Instrument: Zeiss 1540 (RPI cleanroom)

XC2C32A VQG44 sample was decapped in package using sulfuric acid. Not tested post-decap.

Evaporated amorphous carbon using E&ES system. Target thickness was 20 nm, actual value not measured.

Mounted directly to multi-sample holder using carbon tape. All leads appeared well grounded.

Cut through nitride over three wires, then backfilled holes and deposited 20 um x 20 um Pt pads.

Cleaned for ~5 mins in indirect O2 plasma. C residue remained.

Soldered sample to board and tested. Completely dead, not even JTAG IDCODE response. No current measurements collected.

Instrument: Zeiss 1540 (RPI cleanroom)

XC2C32A VQG44 sample was decapped in package using sulfuric acid. Tested and confirmed working post-decap. Soldered short stub wires on to aid test.

Evaporated amorphous carbon using E&ES system. Target thickness was 25 nm. Used bare <110> Si wafer as thickness measurement target, measured ~21 nm by ellipsometry.

Mounted to pin on multi-sample holder with carbon tape. Ensured all critical leads were grounded, most others seemed to make good contact but this wasn't tested.

Cut through nitride over one wire, then backfilled hole and deposited 15 x 15 um Pt pad.

Cleaned for ~15 mins in direct O2 plasma, some C residue remained. Repeated for a total of 45 mins at ~100W (standard photoresist strip process). Several pins were likely floating during this process.

Plugged sample into breadboard and tested. Completely dead, not even JTAG IDCODE response. VCCIO+VCCAUX had notably high leakage (20+ mA).

Current hypothesis: Plasma clean is killing chips, try using flood gun on FEI Versa and not coating.