Transcript Education and Outreach Program

Single Photon Counting Detectors for
Submillimeter Astrophysics:
Concept and Electrical Characterization
John Teufel
Department of Physics
Yale University
Yale:
Minghao Shen
Andrew Szymkowiak
Konrad Lehnert
Daniel Prober
Rob Schoelkopf
NASA/GSFC
Thomas Stevenson
Carl Stahle
Ed Wollack
Harvey Moseley
Funding from NASA Explorer Tech., JPL, GSFC
Overview
• Types of detectors
•Noise and sensitivity in detectors
•What is the Submillimeter?
•The “SQPC” – a high-sensitivity sub-mm detector
•Dark currents and predicted sensitivities of SQPC
• Time scales and saturation effects
• Future Work
Types of Detectors
Coherent
• Measures Amplitude & Phase
• For Narrow-band Signals
• Sensitivity given in Noise
Temperature [K]
• Adds a 1/2 photon of noise
per mode
• Minimum Noise Temperature:
TQ=hf/2k
• Example: a mixer
Incoherent
• Measures only Amplitude
• For Broad-band Signals
• Sensitivity given by NEP
[W/rt(Hz)]
• No fundamental noise limit on
detector
• Ideally limited only by photon
statistics of signal or
background
• Example: a photomultiplier
When to Use an Incoherent Detector
Average occupancy
per mode
<n>
8
RaleighJeans
Wien
n 
6
1
e
hf kTbb
1
4
In the Wien limit:
2
0
0.1
n
2
4
6 8
2
1
hf/kTbb
4
1
6 8
10
1/2 photon per mode of
noise is unacceptable!
Photon Counting in Optical
Background
Radiation
Signal
Source
Photons
nbackground + nsource
n =Rate of
incoming photons
ndark
Rate of detector
false counts
Ntot=(n + ndark)• t
Ntot = Ntot
PMT
Direct Detection with Photoconductor
+
Signal Source
Photons
••+
Background
Radiation, e.g. CMB,
Atmosphere...
Bandpass
Filter, B
Pincident = h (nsignal + nbackground )
NEPbg  nbackground  B
Typical NEPbg ~10 -17 W/ Hz
V
-
What is the Sub-Millimeter?
  100GHz  1THz
  100 m  1mm
E  h 1meV
Infrared
How Many Photons in the Sub-mm “Dark?”
3 K blackbody
10 % BW
single-mode
Photon-counting (background) limit:
NEP ~ h(n )1/2
Future NASA projects need NEP’s < 10-19
W/rt(Hz) in sub-mm !
see e.g. SPECS mission concept, Mather et al., astro-ph/9812454
The SQPC: Single Quasiparticle Photon Counter
Nb antenna
Al absorber
(Au)
•Antenna-coupled Superconducting
Tunnel Junction (STJ)
•Photoconductor direct detector
•Each Photon with 2 Al    2 Nb
excites 2 quasiparticles
~ 1 m
STJ detector
junction
sub-mm
photon
Nb
Al
Al Au
AlOx
Responsivity = 2e/photon = e/ = 5000A/W
What is measured
•Incident photons converted to current
I  2e  n  n false 
Nb antenna
Photocurrent
(Au)
Dark current
Lower Idark=> Higher sensitivity
sub-mm
photon
•Current readout should not
add noise to measurement
STJ detector
junction
V
•FET or RF-SET should have
noise  NTotal
•RF-SET is fast and scalable
Ultimate Sensitivity 
n false  I dark
Integration of RF Circuits, SETs,
and sub-mm Detectors
• 16 lithographic tank
circuits on one chip
• one of four e-beam
fields, with SETs and
SQPC detectors, and
bow-tie antenna
Sensitivity and Charge Sharing with Amplifier
Q ~ 1000 e-
CSTJ ~ 250 fF
FET (2SK152; 1.1 nV, 20 pF)
0.15 e/rt(Hz)
Collects all charge
CSET ~ 1/2 fF
RF-SET (30 nV, ½ fF)
1 x 10-4 e/rt(Hz)
Collects CSET/CSTJ ~ 0.2%
still ~ 3 times better
Either FET or SET can readout STJ @ Fano limit,
But only SET is scalable for > 50-100 readouts
Experimental Set-up and Testing
•Small area junctions fabricated
using double angle evaporation
Bow Tie Antenna
Detector
140 µm
1
µ
m
•Device mounted in pumped
He3 cryostat (T~250mK)
quasiparticle trap
SQUID
loop
junction
antenna
antenna
detector strip
1 µm
Fig. 2. (a) SQPC detector strip and tunnel junctions are located between two halves of a niobium bow-tie
antenna for coupling to submillimeter radiation. A gold quasiparticle trap is included here in the wiring to just
one of two dual detector SQUIDs. (b) Close-up view of detector strip and tunnel junctions made by doubleangle deposition of aluminum through a resist mask patterned by electron beam lithography. Pairs of junctions
form dc SQUIDs, and critical currents can be suppressed with an appropriately tuned external magnetic field.
Detector Junctions form a SQUID
Al/AlOx/Al Junctions: ~ 60 x 100 nm
BX
Current [nA]
Supercurrent Suppression
40
20
0
-20
-40
-400-200 0 200 400
3
2
1
0
0.0 0.5 1.0 1.5 2.0
Magnetic Field [mT]
Current [pA]
Current [nA]
Voltage [mV]
4
3
2
1
0
250
260
270
280
Magnetic Field [mT]
Supercurrent Contributions to Dark Current
Supercurrent
DC Power
•Cooper pair tunneling affects
the subgap current both at zero
and finite voltages
•DC Josephson effect:
IC  I o cos( )
•AC Josephson effect:
I  I C sin( J t )
J 
2eV

Zen
V
RF Power
Zen
Ic sin(J t)
SQPC
PDC  IVbias
1 2
PRF  I C Re[ Z ( J )]
2
I ( B) Re Z ( J ) *
I dark ( B) 
2Vbias
2
c
*Holst et al, PRL 1994
Magnetic Field Dependence of Sub-gap Current
I c2 ( B) Re Z ( J )
I dark ( B) 
2Vbias
Current [pA]
Current [pA]
80
60
40
20
60
40
20
0
0.0
0
0
100
200
300
Voltage [mV]
400
0.5
1.0
1.5
Magnetic Field [mT]
BCS Predictions for Dark Current
8
{
}
eV
Current [nA]
6
T=1.6 K
4
2
0
-2
T=250 mK
-4
-6
-8
-400
-200
0
Voltage [mV]
2
I dark (T ) 
e
eRn


k BT
 eV   eV 
2
 eV    Sinh 
 Ko 

eV  2
2
k
T
2
k
T
 B   B 
200
400
Thermal Dark Current Measurements
BCS Predicts: I dark (T ) e
BCS (357mK)
357 mK
256 mK
10
10
I/500
Current [A]
Current [pA]
20
10
10
10
10
0

k BT
Tc =1.4 K
-9
I @ 50 V
-10
-11
Rn= 13.1 kW
Rn= 9 kW
Rn= 47 kW
-12
-13
2
3
4
5
6
7 8 9
1
0
200
Temperature [K]
400
Voltage [µV]
Imin ~
1
19
2
pA  NEP  10 W
Hz
Recombination and Tunneling Times
Vabs
lead

(large
volume)
absorber
ttunnel ~ VabsRN
thermal trecomb ~ 100 s
@ 0.26 K
x-ray
sub-mm
at low power:
Vabs
1000 m3
0.01 m3
ttunnel << trecomb
RN
½W
50 kW
so quantum efficiency
ttunnel
2 s
2 s
is high
False count rate = Idark/e = 3 MHz for ½ pA
Saturation: Recombination vs. Tunneling
Current
Noise
trec ~ ttunn
I ~ P1/2
Absorber gap reduced
by excess q.p.’s
I~P
NEP ~ P1/4
Idark
NEP ~ P1/2
N~ Id/e
Nsat ~ (tth/ttun) Id/e
Power (P)
(or photon rate, N)
Psat~ 0.02 pW; scales as 1/RN
Demonstration of an RF-SET Transimpedance Amplifier
Input gate
0.5 fF
Trim gate
Electrical Circuit Model and Noise
I 
2eI dark
en
4kT


2
Rb
Z
Total Noise
Amplifier Noise
Johnson Noise
Shot Noise
2
-15
2
10
-16
2
6
4
10
-19
6
4
2
10
0
10
1
10
2
10
Frequency [Hz]
3
6
4
10
SQPC
-18
6
4
6
4
10
en
Johnson Amplifier
Noise
Noise
10
4
V
-20
NEP [W/rt.(Hz)]
Current Noise [A/rt (Hz)]
Shot
Noise
10
Rb
2
Sensitivity :
I
2eI dark
Future Work: Detecting Photons
•Problem: Need to couple known amount of
sub-mm radiation to detector
•Solution: Use blackbody radiation from a heat
source in the cryostat
10
10
10
10
-11
10
-12
10
-13
10
-14
10
-15
2
1
4
6
10
8
10
-9
-10
-11
-12
-13
Blackbody Temperature [K]
Current [A]
Power [W]
10
Cryogenic Blackbody as Sub-mm
Photon Source
•Hopping conduction thermistor
V
•Micro-machined Si for low thermal
conduction
Resistance [W]
W
10
10
10
10
10
8
7
To
~e
6
5
4
1
1 cm
T
10
Temperature [K]
Coming Soon: Photoresponse Measurement
Si Chip with SQPC
T= 1-10K
Quartz
Window
T= 250 mK
Advantages of SQPC
• Fundamental limit on noise = shot noise of dark current
•Low dark currents imply NEP’s < 10-19 W / rt.Hz
•High quantum efficiency – absorber matched to antenna
• High speed – limited by tunneling time ~ sec
• Can read out with FET, but SET might resolve single ’s
• Small size and power (few m2 and pW/channel)
• Scalable for arrays w/ integrated readout
Summary
• When hf>kTbb, a photon counter is preferred
•In the sub-mm, no such detector exists
•The SQPC would be a sub-mm detector
with unprecedented sensitivity
•Contributions to detector noise have been
measured and are well-understood
•Photocurrent measurements in near future