments including the MBAC on the Atacama Cosmology Telescope (ACT), the South Pole Telescope (SPT), SCUBA-2 on the JCMT, EBEX and SPIDER.
A single TES device fabricated by our collaborators at SRON and tested in the ultra-low background test facility at Cardiff. (Click to enlarge)
Researchers at Cardiff have recently led an effort (supported by ESA and STFC) to develop TES technology for use in future Far-Infrared (FIR) space missions such as SPICA-SAFARI and FIRI. One of the challenges in this effort has been developing the first facilities capable of testing ultra-sensitive detectors at FIR wavelengths from 30-200 microns. Another challenge is achieving the sensitivity required for these future instruments which corresponds to a noise equivalent power (NEP) of 10-19 to 10-20 W/√Hz which is approximately two orders of magnitude more sensitive than the previous state of the art TES devices. Based on the experience with TES detectors for astronomy, the group has also developed a TES detector for commercial use marketed by QMC Instruments. This device operates from a cryogen-free 4 K platform and provides high sensitivity and speed over the entire wavelength range from 10 microns to 2 mm.
SSPD chip and close photo of the nanowire meander. Photons are absorbed in the meander section creating a local hot-spot. The increased current density flowing around this hot-spot drives a section of the meander out of the superconducting state creating a voltage pulse across the readout circuit. (Click to enlarge)
The Superconducting Single Photon Detector (SSPD) is an ultrafast (up to GHz) photon counting detector that operates on the basis of a hot electron effect in ultra thin superconducting films. The SSPD consists of a superconducting stripe with submicron width that is cooled to a temperature much lower than critical temperature (TC) and biased with the current Ib close to critical current (IC). An absorbed photon forms a localized hotspot region in the superconducting stripe suppressing superconductivity. The initial hotspot size is determined by the number of factors, namely by the ratio of photon energy and the gap energy of the superconductor, quasi-particle thermalization time and diffusivity of quasi-particles. The hot spot grows due to diffusion of the hot electrons and super-current is expelled from the hotspot volume towards the edges of the strip. If the density of redistributed current exceeds the critical current density, then local superconductivity is destroyed forming a resistive barrier across the width of the device. This resistive patch produces a voltage pulse in the read-out circuit of the detector. The hot spot collapses after the relaxation of the hot electrons and restores superconductivity returning to the initial state of the device.
SSPDs made with superconducting materials with smaller gap energies such as Titanium-Nitride (TiN) are good candidates for the single photon detection in the mid-IR wavelength range. Mid-IR single photon detectors have a number of important applications such as quantum communication and cryptography, quantum computing and mid-IR astronomy. Part of our on-going research is the development of mid-IR nanowire single photon detectors based on small gap energy superconducting films.
Energy level diagrams for a SiCEB. As a biasing voltage is applied hot electrons from the top of the distribution in the semiconductor (Sm) tunnel in to the left hand contact (S) and are replaced by cooler electrons from the right hand tunneling contact. In a detector the bias is selected such that optical power is also needed for tunneling, the greater the power the more tunneling. (Click to enlarge)
Cold Electron Bolometers (CEBs) are a novel type of detector which offer both the advantages of high sensitivity (approximate NEP = 10-18 W/√Hz) and high speed (sub-microsecond time constants). The silicon cold electron bolometer (SiCEB) works on the principal of electron cooling using quantum tunneling between superconducting and semiconducting materials. A bias voltage is applied across a superconductor-semiconductor-superconductor structure. Such an arrangement of superconductors and semiconductors create a potential energy barrier between the material interfaces that electrons can tunnel through provided that there is an energy state on the opposite side of the barrier that the electron can occupy. The difference in energy states between the superconductor and semiconductor allows for higher energy electrons (hot electrons) to tunnel out of the semiconducting material into free states in the superconductor that exist above the superconducting gap energy 2Δ. These hot electrons are replaced with lower energy (cooler) electrons from the superconductor that exist below the gap energy 2Δ thus giving an overall cooling of the device. This cooling effect means that the detector effectively operates at a lower temperature than the cooling platform used to initially cool the device leading to increased sensitivity. Illumination of the semiconductor creates additional hot electrons. This increases the tunneling rate of electrons through the entire structure and hence increases the current read out due to the bias voltage.
Optical micrograph of the first generation prototype device. Credit: H.G LeDuc (JPL) E. Shirokoff (Caltech). (Click to enlarge)
The Cardiff Astronomy Instrumentation Group (AIG) is working in collaboration with several groups in the USA including The California Institute of Technology (Caltech), The NASA Jet Propulsion Laboratory (JPL) and Arizona State University (ASU) to develop an on-chip spectrometer. A spectrometer is an instrument that can measure the optical spectrum (various colours) of a source and are regularly used in astronomy for determining the composition or distance to various astronomical objects. Traditional spectrometers working at millimeter and sub-millimeter wavelengths are bulky and limited in pixel number. Superspec is a planar, superconducting series of filter-banks lithographed on to a silicon substrate feeding an array of ultra-sensitive kinetic inductance detectors. Such a device can be used in place of a traditional spectrometer but is far more compact and is scalable to meet the requirements of future large format spectroscopic cameras. The device works by coupling radiation from a telescope via an antenna. The radiation is passed along a superconducting microwave feed-line where individual frequencies are picked off and coupled to a detector by a superconducting transmission line filter. Development and testing of a proof-of-concept device has been underway since mid-2012, with the first fabrication run in autumn 2012. An optical micrograph of a prototype device is shown in in the above figure. It consists of 73 spectral channels covering the 190-310 GHz band with spectral resolutions ranging from 300-1400.
Four 350 GHz images of a person at 4m range. Here we can clearly see concealed objects in 3 out of the 4 images that show up as dark areas against the bodies natural emission at 350GHz shown in orange.
The Astronomy Instrumentation Group (AIG) and QMC Instruments are working in collaboration to demonstrate a cryogen-free, passive terahertz (THz) imaging system for commercial and industrial applications. A prototype instrument has been built that can image a human being at sub-mm to THz frequencies with the goal to achieve video rate imaging with a temperature resolution of 0.1K or better in each frame.
At THz frequencies, many substances are transparent allowing imaging of objects that would normally be hidden from optical or infrared light. One example of this is shown in the above figure. Using our camera, it is possible to non-invasively reveal the presence of objects on a person that may be behind layers of clothing. Similar but 'active' technologies are used in airport security systems to identify contraband items but these require the illumination of the target with THz radiation. Our system is completely passive and utilises very sensitive cryogenic detectors to pick up a person's natural thermal emissions at THz frequencies.