Silicon Photomultiplier tests in LN, LAr Janicsk o J ozsef, - - PowerPoint PPT Presentation

Silicon Photomultiplier tests in LN, LAr

Janicsk´

• J´
• zsef, Aghaei Khozani Hossein

March 3, 2010

Setup

• LAr

SiPM SiPM VM2000 WLS fiber HPGe detector Dewar

24 cm 8 SiPM DSG 6−fold segmented HPGe 20 m WLS fiber

Little space left for LAr. (X0 = 14 cm, space left 7 cm)

March 3, 2010 1/14

Setup

• 4 x 5m WLS fiber total surface:

600cm2

• 8 x SiPM with 100 pixels
• charge sensitive preamps
• VM2000 foil, home made TPB coating
• VM2000 surface about 3000 cm2
• 18 l active volume (or less), 25kg LAr

March 3, 2010 2/14

Co60, Th228 - background

Plenty of light seen, hundreds of pixels fired, no features in the sum spectrum.

100 200 300 400 500 600

−3

10

−2

10

−1

10 1 10

2

10

Co60 − background Th228 − background

Sum spectra of 8 SiPM’s for Co60 and Th228 external source. (number of pixels fired) March 3, 2010 3/14

Contamination of LAr with Air

The proof that we detect scintillation light:

hps0 Entries 2380800 Mean 7.677 RMS 12.46

• Nb. p.e.

50 100 150 200 250 300 350 400

• 2

10

• 1

10 1 10

2

10

3

10

hps0 Entries 2380800 Mean 7.677 RMS 12.46

hps0

Contamination of LAr with air reduced the detected light intensity by 10X

March 3, 2010 4/14

Anti Compton veto

The HPGe detector used:

• DSG

6 fold segmented detector

• ∼nA leakage current
• ∼10keV resolution

March 3, 2010 5/14

Anti Compton veto

suppressed spectrum for Co60 and for Th228, 1 p.e. threshold

h1

Entries 2.3808e+07 Mean 660.1 RMS 373.4

E [keV] 500 1000 1500 2000 2500 3000 1 10

2

10

3

10

4

10

5

10

h1

Entries 2.3808e+07 Mean 660.1 RMS 373.4

h1

h1

Entries 2.3808e+07 Mean 660.1 RMS 373.4

h1

Entries 4761600 Mean 741.4 RMS 679.3

E [keV] 500 1000 1500 2000 2500 3000 1 10

2

10

3

10

4

10

h1

Entries 4761600 Mean 741.4 RMS 679.3

h1

The flat (Compton) background is suppressed by about a factor two.

March 3, 2010 6/14

DEP

• DEP peak suppressed 3.7 times, gamma line 1.13 times with 1 p.e.

threshold

• The flat background around the DEP peak was suppressed by 2.5 times.

h1

Entries 1.90464e+07 Mean 1617 RMS 67.2

E[keV] 1500 1550 1600 1650 1700 1000 2000 3000 4000 5000 6000 7000 8000

h1

Entries 1.90464e+07 Mean 1617 RMS 67.2

h1

March 3, 2010 7/14

Single segment cut + anti-Compton

Segmented detector + anti-Compton veto Single segment cut with anti-Compton veto together 4 to 5 times suppresion of the flat background.

h1

Entries 1.90464e+07 Mean 892.4 RMS 701.9

E [keV] 500 1000 1500 2000 2500 3000 10

2

10

3

10

4

10

5

10

h1

Entries 1.90464e+07 Mean 892.4 RMS 701.9

h1 E [keV] 1550 1600 1650 1700 1750 1000 2000 3000 4000 5000

All events SSE cut AC cut SSE + AC cut

March 3, 2010 8/14

Next steps

• Anti-Compton veto with SiPM’s was demonstrated

R&D plans for an improved setup:

• Understand the SiPM properties better: Breakdown voltage, gain, dark rate,

Xtalk, afterpulsing

• WLS fiber: BCF-92 ⇒ BCF-91A better match for TPB
• WLS fiber: from ⇒ fiber, from 5.6 to 7.3% trapping efficiency
• Improve TPB coating and try other fluor
• More SiPM’s (?)

March 3, 2010 9/14

SiPM properties

Gain and break down voltage measurment:

• Breakdown voltage = Biasvoltage (Gain = 0)
• Is important to define the overvoltage. VBias(T) = VBreakdown(T) + VOver

Bias [V] 60 62 64 66 68 70 Gain 400 500 600 700 800 900 1000 1100 1200

10 ×

77K 139K 202K 267K 293K SiPM400_Gain-Temp

Temperature [K] 100 150 200 250 300 BreakdownVoltage [V] 58 60 62 64 66 68 70

SiPM400_BDV-Temp

March 3, 2010 10/14

Q.E. & noise v. overvoltage

Q.E. increases with overvoltage, xtalk and afterpulsing as well

ADC counts 2000 4000 6000 8000 100001200014000160001800020000 100 200 300 400 500

hmca_0

• Overvolt. 0.7
• Overvolt. 1.4

Same light intensity, different over voltage March 3, 2010 11/14

Dark Rate v. Temperature

We always knew that the dark rate is small:

Temperature [K] 100 150 200 250 300 dark rate [Hz]

• 2

10

• 1

10 1 10

2

10

3

10

4

10

5

10

0.5pe thr. 1.5pe thr.

⇒ In LN 1 p.e. threshold is reasonable because the dark rate is so small.

March 3, 2010 12/14

WLS Spectral properties

which one is a better match for the TPB emmision spectra? BCF-92 ⇒ BCF-91A expect 20% improvement, from ⇒ fiber another 30%

March 3, 2010 13/14

Last slide

• We are working on it, next experiment with HPGe detector in a few month
• expect 100% improvement to the previous experiment
• for 1000 p.e./MeV new (bigger, better) setup is needed

To be continued ...

March 3, 2010 14/14