RADFET development and calibration
CNES/ESA Final Presentation Days ESTEC, Noordwijk, The Netherlands 28/29 March 2011
Aleksandar Jaksic Devices and Systems for Dosimetry Tyndall National Institute, Ireland
CNES/ESA Final Presentation Days, ESTEC, The Netherlands, 28/03/2011 Outline:
• Basics of RADFET operation • Overview of the work performed • Main results: – Qualification tests and RADFET supply – RADFET calibration – RADFET reader improvements • Future work
CNES/ESA Final Presentation Days, ESTEC, The Netherlands, 28/03/2011 RADFET operating principle:
• Radiation creates electron-hole pairs • Initial recombination of electrons and holes happens • Non-recombined electrons leave the oxide; holes are trapped in the vicinity of the oxide/silicon interface
• RADFET threshold voltage (VT) changes (ΔVT ~ Dose)
CNES/ESA Final Presentation Days, ESTEC, The Netherlands, 28/03/2011 RADFET biasing configurations:
• Irradiation (sense mode): zero current; (B, S and D grounded); G can be:
– Grounded (VIRR=VGS=0V) (option 1)
– Biased (typically VIRR=VGS>0) (option 2) Read-out mode: specified current applied to S=B; G=D grounded
S - Source B - Bulk G - Gate D - Drain
B G
• Irradiation (sense mode) and Read-out mode are the same (option 3)
CNES/ESA Final Presentation Days, ESTEC, The Netherlands, 28/03/2011 RADFET advantages over other dosimeters: • Immediate read-out without destroying the data • Extremely small sensor chip • Very low or zero power consumption • Electronic signal • Low cost (especially of the read-out electronics) Applications: • Nuclear industry and research • Space dosimetry • Radiotherapy • Personal dosimetry [?]
CNES/ESA Final Presentation Days, ESTEC, The Netherlands, 28/03/2011 Overview of main project tasks:
• Supply of RADFETs: – Flight samples for imminent space missions – Experimental samples for space missions and other ESTEC needs • RADFET characterisation and calibration • RADFET reader improvement
CNES/ESA Final Presentation Days, ESTEC, The Netherlands, 28/03/2011 ESAPMOS4 RADFET chip:
• Chip size: 1mm x 1mm • Contains four RADFETs: – two 300/50 devices – two 690/15 devices • Chip types (gate oxide): – 100 nm – 400 nm – 400 nm Implanted (IMPL) – 1 μm – 1 μm Implanted (IMPL)
CNES/ESA Final Presentation Days, ESTEC, The Netherlands, 28/03/2011 RADFET qualification:
• Wafer lot acceptance – Construction analysis of RADFET die with SEM and cross- section • Validation of the assembly – Construction analysis (package) – Residual Gas Analysis – Bond pull test – Die shear test • Assembly and screening • Life test – Burn-in 1000 hr at 125 degrees
CNES/ESA Final Presentation Days, ESTEC, The Netherlands, 28/03/2011 Process flow:
Item No. Description Completed
1 Wafer Lot Acceptance Yes
2 Validation of Assembly Yes
3 Visual inspection to Space Level Yes 4 Pre-cap CSI, including sample bond pull and die shear Yes 5 Temperature cycling Yes 6 Acceleration 30KG Yes 7 Particle Impact Noise Detection (PIND) condition A Yes 8 Pre electrical test and datalog at ambient Yes 9 240hr power burn-in at 125C Yes 10 Post electrical test and datalog at ambient Yes 11 Electrical test and datalog at –55°C Yes 12 Electrical test and datalog at +125°C Yes 13 Fine & Gross Seal Tests Yes 14 External visual Yes Life test: - Sample 10 units - Electrical test and datalog at ambient 15 Yes - Burn-in 1000hr at 125°C - Electrical test and datalog at ambient - External visual
CNES/ESA Final Presentation Days, ESTEC, The Netherlands, 28/03/2011 RADFET supply: • Flight samples: – 4 parts for Proba-II – 16 parts for Alphasat – 20 parts for GlobalStar – 16 parts for O3B (TAS) – 6 parts for BepiColombo • Experimental samples: – 12 parts for Proba-II – 40 parts for high ion irradiations in Juvaskyla – 20 parts for Alphasat – 20 parts for GlobalStar
CNES/ESA Final Presentation Days, ESTEC, The Netherlands, 28/03/2011 Difficulties encountered:
• Originally chosen RADFET batch exhibited VT shift during packaging in 14-pin ceramic packages • Observed behaviour was not always consistent • Detailed tests have shown that the batch in question indeed
exhibited a VT shift owing to a high temperature process during hermetical packaging • Moisture ingression is most probable culprit • Storage of RADFET wafers is critical • RADFET wafers are not passivated – this makes problem worse • Passivation could improve things, however…
CNES/ESA Final Presentation Days, ESTEC, The Netherlands, 28/03/2011 Passivation (Si3N4) and sensitivity/fading: • No effect of passivation to pre-irradiation I-V curves • Effects emerging during and after irradiation
Average Wafer Radiation Sensitivity Avergage Fade
1.0000 3.0000
0.9000
0.8205 0.7792 2.0000 0.8000
0.7000 1.0000
0.6000 0.5712
0.5000 0.0000 W10 (no pass) W9 (pass) W8 (pass,anneal) Fade (%) Fade
Sensitivity (mV/cGy) 0.4000
-1.0000 0.3000
0.2000
-2.0000
0.1000
0.0000 -3.0000 W10 (no pass) W9 (pass) W8 (pass,anneal) Wafer
CNES/ESA Final Presentation Days, ESTEC, The Netherlands, 28/03/2011 RADFETs on NANOSAT 1B:
Heliosynchronous; 98º inclination; Orbit type 690 km height Orbital period 98 minutes Orbits per day 14.7 Life time Minimum 3 years. Up to 5 years Mass 20 kg Length 50 cm
CNES/ESA Final Presentation Days, ESTEC, The Netherlands, 28/03/2011 RADFET board set-up and calibration:
RADFET 400nm calibration function up to 5 krad(Si) 2 1.75 Dose rate: 1.5 130 rad(Si)/h 1.25 1
Voltaje [V] Voltaje 0.75 Δ 0.5 0.25 0 012345 TID [krad(Si)]
RADFET 1μm calibration function up to 5 krad(Si) 6
5
4 Dose rate: 3 130 rad(Si)/h
Voltaje [V] 2 Δ 1
0 012345 TID [krad(Si)]
CNES/ESA Final Presentation Days, ESTEC, The Netherlands, 28/03/2011 Precise calibration and total dose:
Calibration functions up to 100 rad(Si) TID on board NANOSAT1B 100 0.3 1 um; DR = 130 rad/h 90 400nm RADFET 1 um; DR = 660 rad/h 80 1μm RADFET 0.25 70 400 nm; DR = 130 rad/h 0.2 60 400 nm; DR = 660 rad/h 50 0.15 40 Vth [V] Vth TID [rad(Si)] Δ 30 0.1 20 0.05 10 0 0 9 0 0 09 09 09 0 1 10 -10 10 10 0 102030405060708090100 -7- -9- -3- 4 5- 6- -10- -12-09 8- 6- 3- 1-7-1 30 27-8- 24 14-1- 11-2-1011 TID [rad(Si)] 22 19-11-0917 Date
CNES/ESA Final Presentation Days, ESTEC, The Netherlands, 28/03/2011 RADFET reader improvements:
• Current RADFET automated read-out system: – Passive board (in the radiation room) – Controller board (in the radiation room, shielded) – PC with application SW (in the control room, RS-232 connection)
• Proposed modifications: – Change cable between passive and controller board to facilitate placing controller board in the control room – Change connection between controller board and PC to USB to avoid problems with USB/RS-232 adapters – Introduce single mains supply (currently two adapters) – Overhaul application software
CNES/ESA Final Presentation Days, ESTEC, The Netherlands, 28/03/2011 Proposed future work: • Improvement of the RADFET reader • Additional flight and experimental samples • Further RADFET characterisation – Irradiations (Co-60, protons) – Passivation solutions (nitride, CVD oxide, polyamide)
CNES/ESA Final Presentation Days, ESTEC, The Netherlands, 28/03/2011