Laser tests of Wide Band Gap power devices Using Two photon absorption process Frederic Darracq Associate professor IMS, CNRS UMR5218, Université Bordeaux, 33405 Talence, France 1
Outline Two-Photon absorption (TPA) in SiC material Laser parameters and absorption measurement Single Event Effect Testing of Schottky SiC diode Conclusions Perspectives 2
SPA TPA TPA in SiC material 4H - SiC material Single photon (linear) absorption (SPA) Conduction band λ 1 Photon energy E G (3.23 ev) λ 384 nm (UV) λ 384 nm E G =3,23 ev Two-photon absorption (TPA) Photon energy 1.62eV 384 <λ 768 nm (Visible) 384 <λ 1 768 nm Valence band 3
TPA in SiC material TPA : 4H - SiC material - High photon concentration Conduction band 384 <λ 768 nm - Spatially focused beam E G =3,23 ev - Ultrashort pulse duration - Sufficient pulse energy 384 <λ 768 nm Valence band TPA coefficient β SiC unknown 4
Laser Parameters OPA Laser source : Optical Parametric Amplification based system Wavelength : 400 nm < λ 600 nm Pulse duration ~150 fs Pulse Energy 60 nj Pulse repetition rate : 1 Hz 1 khz 5
Transmitted Laser Power(µW) Absorption Measurement Samples : 4H-SiC wafer, 392 µm thick, doping=10 18 cm -3 Non negligible absorption Linear behavior compatible with free carrier absorption Incident Laser Power(µW) Increase of the absorption when the wavelength decreases 6
Single Event Effect Testing of Schottky SiC diode 7
Experimental setup OPA laser Mechanical shutter Pulse Energy setting Power measurement White light xyz translation stages 0.1 µm resolution DUT Microscope objective X100 CCD camera Digital oscilloscope v SEE R 2 C R 1 High Voltage 8
Sample preparation Backside opening Polished Backside chip Plastic Heat sink Rotating Grinding Tool Drilling a hole across the heat-sink, the backside ohmic contact, down to the chip. The opening is not centered on the chip in order to save as much contact as possible 9
Amplitude (mv) Amplitude (mv) Single Event Transients SET amplitude Vs Laser wavelength & energy Sample B2 : V Break =600 V, I max =6A Sample C1 : V Break =600 V, I max =8A 100 10 Sample B2 @ Vr = 50V 481 nm 507 nm 547 nm Substrate thickness=370µm 100 10 Echantillon C1 @ Vr = 50V 425 nm 476 nm 507 nm 547 nm Substrate thickness=380µm 10 15 20 25 30 35 40 45 50 Energie (nj) 10 15 20 25 30 35 40 45 50 Energie (nj) SET Amplitude increases with the pulse energy SET amplitude decreases with the laser wavelength But FCA seems more efficient with low wavelength 10
v SEE (mv) Amplitude (mv) frontside Single Event Transients SET amplitude Vs focusing depth Sample B2 : V Break =600 V, I max =6A Sample C1 : V Break =600 V, I max =8A 12 10 8 6 4 2 0-2 -2 0 2 4 6 8 10 Time (µs) Sample B1 Vr=150V z(µm) 260 286 312 338 364 35 30 25 20 15 10 5 sample C1 @ Vr = 50 V 481 nm 507 nm 0 100 150 200 250 300 350 400 450 Focusing depth (µm) λ=476 nm, E=10 nj E=23 nj Z sensitivity of the TPA technique 11
Amplitude (V) Amplitude (V) Single Event Burnouts Rising the reverse bias leads to an increase of the SET Sample B2 : λ=481 nm, E=41 nj SEB λ=481 nm SEB Time (µs) Reverse Bias (V) When a SEB occurs the supply voltage decreases rapidly because of the high current flowing across the generated defect 12
Reverse Bias (V) Single Event Burnouts Short SET Long SET SEB Pulse Energy (nj) Sample B : Leakage current (V AK =-50 V) - Before SEB : 1 na - After SEB : 2 ma 13
Single Event Burnouts Sample B2 : λ=481 nm, E=23 nj SEB induced at the edge of the cathode contact 14
Conclusion - Demonstration of the ability of TPA process to induce SEB in SiC Schottky diodes with various visible wavelengths - Because of free carrier absorption, moderatively high pulse energy have to be used (several tens of nj) - SET sensitive volume imaging possible 15
Perspectives - Need of low doped wide bandgap material wafer to measure the two photon absorption coefficient - Need to induce SEB signature with an adequate protection of the device in order to measure SEB sensitive volume - Study of SEE sensitivity of SiC and GaN transistors 16