Kinetic Inductance Detectors SRON Stephen Yates Jochem Baselmans

Kinetic Inductance Detectors SRON Stephen Yates, Jochem Baselmans, Andrey Baryshev, Jan Joost Lankwarden, Henk Hoevers. TNO G. Gerini, A. Neto, D. Bekers. Radiation TU Delft R. Barends, J. R. Gao, T. M. Klapwijk. Cardiff University S. Doyle, P. D. Mauskopf, P. Ade Al film 1 Sub-mm 2

SRON: The Netherlands Institute for Space Research • Low Energy Astrophysics; historically: IRAS, ISO-SWS. . . • PI for Herschel HIFI (PI) • ALMA band 9 • • High Energy Astrophysics • • XMM-Newton Sensor Research and Technology TES microcalorimeters for IXO (XEUS) • KIDs research • also HEBs • • Also Earth Orientated Science • Engineering Division - ASIC development • 2

SPICA-SAFARI SPICA Japanese satellite 3. 5 m 4 K mirror (first ever) -> very low photon background European instrument SAFARI FIR-sub-mm FTS imaging spectrometer selected ESA cosmic vision program 2015+ 3 bands Band 1 f/20 implies 31 x 31 mm arrays NEP < 2 10 -19 W/ Hz, 3 d. B roll-off > 20 Hz Instantaneous dynamic range >1000 3

KID: Principle of operation P. Day, et al. , Nature 425, 817 (2003). • • • Superconducting pair breaking detector • • Measure broken Cooper pairs by measuring the Kinetic Inductance • At T<<Tc Superconductor impedance Read out Zs by resonant circuit @ F=2 -8 GHz Combine superconductor in series with C Read-out using phase or amplitude! Zs ~ -i LK Quasiparticles N ~ P / E Superconductor 1 2 EF Photons E > 2 Cooper Pairs 1 2 Superconductor 4

Antenna coupled KID Position Antenna ¼ resonator @Freadout end Most sensitive @ 1 2 RF Photons E > 2 Coupler Readout current Printed antenna @ FRF >> Freadout Most sensitive area CPW ¼ Resonator Antenna does not influence resonator L= 5 mm @ 6 GHz Needs lens! 100 m Al ground plane CPW Through line Bare substrate 1 Readout signal ~4 GHz Coupler Length sets Coupling Q Central conductor 5 2

Measured Beam Pattern X Y 6

SAFARI KID focal plane concept Si microlens array 2 nm roughness 7 nm spherical precision 250 nm position accuracy With markers + Antenna coupled KID design Printed lithographically alignment using backside markers present resonator design fits within F/20 Alignment antenna-lenses within 1 -2 m Identical misalignment for all pixels (if any) No show stoppers expected for F/20 for all SPICA arrays Monolithic Si lens array Many resonators Many resonance features @ different F Radiation 7

Dark NEP • Measurements of noise, responsivity and lifetime => NEP • Lifetime ~ 1 msec • • BW = 160 Hz • Dynamic • P/NEP ~ 10. 000 Hz (KID 43) ~ 3000 Hz (KID 44) NEP~6 10 -19 W/ Hz • • 40 nm. Al on Sapphire, sputtered 100 nm Al on SI, evaporated Range: using θ<90° • Only optical coupling as uncertainty • Calculations ~agree with optical NEP 8

IRAM camera • Take advantage of work for SPICA • Planar antenna experience • Work on electronics • Can support KID option • Demonstrator 1 kpixel • array manufacturer • advise/collaboration readout, electronics Need collaborators • Neél/Rome/Cardiff • 9

Advantages KIDs • • • Cryogenically simple: • 1 coax, cryogenic amplifier for ~10000 pixels Antenna coupling - can have multi band/polarisation or very wideband Have (electrical) NEPs better than required Simple manufacture - high yield (typical~95%) Sensitivity: • vibration insensitive • magnetic field needs to be constant (i. e. SC shield) but doesn't change performance 10

ADR setup Also 3 K SC shield • Vericold GMBH • Pulse tube 3. 5 K • Dual stage ADR • Tmin<50 m. K Sample holder 50 m. K Magnetic shield (bottom part) Nb. Ti Coax Stainless Steel Coax, can use Cu. Ni from 4 K DC blocks 50 m. K 500 m. K link LNA 4 K stage DC block 11

SAFARI KID Readout • • • Þ • Satalite: 20 W , 5 kg Best first order estimate most optimistic design: Power consumption: ~300 W using 180 nm ASIC, ~80 W using 90 nm ASIC • 6000 pixels in 1 GHz BW => Highly questionable!!! • 20 x 50 MHz subbands, 6% dead pixels due to confusion • DAC: 53 d. B SNR, ENOB=8. 6 • 40 analog mixers: 4 -8 kg On chip analog microwave cirquits + optimised digital design needed ASIC design group within SRON Digital experience from Eureca - Xeus 4 -8 GHz BW would be more realistic -> factor 4 in power 2 GHz 2. 05 GHz 2. 1 GHz ~ ~ 10 MHz reference ~ ~ 100 Msample/sec DAC’s DAC 20 m. K 4 K ADC DAC ADC 12

SAFARI KID cryogenic design 242 K – 4 K Sub 1. 7 K single shot Cooler 3 arrays, total 5940 pixels All pixels in 1 pair of cables SPICA has 11 temperature stages 3 m from 242 K – 4. 5 K, ~5 d. B cable loss Sub 1. 7 K Cooler Sorption Cooler + ADR, single shot LNA’s: HIFI heritage T Pload length loss 242 K 100 K 30 m. W 0. 5 m 0. 9 d. B Warm electronics 85 K 60 K 41 K 18 K 4 m. W 3 m. W 2 m. W 1 m. W 0. 5 m 0. 9 d. B 4. 5 K 0. 3 m. W 8 W 0. 5 m tbd 0. 9 d. B 0 3. 6 mm Cu. Ni Ag cladded coax LNA 12 d. B 2 m. W 2. 5 K 1. 7 K 300 m. K 1 W tbd 0 0. 5 W tbd 0 100 m. K 50 n. W 1. 6 mm Nb. Ti Coax LNA 12 d. B 2 m. W 3 arrays in series 13