Imaging and spectroscopy from THz to IR with arrays of cryogenic detectors

1
Imaging and spectroscopy from THz to IR with
arrays of cryogenic detectors

• Phil Mauskopf
• Cardiff University
• Ken Wood
• QMC Instruments, Ltd.

• Low Energy Detector Workshop June 30, 2005

2

• Outline
• 1. Instrument requirements for detector systems
• 2. Existing commercial IR detector systems
• Composite bolometers
• InSb HEBs
• Superconducting HEBs
• 3. New cryogenic detector technologies
• TES bolometers
• KIDs
• SSPDs
• 4. Imaging arrays
• 5. Necessary accessories
• 6. Conclusions

• Low Energy Detector Workshop June 30, 2005

3
Cardiff expertise Instruments (for Astronomy,
etc.) Detector physics ( collaborators) Filters
and optics (spectrometers) Cryogenics How to
design an instrument Define requirements 1.
Angular resolution required? 2. Sensitivity
required - intensity of source? 3. Dynamic range
- minimum vs. maximum signal? 4. Speed of
response required - fastest change in signal? 5.
Frequency bandwidth of source? 6. Frequency
resolution required? 7. Polarisation
discrimination required?

• Low Energy Detector Workshop June 30, 2005

4
1. Angular resolution required ? optics
design Fundamentally limited by diffraction 2.
Sensitivity required ? collecting area, number of
detectors, detector and optics
configuration Fundamentally limited by photon
noise from source 3. Dynamic range ? detector
and readout type 4. Speed of response required ?
detector and readout type 5. Frequency bandwidth
of source ? filtering system 6. Frequency
resolution required ? filtering and detectors 7.
Polarisation discrimination required
5
System components Cryostats (60 mK 4 K),
Filters, Horns, Electronics, etc.
6

• Commercially available cryogenic detector systems
  today
• e.g. From QMC Instruments
• Composite bolometers (4 K and below) with
  semiconductor thermistors
• Bandwidth 90 GHz - gt10 THz (actually up to
  X-ray)
• Detector spectral response defined by filters!
• NEP (4 K) 2 pW/rtHz (measured system optical
  NEP background limited for gt300 K source with
  full bandwidth)
• minimum NEP (4 K) 0.2 pW/rtHz
• f3dB 200 Hz
• ? 1 ms

7
Commercially available cryogenic detector
systems e.g. From QMC Instruments 2.
Semiconductor hot electron bolometers (InSb)
Bandwidth 0 - 500 GHz (no magnetic field) or
FWHM 500 GHz centred at 500 GHz 2.5 THz
(with magnetic field) Detector spectral
response defined by magnetic field NEP (4 K)
1-2 pW/rtHz (measured system optical NEP)
f3dB 1 MHz ? 300 ns
8
Commercially available cryogenic detector
systems e.g. From QMC Instruments 3.
Superconducting hot electron bolometers (Nb,
NbN) Bandwidth 90 GHz - gt 10 THz
Detector spectral response defined by filters!
NEP (4 K) 1-100 pW/rtHz (measured system
optical NEP depends on bandwidth) f3dB 200
MHz (Nb) gt 1 GHz (NbN) ? 1 ns (Nb)
50 ps (NbN)
9

• Next generation of cryogenic detector systems
• What we could have(with infinite funding and
  time)
• More sensitivity single photon counting?
• Faster single photon time resolution? Pulses?
• Imaging
• Spectro-photometry
• Easier to operate (e.g. no cryogens)
• Less expensive to build, more expensive to
  buy(Ken)

10

• New and improved detector and readout
  technologies
• c.f. 2005 Zoology
• 1. Multiplexable bolometers with new types of
  thermistors
• Transition Edge Superconductors SQUIDs
• Ultra-high R silicon thermometers (Gigaohm)
  CMOS
• Kinetic Inductance thermometers HEMTs
• Hot Electron Bolometers HEMTs
• Cold Electron Bolometers quasiparticle
  amplifier
• 2. Semiconductor and superconductor
  photoconductors
• and tunnel junction detectors (I.e. everything
  else)
• BIB Ge and GaAs photoconductors JFET CIA
• Quantum dot photoconductor quantum dot SETs
• Long-wavelength QWIP detectors
• SQPT photoconductor RF SET
• KID direct detector (couple radiation directly)
• SIS/STJ video detector JFET
• Superconducting single photon detector (SSPD)
  HEMT

11

• New cryogenic detector systems (much development
  over the last 5 years)
• 1. TES bolometers replace composite bolometers?
• Nb - 4 K, Mo/Cu, Ti/Au, etc. lt 1 K SiN thermal
  isolation
• Time constant 1 us (1000x faster than
  composite)
• NEP minimum 0.01 pW/rtHz (10x more sensitive)
  (background limited in broad band system still)
• Small arrays possible with simple readout
• Less microphonic than composite bolometers
• State of the art colder, more sensitive TES
  arrays (up to 10000 pixels) with multiplexed
  SQUID readouts now being incorporated in
  astronomical instruments (e.g. ask Duncan)
• Cardiff collaborations with SRON, NIST,
  Edinburgh, Cambridge on TES detectors for
  astronomy
• Cardiff in-house fabrication of 4 K Nb TES
  detectors with SiN

12
(No Transcript)
13
Bolometers at X-ray and IR
?
V , T
G
BOLO
EXT
X-ray
C
?T
BOLO
T
G
T
o
INT
o
? C/G
TIME
?
G
EXT
T
IR
eq
C
T
T
BOLO
G
o
INT
T
o
TIME
Power on here
14
SRON developmentThermal conductance SixNy pixel
support
Slotted membrane Design P 10 pW, aspect ratio
support structure chosen such that G 3 10-10
W/K Tc 100mK K 9.5 10-9 W/K3.6 n 3.6 G
1.1 10-10 W/K Measured G is 3x smaller than
expected!
15

• Upcoming TES arrays
• SCUBA2 10000 pixels Filled array
• APEX 500 pixels Horn coupled
• SPT 500 pixels Horn coupled
• ACT 3000 pixels Filled array
• CLOVER 500 pixels Horn coupled
• OLIMPO 100 pixels Horn coupled
• FIBRE
• POLARBEAR, EBEX
• MAMBOTES
• TESCO
• ASDA
• SAINSBURYS
• MARKS AND TES

16

• New cryogenic detector systems (much development
  over the last 5 years)
• 2. Kinetic Inductance Detectors
• Nb 1 K, Al - 0.1 K no thermal isolation
  needed
• Time constant 1 us (same as InSb)
• NEP minimum 0.01 fW/rtHz (10000x more
  sensitive) (useful for very low background
  environments, e.g. narrow band)
• Arrays possible with simple readout
• Less microphonic than composite bolometers
• State of the art groups testing resonators and
  optical coupling techniques. Can work as optical
  and X-ray detectors as well (e.g. ask Duncan)
• Cardiff collaborations with SRON, Cambridge on
  KIDs
• Cardiff in-house fabrication of 4 K Nb KIDs

17

• New cryogenic detector systems (much development
  over the last 5 years)
• 3. SSPDs replace HEBs?
• NbN meander
• Time constant 3 ps (10x faster than NbN HEB)
• NEP minimum 0.00001 aW/rtHz? (very sensitive)
  (counts individual photons easily at wavelength
  of a few microns)
• State of the art single pixel photon counters
  tested up to wavelengths of 3.5 um
• Cardiff in collaboration with SRON, Moscow,
  Munich, ILTP on SSPD development and testing
• Development and fabrication of SSPDs up to now
  in Moscow

18
Mechanism of SSPD Photon Detection
19
Specifications for some commercially available
single-photon detectors and NbN SSPD at 1.3
micron wavelength
Gated regime with 0.1 per gate after-count
probability. Calculated with 10-4 per gate
probability. Data for a high-speed version
standard devices exhibit 1 105 s-1.
20
NEP of NbN SPD at 1.26 µm.
Noise Equivalent Power (NEP)
21
Two types of THz/FIR focal plane array
architectures
Bare array
Antenna coupled
IR Filter
Filter stack
Bolometer array
Antennas (e.g. horns)
X-misson line
SCUBA2 PACS SHARC2
X-missionline Filters
Detectors
BOLOCAM SCUBA PLANCK
22

• Why antenna coupling vs. filled array
• Minimizes size of detector (like a telescope) ?
• Low heat capacity ? faster
• Low thermal conductivity ? higher sensitivity
• Easier to fabricate ? higher yield
• Room for readout on chip
• Allows on-chip signal conditioning before
  detection
• Microstrip filtering
• Polarization separation and combination
• Beam combination ? interferometry
• BUT Not good for ? ltlt 100 ?m

23
THz planar antennas Quasi-optical (requires
lens) Twin-slot Log periodic Coupling to
waveguide (requires horn) Radial probe Bow tie
24

• Why filled array vs. antenna coupled
• Maximises quantum efficiency over focal plane
• Filled absorber collects all photons ? faster
  mapping
• Detectors spaced by 0.25f? ? Instantaneous full
  sampling
• Easier to fabricate for submm ? higher yield
• Potentially wide band performance - decades
• Multimode

25
NEW SUPERCONDUCTING DETECTOR SYSTEMS

• 1) Use of available TES technology for imaging
  the IR/Sub-mm region NEP ?? 10-17 - 10-18
  W/?Hz
• 2) Development of new detector types with the
  potential of a
• NEP ?lt 10-20 W/?Hz
• Non-equilibrium devices
• Cooper pair breaking in a superconductor at T lt
  Tc read-out by means of
• SSPDs (Moscow, Goltsman)
• Kinetic inductance (Caltech, Zmuidzinas)
• Bolometers
• antenna-coupled TES devices with electron-phonon
  cooling (JPL, Karasik)
• antenna-coupled SINIS devices (Chalmers, Kuzmin)

26
New system design requirements using existing
or new detector technology 1) Imaging e.g.
requires large windows and filters developed at
Cardiff 2) Spectroscopy e. g. FTS, filter
banks, grating 3) Cryogen-free systems -
including sub-K coolers developed at Cardiff
?/4
27
CONCLUSIONS

• Existing detector systems work OK buy one from
  Ken (pW/rtHz sensitivities in LHe systems, broad
  band single pixels, time constants down to 50 ps)
• New detectors promise better sensitivity/speed
• Cryogenic arrays becoming available
• New filters and optics allow flexibility in
  system design
• Cryogen-free systems becoming available soon

28

• Why TES are good
• Durability - TES devices are made and tested for
  X-ray to last years without degradation
• Sensitivity - Have achieved few x10-18 W/?Hz at
  100 mK good enough for CMB and ground based
  spectroscopy
• Speed is theoretically few ?s, for optimum bias
  still less than 1 ms - good enough
• Ease of fabrication - Only need
  photolithography, no e-beam, no glue
• Multiplexing with SQUIDs either TDM or FDM -
  impedances are well matched to SQUID readout
• 1/f noise is measured to be low
• Coupling to microwaves with antenna and matched
  heater
• thermally connected to TES - able to optimize
  absorption and readout separately

29

• Problems with TES detectors
• Saturation - for satellite and balloons.
• Excess noise - thermal and phase transition?
• High sensitivity (NEPlt10-18) requires
  temperatures lt 100 mK
• SQUID readout sensitive to EMI from mechanical
  coolers
• Solutions
• Overcome saturation by varying the thermal
  conductivity of detector - superconducting heat
  link
• Thermal modelling and optimisation
• Reduce slope of superconducting transition
• Better sensitivity requires reduced G - HEBs?
• Magnetic and RF shielding