201223703 六、發明說明： 【發明所屬之技術領域】 本揭示案係關於使用光學監控在研磨上之控制。 【先前技術】 積體電路通常藉由將導電層、半導體層或絕緣層相繼 沉積於矽晶圓上而形成於基板上。一個製造步驟涉及 以下步驟：在非平坦表面上沉積填料層及平坦化填料 層。針對某些應用，平坦化填料層直至暴露圖案化層之 頂端表面。舉例而言’可在圖案化絕緣層上沉積導電填 料層以填充絕緣層尹的溝或洞。在平坦化後，殘留在絕 緣層之突起圖案間之導電層的部分構成通孔、插座和接 線’通孔、插座和接線在基板上的薄膜電路間提供導電 通道。針對其他應用（例如，氧化物研磨），平坦化填料 層直至預定厚度被留在非平坦化表面。此外，光微影技 術通吊需要將基板表面平坦化。 化學機械研磨（Chemical mechanical polishing; CMP) 是一種公認的平坦化方法。平坦化方法通常需要將基板 安裝於承載或研磨頭上。通常將基板之暴露表面抵靠旋 轉研磨墊置放。承載頭在基板上提供可控制負載以推動 基板抵靠研磨墊。通常將一個可研磨之研磨漿供應至研 磨墊之表面。 CMP t所存在的一個問題是決定研磨過程是否已經完 201223703 成’亦即決定是否已經將基板層平坦化至所需的平度或 厚度’或決定何時已移除所需要量之材料。在研磨漿分 布、研磨墊情況、研磨墊與基板之間的相對速度及基板 上負載之變動可以導致材料移除速率之變動。這些變 動，如同基板初始厚度之變動，導致需要到達研磨終點 時間之變動。因此’經常無法僅僅將研磨終點決定為一 個研磨時間的函數。 在一些系統中，在研磨期間，例如，透過研磨塾中的 一個視窗在原處（in-situ)光學監控基板。然而，現存的光 學監控技術也許無法滿足半導體裝置製造商日益增加的 需求。 【發明内容】 由於上述的變動，在不同時間，來自基板之不同個區 域之一個覆蓋層可以被清除。針對一些材料，光學監控 無法可靠的偵測正在研磨之層的厚度。然而，光學監控 經常可以偵測覆蓋層之清除狀況。於清除來自基板不同 個區域之覆蓋層時（亦稱為清除時間），一種改良均勻的 技術為：清除來自基板不同個區域之覆蓋層用以偵測第 基板之不同個區域之覆蓋層清除狀況；和基於第一基板 清除之偵測來調整後續第二基板上之至少研磨壓力。 在一態樣中，在研磨期間，一種控制研磨的方法包含 以下步驟：研磨步驟’研磨第一基板，第一基板在下層 201223703 或層結構上具有覆蓋層；導引步驟，將光束導引至第一 基板上面，光束從第一基板反射以產生反射光束；產生 步驟’在研磨期間產生反射光束強度之連串測量；分類 步驟，將測量分類為群組，每一群組和基板上多個區域 之一個不同區域相關聯；決定步驟，針對每個區域，決 定一個時間，在時間基於來自相關聯群組的測量來清除 覆蓋層；計算步驟，基於在研磨基板期間施加在至少一 個區域之一個壓力、用於至少一個區域之時間、及用於 其他區域之時間，計算至少一個區域之至少—個第二個 經調整研磨壓力；及研磨步驟，使用至少一個調整後的 研磨壓力來研磨第二基板。 實施方式可以包括一個或更多個以下特徵。光束可為 非偏光束。非偏光束可為雷射光。非偏光束可為寬頻可 見光束。覆蓋層可為GST’光束可包含紅外線元件，而 反射光束強度之測量可為反射光束之紅外線元件強度之 測量。覆盍層可為金屬(例如，銅、鋁、鎢、鈕、鈦或鈷卜 光束可包含紅色元件’且反射光束強度之測量可為反射 光束之紅色元件強度之測量。區域可為同心徑向分散狀 區域。研磨可包含以下步驟：使用具有多個室之承載頭 來研磨以將獨立可調整之壓力施加在基板上多個區域。 多個室中之第一室可施加第 域’而多個室中之第二室可 第二區域。決定每個區域中 步驟：決定來自多個區域中 在第一基板的研磨過程中， 一壓力於多個區域之第一區 施加第一壓力於多個區域之 下層暴露的時間可包含以下 201223703 之第一區域之第一時間及決定來自多個區域中之第二區 域之第二時間。基於第一壓力、第—時間、及第二時間， 可計算第一室之至少一個調整後的研磨壓力。第二區域 可為一個最内側之區域或一個最外側之區域。計算調整 後之壓力pi，可包含以下步驟：計算ρι，= ρι*(τι/τ2)， 在此處之P1為第一壓力，T1為第一時間而仞為第二時 間。決定下層暴露之時間可包含以下㈣：決定連串測 量穩定之時間。決定連串測量穩定之時間可包含以下之 步驟：決定由連_之測量所產生之跡線斜率，在預定時 間週期維持在預定範圍内。 在其他態樣中，在研磨期間，一種控制研磨之方法包 含以下步驟：研磨步驟，研磨第一基板，第一基板在下 層或層結構上具有覆蓋層；監控步驟，用原處（in_si^) 監控系統來監控基板以產生連争測量；分類步驟，將測 量分類為群組，每一群組和基板上多個區域之一個不同 區域相關聯’決定步驟，針對每一個區域， 基於來自相 關聯群組之測量，決定覆蓋層清除之時間丨計算步輝， 基於研磨基板期間施加在至少一個區域之壓力、用於至 少一個區域之時間、及用於其他區域之時間，計算至少 一個區域之至少一個第二個經調整研磨壓力；及研磨步 驟’使用至少-個調整後之研磨壓力來研磨第二基板。 實施方式可包含-㈣更多個下列特冑。原處監㈣ 統可包含-個&學監控系、统’光學監控系統將光束導引 至基板上。原處監視系統可包含摩擦感應器。針對每個 201223703 區域，決定下層暴露 時間可包含以下步驟：決定來自 多個區域之第^一 域之第一時間和決定來自多個區域之 第一區域之第二時間。 基於第一壓力、第一時間和第二 時間，可計箕第+ f , 之至個調整後的研磨壓力。計 算調整後的壓力ρι， !刀包含以下步驟：計算P1，= Ρ1*(Τ1/Τ2)，此處的 ρι 是— τ 疋乐歷力，Τ1是第一時間及 Τ2是第二時間。 在另一態樣中，電腦可讀媒體上已健存有指令，指令 在被處理器執行時，使處理器執行運算。 實鈿方式可包含一個或更多個下列的潛在優點。可減 少晶圓内不均勻（Within_wafer η.—⑹mhy WIWNU)。覆蓋層（例如，GST層或金屬層）之清除可實 質上：同時地發生於基板之表面上，此方式可以改善研 磨產量•調整研磨過程以對可消耗生命（例如··研磨 頭，研磨塾或研磨漿）之製程堆積物做出補償。也能改善 不同研承载頭在移除外形上之變化。 在附圖及下列敘述中提出一個或更多個詳細的實施 例。可由敘述、圖示及申請專利範圍中瞭解本案其他態 樣、特徵及優點將》 〜 【實施方式】 在一些半導體晶片製造過程中，覆蓋層（例如，金屬（例 如，銅、鎢、鋁、鈦、钽或鈷）、類金屬或金屬及/或半 201223703 金屬之合金（例如，GeSbTe(鍺、銻及碲之三元化合物， 也稱為GST)))被沉積在圖案化下層或層結構上（例如，一 或更多其他層之堆疊）。該一或更多其他層可包含介電材 料之層（例如，低介電常數之材料及/或低介電常數之蓋 材料）、或障礙金屬之層（例如，氮化组或氮化鈦卜常常 研磨覆蓋層直至覆蓋層被清除，亦即，直至暴露下層或 層結構之頂表面。覆蓋層之部分也許會留在溝、洞等， 溝、洞等由下層或層結橼之圖案所提供。 一般而言，需要完全清除覆蓋層（例如’覆蓋層沒有不 連續區域，覆蓋層覆蓋下層或層結構之頂表面），而且基 本上需要在同一時間跨過基板之表面。完全清除覆蓋層 可避免過度研磨，增加產量及減少晶圓間不均句 (within-wafer non-uniformity; WIWNU)。 覆蓋層研磨之監控所存在一個的問題是：對於—也材 料（例如，高反射率的金屬）而言，基板之光學（例如，光 譜）監控在大量（bulk)研磨期間針對覆蓋層厚度，也許不 會提供有用的資訊。不受限於任何特定理論，通常使用 波長來做光學監控，覆蓋層材料之消光係數（extincti⑽ coefficient)也許夠高，使在大量研磨期間厚度減少時， 反射率不會明顯改變。因而，光學監控在一些材料之大 量研磨期間也許不適用於研磨參數之原處回饋控制。雖 然當清除覆蓋層及暴露下層時，覆蓋層之反射率可以改 變，但到了偵測到光學行為中這樣的改變時，針對正在 研磨之基板之研磨速率之原處調整可能已經太晚以致無 201223703 法改善跨過基板清除時間的不均勻了。 此外，針對這些相同材料中的 二渦電流監控在偵 測覆蓋層之清除還不夠有效率。 半舉例而言，障礙金屬（例 如’氮化鈦及氮化钽），因為# 马他們本質上的高電阻率，障 礙金屬無法提制電流好的訊號。針對其他金屬（例如， 銅’銘和鶴），渦電麵心無法_留在基板上之不連 續金屬薄膜中獲得訊號， 祕 和復盍層清除之偵測也因此而 變得不可靠。 然而對於許多製程而言，霜 復盖層具有一個與下層或層 結構不同之反射率。不受限於任何特定理論，這也許因 :覆蓋層與下層具有-個不同(例如，較高)的消光係 數。在這些情況下，當杳 … 滑除覆蓋層和暴露下層時，反射 率或來自基板之反射光譜應該會改變。 下文討論之技術用來改善清除覆蓋層的不均句狀況， 此技術為：光學偵測第一基板之多個不同區域中之覆蓋 層之清除’和基於在第一其杯p店如 土板上偵測到的清除時間，調 整後續第二基板之至少一彳 ^個&域上之—個壓力，使得在 具有調整的情況會比古 嘗比在不具有調整的情況下，在更接近 相同的時間（例如，基， 同時間）元成（occur)清除。 雖然此技術可以胜a丨田，、，$ π 特別用以處理上述問題，即使光學監 控在大量研磨期間可槎 权供針對覆蓋層厚度的有用資 訊，此技術仍可有—舻庙 般的應用。在這個情況下，此技術 還有其他優點，例如，+ π β u ,, 灼如，在不同材料間控制製程的一致性， 或計算負載的減少『田& Μ也，、主^ (因為偵測清除比決定厚度較不會有 10 201223703 大虿S十异）。因此，此技術可+ 4 J用在其他半透明材料（例如， 半透明金屬（例如，GST))。廿 )尤其，此技術是關於當清除 層時，反射軌跡之行為是否 ~疋古有明顯的改變。 第1圖是化學機械研磨 W怨1站150之示意性橫截面圖，可 操作化學機械研磨站15〇 乂研磨基板10。研磨站150包 含可旋轉圓盤狀平台24，/-!· 在可旋轉圓盤狀平台24上安 裝研磨墊30。可操作平a 一 十〇 24以繞著軸25旋轉。舉例而 言，馬達（未圖示）可以轉動驅動輪27來旋轉平自Μ。舉 例而言，研磨墊3 0可藉由魟金府 猎由黏著層以分開地的方式緊固於 平口 24 _L 研磨墊3〇磨損時，可分離及替換研磨塾 30。研磨墊30可以是具有外部研磨層32和較軟支持層 34之雙層研磨墊。 研磨站150彳以包含組合的研磨浆/水洗臂(Hnse 叫39。在研磨過程中，可操作臂39來分散研磨漿38(例 如，-種具有研磨顆粒之液體)。或者，研磨站22包含 研磨料，可操作研磨料Μ散研ϋ㈣研磨塾30 上。 研磨站150也包含承載頭80，可操作承載頭80以使 基板1〇抵靠研磨墊30固定。承載頭8Μ皮懸掛在支持纤 構上(例如，迴轉料架154)’且承載頭8〇藉由承載驅動 ，74連接到承載頭旋轉馬達％，以使承載頭⑼可以繞 著轴7丨旋轉。此外，承载頭8() ^在支持結構構成的 徑向槽1M則向擺動（⑽腕e)。在操作上，平纟24圍繞 著平㈣中心軸25旋轉’而承裁頭8〇繞著承載頭的中 201223703 〜軸71旋轉而且側向地移動跨過研磨墊3〇之頂表面。 >考第2圖，為了施加獨立可控制之壓力到基板上多 個區域（例如，同心區域），承载頭8〇可以包含多個室。 在一個實施中，承載頭80包含外殼302、基座總成3〇4、 環架結構306(可將環架結構視為基座總成304的一部 分）、負載室308、固定環310、及基板支持總成32〇,基 板支持總成320包含彈性膜326 ’彈性膜320定義多個 獨立可加壓室（例如，内室330、中間室332、中間室334、 中間室336、及外室338)。這些加壓室控制彈性膜同心 區域上之壓力’因此這些加壓室可以在基板10同心部分 上提供獨立的壓力控制。在一些實施中，承載頭8〇包含 五個室且在每一室中具有壓力調節器。舉例而言，參考 第3B圖’五個室330、332、334、336、及338可以控 制施加在基板1〇上之五個同心區域Zi、Z2、Z3、Z4、 及Z5之壓力。 參考第1圖，研磨站150也包含光學監控系統，光學 監控系統可用來決定如以下討論之研磨終點。光學監控 系統包含光源5 1和光偵測器52。光從光源5 1經過，穿 過磨研墊30 ’照射基板1 〇及從基板後端反射以穿過磨 研墊’及前進至光偵測器52 » 藉由包含孔（亦即，穿越墊的一個洞）或固態視窗36， 提供了穿越研磨墊30之光的存取。雖然在一些實施中， 固態視窗36可以被支撐在平台24上及投射至研磨墊30 中的孔，固態視窗可以被緊固於研磨墊30。在一些實施 12 201223703 方式中，固態視窗36被緊固於研磨墊30且固態視窗36 是聚胺基甲酸乙脂視窗。研磨墊30經常被放在平台24 上，使得孔或視窗覆蓋光學頭53，光學頭53被安置在 平台24之頂表面之凹槽26中。因此，光學頭53具有透 過孔或視窗到正在研磨之基板之光的存取。 可用分叉光纜58來將光從光源5 1傳輸到視窗36 ,並 將光從視窗36反射回光偵測器52。分又光纜58可包含 「主幹」58a和兩個「分支」58b和58c。 如上文所提到的，平台24包含凹槽26，在凹槽26 内設有光學頭53。光學頭53緊固於分叉光纖電纜58之 主幹58a的一端’分叉光纖電纜58被配置來將光傳遞到 達正在研磨之基板表面’及分又光纖電纜58被配置來傳 遞來自正在研磨之基板表面之光。光學頭53可包含一個 或更多個透鏡或視窗’該視窗覆蓋分又光纖電缓58之一 端。或者’光學頭53可以只緊固於主幹58a之端，主幹 58a鄰近研磨基板中之視窗。舉例而言，如果有需要的 話，可以自凹槽26移除光學頭53以執行預性防或矯正 性維護。 平台24包含可移除原處監控模組5〇。原處監控模組 50可包含一個或更多個以下元件：光源5卜光源偵測器 52、及用以從光源5 1和光源偵測器52送出及收集到之 訊號之電路系統。舉例而言，偵測器52之輸出可以是數 位電子訊號，該數位電子訊號經過在驅動轴27中的旋轉 耦合器（例如，集電環），至用在光監控系統之控制器9〇。 13 201223703 同樣的，可以根據數位電子訊號中控制命令之反應打開 或關閉光源，數位電子訊號從控制器9〇經過、穿越旋轉 耦合器到達模組5 0。 原處監控模組也可以緊固於分叉光纖電纜58之分支 部分59b及58c的各自端。可操作光源來傳輪光，光被 傳遞經過分支58b，且將光自位於光學頭53之主幹5以 之端離開，並將光照射在正在研磨之基板上面。在位於 光學頭53中之主幹58a之端，接收自基板反射之光，及 將光經由分支58c傳遞至光偵測器52。 在一些實施方式中’分叉光纖電纜58是光纖束。光纖 束包含光纖之第一群組及光纖之第二群組。連接第一群 組中之光纖以傳送光從光源51到正在研磨之基板表 面。連接第二群組中之光纖以接受來自正在研磨之基板 表面反射之光，並傳遞接受到的光到光偵測器。可以布 置光纖，使第二群組中之光纖形成與χ相似的形狀，該 與X相似的形狀被集中在分叉光纖之縱軸上（可視為在 分又光纖電纜58之交又部分或者，可以實施其他布 置。舉例而§，在第二群組中之光纖可以形成與V相似 的形狀，該等與V相似的形狀彼此呈鏡像。一個適合的 分叉光纖可從 Verity Instruments，Inc. of Carrollton, Tex 取得。 在研磨塾30之視窗36與靠近研磨墊3〇之視窗刊之 分叉光纖電纜58之主幹58a之端之間，經常有理想距 離。舉例而§，理想距離可由經驗法則決定且被視窗36 201223703 之反射、從分叉光纖電纜射出之光束之形狀、及光束到 正在監控之基板之距離等所影響。在―些實施方式中， 安置分叉纖維電纜為：使接近視窗36之端，在實際上沒 碰到視窗36的情況下能盡可能的靠近視窗36之底部。 採取此實施方式時’研磨站15〇可以包含機械裝置（例 如，當作部分的光學頭53)，可操作機械裝置以調整在分 叉光纖電纜58之端與研磨墊視窗36之底表面間之距 離。或者，將分叉光纖電纜之接近端嵌入至視窗36之中。 如果覆蓋層是GST，那麼可選擇發射在接近紅外線範 圍之光（例如，單色光（例如，波長約丨3微米之光））的光 源51。或者，可以配置光源51來發射窄帶寬之光（例如， 約1.3微米）。或者，可以配置光源51來發射在接近紅外 線範圍之寬帶寬之光（例如，包含約丨3微米之光），及可 以配置偵測器52來偵測較窄帶寬之光（例如，約13微 米），或偵測器5 2可以是一個光譜儀，可配置光譜儀以 使用來自接近紅外線範圍（例如，約丨3微米）之強度測 量。在一些實施中，光源44發射波長範圍在2到5微米 之光，波長範圍在2到5微米之光適用於GST厚度測量。 在一些實施中，光源44發射波長範圍在1 〇微米内之光， 波長範圍在10微米内之光適用於GST結構相測量。 如果覆蓋層是另一金屬（例如，銅、鎢、鋁、鈦或鈕）， 或障礙金屬（例如，氮化鈦或氮化鈕），那麼可以選擇發 射可見光範圍内之光（例如，紅光）的光源5卜光源51可 以發射單色光（例如，波長介於650到670奈米之光）。 15 201223703 或者，可以配置光源5 i來發射窄帶寬之光(例如，約㈣ 到670不米之光）、或可見光範圍内寬帶寬之光。可配置 偵測器52則貞測跨過—些或全部可見光範圍之反射光 束之全部強度(例如，偵測器可以是簡單的光二極體，該 光二極體可在可見光範圍内操作，或偵測器可被配置來 加總跨過波長頻帶強度之光譜儀），或可配置偵測器52 以偵測實質上單一波長之反射光束強度(例如，偵測器可 以疋光一極體，該光二極體實質上在單一波長操作），或 偵測器52可以是光譜儀，可配置光譜儀以使用來自已偵 測光譜之單-波長強度之測量。偵測$ 52可對紅光(例 如，波長介於650到670奈米之紅光）較敏感。 將光源5 1和光源偵測器52連接到控制器9〇以控制光 源51和光源52之操作，並接收光源5丨及光源52之訊 號。舉例而言，關於控制光源51和光源偵測器52之操 作，控制器90可以藉著平台24的旋轉來同步光源5丨的 啟動。如第3A圖圖示，控制器9〇可使光源51發射連串 閃光，連串閃光只在基板丨〇通過原處監控模組之前開 始、並只在基板10通過原處監控模組之後結束。所圖示 之點701到711中的每個點代表來自原處監控模組之光 照射與反射的一個位置《或者，控制器可以使光源 51持續地發射光，光只在基板丨〇通過原處監控模組之 前開始、並只在基板1 〇通過原處監控模組之後結束。雖 然沒有圖不，每次基板1 〇通過監控模組時，監控模組與 基板10間的校準可以和上次基板10通過原處監控模組 201223703 時一者間的校準不同《在基板1 〇 一次旋轉的期間，可從 來自基板10上不同角度位置，如同在基板1〇上不同的 徑向分散狀位置，來獲得強度測量。換言之，可以從接 近基板10中心的位置獲得一些強度測量，且可以從邊附 近獲得一些強度測量。控制器可將來自基板10之強度測 量分類為群組，該群組對應於同心徑向分散狀區域（例 如’藉由計算強度測量之位置和板中心之間的距離）。徑 向分散狀區域可以對應於承載頭80上不同可控制之區 域。舉例而言，參考第3Β圖’可將強度測量分類為群組， 該群組對應於基板1 〇上同心徑向分散狀區域Ζ1、Ζ2、 Ζ3、Ζ4、及Ζ5。三、四、五、六、七或更多區域可在基 板10的表面上界定。 關於接收訊號’舉例而言，控制器90可以接收具有光 之強度之訊號，該光是由光偵測器52所接收。控制器 90可以處理訊號到暴露下層為止’且使用這資訊來調整 研磨參數（例如，承載頭室之室之壓力）以改善研磨不均 勻。 參考第4圖，針對給定徑向分散狀區域，來自感應器 的一系列掃描之連串測量產生強度跡線800，強度跡線 800是時間或平台旋轉次數之函數。如圖所示，針對GST 層之研磨’從基板10反射之光的強度隨著研磨過程逐漸 成形’從基板10反射之光之強度通過一個或更多個波峰 802及/或波谷804,然後從基板10反射之光之強度在高 原806處變得穩定。第4圖所示強度跡線800只是為了 17 201223703 拿來作說明’及第4圖所示強度跡線800具有許多其他 形狀’在GST研磨期間所產生的強度跡線將具有共同特 徵’此共同特徵為在變動之初期衰減後，強度跡線於高 原806處穩定。 不受限於任何特定理論，當GST層正在研磨時，GST 層的厚度改變。GST層厚度的改變導致來自gst層表面 之反射光和任何下層之間干擾的變化，干擾的變化導致 反射光強度的變化。然而，一旦暴露GST層下面之層， 訊號主要來自下層之反射，且反射訊號強度變得穩定。 藉由偵測強度跡線何時穩定，控制器可以決定清除GST 層及暴露下層之時間。強度跡線穩定之偵測可包含以下 步驟·偵測跡線之斜率在某一臨界值時間周期8丨〇處， 疋否維持在預定範圍内（接近零斜率）。強度跡線穩定偵 測也可包含以下步驟：在臨界值時段，偵測跡線數值是 否維持在一範圍内（該範圍係相對於初始時段之數值被 設定）。 如上上所提到的，來自光學感應器之強度測量可被分 類為不同徑向分散狀區域。分類動作允許針對每一個徑 向分散狀區域產生一個個別的強度跡線。舉例而言，如 第5圖所不，如果強度跡線被分成五個徑向分散狀區 域然後可以產生五個相對應跡線（例如，跡線8〗〇、812、 814、816、及 818) 〇 參考第6圖’針對覆蓋層之研磨，覆蓋層是反射性材 料（例如’金屬（例如’鋼、鶴、銘、欽或组）），從來自感 18 201223703 應器-系列掃描之連串測量產生多個強度跡線，強度跡 線是時間或平台旋轉次數之函數。舉例而言，如果強户 跡線被分成五個徑向分散狀區域，則可以產生五個相二 應跡線（例如，跡線910、912、914、916、及918卜— 般而言’針對每條相對應跡線，相對應跡線之強度在初 始商原922維持相對穩定。然而，當清除覆蓋層及暴露 下層時，強度跡線具有快速下降現象924。一旦完全暴 露下層之頂表面，強度跡線穩定於第二高原9%。藉由 偵測何時強度跡線下降然㈣度料變得敎，㈣器 可以決定在基板之每—個區域中清除反射性覆蓋層及暴 露下層之時間。 參考第7圖將解釋一種研磨的方法。使用具有多個可 控制區域之承載頭及針對每_個區域使用預設麼力來研 磨具有覆蓋層之第一基板（步驟91〇)。舉例而言，如果有 五個區域’則在基板上之個別區域Z卜Z2、Z3、Z4、及 Z5(見第3B圖）施加由承載頭之五個室所造成之壓力 PI、P2、P3、P4、及p5。在研磨期間，使用光學監控系 統來在原處監控覆蓋層。將來自監控系統之強度測量分 類為群組，該群組對應於徑向分散狀區域，並針對每一 個徑向分散狀區域，可基於來自相對應群組之測量來計 算凊除覆蓋層以暴露下層的時間（步驟92〇)。舉例而言， 參考第5圖’針對各自區域Z1、Z2、Z3、Z4和&，會 得到清除時間T1、T2、T3、丁4和T5,可以產生五個強 度跡線810、812、814、816和8 18(見第3Β圖）。舉例而 19 201223703 S，參考第6圖，針對各自區域Z1、Z2、、及201223703 VI. Description of the Invention: [Technical Field of the Invention] The present disclosure relates to the control of polishing using optical monitoring. [Prior Art] An integrated circuit is usually formed on a substrate by successively depositing a conductive layer, a semiconductor layer or an insulating layer on a germanium wafer. A fabrication step involves the steps of depositing a filler layer and a planarization filler layer on a non-planar surface. For some applications, the filler layer is planarized until the top surface of the patterned layer is exposed. For example, a conductive filler layer can be deposited over the patterned insulating layer to fill the trench or hole of the insulating layer. After planarization, portions of the conductive layer remaining between the raised patterns of the insulating layer constitute vias, sockets and wires' vias, sockets and wires provide conductive paths between the thin film circuits on the substrate. For other applications (e.g., oxide milling), the filler layer is planarized until a predetermined thickness is left on the non-planarized surface. In addition, photolithography requires planarization of the substrate surface. Chemical mechanical polishing (CMP) is a well-known method of planarization. The planarization method typically requires mounting the substrate on a load bearing or grinding head. The exposed surface of the substrate is typically placed against a rotating polishing pad. The carrier head provides a controllable load on the substrate to urge the substrate against the polishing pad. A grindable slurry is typically supplied to the surface of the polishing pad. One problem with CMP t is to determine if the grinding process has been completed. It is to determine whether the substrate layer has been flattened to the desired flatness or thickness or to determine when the required amount of material has been removed. Variations in material removal rate can result from variations in slurry distribution, polishing pad conditions, relative speed between the polishing pad and the substrate, and variations in load on the substrate. These changes, like variations in the initial thickness of the substrate, result in a change in the time required to reach the end of the polishing. Therefore, it is often impossible to determine the grinding end point as a function of the grinding time. In some systems, the substrate is optically monitored in-situ during polishing, for example, through a window in the polishing crucible. However, existing optical monitoring technologies may not meet the increasing demands of semiconductor device manufacturers. SUMMARY OF THE INVENTION Due to the above variations, a cover layer from a different area of the substrate can be removed at different times. For some materials, optical monitoring does not reliably detect the thickness of the layer being ground. However, optical monitoring often detects the removal of the overlay. In the process of removing the cover layer from different regions of the substrate (also referred to as the clearing time), an improved uniform technique is to remove the cover layer from different regions of the substrate for detecting the cover layer removal condition of different regions of the substrate. And adjusting at least the polishing pressure on the subsequent second substrate based on the detection of the first substrate removal. In one aspect, during the grinding, a method of controlling the grinding comprises the steps of: grinding the first substrate, the first substrate having a cover layer on the lower layer 201223703 or the layer structure; and the guiding step of guiding the light beam to Above the first substrate, the light beam is reflected from the first substrate to generate a reflected beam; a step of generating a series of measurements of the intensity of the reflected beam during the grinding process; a classification step of classifying the measurements into groups, each group and multiple substrates Determining a different region of the region; determining a step for each region, determining a time, clearing the overlay based on measurements from the associated group at a time; calculating a step based on applying one of the at least one region during the polishing of the substrate Calculating at least one second adjusted grinding pressure of at least one region, and at least one adjusted grinding pressure to grind the second, the pressure, the time for the at least one region, and the time for the other regions Substrate. Embodiments may include one or more of the following features. The beam can be a non-polar beam. The non-polarizing beam can be laser light. The non-polar beam can be a broadband visible beam. The cover layer can be a GST' beam that can contain an infrared component, and the measured beam intensity can be measured as the intensity of the infrared component of the reflected beam. The cover layer can be a metal (for example, a copper, aluminum, tungsten, button, titanium or cobalt beam can contain a red component' and the measured intensity of the reflected beam can be a measure of the intensity of the red component of the reflected beam. The region can be concentric radial Dispersed regions. Grinding can include the steps of: grinding using a carrier head having a plurality of chambers to apply independently adjustable pressure to a plurality of regions on the substrate. The first chamber of the plurality of chambers can be applied to the first domain while in the plurality of chambers The second chamber can be a second region. The step of determining each region is: determining that the first region of the plurality of regions applies a first pressure to the plurality of regions during the grinding process of the first substrate from the plurality of regions The time of exposure of the lower layer may include the first time of the first region of 201223703 and the second time of determining the second region from the plurality of regions. Based on the first pressure, the first time, and the second time, the first time may be calculated At least one adjusted grinding pressure of the chamber. The second region may be an innermost region or an outermost region. Calculating the adjusted pressure pi may include the following steps Calculate ρι,= ρι*(τι/τ2), where P1 is the first pressure, T1 is the first time and 仞 is the second time. The time for determining the lower layer exposure may include the following (4): determining the stability of the series of measurements Time. The time for determining the stability of the series of measurements may include the steps of: determining the slope of the trace produced by the measurement of the connection, which is maintained within a predetermined range for a predetermined period of time. In other aspects, during the grinding, a control The method of grinding comprises the steps of: grinding a first substrate, the first substrate has a cover layer on the lower layer or the layer structure; monitoring step, monitoring the substrate with the in-situ (in_si^) monitoring system to generate a continuous measurement; Steps, classifying the measurements into groups, each group and a different region of the plurality of regions on the substrate are associated with a 'decision step, for each region, determining the time for the overlay to be cleared based on measurements from the associated group丨Calculating the step, based on the pressure applied to at least one region during the polishing of the substrate, the time for at least one region, and the time for other regions At least one second adjusted grinding pressure of at least one region; and a grinding step 'grinding the second substrate using at least one adjusted grinding pressure. Embodiments may include - (d) more of the following characteristics. The system can include a & monitoring system, the optical monitoring system directs the beam onto the substrate. The in situ monitoring system can include a friction sensor. For each 201223703 area, determining the lower exposure time can include the following steps: The first time from the first domain of the plurality of regions and the second time from the first region of the plurality of regions. Based on the first pressure, the first time, and the second time, the + f is determined The adjusted grinding pressure. Calculate the adjusted pressure ρι, ! The knife consists of the following steps: Calculate P1, = Ρ1*(Τ1/Τ2), where ρι is — τ 疋 Leli, Τ1 is the first time and Τ 2 is the second time. In another aspect, instructions are stored on the computer readable medium, and the instructions, when executed by the processor, cause the processor to perform operations. A practical approach may include one or more of the following potential advantages. In-wafer non-uniformity can be reduced (Within_wafer η.—(6)mhy WIWNU). The removal of the cap layer (eg, the GST layer or the metal layer) can be substantially simultaneous: on the surface of the substrate, which can improve the grinding yield. • Adjust the grinding process to be life-consuming (eg, grinding head, grinding 塾) Or the process deposits of the slurry) are compensated. It can also improve the change of the shape of the different research carrier heads. One or more detailed embodiments are set forth in the drawings and the description below. Other aspects, features, and advantages of the present invention can be understood from the scope of the description, drawings, and claims. [Embodiment] In some semiconductor wafer manufacturing processes, a cover layer (for example, metal (for example, copper, tungsten, aluminum, titanium) , bismuth or cobalt), metalloid or metal and/or semi-201223703 metal alloy (for example, GeSbTe (ternary compound of lanthanum, cerium and lanthanum, also known as GST))) is deposited on the patterned lower layer or layer structure (for example, a stack of one or more other layers). The one or more other layers may comprise a layer of dielectric material (eg, a material of low dielectric constant and/or a cap material of low dielectric constant), or a layer of barrier metal (eg, nitrided or titanium nitride) The cover layer is often ground until the cover layer is removed, that is, until the top surface of the lower layer or layer structure is exposed. The portion of the cover layer may remain in the grooves, holes, etc., and the grooves or holes are formed by the underlying layer or the layered pattern. In general, it is necessary to completely remove the cover layer (for example, 'the cover layer has no discontinuous area, the cover layer covers the lower surface or the top surface of the layer structure), and basically needs to cross the surface of the substrate at the same time. Completely remove the cover layer. Over-grinding can be avoided, yield and inter-wafer non-uniformity (WIWNU) can be avoided. One problem with overlay layer monitoring is: for - also materials (eg, high reflectivity metals) In terms of optical (eg, spectral) monitoring of the substrate for bulk thickness during bulk polishing, may not provide useful information. The wavelength is usually used for optical monitoring, and the extinction coefficient (extincti(10) coefficient) of the cover material may be high enough so that the reflectance does not change significantly when the thickness is reduced during a large amount of grinding. Therefore, optical monitoring is performed during a large amount of grinding of some materials. It may not be suitable for the original feedback control of the grinding parameters. Although the reflectivity of the cover layer can be changed when the cover layer is removed and the lower layer is exposed, the grinding of the substrate being ground is detected when such a change in optical behavior is detected. The original adjustment of the rate may be too late so that the 201223703 method does not improve the unevenness of the removal time across the substrate. In addition, the detection of the second eddy current monitoring in these same materials is not efficient enough to detect the removal of the cover layer. In terms of barrier metals (such as 'titanium nitride and tantalum nitride), because of their high resistivity in nature, barrier metals cannot extract good current signals. For other metals (for example, copper 'ming and crane') , the eddy current face can not be left in the discontinuous metal film on the substrate to obtain the signal, secret and retanning layer The detection of the removal is therefore unreliable. However, for many processes, the frost overlay has a different reflectivity than the underlying or layer structure. It is not limited to any particular theory, perhaps due to: overlay and The lower layer has a different (eg, higher) extinction coefficient. In these cases, when the cover layer is removed and the lower layer is exposed, the reflectance or reflectance spectrum from the substrate should change. Improving the condition of the non-uniform sentence of the clearing cover layer, the technology is: optically detecting the clearing of the cover layer in a plurality of different areas of the first substrate' and based on the detection detected on the first cup of the shop, such as the soil plate Time, adjusting the pressure on at least one of the subsequent & second regions of the second substrate such that the adjustment has a situation that is closer to the same time than if the adjustment is not performed (eg, Base, at the same time) Yuan Cheng (occur) clear. Although this technology can win a field, and $ π is especially used to deal with the above problems, even if optical monitoring can provide useful information for the thickness of the overlay during a large amount of grinding, this technology can still have a temple-like application. . In this case, there are other advantages to this technique, such as + π β u , , burning, control of process consistency between different materials, or reduction of computational load. "Field & Μ also, main ^ (because The detection clearing ratio is less than the thickness of 10 201223703. Therefore, this technique can be used for other translucent materials (eg, translucent metals (eg, GST)).廿 ) In particular, this technique is about whether the behavior of the reflection trajectory changes significantly when the layer is removed. Fig. 1 is a schematic cross-sectional view of a chemical mechanical polishing W station 1 station 150, which can be operated by a chemical mechanical polishing station. The polishing station 150 includes a rotatable disc-shaped platform 24,/-!. A polishing pad 30 is mounted on the rotatable disc-shaped platform 24. It can be operated by a flat ten to rotate around the axis 25. For example, a motor (not shown) can rotate the drive wheel 27 to rotate the pedal. For example, the polishing pad 30 can be separately and securely fastened to the flat mouth by the adhesive layer by the adhesive layer. The polishing pad 30 can be separated and replaced when the polishing pad 3 is worn. The polishing pad 30 can be a dual layer polishing pad having an outer abrasive layer 32 and a softer support layer 34. The grinding station 150 is configured to include a combined slurry/water wash arm (Hnse is 39. During the grinding process, the arm 39 can be operated to disperse the slurry 38 (eg, a liquid having abrasive particles). Alternatively, the polishing station 22 includes The abrasive material is operable to grind the abrasive material on the polishing crucible 30. The grinding station 150 also includes a carrier head 80 that is operable to secure the substrate 1 to the polishing pad 30. The carrier head 8 is suspended from the support fiber. Constructed (eg, revolving rack 154)' and the carrier head 8 is coupled to the carrier head rotation motor % by load bearing 74 to enable the carrier head (9) to rotate about the shaft 7 此外. In addition, the carrier head 8 () ^ The radial groove 1M formed in the support structure is swung ((10) wrist e). In operation, the flat 24 rotates around the flat (four) central axis 25 and the receiving head 8 wraps around the middle of the carrier head 201223703 ~ axis 71 rotates and moves laterally across the top surface of the polishing pad 3>> Figure 2, in order to apply independently controllable pressure to multiple regions on the substrate (e.g., concentric regions), the carrier head 8 can contain Multiple chambers. In one implementation, the carrier head 80 includes a housing 30 2. Base assembly 3〇4, ring frame structure 306 (the ring frame structure can be regarded as a part of the base assembly 304), the load chamber 308, the fixing ring 310, and the substrate support assembly 32〇, the total substrate support The 320 includes an elastic film 326 'The elastic film 320 defines a plurality of independently pressurizable chambers (eg, the inner chamber 330, the intermediate chamber 332, the intermediate chamber 334, the intermediate chamber 336, and the outer chamber 338). These pressurized chambers control the elastic membrane The pressure on the concentric regions' thus these pressurized chambers can provide independent pressure control on the concentric portion of the substrate 10. In some implementations, the carrier head 8 includes five chambers and has a pressure regulator in each chamber. Referring to FIG. 3B, the five chambers 330, 332, 334, 336, and 338 can control the pressures of the five concentric regions Zi, Z2, Z3, Z4, and Z5 applied to the substrate 1A. Referring to FIG. 1, grinding Station 150 also includes an optical monitoring system that can be used to determine the polishing endpoint as discussed below. The optical monitoring system includes a light source 51 and a light detector 52. Light passes from source 51 through the polishing pad 30' to illuminate the substrate. 1 〇 and reflected from the back end of the substrate Over-grinding pad' and advancing to photodetector 52 » provides access to light passing through polishing pad 30 by including a hole (ie, a hole through the pad) or solid-state window 36. Although in some implementations The solid state window 36 can be supported on the platform 24 and projected into the aperture in the polishing pad 30, and the solid state window can be secured to the polishing pad 30. In some implementations 12 201223703, the solid state window 36 is secured to the polishing pad 30 And the solid state window 36 is a polyethylene glycol window. The polishing pad 30 is often placed on the platform 24 such that the aperture or window covers the optical head 53, and the optical head 53 is placed in the recess 26 of the top surface of the platform 24. Therefore, the optical head 53 has access to light passing through the hole or window to the substrate being polished. The split cable 58 can be used to transmit light from the source 51 to the window 36 and reflect light from the window 36 back to the photodetector 52. The split cable 58 can include a "trunk" 58a and two "branches" 58b and 58c. As mentioned above, the platform 24 includes a recess 26 in which an optical head 53 is disposed. The optical head 53 is fastened to one end of the trunk 58a of the bifurcated fiber optic cable 58 'the bifurcated fiber optic cable 58 is configured to transmit light to the surface of the substrate being ground' and the fiber optic cable 58 is configured to pass the substrate from the substrate being ground The light of the surface. The optical head 53 can include one or more lenses or windows that cover one end of the fiber optic cable 58. Alternatively, the optical head 53 may be fastened only to the end of the stem 58a adjacent to the window in the polishing substrate. For example, the optical head 53 can be removed from the recess 26 to perform pre-emptive or corrective maintenance, if desired. The platform 24 includes a removable in situ monitoring module 5〇. The original monitoring module 50 can include one or more of the following components: a light source 5, a light source detector 52, and circuitry for transmitting and collecting signals from the light source 51 and the light source detector 52. For example, the output of the detector 52 can be a digital electronic signal that passes through a rotary coupler (e.g., a slip ring) in the drive shaft 27 to a controller 9 of the optical monitoring system. 13 201223703 Similarly, the light source can be turned on or off according to the reaction of the control commands in the digital electronic signal, and the digital electronic signal passes from the controller 9 through the rotary coupler to the module 50. The original monitoring module can also be fastened to the respective ends of the branch portions 59b and 58c of the bifurcated fiber optic cable 58. The light source is operable to transmit the light, the light is transmitted through the branch 58b, and the light exits from the end of the main body 5 of the optical head 53, and the light is irradiated onto the substrate being polished. At the end of the stem 58a in the optical head 53, the light reflected from the substrate is received, and the light is transmitted to the photodetector 52 via the branch 58c. In some embodiments the 'forked fiber optic cable 58 is a bundle of fibers. The fiber bundle includes a first group of fibers and a second group of fibers. The fibers in the first group are connected to transmit light from the source 51 to the surface of the substrate being polished. The fibers in the second group are connected to receive light reflected from the surface of the substrate being polished and the received light is delivered to the photodetector. The fibers can be arranged such that the fibers in the second group form a shape similar to that of the crucible, the shape similar to X being concentrated on the longitudinal axis of the bifurcated fiber (as may be seen as part of the intersection of the fiber optic cable 58 or Other arrangements may be implemented. By way of example, the fibers in the second group may form a shape similar to V, which are mirror images of each other. A suitable bifurcated fiber is available from Verity Instruments, Inc. of Carrollton, Tex. Between the window 36 of the abrasive 塾 30 and the end of the trunk 58a of the bifurcated fiber optic cable 58 near the window of the polishing pad 3, there is often an ideal distance. For example, §, the ideal distance can be determined by empirical rules. And is affected by the reflection of the window 36 201223703, the shape of the beam emitted from the bifurcated fiber optic cable, and the distance of the beam from the substrate being monitored. In some embodiments, the bifurcated fiber cable is placed to: approach the window 36 At the end, it can be as close as possible to the bottom of the window 36 without actually touching the window 36. When this embodiment is adopted, the 'grinding station 15' can contain mechanical devices ( For example, as part of the optical head 53), a mechanical device can be operated to adjust the distance between the end of the bifurcated fiber optic cable 58 and the bottom surface of the polishing pad window 36. Alternatively, the proximal end of the bifurcated fiber optic cable can be embedded into the window. 36. If the cover layer is GST, then a light source 51 that emits light in the near infrared range (eg, monochromatic light (eg, light having a wavelength of about 3 microns)) may be selected. Alternatively, the light source 51 may be configured to emit A narrow bandwidth of light (e.g., about 1.3 microns). Alternatively, light source 51 can be configured to emit light in a wide bandwidth near the infrared range (e.g., containing about 3 microns of light), and detector 52 can be configured to detect Measure a narrower bandwidth light (eg, about 13 microns), or the detector 52 can be a spectrometer that can be configured to use intensity measurements from a near infrared range (eg, about 3 microns). In some implementations. Light source 44 emits light having a wavelength in the range of 2 to 5 microns, and light having a wavelength in the range of 2 to 5 microns is suitable for GST thickness measurement. In some implementations, source 44 emits at a wavelength in the range of 1 〇 micron. Light, light with a wavelength in the range of 10 microns is suitable for GST structural phase measurements. If the cover layer is another metal (for example, copper, tungsten, aluminum, titanium or button), or a barrier metal (for example, titanium nitride or nitrogen) The light source 51 can emit a monochromatic light (for example, light having a wavelength between 650 and 670 nm). 15 201223703 Alternatively, it is possible to select a light source 5 that emits light in the visible range (for example, red light). The light source 5 i is configured to emit light of a narrow bandwidth (for example, about (4) to 670 meters) or a wide bandwidth of light in the visible range. The configurable detector 52 detects reflections across some or all of the visible range. The full intensity of the beam (for example, the detector can be a simple photodiode, the photodiode can operate in the visible range, or the detector can be configured to add a spectrometer across the intensity of the wavelength band), or The detector 52 is configured to detect the intensity of the reflected beam at substantially a single wavelength (eg, the detector can be dimmed with a polar body, the photodiode operates substantially at a single wavelength), or the detector 52 can be a spectrometer. Counter spectrometer has to detect from the use of a single spectral measurement - measured wavelength intensity. A detection of $52 can be sensitive to red light (for example, red light with a wavelength between 650 and 670 nm). The light source 51 and the light source detector 52 are connected to the controller 9 to control the operation of the light source 51 and the light source 52, and receive signals from the light source 5 and the light source 52. For example, with respect to controlling the operation of light source 51 and light source detector 52, controller 90 can synchronize the activation of light source 5 by the rotation of platform 24. As shown in FIG. 3A, the controller 9 can cause the light source 51 to emit a series of flashes, and the series of flashes are started only before the substrate passes through the original monitoring module, and only after the substrate 10 passes through the original monitoring module. . Each of the points 701 to 711 shown represents a position of light illumination and reflection from the original monitoring module. Alternatively, the controller can cause the light source 51 to continuously emit light, and the light passes through the substrate only. The monitoring module starts before and ends only after the substrate 1 passes through the original monitoring module. Although there is no picture, each time the substrate 1 passes through the monitoring module, the calibration between the monitoring module and the substrate 10 can be different from the calibration of the previous substrate 10 through the original monitoring module 201223703 "on the substrate 1" During one rotation, intensity measurements can be taken from different angular positions from the substrate 10, as in different radially dispersed positions on the substrate 1〇. In other words, some intensity measurements can be taken from a position near the center of the substrate 10, and some intensity measurements can be obtained from the vicinity. The controller can classify the intensity measurements from the substrate 10 into groups corresponding to concentric radially dispersed regions (e.g., 'the distance between the position measured by the intensity and the center of the plate). The radially dispersed regions may correspond to different controllable regions on the carrier head 80. For example, the intensity measurements can be classified into groups according to FIG. 3', which corresponds to the concentric radially dispersed regions Ζ1, Ζ2, Ζ3, Ζ4, and Ζ5 on the substrate 1. Three, four, five, six, seven or more regions may be defined on the surface of the substrate 10. Regarding the received signal', for example, the controller 90 can receive a signal having the intensity of light that is received by the photodetector 52. The controller 90 can process the signal until the lower layer is exposed' and use this information to adjust the grinding parameters (e.g., the pressure of the chamber carrying the head chamber) to improve uneven grinding. Referring to Figure 4, for a given radially dispersed region, a series of measurements from a series of scans of the sensor produces intensity traces 800 which are a function of time or number of revolutions of the platform. As shown, the grinding of the GST layer 'the intensity of the light reflected from the substrate 10 is gradually shaped as the grinding process is formed'. The intensity of the light reflected from the substrate 10 passes through one or more peaks 802 and/or troughs 804, and then from The intensity of the light reflected by the substrate 10 becomes stable at the plateau 806. The intensity trace 800 shown in Figure 4 is only for the purpose of the description of the 2012 2012 703 and the intensity trace 800 shown in Figure 4 has many other shapes. The intensity traces produced during GST grinding will have the same characteristic 'this common The characteristic is that after the initial decay of the change, the intensity trace is stable at the plateau 806. Without being bound by any particular theory, the thickness of the GST layer changes as the GST layer is being ground. The change in the thickness of the GST layer causes a change in the interference between the reflected light from the surface of the gst layer and any underlying layer, and the change in the interference causes a change in the intensity of the reflected light. However, once the layer under the GST layer is exposed, the signal is mainly from the reflection of the lower layer, and the intensity of the reflected signal becomes stable. By detecting when the intensity trace is stable, the controller can determine when to clear the GST layer and expose the underlying layer. The detection of the intensity trace stability may include the following steps: detecting the slope of the trace at a critical time period of 8 ,, whether or not it is maintained within a predetermined range (near zero slope). The intensity trace stabilization detection may also include the step of detecting whether the trace value is maintained within a range (the range is set relative to the initial time period) during the threshold period. As mentioned above, the intensity measurements from the optical sensor can be classified into different radially dispersed regions. The categorization action allows for the creation of an individual intensity trace for each radial dispersed region. For example, as shown in FIG. 5, if the intensity trace is divided into five radially dispersed regions, then five corresponding traces can be generated (eg, traces 8 〇, 812, 814, 816, and 818). 〇 Refer to Figure 6 for grinding of the cover layer, the cover layer is a reflective material (such as 'metal (such as 'steel, crane, Ming, Chin or group)), from the sense 18 201223703 reactor - series of scans The string measurement produces multiple intensity traces, which are a function of time or number of platform rotations. For example, if the strong household trace is divided into five radially dispersed regions, then five phase two traces can be generated (eg, traces 910, 912, 914, 916, and 918 - generally speaking' For each corresponding trace, the intensity of the corresponding trace remains relatively constant at the initial quotient 922. However, when the overlay is removed and the underlayer is exposed, the intensity trace has a rapid drop 924. Once the top surface of the lower layer is completely exposed The intensity trace is stable at 9% of the second plateau. By detecting when the intensity trace drops (4) the material becomes helium, the (four) device can determine to remove the reflective cover layer and expose the lower layer in each of the regions of the substrate. Time. A method of grinding will be explained with reference to Fig. 7. A carrier having a plurality of controllable regions is used and a first substrate having a cover layer is ground using a predetermined force for each of the regions (step 91). In the case of five zones, the pressures PI, P2, P3, P4 caused by the five chambers of the carrier head are applied to individual zones Z, Z2, Z3, Z4, and Z5 on the substrate (see Figure 3B). And p5. During grinding, The optical monitoring system is used to monitor the overlay in situ. The intensity measurements from the monitoring system are classified into groups, which correspond to radially dispersed regions, and for each radially dispersed region, based on corresponding segments The set of measurements is used to calculate the time to remove the overlay to expose the underlying layer (step 92A). For example, referring to Figure 5, for the respective regions Z1, Z2, Z3, Z4, and &, the clearing times T1, T2 are obtained. , T3, D4, and T5, can produce five intensity traces 810, 812, 814, 816, and 8 18 (see Figure 3). For example, 19 201223703 S, refer to Figure 6, for the respective regions Z1, Z2 ,and
及T 5，可以產 及918(見第3B Z5，會得到清除時間 生五個強度跡線91〇 圖）。 ΤΙ ' T2、T3、T4、 、912 、 914 、 916 、 回到第7圖，接著’針對承載頭之可控制區域的至少 一者計算調整後的研磨壓力（步驟930)。可以在一簡單 Prestonian模型下計算壓力以調整研磨壓力，使得幾^在 同一時間清除每一個區域。可以選擇一個區域（例如，最 内側區域Z5)來當作參考區域。針對每個其他區域，藉 由預設壓力與區域之研磨時間速率相乘來計算調整後之 壓力，該區域被調整至參考區域之研磨時間。舉例而言， 區域Zl、Z2、Z3、Z4之調整後的壓力Ρ1ι、 Ρ2·、 P3,、 P4’ 可被計算為 P1，=P1*(T1/T5)，Ρ2·=Ρ2*(Τ2/Τ5)， Ρ3’=Ρ3*(Τ3/Τ5)和 Ρ4’=Ρ4*(Τ4/Τ5)。或者，可選擇最外側 區域、或中間區域中之一個區域當作參考區域。然後用 調整後之研磨壓力來研磨後續基板（步驟940)。 可用光學監控系統來監控後續基板之研磨；針對各自 區域Zl、Ζ2、Ζ3、Ζ4、及Ζ5，可決定一組新的清除時 間ΤΙ、Τ2、Τ3、Τ4、及Τ5(步驟910);且可由先前經計 算調整後之壓力來計算一組新的調整後之壓力，該新的 調整後之壓力被使用來當作新的預設壓力（步驟920);及 使用調整後之壓力之新組研磨其他基板。更一般性地 說，光學監控系統可以執行疊代回饋方法，在疊代回馈 方法中’針對每一接著要研磨之基板，使用前個基板之 20 201223703 清除時間與屢力以 < 曾 力以计异接著要研磨之基板之調整後之壓 b外可基於針對多個先前基板之清除時間及/或壓 權移動平均，而非只基於前一個基板之清除時間 u力來計算調整後之壓力，來計算調整後之壓力。 在-些實施方式中，在預訂量的時間内，在第一平台 上研磨具有覆蓋層之基板以達到大量清除，然後利用； 7圖描述之技術在第二平台上研磨。如果覆蓋層是導電 材料’則在第一平台上，可用渦電流監控系統來監控覆 蓋層之厚度。換言之，在第一平台上研磨基板；使用渦 電流監控系統在原處監控導電層之厚度；將來自第一監 控系統之厚度測置分類為群體，群體對應至徑向分散狀 區域；及控制器針對覆蓋層之每一個徑向分散狀區域計 算預估時間以達到目標厚度。計算—個或更多個調整後 之壓力，使基板之徑向分散狀區域將比在沒有調整壓力 之it况下，在更接近相同時間達到目標厚度，並施加調 整後之壓力以達到第一平台上之基板的完全研磨。然後 在第二平台上研磨基板；使用光學監控系統在原處監控 覆蓋層之厚度；將來自第二監控系統之強度測量分類為 群組，該群組對應於徑向分散狀區域；控制器偵測強度 跡線到達高原之時間，強度跡線到達高原表示覆蓋層已 被清除，·且計算一個或更多個調整後之壓力，該一個或 更多個調整後之壓力被用以在第二平台上研磨後續基 板。 雖然上述之討論集中在使用光學技術來偵測清除，光 21 201223703 學技術過程可應用在其他原處監控技術，原處監控技術 可债測覆蓋層之清除狀況（例如，局部表面磨擦力感測 (例如’如美國專利號7，153,818所述，美國專利號 7，153,818已併入此案以供參考或應用在渦電流監控 (假設用來偵測清除之渦電流監控技術之精確度最終已 獲得改善）。 在本說明書中之實施及描述的所有功能性操作可實施 在數位電子電路中或電腦軟體、韌體或硬體中，電腦軟 體、韌體或硬體包括^於本說明#巾之結構構件及結 構均等物，或實施及所有功能性操作可實施在數位電子 電路與電腦軟體、韌體或硬體的組合中。可實作本說明 書所描述之實施方式為一個或更多個電腦程式產品，亦 即’具體地實施-個或更多個電腦程式於資訊載體（例 如’在機器可讀儲存裝置或可達成傳播訊號之物質中）， 電腦程式可用資料處理設備來執行或控制資料處理設備 4如T程式化處理器、電腦、或多個可程式化處理器 或多個電腦）之操作。 β 1腦程式（亦稱為程式、軟體、軟體應用或程式； i何形式之程式語言編寫，任何形式之程式語 含編譯後或解譯後 佑1 曼之扣5具電腦程式可用任何形 佈署，該任何形式的佈署包 紐、匕3作為獨立程式或作< 分· < 使用在計算環境中之其1 70。電腦程式不一定晷槲庙 疋要對應於檔案。雷聪兹彳可， 於保留其他程千電腌程式可Ui 、他程式或資料之檔案的—部分，專用於所i 22 201223703 式之單個或多個同一類型的檔案（例如儲存—個或更多 個模組、副程式、或程式碼的部分）中。—個電腦程气可 經佈署在一個站點或分散在多個站點處且由通訊網路互 連的一台或多台電腦上執行。 本說明書中描述之製程及邏輯流程可由執行一個戋更 多個電腦程式之一或更多可程式處理器來執行，以藉由 對輸入資料操作及產生輸出來執行功能。製程及邏輯流 程亦可由專用邏輯電路來執行，且設備亦可實施為專用And T 5, can produce 918 (see 3B Z5, will get the clearing time to produce five intensity traces 91 〇). ΤΙ ' T2, T3, T4, 912, 914, 916, returning to Figure 7, and then adjusting the adjusted grinding pressure for at least one of the controllable regions of the carrier head (step 930). The pressure can be calculated in a simple Prestonian model to adjust the grinding pressure so that each area is cleared at the same time. An area (for example, the innermost area Z5) can be selected as the reference area. For each of the other zones, the adjusted pressure is calculated by multiplying the preset pressure by the grinding time rate of the zone, which is adjusted to the grinding time of the reference zone. For example, the adjusted pressures Ρ1, Ρ2·, P3, and P4' of the zones Z1, Z2, Z3, Z4 can be calculated as P1, = P1*(T1/T5), Ρ2·=Ρ2*(Τ2/ Τ5), Ρ3'=Ρ3*(Τ3/Τ5) and Ρ4'=Ρ4*(Τ4/Τ5). Alternatively, one of the outermost area or the middle area may be selected as the reference area. The subsequent substrate is then abraded with the adjusted polishing pressure (step 940). An optical monitoring system can be used to monitor the grinding of subsequent substrates; for each of the regions Z1, Ζ2, Ζ3, Ζ4, and Ζ5, a new set of clearing times ΤΙ, Τ2, Τ3, Τ4, and Τ5 can be determined (step 910); A new set of adjusted pressures is previously calculated by calculating the adjusted pressure, the new adjusted pressure being used as a new preset pressure (step 920); and a new set of grinding using the adjusted pressure Other substrates. More generally, the optical monitoring system can perform an iterative feedback method. In the iterative feedback method, 'for each substrate to be ground, use the previous substrate 20 201223703 to clear the time and the force to < The adjusted pressure b of the substrate to be ground may be based on the removal time and/or the weighted moving average for the plurality of previous substrates, rather than calculating the adjusted pressure based solely on the removal time u force of the previous substrate. To calculate the adjusted pressure. In some embodiments, the substrate having the overlay is ground on the first platform for a predetermined amount of time to achieve substantial removal, and then ground on the second platform using the technique described in Figure 7. If the cover layer is a conductive material, then on the first platform, an eddy current monitoring system can be used to monitor the thickness of the cover layer. In other words, the substrate is ground on the first platform; the thickness of the conductive layer is monitored in situ using an eddy current monitoring system; the thickness measurement from the first monitoring system is classified into a population, the population corresponds to a radially dispersed region; and the controller is directed to The estimated time is calculated for each radially dispersed region of the overlay to achieve the target thickness. Calculating one or more adjusted pressures such that the radially dispersed region of the substrate will reach the target thickness at a more nearly the same time than without adjusting the pressure, and the adjusted pressure is applied to reach the first Complete grinding of the substrate on the platform. And then grinding the substrate on the second platform; monitoring the thickness of the overlay layer in situ using an optical monitoring system; classifying the intensity measurements from the second monitoring system into groups, the group corresponding to the radially dispersed regions; When the intensity trace reaches the plateau, the intensity trace reaches the plateau indicating that the overlay has been removed, and one or more adjusted pressures are calculated, the one or more adjusted pressures being used on the second platform The subsequent substrate is ground. Although the above discussion focuses on the use of optical techniques to detect cleanup, the technical process can be applied to other in situ monitoring techniques, and the original monitoring technology can detect the removal of the overlay (eg, local surface friction sensing). (For example, as described in U.S. Patent No. 7,153,818, U.S. Patent No. 7,153,818, which is hereby incorporated hereinby incorporated herein by reference in its entirety in its entirety herein in Improved.) All functional operations implemented and described in this specification can be implemented in digital electronic circuits or in computer software, firmware or hardware, including computer software, firmware or hardware. Structural members and structural equivalents, or implementations and all functional operations may be implemented in a combination of digital electronic circuitry and computer software, firmware or hardware. The embodiments described herein may be implemented as one or more Computer program product, ie 'specifically implemented - one or more computer programs on information carriers (eg 'in machine readable storage devices or achievable communication The computer program can be used to execute or control the operation of a data processing device such as a T-programmed processor, a computer, or a plurality of programmable processors or a plurality of computers. Also known as a program, software, software application or program; i written in any form of programming language, any form of programming containing compiled or interpreted after the 1 man's deduction 5 computer programs can be deployed in any form, any form The layout of the package, 匕3 as a stand-alone program or as a < points · < used in the computing environment of its 1 70. Computer programs are not necessarily 晷槲 temple 疋 to correspond to the file. Lei Cong 彳 可, reserved The other parts of the file, Ui, the program or the file of the data, are dedicated to the single or multiple files of the same type (such as storage - one or more modules, sub-programs, etc.) Or part of the code. - A computer program can be deployed on one or more computers that are deployed at one site or across multiple sites and interconnected by a communications network. Process and logic The process may be performed by executing one or more computer programs or one of the programmable processors to perform functions by operating on the input data and generating output. The process and logic flow may also be performed by dedicated logic circuits, and the device Can also be implemented as dedicated
Programmable 積體電路 邏輯電路（例如’現場可程式閘陣列（fieid gate array; FPGA)或特殊應用 (application-specific integrated circuit; ASIC)) 〇 上述研磨設備及方法可應用於各種研磨系統中。研磨 塾或承載頭或研磨墊及承載頭兩者皆可移動以提供在研 磨表面與晶圓間的相對運動。舉例而言，平台可繞執道 運行而非旋轉。研磨墊可為緊固至平台之圓形（或某二 其他形狀）墊。終點谓測系統之一些態樣可適用於線性 研磨系統（例如，在研磨糸絲由，m α 研磨糸，，充中，研磨墊為線性移動之連 續帶或捲盤帶）。研磨層可 π微層j马‘準（例如，具有或不具有 填料之聚氛基甲酸醋）研磨材料、軟材料或固定研磨材 料使用相對定位之術語；相對定位應理解為可將研磨 表面及晶圓固持在垂直方向或-些其他方向上。 術語「資料處理設備」包含所有設備、裝置和處理資 料的機器’舉例來說，該所右讯 換％ 兄顯有叹備、裝置和處理資料的 匕3可程式處理器、電腦、或多個處理器或多個電 23 201223703 腦。除了硬體外’資料處理設備亦可包含創造在所述電 腦程式上之執行環境的程式碼（例如，組成處理器勒體之 程式碼、協定堆疊、資料庫管理系統、操作系統或上述 組成處理器㈣之程式碼、協^堆疊、資料管理系統、 操作系統之一個或更多個之組合。 舉例而言，適用於電腦程式執行之處理器包含一般用 途微處理器與特殊用途微處理器兩者，及任何種類數位 電腦之任何一個或更多個處理器。一般而言，處理器將 從唯讀記憶體或隨機存取記憶體或唯讀記憶體及隨機存 取記憶體兩者接收指令和資料。電腦中主要的元件是用 來執行指令的處理器及一個或更多個用來儲存指令及資 料的記憶體裝置…般而言，電腦將也包含—個或更多 個用以儲存資料之大量儲存裝置；或電腦將也可操作地 耦合以從-個或更多個用以儲存資料之大量儲存裝置接 收資料；或電腦將也可操作地耦合以傳輸資料至一個或 更多個儲存資料之大量儲存裝置；或電腦將也可操作地 輕合以接收資料從/及傳輸資料至一個或更多個儲存資 料之大I儲存裝置（例如，磁碟、磁光碟或光碟）。然而， 電腦不需要擁有這些裝置。 人適用於儲存電腦程式指令及資料之電腦可讀媒體包 含所有形式之非揮發性記憶冑、媒體及記憶體裝置，舉 ’來說非揮發性§己憶體、媒體及記憶體裝置包含半導 體记憶裝置（例如’可抹拭唯讀記憶體(EPR〇M)，電子可抹 讀記憶體（EEPR0M)及快閃記憶體裝置）、磁碟（例如/内 24 201223703 P硬碟或可移除式硬碟）、磁光碟及唯獨型光碟（CD 和僅可讀DVD $ (DVD_R〇M)。處理器和記憶體可由特 殊用途邏輯電路來補充或由特殊用途邏輯電路來組合。 為了提供與使用者的互動，本說明書中所敘述技術主 題之貫施例可在具有顯示裝置（例如，陰極射線管顯示器 (eath〇de ray tube ’ CRT)或液晶顯示器（liquid crystal display ’ LCD))之電腦上實施以顯示資訊給使用者，且 電腦包含鍵盤和指標裝置（例如，滑鼠或執跡球），藉由 滑鼠或轨跡球，使用者可以提供輸入給電腦。其他種類 的裝置同樣可用來提供與使用者之互動，例如，使用者 提供的反饋可以是任何形式的感測回饋（例如，視覺回 饋、聽覺回饋或觸覺回饋）；及可接收來自使用者之包含 聽覺的、口語的或觸覺的任何形式之輸入。 本說明書包含許多實施，然而這些實施例不是用來限 制作任何發明範圍或限縮申請專利範圍，而是作為技術 特徵的描述，技術特徵可用來具體化特定發明之特定實 施例。本說明書中所描述在不同實施例上下文中之某些 技術特徵也可組合以實施在單一實施例。 相反的’在單一實施例上下文所描述之不同技術特徵 也可在多個不同實施例中個別實施或在任何適當的實施 例次組合中實施。此外，雖然技術特徵可如同上面所描 述在某些實施例組合中實施，甚至一開始就如技術特徵 所宣稱，但已旦稱組合的一個或更多個技術特徵在一此 情況下可以從組合令去除，及也許可以將已宣稱組合導 25 201223703 引至一次組合或—次組合之變化。 @樣地’當操作在圖示裡以特定順序摇述’這不應被 解讀為刼作必須以圖示所示特定順序或連續的序列來執 打，或不應被解讀為必須執行所有圖示之操#以達到想 要的’果。在某些情況下，多工及平行化處理也許更為 有利。此外，上述實施例中不同系統元件的分離不應被 解讀為在所有的實施例中都需要如此之分_，及上述實 施例中不同系統元件的分離應被解讀為描述之程式元件 和系統通常應被整合在單一軟體產品中或包在多個軟體 產品中。 【圖式說明】 第1圖圖示研磨站之示意性橫截面圖。 第2圖圖示承載頭之示意性橫截面圖。 第3A圖圖示在平台上一基板之俯視圖及發生測量之 位置。 第3B圖圖示基板上之同心區域。 第4圖圖示訊號強度之示意性範例圖，該示意性範例 圖為正在研磨之基板（基板包含GST層）之時間函數。 第5圖圖不訊號強度之示意性範例圖，訊號強度來自 正在研磨之基板的多個區域，該正在研磨之基板包含 GST 層。 第6圖圖示訊號強度之示意性範例圖，訊號強度來自 26 201223703 正在研磨之基板的多個區域，該正在研磨之基板包含金 屬層。 第7圖圖示研磨過程之流程圖。 以相同元件符5虎表示各圖式中的相同元件。 【主要元件符號說明】 10 基板 24 平台 25 中心轴 26 凹槽 27 驅動軸 30 研磨塾 32 研磨層 34 支持層 36 固態視窗 38 研磨漿 39 臂 50 原處(in-situ)監控模組 51 光源 52 光偵測器 53 光學頭 58 分叉光纜 58a 主幹 58b 分支 58c 分支 71 中心軸 74 承載驅動轴 76 承載頭旋轉馬達 80 承载頭 90 控制器 150 化學機械研磨站 154迴轉料架 302 外殼 304 基座總成 306 環架結構 308 負載室 27 201223703 310 固定環 320基板支持總成 326 彈性膜 330内室 332 中間室 334中間室 336 中間室 338外室 701 點 702點 703 點 704點 705 點 706點 707 點 708點 709 點 710點 711 點 800強度跡線 802 波峰 8 04波谷 806 南原 810臨界值 812 跡線 8 1 4跡線 816 跡線 818跡線 910 研磨初始基板及在原處監控研9 12跡線 磨 914 跡線 916跡線 918 跡線 920決定多個區域覆蓋層清除之時 間 922 初始高原 924快速下降 926 第二高原 930決定至少一個區域研磨壓力之 調整 940使用調整後之研磨壓力研磨下T1 時間 28 201223703 Τ2 Τ4 Ζ1 Ζ3 Ζ5 個基板 時間 Τ3 時間 時間 Τ5 時間 同心區域 Ζ2 同心區域 同心區域 Ζ4 同心區域 同心區域 29Programmable Logic circuits (such as 'fieid gate arrays (FPGAs) or application-specific integrated circuits (ASICs)) 〇 The above grinding equipment and methods can be applied to various grinding systems. Both the abrasive crucible or the carrier head or the polishing pad and the carrier head are movable to provide relative motion between the polishing surface and the wafer. For example, the platform can run around the road instead of spinning. The polishing pad can be a circular (or some other shape) pad that is fastened to the platform. Some aspects of the end point prediction system can be applied to linear grinding systems (for example, in the case of grinding crepe, m α grinding 糸, filling, the polishing pad is a linear moving continuous belt or reel belt). The abrasive layer may be π microlayers (for example, aramid vinegar with or without filler) abrasive materials, soft materials or fixed abrasive materials using relative positioning terms; relative positioning is understood to mean the abrasive surface and The wafer is held in the vertical direction or in some other direction. The term "data processing equipment" encompasses all equipment, devices, and machines that process data. For example, the right-hander has a sigh, device, and processing data, a computer, or multiple Processor or multiple electric 23 201223703 brain. In addition to the hard-body's data processing device, the code may be included to create an execution environment on the computer program (eg, a program code that constitutes a processor, a protocol stack, a database management system, an operating system, or a component processor as described above). (4) A combination of one or more of a program code, a stacking, a data management system, and an operating system. For example, a processor suitable for computer program execution includes both a general purpose microprocessor and a special purpose microprocessor. And any one or more processors of any kind of digital computer. Generally, the processor will receive instructions and from both read-only memory or random access memory or both read-only memory and random access memory. The main components of a computer are the processor used to execute the instructions and one or more memory devices used to store the instructions and data. In general, the computer will also contain one or more storage data. a plurality of storage devices; or the computer will also be operatively coupled to receive data from one or more of a plurality of storage devices for storing data; or a computer A plurality of storage devices that are also operatively coupled to transmit data to one or more stored materials; or the computer will also be operatively coupled to receive data from/and transmit data to one or more stored data. Storage devices (eg, magnetic disks, magneto-optical disks or optical disks). However, computers do not need to have such devices. Computer-readable media for storing computer program instructions and data contain all forms of non-volatile memory, media and memory. Body device, for non-volatile § memory, media and memory devices including semiconductor memory devices (such as 'erasable read-only memory (EPR〇M), electronically readable memory (EEPR0M) And flash memory devices), disks (eg / 24 201223703 P hard drive or removable hard drive), magneto-optical discs and only discs (CD and read-only DVD $ (DVD_R〇M). The memory and the memory may be supplemented by special purpose logic circuits or by special purpose logic circuits. In order to provide interaction with the user, the embodiments of the technical subject matter described in the present specification may have a display device For example, a computer of a cathode ray tube display (CRT) or a liquid crystal display (LCD) is implemented to display information to a user, and the computer includes a keyboard and an indicator device (eg, a mouse or By the mouse or the trackball, the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user. For example, the feedback provided by the user can be any form of sensing. Feedback (eg, visual feedback, audible feedback, or tactile feedback); and can receive any form of input from the user that includes hearing, speaking, or tactile. This specification contains many implementations, however these embodiments are not intended to be limiting The scope of the invention is intended to be limited by the scope of the invention, and the description of the technical features can be used to exemplify specific embodiments of the invention. Certain technical features described in this specification in the context of different embodiments can also be combined to be practiced in a single embodiment. The different technical features described in the context of a single embodiment can also be implemented individually in a plurality of different embodiments or in any suitable embodiment combination. Moreover, while the technical features may be implemented in combinations of certain embodiments as described above, even as originally stated as technical features, one or more of the technical features that have been combined may in this case be combinable The removal, and perhaps the change of the claimed combination 25 201223703 to a combination or a combination of times. @样地 'When the operations are swayed in a specific order in the diagrams', this should not be interpreted as a must-have in a specific order or a sequential sequence as shown, or should not be interpreted as having to perform all the diagrams. Show the fuck # to achieve the desired 'fruit. In some cases, multiplex and parallelization may be more advantageous. Moreover, the separation of different system components in the above embodiments should not be construed as requiring such a division in all embodiments, and the separation of different system components in the above embodiments should be interpreted as describing the programmed components and systems. Should be integrated in a single software product or packaged in multiple software products. BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a schematic cross-sectional view of a polishing station. Figure 2 illustrates a schematic cross-sectional view of the carrier head. Figure 3A shows a top view of a substrate on the platform and the location where measurements are taken. Figure 3B illustrates concentric regions on the substrate. Figure 4 illustrates a schematic example diagram of signal strength, which is a time function of the substrate being ground (the substrate contains the GST layer). Figure 5 is a schematic illustration of the intensity of the signal, the signal intensity being from multiple regions of the substrate being polished, the substrate being ground containing the GST layer. Figure 6 illustrates a schematic example of signal strength from 26 201223703 multiple regions of the substrate being ground, the substrate being ground containing a metal layer. Figure 7 illustrates a flow chart of the grinding process. The same components in each drawing are denoted by the same component 5 tiger. [Main component symbol description] 10 Substrate 24 Platform 25 Central axis 26 Groove 27 Drive shaft 30 Grinding 塾 32 Grinding layer 34 Support layer 36 Solid window 38 Grinding slurry 39 Arm 50 In-situ monitoring module 51 Light source 52 Photodetector 53 optical head 58 bifurcated cable 58a trunk 58b branch 58c branch 71 central shaft 74 carrying drive shaft 76 carrier head rotary motor 80 carrier head 90 controller 150 chemical mechanical polishing station 154 rotary rack 302 housing 304 base total 306 ring frame structure 308 load chamber 27 201223703 310 fixed ring 320 substrate support assembly 326 elastic film 330 inner chamber 332 intermediate chamber 334 intermediate chamber 336 intermediate chamber 338 outer chamber 701 point 702 points 703 points 704 points 705 points 706 points 707 points 708 points 709 points 710 points 711 points 800 intensity traces 802 wave peaks 8 04 troughs 806 Nanyuan 810 threshold 812 traces 8 1 4 traces 816 traces 818 traces 910 grinding the initial substrate and monitoring in the original 9 9 trace grinding 914 Trace 916 Trace 918 Trace 920 determines the time for multiple area coverage to be cleared 922 Initial Plateau 924 Rapid Drop 926 Second Plateau 930 determines the adjustment of at least one region of the grinding pressure. 940 uses the adjusted grinding pressure to grind the T1 time. 28 201223703 Τ2 Τ4 Ζ1 Ζ3 Ζ5 substrates Time Τ3 Time Time Τ5 Time Concentric area Ζ2 Concentric area Concentric area Ζ4 Concentric area Concentric area 29