Advances in multi-pixel Geiger mode APDs (Silicon...
Transcript of Advances in multi-pixel Geiger mode APDs (Silicon...
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 1
Advances in multi-pixel Geiger mode
APDs (Silicon Photomultipliers).
Yuri Musienko
Northeastern University, Boston
&
INR, Moscow
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 2
Outline
• SiPM: principle of operation
• SiPM parameters, important for HEP
applications
• New developments
• Radiation hardness
• Summary
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 3
APD’s operated in Geiger mode
High gain operate APDs over breakdown Geiger mode APDs
Single pixel Geiger mode APD’s developed long time ago
( see for example: R.Haitz et al, J.Appl.Phys. (1963-1965))
Planar APD structure Passive quenching circuit
• Single pixel devices are not capable of operating in multi-photon mode
• Sensitive area is limited by dark count and dead time (few mm2 Geiger mode APD can
operate only at low temperature, needs “active quenching”)
Solution: Multipixel Geiger mode APD (MPGM APD)
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 4
First design (MRS APD, 1989) Few % photon detection efficiency for red light was measured with 0.5x0.5 mm2 APD. Good pixel-to-pixel uniformity. Small geometrical efficiency. Very low QE for green and blue light.
LED pulse spectrum(A. Akindinov et al., NIM387 (1997) 231)
The very first metall-resitor-smiconductor APD (MRS APD) proposed
by A. Gasanov, V. Golovin, Z. Sadygov, N. Yusipov (Russian patent
#1702881, from 10/11/1989 ). APDs up to 5x5 mm2 were produced by
MELZ factory (Moscow).
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 5
Developers and producersSince 1989 many GM APD structures were developed by
different developers:
• CPTA/Photnique (Moscow/Geneva)
• Zecotek (Singapur)
• MEPhI/Pulsar (Moscow)
• Hamamatsu Photonics (Hamamatsu, Japan)
• SensL (Cork, Ireland)
• RMD (Boston)
• MPI Semiconductor Laboratory (Munich, Germany)
• FBK-irst (Trento, Italy)
• ……
Every producer invented their own name for this device:
MRS APD, MAPD, SiPM, SSPM, SPM, G-APD …
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 6
Structure and principles of
operation
Picture from talk of E. Grigoriev at Como 2001
• Geiger avalanche is quenched by an individual pixel resistor (from 100kΩ to several MΩ).
• It contains 100 20 000 pixels/mm2 , made on common substrate and connected together
• Each pixel works as a binary device
• MGAPD is pixellated silicon avalanche photodiode operated in Geiger mode (~10-20% over breakdown voltage)
• For small light pulses (Ng<<Npixels) device as a whole works as an analog detector
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 7
SiPMs pixel-to-pixel uniformity
Green-red light sensitive APD, low amplitude
light signals, U=43V, T=-28 C
0
500
1000
1500
2000
0 500 1000 1500 2000
ADC ch#
Co
un
ts
MPGM APDs/SiPMs have very good pixel-to-pixel signal uniformity. Pedestal is
well separated from the signal produced by single fired pixel Q1 .
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 8
Parameter definition: Gain
Each pixel works as a digital device – 1,2,3... photons produce the same
signal Q1=Cpixel*(V-Vb) (or Single Pixel Charge).
Multi-pixel structure works as a linear device, as soon as Npe=Ng*QE<<N0
, N0 – is a total number of pixels/device
Measured charge :
Qoutput=Npe*Gain ,
It was found by many groups that : Gain ≠ Q1 ,
More than 1 pixel is fired by one primary photoelectron!
Gain=Q1*np ,
where np is average number of pixels fired by one primary photoelectron.
There are 2 reasons for this discrepancy:
- optical cross-talk between pixels
- after-pulsing (one pixel can be fired more than 1 time during light flesh)
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 9
Gain and Single Pixel Charge
)(1 Bpix VVCQ
fired
pixelsNQM 1
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 10
Photon detection efficiency
Photon detection efficiency (PDE) is the probability to detect single photon when threshold is < Q1 .
It depends on the pixel active area quantum efficiency (QE), geometric factor and probability of
primary photoelectron to trigger the pixel breakdown Pb (depends on the V-Vb !!, Vb – is a breakdown
voltage) .
PDE (l, U,T) = QE(l)*Gf*Pb(U,T)
0
5
10
15
50 55 60 65
PD
E(5
15
nm
) [%
]
Bias [V]
MEPhI/PULSAR APD
T= 22 C
T=-28 C
MEPhI/PULSAR APD, U=57.5 V, T=-28C
0
500
1000
1500
2000
2500
3000
3500
4000
4500
0 200 400 600 800 1000ADC, ch.
Co
un
tsTo determine <Npe> in light pulse one can use well known property of the Poisson distribution :
<Npe> = - ln(P(0))
Average number of photons <Ng> in LED pulse can be measured using calibrated photo-sensor . Then:
PDE(l) = <Npe>/ <Ng>
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 11
Breakdown initiation probability
Because of the higher ionization coefficient, the electron triggering probability
is always higher than that of holes
Ionization coefficients for electrons and
holes in silicon
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 12
Geometric factor
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 13
Structure for green/red light
B. Dolgoshein et. al., “An advanced study of
silicon photomultiplier”, ICFA-2001
MEPhI/PULSAR APD, T=22C, U=59 V
0
2
4
6
8
10
12
400 450 500 550 600 650 700 750 800
Wavelength [nm]
PD
E [
%]
(Y. Musienko, Beaune-05)
0
10
20
30
40
50
60
350 400 450 500 550 600 650 700 750 800
PD
E [
%]
Wavelength [nm]
T=22 CCPTA/Photonique APD
(Y. Musienko, PD-07, Kobe)
Absorption length for light in silicon
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 14
Improved blue sensitivity
Shallow-Junction SiPM
13
14
15
16
17
18
19
20
0 0.2 0.4 0.6 0.8 1 1.2 1.4
depth (um)
Do
pin
g c
on
c. (1
0^
) [1
/cm
^3
]
0E+00
1E+05
2E+05
3E+05
4E+05
5E+05
6E+05
7E+05
E f
ield
(V
/cm
)
Doping
Field
n+ p
(G. Pauletta: PD07, Kobe, Japan)0.00E+00
2.00E+00
4.00E+00
6.00E+00
8.00E+00
1.00E+01
1.20E+01
1.40E+01
1.60E+01
350 400 450 500 550 600 650 700 750 800
Wavelength (nm)
PD
E (
%)
36V
36.5V
37V
37.5V
38V
DV=2V
2.5V
3.5V3V
4V
To improve sensitivity for blue/UV light structure with shallow junction (~100 nm) was produced
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 15
Blue sensitive G-APD
p+
p--epi
n++-subst.
K.Yamamoto, PD-07, Kobe
(optical crosstalk and after-pulses were not
taken into account)
0
5
10
15
20
25
30
35
350 400 450 500 550 600 650 700 750 800
PD
E [
%]
Wavelength [nm]
T=22 C HPK S10362-11-050C
(Y. Musienko, PD-07, Kobe)
Measured using t technique described in
NIMA 567 (2006) 57
Another solution: “reversed” APD structure. In “reversed”
structure electrons initiate the avalanche breakdown
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 16
Optical cross-talk
Hot-carrier luminescence:
105 carriers produces ~3 photons with
an wavelength less than 1 mm
Increases with the gain !
Solution: optically separate pixels with grooves
Optical cross-talk causes adjacent pixels to be fired increases gain
fluctuations increases noise and excess noise factor !
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 17
Single electron spectrum
SES CPTA APD, U=42 V, T=-28 C
1
10
100
1000
10000
200 300 400 500 600 700
ADC ch.
Co
un
ts
When V-Vb>>1 V typical single pixel signal resolution is better than 10%
(FWHM)). However an optical cross-talk results in more than one pixel fired by
single photoelectron. This results in deterioration of SiPM SES and …
SES MEPhI/PULSAR APD, U=57.5V, T=-28 C
1
10
100
1000
10000
0 100 200 300 400 500
ADC ch.
Co
un
ts
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 18
Excess Noise Factor
MEPhI/PULSAR APD
0
0.5
1
1.5
2
2.5
3
0 0.5 1 1.5 2 2.5 3Single Pixel Charge*106
Nu
mb
er
of
fire
d
pix
els
T= 22 C
T=-28 C
2
2
1M
F M
… and in an increase of the SiPM excess noise factor
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 19
The dark rate of the SiPM for different gains in
dependence on the level of the threshold
0 2 4 6 8 10 12 14 1610
-1
100
101
102
103
104
105
106
dark
rate
, H
z
Threshold, pixels
gain 7*105
gain 1*106
gain 1.3*106
Optical cross-talk increases
the dark count at high
electronics thresholds
(E.Popova, CALICE meeting)
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 20
Gain is voltage and temperature
dependent
MEPhI/PULSAR APD
0
2
4
6
8
50 55 60 65Bias [V]
Gain
*106
T= 22 C
T=-28 C
MEPhI/PULSAR APD
0
0.5
1
1.5
2
2.5
3
50 55 60 65Bias [V]
Sin
gle
Pix
el
Ch
arg
e*1
06 T= 22 C
T=-28 C
MEPhI/PULSAR APD
0
0.5
1
1.5
2
2.5
3
0 0.5 1 1.5 2 2.5 3Single Pixel Charge*106
Nu
mb
er o
f fi
red
pix
els
T= 22 C
T=-28 C
SiPM gain and PDE are temperature dependent
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 21
Response temperature
sensitivityCPTA APD
0
50
100
150
200
250
300
350
400
30 32 34 36 38 40 42 44
Bias [V]
Sig
na
l a
mp
litu
de
[A
DC
ch
.]
T=-25 C
T= 22 C
Hamamatsu MPPC
0
20
40
60
80
100
120
140
160
180
200
66.5 67 67.5 68 68.5 69 69.5 70 70.5 71
Bias [V]
Am
plitu
de [
AD
C c
h.]
T=-25 C
T= 22 C
CPTA/Photnique:
dVB/dT=-20 mV/C
Hamamatsu:
dVB/dT=-50 mV/C
LED signal was measured in dependence on bias at 2
temperatures. During low temperature measurements
(T=-25 C) G-APDs were placed inside commercial
freezer (LED was kept at room temperature)
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 22
Temperature coefficient
CPTA APD
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
34 35 36 37 38 39 40 41 42 43
Bias [V]
-1/A
*dA
/dT
[%
]
S10362-11-050C HPK MPPC
0
2
4
6
8
10
12
14
16
69 69.2 69.4 69.6 69.8 70 70.2 70.4 70.6
Bias [V]
-1/A
*dA
/dT
[%
]
kT=dA/dT*1/A, [%/C]
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 23
Optical cross-talk reduction
To reduce optical cross-talk CPTA was the first to introduce trenches separating the neighbouring pixels (E. Grigoriev Como 2001)
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 24
SiPMs with reduced optical
cross-talkSiPM with trenchesSiPM without trenches
MEPhI/PULSAR APD
0
0.5
1
1.5
2
2.5
50 55 60 65Bias [V]
Exce
ss N
oise
Fac
tor T= 22 C
T=-28 C
CPTA APD
0.9
0.95
1
1.05
1.1
1.15
1.2
30 32 34 36 38 40 42 44
Bias [V]
F
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 25
The dark rate of the SiPM trenches in
dependence on the discriminator threshold
0.1
1
10
100
1000
10000
0 1 2 3
Dark
Co
un
t [k
Hz]
Threshold [fired pixels]
36V
33 V
CPTA/Photonique SSPM with trenches
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 26
SiPM timing
Measured with MEPhI/Pulsar
SiPM using single photons (B.
Dolgoshein, Beaune-02)
SiPMs have excellent timing properties
Measured with 100 mm SPAD
using single photons
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 27
Linearity
(B. Dolgoshein, TRD05, Bari)
This equation is correct for
light pulses which are shorter
than pixel recovery time , and
for an “ideal” SiPM (no cross-
talk and no after-pulsing)
Linearity of SiPM is determined by its total number of pixels
In the case of uniform illumination:
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 28
Micro-pixel APDs with large
dynamic range
Z. Sadygov, Beaune-05
M.Golubeva et.al. LONGITUDINALLY SEGMENTED
LEAD/SCINTILLATOR HADRON CALORIMETER WITH
MICROPIXEL APD READOUT (this conference)
Micro-well structure with multiplication
regions located in front of wells at 2-3 mm
depth was developed by Z. Sadygov.
MAPDs with 10 000 – 15 000 pixels/mm2
were produced. Such devices are good for
calorimetry applications.
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 29
After-pulsing
Solution: “cleaner” technology or longer pixel recovery time
-0.35
-0.3
-0.25
-0.2
-0.15
-0.1
-0.05
0
0.05
-1.0E-08 1.0E-08 3.0E-08 5.0E-08 7.0E-08
Time (s)
Vo
lta
ge
(V
)
Events with after-pulse measured on a
single micropixel.
y = 0.0067x2 - 0.4218x + 6.639
y = 0.0068x2 - 0.4259x + 6.705
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16
31 32 33 34 35 36
Voltage (V)
Afte
rpu
lse
/pu
lse
Tint = 60ns
Tint = 100ns
After-pulse probability vs bias
C. Piemonte: June 13th, 2007, Perugia
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 30
Large area SiPMs
1x1mm 2x2mm 3x3mm (3600 cells) 4x4mm (6400 cells)
FBK-irst SiPMs, C. Piemonte: June 13th, 2007, Perugia
SiPMs with up to 3x3 mm2 area produced by many companies: Hamamatsu,
CPTA/Photonique, Pulsar, Zecotek, SensL, FBK-irst
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 31
Radiation hardness studies
Motivation: SiPMs will be used in HEP experiments
Radiation may cause:
• Fatal SiPM damage (SiPM can’t be used after certain
absorbed dose)
• Dark current and dark count increase (silicon …)
• Change of the gain and PDE vs. voltage dependence
(SiPM blocking effects due to high induced dark carriers
generation-recombination rate)
• Breakdown voltage change
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 32
Radiation damage of MPPCs using
gammas from Co60
Matsubara, PD-07, Kobe
Infrared emission (similar
effects were seen with
irradiated “linear” APDs
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 33
Radiation damage of SiPMs using protons
200 MeV protons, M.Danilov
arXiv:0704.3514v1
Protons like 1 MeV neutrons produce
defects inside silicon.
Increase of the dark current:
Id~a*F*V*M*k,
a – dark current damage constant [A/cm];
F – particle flux [1/cm2];
V – silicon active volume [cm3]
M – SiPM gain
k – NIEL coefficient
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 34
Irradiation studies at PSI (82 MeV protons)
• 23 SiPMs from 4 different producers irradiated at PSI last week
• 1*1010 400 MeV/c (82 MeV kinetic energy) protons/cm2 in 4 steps
•NIEL factor is ~2 times of 1 MeV neutrons, Total flux: 1*1010 protons/cm2, equivalent to ~2*1010 1 MeV neutrons/cm2
Gain, PDE, Id, Dark count vs. voltage were measured before irradiation
0
2
4
6
8
10
12
0 1 2 3 4 5
Dar
k C
urr
en
t [m
kA]
Irradiation #
"HPK_1mm_#535" (U=69.5V)
"HPK_1mm_#535"(U=70.1V)
"HPK_1mm_#535"(U=70.7V)
CPTA#3
0
5
10
15
20
25
30
30.5 31 31.5 32 32.5 33 33.5 34 34.5
Bias [V]
PD
E(5
15n
m)
[%]
before irr.
after irr.
FBK_K1
0
5
10
15
20
25
30
31 31.5 32 32.5 33 33.5 34 34.5
Bias [V]
PD
E(5
15n
m)
[%]
before irr.
after irr.
Hamamatsu #534
0
5
10
15
20
25
30
69 69.2 69.4 69.6 69.8 70 70.2
Bias [V]
PD
E(5
15
nm
) [%
]
before irr.
after irr.
Actve area of SiPMs ~1 mm2
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 35
Dark Count vs. bias 2 month after irradiationHamamatsu #534
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
69 69.2 69.4 69.6 69.8 70 70.2
Bias [V]
Da
rk C
ou
nt
[kH
z] before irr.
after irr.
FBK_K1
0
5000
10000
15000
20000
25000
30000
35000
40000
31 31.5 32 32.5 33 33.5 34 34.5
Bias [V]
Da
rk C
ou
nt
[kH
z]
before irr.
after irr.
CPTA#3
0
5000
10000
15000
20000
25000
30000
35000
40000
30.5 31 31.5 32 32.5 33 33.5 34 34.5
Bias [V]
Dark
Co
un
t [k
Hz]
before irr.
after irr.
Dark count increase
0
200
400
600
800
1000
1200
1400
1600
0 5 10 15 20 25 30
PDE(515nm)
Da
rk C
ou
nt/
PD
E(5
15n
m) CPTA#3
FBK_K1
Hamamatsu #534
0
5
10
15
20
25
0 10 20 30Eff
ec
tive
Th
ick
nes
s [
mm
]
PDE(515nm)
Effective thickness (aSi= 4*10-17 A/cm)
CPTA#3
FBK_K1
Hamamatsu #534
INSTR-08, Novosibirsk, 3.03.2008 Y. Musienko ([email protected]) 36
Summary
Significant progress in development of SiPMs during recent 2-3 years:
• High PDE~30-40% for blue-green light (CPTA/Photonique, Hamamatsu)
• Reduction of dark count at room temperature (~100-300 kHz, Hamamastu)
• Low cross-talk (<5-10%, CPTA/Photonique, FBK)
• Low temperature coefficient (~0.3%/C, CPTA/Photonique)
• Fast timing (~50 ps (RMS) for single photons)
• Large dynamic range (10 000 – 15 000 pixels/mm2, Dubna/Zecotek)
• Large area (3x3 mm2, CPTA/Photonique, Hamamatsu, FBK, SensL,
Dubna/Zecotek…)
All this (together with understanding of radiation hardness issues) makes
these devices excellent candidates for application in HEP experiments
(see presentations of T.Ijima, P.Krizhan, A.Reshetin, D.Renker,
V.Rusinov, S.Schuwalow, Yu.Kudenko , A.Ivashkin…)