Just when you think going to be a peaceful weekend, QC detected with suspect undercut. Inspected, whole lot affected with undercut on capacitor. Defect looks same for all 25pcs and is affecting the whole wafer.
Etch capacitor done in WB11 @er:878 A/min. No machine down. No abnornal delay in WSO. Lot ran before: W2543 already terminated without hold. after: W15V24 also no hold.
Mask capacitor have down time when running this lot in WW54 due to machine hang. Checked with masking engr but say the machine problem is vacuum, and will not cause overexpose on wafer.
Checked with G2 bayleader and G3 bayleader, they used the fix time for this prod for this capacitor lvl, which is 11m. As per G2 bayleader, if ER at high side, and used 11m will see high undercutut, thus they use 10.5min.
As per G3 bayleader, she always used 11m, she do see slightly high under cut for some device(she forget de device) but not as bad as the current affected lot.
Asked G4 Operator Wanita both uses etch timing based on incoming thickness calculation. Usually etch time is around 10.5min.
She was on OT tonight, she said all the while she is following the r/s timing (11min) for this product/level. When told her about the lot history on bench, she said "maybe" wrong etch timing used.
Checked the r/s for of the affected lot + other lots that she etched, nothing written on r/s. The 15.5min lots ran before and after the affected lot, already terminated.
Tagged WW1YK8 (same pdt), measured incoming thickness ~7200-7300 and took 1pc LM19 to etch 15.5min, the rest did std 11min. Seen higher undercut for LM19. Refer to photos. Adv hold this lot after metal etch to check.
Pls help to follow up HIF w/ JJ since delay time from etch to semitool very long ~51mins (13:47-14:38)
Interviewed JJ with Leng, accor to JJ, she followed procedure and sometimes 11mins can see slightly bigger undercut thus reduced time to 10.5mins. She was also the one who inspect the lot aft etch, to her the undercut is still acceptable thus move out the lot. She recalled that she changed the timing to 11mins for WB11 when she wants to etch the lot. The 51mins delay frm etch to srd is due to her break time, no one to transfer the lot to srd
Help tag same prdt after metal etch strip resist/freckle & take photo for similar structures to compare undercut . Previous tag lot WYK8 now at clmode after etch contact,to take photo after ewmetal strip resist & to compare undercut of lm19 & lm13 . N3 plan to discuss w/ device on this issue
