← 回 黃光
I 冊 · 黃光

Photolithography Daily Abbreviations

黃光每天會用到的縮寫
共 16 個字
— 縮寫表 · 共 16 個字 —
CDCritical Dimension
字母分開念:C-D· /siː diː/
關鍵尺寸 — 線寬或 contact 直徑等最敏感的量測值。黃光對這個錙銖必較。
The CD on this lot is shifted 1.5 nm from target.
DICDDevelop Inspect Critical Dimension
字母分開念:D-I-C-D· /diː aɪ siː diː/
顯影後量測的 CD,對應到最後 etch 完的 FICD。
DICD looks fine; let's see if FICD also passes.
FICDFinal Inspect Critical Dimension
字母分開念:F-I-C-D· /ɛf aɪ siː diː/
蝕刻後最終的 CD,客戶實際拿到的尺寸就是這個。
FICD drift suggests an etch bias problem, not photo.
DOFDepth of Focus
字母分開念:D-O-F· /diː oʊ ɛf/
焦深 — 製程窗口的關鍵指標。DOF 不夠 wafer 邊緣就會走樣。
We're running out of DOF margin on the deeper layers.
BARCBottom Anti-Reflective Coating
念整個字「巴克」· /bɑːrk/
底部抗反射層 — 在光阻下方,壓住反射造成的 standing wave。
Try a thicker BARC to kill the swing curve.
TARCTop Anti-Reflective Coating
念整個字「踏克」· /tɑːrk/
頂部抗反射層 — 蓋在光阻上方,薄薄一層。較少用。
TARC adds CD uniformity but costs an extra coater step.
PEBPost Exposure Bake
字母分開念:P-E-B· /piː iː biː/
曝後烘烤 — 觸發化學放大型光阻反應,溫度差 0.3°C 都會影響 CD。
PEB hotplate failed temp uniformity check — hold the lot.
OPCOptical Proximity Correction
字母分開念:O-P-C· /oʊ piː siː/
光學鄰近修正 — 把光罩圖形預先「變形」以補償繞射誤差。
The OPC model needs re-calibration after the new resist swap.
NANumerical Aperture
字母分開念:N-A· /ɛn eɪ/
數值孔徑 — 機台解析度的天花板。NA 越大越能畫小,但 DOF 變差。
High-NA EUV can resolve 8 nm pitch but DOF is brutal.
DUVDeep Ultraviolet
字母分開念:D-U-V· /diː juː viː/
深紫外光 — 含 KrF (248nm) 和 ArF (193nm),目前主流製程。
We're still using DUV immersion for non-critical layers.
EUVExtreme Ultraviolet
字母分開念:E-U-V· /iː juː viː/
極紫外光 (13.5 nm) — 7nm 以下節點主力。ASML 一家獨佔。
EUV pellicle inspection caught a particle — pull the reticle.
LWRLine Width Roughness
字母分開念:L-W-R· /ɛl dʌbəl juː ɑːr/
線寬粗糙度 — 線條兩邊起伏的振幅,影響 device 性能。
LWR on EUV lines is 3.2 nm; spec is 2.5.
LERLine Edge Roughness
字母分開念:L-E-R· /ɛl iː ɑːr/
線邊緣粗糙度 — 跟 LWR 相關,看的是單邊。
Reduce LER by tuning developer temperature, not exposure.
OverlayOverlay (Layer-to-Layer Alignment Error)
念整個字「歐佛累」· /ˈoʊvərleɪ/
層間對位誤差 — 上下兩層的位置偏差,nm 級。
Overlay excursion on tool LITHO-07 — escalate to PIE.
ReticleReticle (Photomask)
念整個字「瑞踢扣」· /ˈrɛtɪkəl/
光罩 — 一片要價數十萬美元(EUV 上百萬)。是黃光最珍貴的耗材。
Pellicle on reticle M-1837 has a defect — swap to backup.
ScannerScanner (Lithography Tool)
念整個字「史坎勒」· /ˈskænər/
掃描式曝光機 — 取代古早 stepper,一次掃過整片 die。ASML / Nikon / Canon。
Yield drop correlates with scanner chuck temperature drift.
動詞搭配 · Verb Collocations · 共 10 條 —
CD shifts by N nmCD + shifts / drifts / moves by [N] nm

CD 偏移 N nm。shift 中性,drift 暗示緩慢的飄移,move 較口語。Conf call 上最常用。

  • CD shifted by 1.5 nm from target on lot Q3-218 — investigating.
  • We've seen CD drift across the past four lots on scanner LITHO-07.
overlay excursion at tooloverlay + excursion / drift / spike at [tool]

對位異常 / 飄移 / 跳動發生在某機台。Excursion 是黃光部門最敏感的詞,出現基本要 escalation。

  • Overlay excursion at LITHO-07 — escalate to PIE immediately.
  • Overlay drift trending up on the new M3 layer — monitor next 5 lots.
reticle is swapped / pulled / inspectedreticle + is swapped / pulled / inspected / qualified

光罩被換 / 抽出 / 檢查 / 認證。Pellicle 有缺陷就要 pull reticle,EUV reticle pull 一次成本驚人。

  • Reticle M-1837 was pulled for pellicle inspection after a particle alarm.
  • Swap to the backup reticle — primary is being inspected.
scanner goes down / is recoveredscanner + goes down / is recovered / is qualified

曝光機停機 / 復機 / 認證。Scanner 是 fab 最貴的單體機台,downtime 直接是 WIP 災難。

  • Scanner LITHO-07 went down at 03:00 for chuck swap; ETA recovery 18:00.
  • Tool is recovered and qualified — back online for production lots.
DOF margin tightens / improvesDOF margin + tightens / shrinks / improves / opens up

焦深餘裕變窄 / 改善。先進製程 DOF 永遠不夠,每代節點都在跟 DOF 角力。

  • DOF margin is tightening at the deeper contact layers.
  • After resist swap, DOF margin opened up by 15%.
exposure dose is tuned / dropped / increasedexposure dose / focus + is tuned / dropped / increased

曝光劑量被調整 / 降低 / 拉高。Dose 跟 focus 是 scanner 兩個最基本的旋鈕。

  • Exposure dose was bumped up 3 mJ to recover the CD target.
  • We tuned dose and focus together; CD is back on target.
layer is qualified / released / holdlayer + is qualified / released / locked / held

層別認證 / 放行 / 鎖定 / 暫停。一個新 layer release 要過完整 Qual 流程。

  • M3 layer is qualified after the new OPC model passed.
  • Layer is held pending overlay re-cal on tool LITHO-09.
pellicle is replaced / damagedpellicle + is replaced / damaged / inspected / cleaned

防塵膜被換 / 損壞 / 檢查 / 清潔。EUV pellicle 是耗材,DUV pellicle 一片也要幾萬美金。

  • Pellicle replaced on reticle M-1837 — back to production.
  • Pellicle damage detected during routine inspection — pull and rebuild.
resist is coated / developed / strippedresist / photoresist + is coated / developed / stripped

光阻被塗佈 / 顯影 / 剝除。三個是 photo flow 的核心動詞。

  • Resist coated thinner than spec — adjust the spin speed.
  • Develop is fine but resist is leaving residue at the bottom.
litho is the bottlenecklitho / photo + is the bottleneck / drives the WIP

黃光成為瓶頸 / 主導 WIP。一台 scanner 停會堆 wafer。

  • Litho is the bottleneck this week — three scanners under PM.
  • Photo drives the WIP pattern on this product line.
句型 · Sentence Patterns · 共 5 條 —

deviation report · CD excursion

We observed a CD shift of N nm on lot lot_id starting at the layer step.

我們在 {批號} 從 {層別} 開始觀察到 CD 偏移 {N} nm。Photo deviation report 的標準開場。

  • We observed a CD shift of 1.8 nm on lot Q3-218 starting at the M2 photo step.
  • We observed a CD shift of -2.3 nm on lot 1842 starting at the contact layer.

8D · overlay excursion

The overlay excursion is attributed to root_cause, confirmed by evidence.

對位異常歸因於 {根因},以 {證據} 確認。8D D4 標準寫法,光罩或 chuck 是常見嫌犯。

  • The overlay excursion is attributed to scanner chuck temperature drift, confirmed by SPC trend and chamber correlation.
  • The overlay excursion is attributed to reticle stage misalignment, confirmed by the alignment metrology data.

daily review · scanner downtime

Scanner tool_id is down for reason; estimated recovery time; WIP impact: N lots.

{機台} 因 {原因} 停機;預計復機 {時間};WIP 衝擊 {N} 批。Photo 主管早會的標準報告。

  • Scanner LITHO-07 is down for chuck swap; estimated recovery 18:00; WIP impact: 12 lots.
  • Scanner LITHO-09 is down for pellicle replacement; estimated recovery 4 hours; WIP impact: 3 lots.

PIE disposition · litho

Recommend action on tool_or_recipe pending check.

建議對 {機台/配方} 做 {動作},等待 {確認} 結果。Photo 對 scanner / recipe disposition 的標準建議句。

  • Recommend recipe re-cal on LITHO-07 pending the OPC model verification.
  • Recommend hold on M3 layer pending overlay metrology cross-check.

change management · resist / OPC

After the change on date, we are re-qualifying scope per the FMEA.

在 {變更} 於 {日期} 之後,我們依 FMEA 重新認證 {範圍}。Photo 變更管理用語。

  • After the resist supplier swap on 03/12, we are re-qualifying all critical layers per the FMEA.
  • After the OPC model update on 04/01, we are re-qualifying the contact layer per the FMEA.
真實場景 · Real Scenarios · 共 3 條 —

Email — Scanner Downtime Notice

Photo 部門早班負責人 6:30 AM 發給 fab 全 OPS 的 broadcast,scanner 突然 down。

Subject: [HEADS UP] Scanner LITHO-07 down — chuck swap in progress

All,

LITHO-07 went down at 04:55 this morning. Initial diagnosis is chuck temperature drift exceeding the recovery window — requires a chuck swap.

ETA recovery: 18:00 today (about 13 hours total downtime).

Impact:
- 12 lots currently in WIP queue rerouted to LITHO-09 and -11.
- Critical layer M2 on Product X delayed by ~6 hours.
- No customer-facing impact expected.

Equipment team is on site. I'll send a recovery update at 12:00 and again when tool is back qualified.

Thanks,
Photo OPS

相關縮寫:scanner · WIP · qualification

Conf Call — CD Shift Root Cause

PIE 主持 yield review,Photo 部門被點名解釋 CD shift 根因。Photo 工程師發言 90 秒。

Thanks. Quick recap on the CD shift situation on Product X.

We started seeing CD drift on the M2 photo step about two weeks ago. SPC tagged it on day three; we held the next two lots and ran a focused investigation.

Three things we found:

First, the drift correlates strongly with scanner LITHO-07 chamber B — Pearson r is 0.88 across 22 data points. LITHO-09 and -11 are clean.

Second, the root cause is reticle stage temperature drift after the routine PM on 03/12. The thermal recovery curve didn't return to baseline within the qualification window.

Third, the containment action is recipe re-cal on LITHO-07 and a tighter thermal qualification spec going forward. Recipe was re-cal'd Tuesday; we're monitoring SPC for the next 48 hours.

Long-term, we're updating the FMEA to include this thermal failure mode and adding an in-line thermal monitor.

Open to questions.

相關縮寫:CD · SPC · FMEA · scanner · PIE

Deviation Report — Overlay Excursion (Section 4)

正式 deviation report 中 Section 4 對 overlay 異常的根因說明段落。

4.0 Root Cause Analysis — Overlay Excursion at LITHO-07

4.1 Symptom
Lot Q3-241 showed overlay error of 3.8 nm (spec 3.0 nm) on the M3 photo step. SPC alarmed at 02:14; lot held within 30 minutes.

4.2 Investigation Path
- Chamber correlation: error correlates with LITHO-07 chuck A (Pearson r = 0.93, n = 18). Chuck B and other scanners clean.
- Pareto: 70% of overlay deviation explained by stage Z-axis temperature drift since the 03/12 PM.
- Reticle alignment metrology confirmed reticle is within spec; not a mask issue.

4.3 Root Cause Statement
The overlay excursion is attributed to incomplete thermal recovery of LITHO-07 chuck A after the 03/12 PM, based on the SPC trend, chamber correlation, and stage thermal logs.

4.4 Containment
- All affected lots (4 lots, 96 wafers) on M3 layer held at WAT.
- Scanner LITHO-07 chuck A re-qualified with extended thermal soak (4 hours vs 2 hours previous).
- 100% overlay metrology on the next 10 lots through this tool.

相關縮寫:overlay · SPC · scanner · WAT · PM

讀完了?來測驗看看記住了沒。

開始測驗 → (16 題)