Aston™ 質譜儀設備與工藝協同優化 EPCO瀏覽數: 843
Aston™ 質譜分析儀設備與工藝協同優化 EPCO: 380 億美元的製造優化機會
先進的工藝需要設備和工藝協同優化 EPCO. 麥肯錫公司 McKinsey & Co. 在2021年發表的一篇論文表明, 利用人工智慧 AI 和機器學習 ML 進行半導體製造優化, 通過提高產量和輸送量, 有望節省 380億美元的成本.
Advanced processes now require Equipment and Process Co-Optimization (EPCO). A 2021 paper by McKinsey & Co. demonstrated that semiconductor manufacturing optimization, using artificial intelligence (AI) & machine learning (ML), represents a $38B cost saving opportunity through improved yields and increased throughput.
麥肯錫強調, 幫助企業實現這些好處的最大干預點是調整工具參數, 使用當前和以前步驟的即時工具感測器資料, 使 AI/ML 演算法優化工藝操作之間的非線性關係.
McKinsey highlighted the single biggest intervention point to help companies realize these benefits is adjustment of tool parameters, using live tool sensor data from current and previous steps to enable AI/ML algorithms to optimize the nonlinear relationship between process operations.
成功部署 AI/ML 的關鍵是可操作的即時資料. 伯東公司日本 Atonarp Aston™ 質譜儀的原位即時分子診斷和雲連接資料是實現這一能力的關鍵技術, 從而解鎖半導體設備與工藝協同優化的潛力.
Key to successful AI/ML deployment is actionable real-time data. Aston in-situ real-time molecular diagnostics and its cloud connected data are key technologies enabling this capability to unlock the potential for semiconductor EPCO.
問題
隨著工藝節點的縮小, 影響工藝良率的新變數出現, 挑戰了已建立的 Copy Exactly！ 方法論. 其中一些可能影響工藝性能的關鍵變數包括局部虛擬真空洩漏, 細微的反應氣體分壓變化, 由於泵送性能變化導致的晶片表面飽和, 由於晶片溫度變化導致的表面反應性, 腔室清潔終點和腔室老化曲線.
As process nodes have shrunk, new variables have emerged that affect process yield, challenging the established Copy Exactly! methodologies. Some of these critical variables that can affect process performance include localized virtual vacuum leaks, subtle reaction gas partial pressure variations, wafer surface saturation due to changes in pumping performance, surface reactivity due to changing wafer temperature, chamber clean end point, and chamber seasoning profile.
其他挑戰, 如層間粘附, 300mm 晶圓機械應力, 新的原子級沉積和蝕刻化學, 特殊的低電阻接觸和填充金屬, 嚴格的交叉污染協議和最大化輸送量, 都需要更深入地瞭解工藝和設備的相互作用, 優化諸如此類的先進工藝現在需要更高精度的計量工具, 增加了 Copy Exactly! 方法學協議的原位分子複雜性.
Additional challenges such as inter-layer adhesion, 300mm wafer mechanical stresses, new atomic level deposition and etch chemistries, exotic low resistance contact and fill metals, stringent cross contamination protocols and maximizing throughput all require greater insight to how the process and equipment are interacting. Optimizing advanced processes such as these now demand higher accuracy metrology tools and add a new layer of in-situ molecular complexity to the Copy Exactly! methodology protocols.
伯東公司日本 Aston™ 質譜儀提供設備與工藝協同優化解決方案: 原位,即時資料
半導體程序控制 FAB 環境中的資料主要分為三種類型:
1. 在工藝工具上即時獲取的現場資料
2. 處理步驟後測量結果的線上資料（通常立即）
3. 參數或 Fab 後資料（用於晶圓生產線良率和晶圓出貨驗收標準）
There are three main types of data in the semiconductor process control FAB environment:
1. In-Situ data taken real time on the process tool
2. In-line data to measure results (usually immediately) after a processing step
3. Parametric or post-Fab data (used for wafer line yield and wafer ship acceptance criteria)
此外, 這三個主要資料可以進一步分為三個子類型
1. 目標資料, 即作為配方一部分的工具所針對的目標, 例如, 目標溫度: 327 °C, 目標 SiF4 摩爾濃度: 100 mol/l
2. 測量資料, 即在給定情況下測量的資料, 例如, 測量溫度 9 °C, 實際 CF4 摩爾濃度: 0.097 mol/l
3. 資訊資料, 例如晶圓批號: 8F2342G, 設備序號和腔室: 32FF4567-4
Additionally, these three main data can be further categorized into three sub-types
1. Target Data i.e., what the tool was targeting as part of the recipe e.g., target temperature: 327 °C, target SiF4 molar concentration: 100 mol/l
2. Measured Data i.e., what is measured at a given instance e.g., measured temperature 9 °C, actual CF4 molar concentration: 0.097 mol/l
3. Informational Data i.e., wafer lot number: 8F2342G, equipment serial number & chamber: 32FF4567-4
在分子水準上測量原位即時資料可以真正洞察過程是如何設置和進行的, 提供豐富, 可操作和有影響力的資料. 反應物, 副產物和分壓濃度可以被識別和量化, 允許動態程序控制, 以確保對給定過程模組在運行到運行, 腔室到腔室, 工具到工具之間進行嚴格的平均和標準差控制 -工具, 甚至網站到網站. 管理整體複雜的半導體工藝控制和生產線良率首先要嚴格控制各個工藝步驟, 並確保低可變性和嚴格的統計工藝控制 SPC.
Measuring in-situ, real-time data at the molecular level gives true insight to how the process is setup and proceeding, offering rich, actionable, and impactful data. Reactants, by-products, and partial pressure concentrations can be identified and quantified, allowing for dynamic process control to ensure tight mean and standard deviation control for a given process module across run-to-run, chamber-to-chamber, tool-to-tool, and even site-to-site. Managing overall complex semiconductor process control and line-yield starts with having tight control on individual process steps and ensuring low variability and tight statistical process control (SPC).
伯東公司 Aston™ 質譜的設計初衷是為了滿足原位分子分析的需求, 從而實現 EPCO, Aston 強大的即時原位分子感測器解決方案具有許多先進的性能優勢, 包括:
Aston from Atonarp was designed from the ground up to meet the needs of in-situ molecular analysis to enable EPCO. The advanced performance benefits available with Aston’s robust, real-time, in-situ molecular sensor solutions are many, including:
• 準確的即時終點檢測 Accurate real-time based end point detection
• 逐次運行和即時 EPCO Run-to-run and real time EPCO
• 參數調整 Parameter adjustments
• 機器學習, 人工智慧 Machine learning, artificial intelligence
• 過程統計程序控制和偏差識別 Process statistical process control and deviation identification
• 生產線良率根本原因分析 Line yield root-cause analysis
• 優化的預防性維護 Optimized preventative maintenance
• 跟蹤重要工具或流程 Tracking to a golden tool or process
|
Aston™ 質譜儀特點 |
應用 |
|
1. 耐腐蝕性氣體 Corrosive Gas Resistant |
1. 介電蝕刻: Dielectric Etch |
若您需要進一步的瞭解 Atonarp Aston™ 在線質譜儀詳細資訊或討論, 請參考以下聯絡方式：
上海伯東: 葉小姐 臺灣伯東: 王小姐
T: +86-21-5046-3511 ext 107 T: +886-3-567-9508 ext 161
F: +86-21-5046-1490 F: +886-3-567-0049
M: +86 1391-883-7267 ( 微信同號 ) M: +886-975-571-910
qq: 2821409400
現部分品牌誠招合作代理商, 有意向者歡迎聯絡伯東 王小姐 +886-975-571-910
伯東版權所有, 翻拷必究
E-mail contact
+886-2-8772-8910