Investigation of radiation hard sensor materials

Apr 21, 2023 Actual problems of microworld physics. Gomel 2009

. Investigation of radiation hard sensor materials

K. Afanaciev on behalf of FCAL collaboration

The Xth International School-Seminar The Actual Problems of Microworld Physics

Gomel, Belarus 2009


Very Forward Region of the ILC Detector

Interaction point

• The purpose of the instrumentation of the very forward region is:

– Hermeticity: increase the coverage to polar angles > 5mrad

– Fast beam diagnostics

EM calorimeter with sandwich structure:

30 layers of 1 X0

3.5mm W and 0.3mm sensor Angular coverage from 5 mrad to 45 mrad Moliére radius RM ≈ 1cm Segmentation between 0.5 and 0.8 x RM

BeamCal


The Challenges for BeamCal

e+e- pairs from beamstrahlung are deflected into the BeamCal

15000 e+e- per BX => Edep 10 – 20 TeV

~ 5 MGy per year strongly dependent on the beam and magnetic field configuration

=> radiation hard sensors

Detect the signature of single high energetic particles on top of the background.

=> high dynamic range/linearity

e- e+

Creation of beamstrahlung at the ILC

≈ 1 MGy/y

≈ 5 MGy/y

For 1 layer, per cell

Bethe-Heitler process


Irradiation facility

Superconducting DArmstadt LINear ACceleratorTechnical University of Darmstadt

Irradiation up to several MGy using the injector line of the S-DALINAC:8 and 10MeV electrons, beam currents from 2 to 100 nA

corresponding to doserates about 10 to 600 kGy/h


.

Investigated materials

Gallium arsenide (GaAs), Polycrystalline CVD (chemical vapour deposition) Diamond (pCVD) Single crystall CVD Diamond (sCVD) Silicon (common detector-grade Si for comparison)

GaAs Si DiamondDensity 5.32 g/cm3 2.33 3.51

Pair creation E 4.3 eV/pair 3.6 13 Band gap 1.42 eV 1.14 5.47 Electron mobility 8500 cm2/Vs 1350 2200

Hole mobility 400 cm2/Vs 450 1600 Dielectric const. 12.85 11.9 5.7 Radiation length 2.3 cm 9.4 18.8

Ave. Edep/100 m(by 10 MeV e-) 69.7 keV 53.3 34.3

Ave. pairs/100 m 13000 9200 3600Structure p-n or insul. p-n insul.


Methodology. Irradiation

collimator (IColl) sensor box (IDia, TDia, HV)

•Irradiation under bias voltage•Monitoring of beam and sample currents, sample temperature

exit windowof beam line

Faraday cup (IFC, TFC)


Methodology. CCD Setup

typical spectrum of an E6 sensor

Sr90 source

Preamplifier

Sensor box

Trigger box

& Gate

PA

discr

discr

delay

ADC

Sr90

sample

Scint.

PM1

PM2


pCVD Diamond Detector

• pCVD diamonds are an interesting material:– radiation hardness– nice properties like: high mobility,

low εR = 5.7, thermal conductivity– availability on wafer scale– Solid-state ionisation chamber (no p-

n junction)

• Samples from two manufacturers were investigated: – Element SixTM (ex-DeBeers)– Fraunhofer Institute for Applied

Solid-State Physics – IAF • 1 x 1 cm2 • 200-500 μm thick

(typical thickness 300μm)• Ti(/Pt)/Au metallization

(courtesy of IAF)

(courtesy of IAF)


After absorbing 7MGy: CVD diamonds still operational.

pCVD Diamond. Irradiation

A number of samples from two producers were irradiated

Typical behaviour:Increase in CCD at low dose =>pumping - i.e. filling of the traps

Then gradual decrease of efficiency with dose

%CCD of the pumped value vs dose


~ -80%

Signal decreased by 80 % after absorbed dose of about 7 MGyBut signal is still visible.Slight increase in current, but still in pA range

pCVD Diamond. Irradiation results

E6_4 sample from Element 6, 500 μm


pCVD Diamond. Irradiation results

Signal from sample decreases with time after irradiationfurther decrease after ‘purging’ trapping levels with UV light

Signal increases back after irradiation with small dose - ‘pumping’

FAP5 sample from IAF, 500 μm


sCVD Diamond detector

Single crystal CVD (chemical vapour deposition) diamond

CVD growth on top diamond substrate + Low defect content, very good detector properties

- Small area (up to 5x5 mm), very high price

Sample produced by Element Six 5x5 mm, 320μm thickness initial charge collection efficiency about 100% (CCD 320μm)


sCVD Diamond. Irradiation results

Irradiation to 10 MGyCCE dropped to 10%of the initial value

No visible annealing in 18 month

No significant increase in the dark currentafter the irradiation (still in pA range)


sCVD Diamond. Switched HV

Charge trapping leads to the decrease in the efficiencydue to nonuniform charge distribution=> switched HV supply => more uniform charge distribution

CCD converge to about 1.5x higherand no time dependence


Diamond Detector. Linearity

Testbeam at CERN Proton Synchrotron.4 GeV hadronsResponse to 10 ns spillsof variable intensity

linear response for particle fluxesfrom 2x104 MIP/cm2 up to 3x106 MIP/cm2


GaAs Detector

Supplied by FCAL group at JINRProduced by Siberian Institute of Technology, Tomsk

Sample is semi-insulating GaAs doped by Sn(shallow donor) and compensated by Cr (deep acceptor). This is done to compensate electron trapping centers EL2+ and provide i-type conductivity.

Sample works as a solid state ionisation chamberStructure provided by metallisation (similar to diamond)

500 m thick detector is divided into 87 5x5 mm padsand mounted on a 0.5mm PCB with fanout

Metallisation is V (30 nm) + Au (1 m)

Same material - simple 4 pad structure


GaAs. Signal

Clear separation of peaks from Sr90 source

S8 pad4ring 4

S8 pad4ring 6

Quite homogeneous response over different pads

Saturation of signal @ about 200V bias

Collection efficiency 60%


GaAs. Irradiation results

Samples 7&8, pad4, ring6 @ 200V

Preliminary

Sample 7 Sample 8

Results: CCE dropped to about 6% from 60% (by ~ 90%) after 1.5 MGythis corresponds to signal size of about 2000 e-



Dark current increased 2 times (from 0.4 to 1 A @ 200V)

Signal is still visible for an absorbed dose of about 1.5 MGy


GaAs. Second testbeam

A set of GaAs samples with different doping concentrations was irradiated

Batch # Shallow donor type Concentration, cmcm-3-3

1 Te (1-1.5)*1017

2 Te (5-6)*1016

3 Sn (1-3)*1016

Thicknesses 150 – 200 µm

Metallization:V (30 nm) + Au (1 µm) from both sides



A set of GaAs samples with different doping concentrations was irradiated

Batch with lowest shallow donor concentration shows best radiation hardness

Batch 3 - 90% dropin CCE @ 1.2 MGy


Silicon. Comparison

For comparison purposes two silicon detectors were irradiated(detector grade by Micron)

Degradation of signal starts at doses 40 kGyAt higher doses dark current and noise levelincreases dramatically (from 10 nA to A range).No clear signal visible


Conclusion

At the moment CVD diamond and GaAs prove to be sufficientlyradiation hard for application in BeamCal

Inconsistent results from different diamond samples could be a problemPumping-depumping behavior could be a problem for calorimetry

For Electromagnetic irradiation:

• pCVD Diamond operational up to 7 MGy• sCVD Diamond operational up to 10 MGy• Semiinsulating GaAs operational up to 1.5 MGy

Diamond detector have linear response for particle fluxesfrom 2x104 MIP/cm2 up to 3x106 MIP/cm2


Future work

A new testbeam is planned for this year

New GaAs samples are being produced by JINR FCAL groupWe are going to test samples with different donor and acceptorconcentrations.

More tests are needed to understand the possible application ofswitched HV to improve the diamond detector efficiency

Possible new materials?

Thank you for your attention


Backup slides


Irradiation facility

Documents

Investigation of radiation hard sensor materials