Radiation damage studies in LGAD detectors from recent CNM and FBK run - PowerPoint Presentation

1. Radiation damage studies in LGAD detectors from recent CNM and FBK runGregor Kramberger(a), Mar Carulla Areste(b), Emanuele Cavallaro(c), V. Cindro(a), I. Mandić(a), M. Petek(a)(a) Jožef Stefan Institute, Ljubljana(b) CNM, Barcelona(c) IFAE, Barcelona

2. CNM Run 6827 (Epitaxial devices, 100 Wcm, 50 mm thick)LGAD samples of low boron concentration in multiplication layer – gain of 7 reached only at high bias voltages Control PIN samples with no multiplication layerExcellent high voltage toleranceCNM Run 9088 (SOI devices, high-resistivity, 45 mm thick)Three different multiplication layer doping concentrationsW3 – Dose=1.8e13 cm-2W5 and W7- dose =1.9e13 cm-2 (most studied)W11 – Dose=2.0e13 cm-2 Three different device structuresControl PIN diodes produced along the LGAD15.11.2016G. Kramberger,"Update of the LGAD radiation hardness studies", HGTD meeting, CERN2CNM SamplesHGTD4x2x2 mm2LGB3x3 mm2LGA1.3x1.3 mm2

3. Similar device performance regardless of the device typeTypical break down in CCE measurements ~300 V (W3), ~260 V(W5) and 90V (W11)28/10/2016G. Kramberger, "Status of thin CNM LGAD radiation hardness studies" , HGTD detector meeting3Non-irradiated devices CNM-R9088T=20oCT=20oCMost probable signal ~2800 e for non-irradiated detector – 45 mm thickRatio most probable/mean=0.7 – slightly less was expected.

4. Gain degrades, but follows the expectations (several new fluences added)“Breakdown” of the device is shifting to higher bias voltages with irradiation for W515.11.2016G. Kramberger,"Update of the LGAD radiation hardness studies", HGTD meeting, CERN4Signal/gain after neutron irradiations (R9088)G~10G~60G~20G~41.) at fluences < 1e15 cm-2effective acceptor removal in multiplication layer reduces the gain, which appears as soon as the multiplication layer is depleted. 2.) Smaller Neff in multiplication layer leads to smaller slope of the Q-V dependence.3.) at fluences of 1e15-2e15 cm-2the multiplication is visible only at higher bias voltages – up to few 100 V collected charge similar to pin diode - the difference between LGAD and PIN becomes small.T=-10oCDose = 1.921G~10G~60G~20G~6T=-10oC

5. At high fluences F>2e15 cm-2 the behavior is the same for all samples Regardless of initial doping concentration .Regardless of p+ layer doping .Regardless of annealing behavior (needs to be verified by several samples – so far only for pin diode).It seems that the influence of fluence on Q-V is very small.LGAD are advantageous (CCE) for high gain device up to 2e15 cm-215.11.2016G. Kramberger,"Update of the LGAD radiation hardness studies", HGTD meeting, CERN5Signal after neutron irradiations for thin LGADsG=8.4The last point is always taken just before the detector exhibits a “soft break down”:Irradiation shifts the breakdown voltage to higher values.The rise of the charge is associated with the rise of the current and noise (system dependent)

6. Signal – does initial doping matter (R6827)?Simple calculation of gain from field qualitatively describes the gain at very high fluences!G. Kramberger, "Status of thin LGAD radiation hardness studies" , HGTD meeting, CERN613/9/2016T=-10oChttps://indico.cern.ch/event/436057/contributions/1921349/subcontributions/173698/attachments/1260164/1862618/RadiationHardnessOfHGTD-3.pdfcalculationpredictionR6827The effect of multiplication layer is also visible for R6827 at 1e15 cm-2It is gone at 3e15 cm-2 , where there is no difference between PIN and LGAD (low dose device)Conclusions:Initial doping seem to be important at lower fluences, where higher gain can be achieved at bias voltage (benefits in power and safety)At higher fluences any small difference that remains can be compensated by slightly higher voltage.

7. GeometrySmall 0.5 mm2 Multi guard ring structureNot all small diodes were diced (multiple samples)Two wafers W3 and W10LGAD of different gains for W10 (Split 4), W3(Split 2)PIN diodeDoping profiles – deep dopingFBK sensorsIs such profile more/less radiation hard?28.10.2016G. Kramberger, First look at the radiation hardness of FBK LGADs, HGTD detector meeting7

8. FBK TCT measurements – Pulses15.11.2016G. Kramberger,"Update of the LGAD radiation hardness studies", HGTD meeting, CERN8Vbias=300 VT=20oCGTCT=14Vbias=160VT=20oCGTCT=6.8ILGADIPINIPINILGADW3 – wafer W10 – wafer Vbias=160VT=20CGTCT=3.49Vbias=300VT=20CGTCT=3.84IPINILGADIPINILGADPIN diodeLGAD dioderedlaserredlaser

9. Back illumination – clear separation of primary and multiplied charge (calculation of gain possible)The space charge seems to change as a function of voltageAt lower voltages the space charge seems to be positive (see the slope of the charge)At higher voltages the space charge seems to be negativeThe same observation for front illumination and Q-V for back illumination.TCT (W3 samples)backill.frontill.28.10.2016G. Kramberger, First look at the radiation hardness of FBK LGADs, HGTD detector meeting9

10. Huge gain for W10; G>100Same observation of space charge changing with bias voltageAt larger gain the temperature dependence is much more pronounced TCT (W10 samples)G=7G=7Vbias=160 VT=20C28.10.2016G. Kramberger, First look at the radiation hardness of FBK LGADs, HGTD detector meeting10

11. GTCT=7@160VGTCT~200 @ 480VLarge gain of the device at high bias voltages!28.10.2016G. Kramberger, First look at the radiation hardness of FBK LGADs, HGTD detector meeting11TCT (W10 samples)FBKW1013ABC

12. Several features not really understood:Sensitivity to DC illumination (not seen in CNM samples) – indication of deep defects?Deep defects would also explain changes of space charge with voltageUnknown features in the signal for W10 – unclear if real physics (e.g. illumination of guard rings) of “feature of setup”TCT for W10 – changes in space chargea change due to DC ill.not observed for CNMDC – illuminationNo DC illuminationDC – illuminationNo DC illuminationFront ill.Vbias=100 VT=20C28.10.2016G. Kramberger, First look at the radiation hardness of FBK LGADs, HGTD detector meeting12

13. 15.11.2016G. Kramberger,"Update of the LGAD radiation hardness studies", HGTD meeting, CERN13FBK 90Sr measurements – charge collection15000 e450 V, @ 20oC147 V, @ 20oCFBKW315ABFBKW318ABFBKW10F1AFBKW10F1ASmall sample size is a problem as trigger purity is affected:Noise peak – events missing the sensorGuard ring collection – capacitevly transferred charge to the connected electrodeActive area – multiplication Huge signals saturate the amplifier for W10 – break down ~500-600V

14. TCT – W10 (2e14 cm-2) W10293 K-10C 20CBack illumination28.10.2016G. Kramberger, First look at the radiation hardness of FBK LGADs, HGTD detector meeting14W320C-10C 20CBack illumination

15. FBK - CCE with 90Sr after 2e14 cm-2T=20CLarger gain loss for GMIP wrt. to GTCT – under investigation Similar performance as CNM detectors15.11.2016G. Kramberger,"Update of the LGAD radiation hardness studies", HGTD meeting, CERN15GMIP~4, GTCT~6

16. First FBK production of LGAD with deep implants were produced.TCT and CCE measurements showed very high multiplication and excellent break down performance before irradiation.After irradiation loss of gain is comparable to CNM devices – a more systematic study is ongoing.Conclusions28.10.2016G. Kramberger, First look at the radiation hardness of FBK LGADs, HGTD detector meeting16