Study of a Thermal Annealing Approach for Very High Total Dose Environments S. Dhombres 1-2, J. Boch 1, A. Michez 1, S. Beauvivre 2, D. Kraehenbuehl 2, F. Saigné 1 RADFAC 2015 26/03/2015 1 Université Montpellier, IES (CC 05006) - UMR UM2/CNRS 5214-860, Rue de St Priest, 34095 Montpellier, France 2 Systheia, 672 rue du Mas de Verchant, 34000 Montpellier, France 1
1. Context Issue Environment : Civil Nuclear Power (Power plant) and Space Cameras : deploying solar panels, observation, monitoring,... Restrictive radiative environment: ions, protons, electrons, neutrons Frame and electrical parameters Degradation ( > 100 krad ) Limited lifetime Human access/repair impossible : high doses, in flight,... How to extend lifetime of CMOS APS image sensor in harsh environment until 300 krad then 1Mrad and more? 2
1. Context Work objectives Main objective : extend CMOS APS image sensor lifetime in harsh environment Develop thermal annealing method [1] Heat to recover initial parameters Apply this method at regular time to recover frame Parameter degradation Acceptable level Define fundamental parameters : Bias, temperature, annealing duration Thermal annealing Time [1] F. Roig, L. Dusseau, J. Boch, et al., «A Thermal Annealing Approach to Extend Metal Oxide Semiconductor Devices Lifetime Exposed to Very High Dose Levels», IEEE RADECS, Sept 2012, Biarritz (France). 3
1. Context Radiation-induced effects on image sensor Typical X-Rays degradation Main effects on CMOS APS - Increase of dark current Increase of «hot pixels» Before irradiation Time After 250 krad - Random Telegraph Signals noise - Reduced of responsivity Dark current : After 325 krad After 625 krad Hot pixel = pixel with high dark current - Leakage current in the absence of light 4
Image sensor principle Architecture CMOS image sensor : pixel matrix Active-Pixel Sensor (APS) : photo-detector + active amplifier Pixel : place of conversion photons electrons by photoelectric effect Clock Row decoder j Column readout circuit Column decoder i Image sensor architecture «Pinned-PhotoDiode 4T APS» 5
1 st Test of Thermal annealing Experiments Irradiation with 22 krad step Condition tests : all pins grounded Continuous thermal annealing at High Temp. Condition tests : all pins grounded : 01h00 200 C : 02h30 200 C : 03h30 200 C : 06h30 200 C : 71h30 25 C High temperature (> 150 C : several minutes) allows to recover faster than low temperature (<150 C : several hours) 6
Reliability in time X- Rays Experiments Sensor A without annealing : - Irradiation by 25 krad step - All pins grounded Sensor B with annealing (175C/30 minutes) : - Irradiation by 25 krad step - All pins grounded Dark current average [DN] 1100 1000 900 800 700 600 500 400 300 200 100 0 Estimated deposited dose (krad) 25 75 125 175 225 275 325 375 425 475 525 575 625 675 725 775 825 875 925 975 1025 Sensor with annealing After anneal. (High Temp) Irradiation (Before annealing) Sensor without annealing 0 1 2 3 4 5 6 7 8 9 10111213141516171819202122232425262728293031323334353637383940 Cycle number At 175 o C/30min : 40 irradiation and annealing cycles and 1 Mrad!! Acceptable level [2] S. Dhombres, A. Michez, J. Boch, and al., «Study of a thermal annealing approach for very high total dose environments», IEEE Trans. Nucl. Sci., Dec 2014 7
Reliability in time Experiments Sensor A without annealing Sensor B with annealing Before irradiation Before irradiation After 200 krad After 200 krad After 600 krad Unusable After 600 krad Usable 8
Keys Parameters of Thermal Annealing X- Rays Experiments To keep an acceptable level: Increase of Temperature Increase of Thermal Annealing Duration Decrease of Dose Step Parameter degradation Acceptable level Thermal annealing Time 9
Thermal annealing protocol IRF130 power MOSFETs Annealing is also sensitive to electric field CMOS APS image sensor = complex component Tested on IRF130 N-power MOSFET and vertical structure. Method development and Better understanding of mechanisms 10
Thermal annealing protocol Negative bias during thermal annealing Negative bias during annealing Drain current Id (A) 17 Pre-irrad 17 17_IRRAD_12V Irrad 10krad/12V 17_REGEN- Anneal. -20V - - 003h 20V_3H 17_REGEN_- Anneal. -20V - - 271h 20V_271H 17_REGEN_- Anneal. -20V -- 470h 20V_470H 17_REGEN_- Anneal. -20V - 662h 20V_662H 0,011 0,01 0,009 0,008 0,007 Anneal. 0,006 0,005 0,004 0,003 Irradiation 0,002 0,001 0-5 -4-3 -2-1 0 1 2 3 4 5 Gate voltage Vgs (V) Dose rate : 108,1 rad/s Bias during irradiation 10 krad : +12V Bias during annealing at 100 C : -20V Incomplete recovery Saturation recovery 11
Thermal annealing protocol Positive bias during thermal annealing Positive bias during annealing Drain current Id (A) Pre-irrad 19 19 Irrad 10krad/12V 19_IRRAD_12V Anneal. 19_REGEN+20V_3H - 003h 19_REGEN+20V_27H Anneal. 20V - 027h 0,011 0,01 0,009 0,008 0,007 0,006 0,005 0,004 0,003 Annealing Irradiation 0,002 0,001 0-5 -4-3 -2-1 0 1 2 3 4 5 Gate voltage Vgs (V) Dose rate : 108,1 rad/s Bias during irradiation 10 krad : +12V Bias during annealing 100 C : +20V Complete recovery after 27h For 0V, -20V bias recovery > 1000h! Bias/Duration effects 12
Trapping/Detrapping Temperature effect 2 - - E C E ph = h.υ Photon 1-3 4-5 + 5 - + 4 3 E t_n E t_p E V Traps levels + + 2 + 1 Radiation-induced electron-hole pairs generation 2 Carriers drift-diffusion in their respective allowed band 3 Free carriers trapping 4 Recombination of trapped carriers by free carriers of opposite type 5 Thermal reemission of trapped carriers to their respective allowed band 13
Bias effect Poole-Frenkel and tunneling effects Poole-Frenkel effect 1 E' C Tunneling effect E C Φ 2 ΔΦ n x Al SiO2 - Si Φ b Ec Ev kt E trap - Electric field 1 Electric field curves conduction band 2 Barrier potential lowering for electrons 1 Electric field curves conduction band 2 Apparent thinning barrier potential lowering for electrons 14
Conclusion / Perspectives Conclusion Heating image sensor allows to : recover degraded electrical parameters (dark current) keep an usable frame after 1Mrad and more. 40 irradiations/annealing cycles were supported at high temperature : reliability in time No premature breakdown observed due to thermal cycles if temperature < 150 C Perspectives On going : Understanding of physical mechanisms with TCAD ECORCE (Pixel modeling in 2D). Law extraction depending of thermal annealing duration and temperature 15
Thank you for your attention Any questions? 16
Dark current mechanisms Origin Measurement in complete dark without illumination. Dark count = an electron is generated and pixel electronic considers it as an incident photon 4 main mechanisms : - Surface generation from trap states at the Si-SiO2 interface - Trap-assisted thermal generation of charge in the space charge layer - The diffusion current due to thermal generation of charge in the bulk - Crosstalk between neighboring pixels during charge collection, storage, and possibly readout 17