200525017 九、發明說明： 【發明所屬之技術領域】 本發明係針對用於在諸如玻璃、半導體、介電質、金屬 及其複合物、磁質儲存媒體與積體電路之項目（item)上創造 光滑、超平之表面的化學機械研磨。 【先前技術】 化學機械研磨（CMP)已成功地用於平面化金屬與介電薄 膜在似乎可能之平面化機制中，吾人認為研磨處理在 研磨漿存在的情況下涉及墊材料與晶圓表面上之高點之間 的緊毪接觸。在遠情況下，自正被研磨之晶圓表面與研磨 漿間之反應而產生的腐蝕材料藉由在墊_晶圓介面處進行 剪切而被移除。墊材料之彈性特性會顯著影響最終平面度 與研磨率。又，彈性特性是内在聚合物與其發泡結構之函 數。 在歷史上，基於聚胺基甲酸酯之墊已因其高強度、硬度、 模數與斷裂時之高伸長率而用於CMp。雖然該等塾可達成 良好的均句性與有效的構形減少，但是其快速且均勻地移 除表面材料之旎力作為使用函數而快速地下降。為基於聚 胺基甲酸S旨之墊而觀察之作為時間函數之材料移除率的下 降已被歸因於該等研磨塾在臨界剪切條件下之機械回應的 變化。通常，咸信基於聚胺基甲酸酯之CMP墊之功能性的 抽耗是歸ID於自用於研磨中之研磨漿與塾間之相互作用之 墊分解。 此外’为解在聚胺基甲酸酯墊之情況下產生了表面内與 96108.doc 200525017 表面自身之表面修改，其可對均勻 此^ 』所靨有害。或者，在某 2形下，用於CMP研磨墊之材料的表面修改可改良應用 效恥。“，該等修改可能為臨時的，因此需要經常替換 或撤消CM—。通常，聚胺基甲酸㈣在研磨之前亦需要 一磨合時期（break-in peri〇d)，除了在使科期後之修復盘 撤消之外。當研磨設備在空閒模式中日夺，亦經常有必要保 持傳統墊潮濕。當使用聚胺基甲酸醋或類似的習知墊時， 該等特徵不良地降低了 CMP之總效率。 因此，吾人所需的為一種改良的CMP墊，其能夠在CMp 期間提供高度平坦之表面並具有改良之壽+，而不會經歷 上述問題。 【發明内容】 為處理先前技術之上述缺陷，本發明在一實施例中提供 一種供化學機械研磨用之研磨墊。該研磨墊包括一包括熱 塑性泡沫基板之研磨體。該熱塑性泡沫基板具有一包括凹 面泡孔之表面。一塗覆於遠專凹面泡孔之一内表面的研磨 劑包括一包含碳化物或氮化物之無機金屬氧化物。 本發明之另一實施例係針對一種用於製備研磨墊之方 法。該方法包括在熱塑性泡沫基板内曝露封閉泡孔，以提 供一包括凹面泡孔之基板表面。該方法進一步包含以一包 括無機金屬氧化物之研磨劑來塗覆該等凹面泡孔之一内表 面’其中碳化物與氮化物在塗覆期間被併入於該無機金屬 氧化物中。 【實施方式】 96108.doc 200525017 本發明受益於先前未被認識之優勢，其使用熱塑性聚合 物作為基板，用於在凹面泡孔上沈積研磨劑之均勻冷層。 已發現凹面泡孔之内表面可形成極好的用於容納研磨劑之 均勻圖層的容器。假設凹面泡孔之中心作為一用於塗覆之 極好的喊核點’因為該泡孔在該中心處之表面能量為最 低。咸信在此位置處之初始塗覆促進了藉由研磨劑對凹面 泡孔之内表面進行之均勻覆蓋，藉此促進了具有該表面之 墊之研磨效能。 本發明之研磨劑包括一包含氮化物或碳化物之無機金屬 氧化物。使用该專無機金屬氧化物來塗覆墊表面亦可有利 地使研磨墊表面為永久親水性的。較佳地，無機金屬氧化 物具有將鼠化物或奴化物併入晶格中之原子晶格。該等研 磨劑表面塗層之使用藉由修改研磨墊之表面機械特性而增 強了研磨。 舉例而言，改變特定無機金屬氧化物之氮化物或碳化物 含量將允許精密地調整研磨墊表面之機械特性，以便匹配 正被研磨之表面的機械特性。又，使研磨墊表面之機械特 性與正被研磨之表面的機械特性相匹配會改良研磨率選擇 性並減少處理誘導之缺陷，如刮擦。藉由利用由電漿增強 化學氣相沈積（PECVD)表面塗層與在熱塑性基板表體中之 一級熱誘導之反應所產生的熱塑性聚合物基板改變實現了 調整表面機械特性。 本發明之一實施例為一用於化學機械研磨半導體裝置之 研磨塾。圖1呈現本發明之例示性研磨墊丨〇〇。研磨墊丨 96108.doc 200525017 包括研磨體110。研磨體110包括一具有一包括凹面泡孔125 之表面丨20之熱塑性泡沫基板U5e塗覆凹面泡孔125之内表 面135之研磨劑13G包括—包含碳化物或氮化物之無機金屬 氧化物。 研磨劑130包括由一或多種無機金屬氧化物組成之陶瓷 化合物，言亥-或該等多種無機金屬氧化物藉由在電漿增強 化學氣相沈積（PECVD)處理中於熱塑性泡珠基板ιΐ5表面 120上接枝一級反應物而形成。如下文進一步說明，在本發 月中可改iPECVD處理以促進碳化物或氮化物在無機金 屬氧化物中之包含。 將碳化物或氮化物中之—或兩者併人無機金屬氧化物之 晶格中係較佳的。舉例而言’當無機金屬氧化物包括二氧 化夕時則Ba格内可包括結構為聚合之si_〇_si之石夕酸鹽， 該等結構具有四面體與失真四面體組態。氮化物可包ς被 併入該等晶格中之氮化石夕。或者，當無機金屬氧化物包括 氧化鈦時，則氮化物可包括被併人氧化鈦晶格中之氮化 欽。類似地’可將碳切與碳化鈦併人研磨劑13〇中，其無 機孟屬氧化物分別包括二氧化石夕與氧化欽。在一些較佳實 &例中’諸如鼠化石夕之氮化物包括約10莫耳百分數之研磨 劑1 3 0，而在盆夕眚始么丨+ 牡，、匕貫施例中，諸如碳化矽之碳化物包括約⑺ 莫耳百分數之研磨劑。在-些較佳實施例中，研磨齊"3〇 可^括以該等濃度之氮化物與碳化物兩者。 >田ECVD處理被延長至更長時期時，研磨劑丄⑽中之石夕 知展度降低’導致了氧_矽之比率降低。在研磨墊_之一 96108.doc 200525017 些有利實施例中，其中研磨劑130包括二氧化矽，〇:Si比率 為至少約8:1，且在一些情況下，其為至少約9.9:1。 將氮化物與碳化物包含在研磨劑130中提供了改變研磨 墊100之機械特性之額外之以前未被認識之方法，並藉此改 變墊之研磨特性。在一些較佳實施例中，研磨墊1 〇〇具有大 於60 KPa之硬度，且更佳大於70 KPa。在其它較佳實施例 中，研磨墊100具有在約62 KPa與約70 KPa之間的硬度。在 某些較佳實施例中，研磨墊100具有約大於約3 MPa之彈性 模數，且更佳為4 MPa或更大。在其它較佳實施例中，研磨 墊具有分別在約4 MPa與8.2 MPa之間的彈性模數。在另外 其它實施例中，研磨墊100具有約〇·4 MPa或更大之損耗模 數’且更佳為至少約0.6 MPa。 如美國專利6,579,604與美國申請案號第1〇/241，〇74號（其 、引用的方式併入本文中）中所揭示，可自在處理中 用作二級反應物之各種含氧有機金屬化合物中產生研磨劑 1 3 0之無機金屬氧化物。舉例而言，二級電漿混合物可包含 1太錳或鈕之過渡金屬。然而，能夠形成揮發性有機 至屬化合物(如含有一或多個氧原子之金屬酯)且能夠被接 枝至聚合物表面之任何金屬元素都是合適的。矽亦可用作 有機金屬二級電漿混合物之金屬部分。在該等實施例中， 有機金屬試劑之有機部分可為_、乙酸§旨或烧氧基斷片。 在車父佳實施例中，研磨劑13〇之無機金屬氧化物包括：石夕氧 =或鈦氧化物’分別例如二氧化石夕或二氧化欽；四乙氧 烷聚合物；或鈦烷氧化物聚合物。 96108.doc 200525017 其它二級電漿反應物包含臭氧、烷氧基矽烷、水、氨、 醇類、礦油精或過氧化氫。在一些較佳實施例中，- 一級電 漿反應物包括：鈦酯；鈕烷氧化物，包含其中燒氧化物部 分具有1-5個碳原子之鈕烷氧化物；乙酸錳水溶液；溶解於 礦油精中之猛烧氧化物；乙酸猛；乙醯基丙g同酸鐘；銘产 氧化物；烷氧基鋁酸鹽；氧化鋁；鍅烷氧化物，其中燒氧 化物具有1 -5個破原子；烷氧基锆酸鹽；乙酸鎂；及乙酿笑 丙_酸鎂。二級電漿反應物亦涵蓋其它實施例，例如，院 氧基矽烷與臭氧、烷氧基矽烷與氨、鈦酯與水、鈦g旨與醇 類、或鈦醋與臭氧。 熱塑性泡沐基板11 5之一些較佳實施例包括交聯聚稀 烴’如聚乙烯、聚丙烯及其組合。在某些較佳實施例中， 熱塑性泡沫基板115包括交聯均聚物或共聚物之封閉泡孔 泡珠。包括聚乙烯（PE)之封閉泡孔泡沫交聯均聚物之實例 包含：來自 Voltek(Lawrence，MA)之 VolaraTM 與 VolextraTM ; 來自 JMS Plastics Supply，Inc.(Neptune，NJ)之 Aliplast™;或 Senflex T-Cell™(R0gers Corp·，Rogers，CT)。包括聚乙烯與 乙稀乙酸乙烯酯（EVA)之交聯共聚合物之封閉泡孔泡沫的 貫例包含：VolextraTM(來自 Voltek Corp.); Senflex EVAtm(來自 R〇gers c〇rp·);及 J-foamTM(來自 jms Plastics JMS Plastics Supply，Inc.)。 在其它較佳實施例中，封閉泡孔熱塑性泡沫基板115包括 交聯乙烯乙酸乙烯酯共聚物與低密度聚乙烯共聚物（即，較 佳在約0.1與約〇·3 gm/cc之間）的摻合物。在另外其它有利的 96108.doc •10- 200525017 實施例中，摻合物具有在約1:9盘約@ ^ w 〃、αν·1之間的乙烯乙酸乙烯 酉曰·？乙稀重量比率。在某些較佳實施例中，摻合物包括在 自約5至約45 Wt%範圍内之EVA，較佳為自約6至約Μ 秦且更佳為自約12至約24 wt%。吾人認為該等摻合物 有盈於對熱塑性泡沫基板! 15之封閉泡孔14〇的所需產生， 其具有小尺寸（例如，直徑在約1〇與約5〇〇微米之間，且更 佳在約5G至15G微米之間）。在另外更佳實施财，換合物 具，在約〇·6··9·4與約u··8·2之間的乙烯乙酸乙稀酯··聚乙稀 重里比率。在甚至更佳的實施例中，摻合物具有在約〇.6··9·4 與約1.2:8.8之間的乙烯乙酸乙烯酯:聚乙烯重量比率。 如圖1中之進一步說明，#塑性泡珠基板u5包括封閉泡 此處所使用之術語封閉泡孔丨4〇係指由基板11 $内被 氣或用作喷％劑之其它氣體（如氮或氦）佔據之膜所界定 、何谷積封閉泡孔140形成一在削切基板丨丨5時所形成 之大體上凹面泡孔135。然而，凹面泡孔135無需具有光滑 或弓曲2。相反，凹面泡孔丨35可具有不規則的形狀與尺 寸，諸如熱塑性泡沫基板11 5之組份與用於製備熱塑性泡 沬基板115之程序的若干因素可影響封閉泡孔14〇與凹面泡 孔135之形狀及尺寸。 ρ如圖1中之進一步說明，熱塑性泡沫基板115可耦接至任 選的背板材料145。在一些較佳實施例中，背板材料145為 门J丨生的。剛性背板有利地限制了泡沫在研磨期間的可壓縮 11與伸長率’其又降低了在藉由CMP之金屬研磨期間的腐 餘”表面凹陷效應。在一些情況下，背板材料丨45包括高密 96108.doc 200525017 又“稀(即’大於約0.98 gm/cc)，且更佳為濃縮之高密度 聚乙稀。纟某些情況下，藉由使用諸如環氧樹脂或孰習此 項技術者所熟知之其它材料之習知黏著劑15〇的化學鍵社 來達錢熱塑性、泡泳基板115之輕接。在—些較佳實施: 中，藉由將熔融背板材料145擠壓塗覆至熱塑性 115上來達成耦接。在另外其它較佳的實施例中，背板材料 145熱焊接至熱塑性泡沫基板11 5。 本發明之另一態樣係一用於製備供化學機械研磨用之研 磨墊的方法。圖2至圖4呈現一製備研磨墊2〇〇之例示性方法 中之所選步驟。可將研磨墊及其組件部分之任何實施例(包 含上述一級與二級電漿反應物）併入製備研磨墊2 〇 〇之方法 中0 現轉向圖2，其顯示在曝露研磨墊之熱塑性泡珠基板 2 2 0内之封閉泡孔2丨〇以提供一包括凹面泡孔2 4 〇之基板表 面230之後之部分構造的研磨墊2〇〇。凹面泡孔24〇藉由削切 而形成於基板之表面230上。此處所用之術語削切意指用以 切去基板220之一表面薄層以便曝露熱塑性泡沫基板22〇内 之凹面泡孔240的任何處理。可藉由使用一般技術者所熟知 之任何習知技術來達成削切。 圖3及圖4說明表面塗覆處理中之所選階段。首先轉向圖 3，其說明在將基板表面23〇曝露於初始電漿反應物、接著 在PECVD處理中曝露於二級電漿反應物之後之部分完成的 研磨墊2 0 0。 曝路於初始電漿反應物形成了熱塑性泡沫基板22〇之修 96108.doc -12- 200525017 改的表面310。小心地控制電漿撤消之條件與持續時間以避 免對熱塑性泡沫基板220造成過度損害係重要的。舉例而 言，過高或未被控制之輻射流（radio flow)放電電極溫度可 導致熱塑性泡沫基板220溶融、彎曲或斷裂。在該方法之一 些較佳實施例中，輻射流放電電極溫度維持在約2〇。(：與1〇〇 °C之間’且更佳在約30°C與5CTC之間。在一些情況下，使 用在約250與約1000瓦特之間之射頻（RF)操作功率，且更佳 在約300瓦特與400瓦特之間。在某些較佳實施例中，初始 電漿反應物包括情性氣體’諸如氖，且更佳為氬或氦。在 一些情況下，曝露於初始電漿進行約1秒與約6〇秒之間，且 更佳為約30秒。在一些實施例中，pECvD反應腔室維持在 約300 mToir與約400 mT〇rr之間，且更佳為約35〇 mT〇rr。 圖3亦顯示在將修改的表面3 10曝露於二級電漿反應物之 後之部分完成的研磨墊200。在一些較佳實施例中，二級電 漿反應物包括四乙氧基矽烷（TEOS)或鈦烷氧化物 (TYZOR)。在一些情況下，二級電漿反應物亦包含第一電 漿反應物，如與氦或氬氣體混合之TEOS或TYZOR氣相。曝 露於二級電漿反應物會導致二級電漿反應物接枝至修改的 表面3 1 0，以形成一包括無機金屬氧化物之研磨劑320。研 磨劑320塗覆凹面泡孔240之内表面330。 此外，小心地控制對二級電漿反應物之曝露的條件與持 續時間，以避免損害熱塑性泡珠基板220或研磨劑320，並 以達成研磨劑320之長久塗覆。在一些實施例中，pecvd 反應腔室維持在約300 mTorr與約400 mTorr之間，且更佳為 96108.doc -13- 200525017 約350 mToir。在一些情形下，輻射流放電電極溫度維持在 約20°C與100t之間，且更佳為約30°C與50。(：之間。在一此 情況下，使用在約50與約500瓦特之間之111?操作功率，且更 佳為在約250瓦特與約350瓦特之間。 現轉向圖4，其說明在曝露於二級電漿反應物至少約3〇 分鐘之時期之後之部分完成的研磨墊2〇〇。在一些較佳實施 例中，曝露於二級電漿反應物持續在約3〇分鐘與約6〇分鐘 之間。在其它較佳實施例中，曝露於二級電漿反應物持續 在約30分鐘與約45分鐘之間。該等曝露時期有利地增強了 氮化物或奴化物、或該等兩者併入研磨劑3 2 〇之無機金屬氧 化物中。在該方法之一些實施例中，熱塑性泡沫基板22〇 之封閉泡孔410的内部包括氮氣體。該氮氣體可與二級電栽 反應物進行反應以形成氮化物。在該方法之其它實施例 中’熱塑性泡沫基板220之至少一部分與二級電漿反應物進 行反應以形成碳化物。例如，在一些實施例中，熱塑性泡 沫基板220自修改的表面310之約1微米深度内之碳基物質 可與二級電漿反應物進行反應。 如上文所論述與隨後之實例部分中之進一步說明，藉由 PECVD對研磨劑320之沈積修改了某些熱塑性泡沫23〇(如 聚烯fe泡沫）之表面特性。藉由研磨劑320對熱塑性泡沫表 面3 10進行之高達約3〇分鐘之表面塗覆藉由一機制而發 生。然而’在該時期之後，表面塗覆藉由一不同的機制而 發生。以此方式又導致產生了 一在表面微機械與化學反應 中具有獨特差異（視塗覆時間而定）之研磨墊表面41 〇。 96108.doc -14- 200525017 在些6況下，當塗覆時間增加，溫度熱塑性泡沫基板 220之溫度增加。以此方式又導致了用於使基板22〇發泡且 位於基板240之封閉泡孔41〇中之氮氣體之除氣。在一些實 施例中，例如其中研磨劑32〇包括二氧化矽，除氣之氮與研 磨墊表面上之矽物質反應，其導致了在自以〇4至以3凡之化 學叶ΐ轉化中形成S“N4物質。當然，在研磨劑包括諸如氧 化鈦之其它無機金屬氧化物之實施例中可發生類似反應。 類似地，在長塗覆時間時，對熱塑性泡沫基板24〇表面3iq 之離子轟擊會在墊2〇〇之表面31〇上產生可感知之碳基量。 在一些實施例中，例如其中研磨劑32()包括二氧化矽，該等 基與矽物質進行反應以形成碳化矽（sic)，其隨後被併入塗 覆墊200之研磨劑320中。 與起初熱塑性泡沫基板或經受短暫塗覆時期之基板相 比，諸如S“N4與SiC之物質併入研磨劑32〇中修改了研磨墊 2〇〇之特性，如增強了其剛性、硬度且改變了其彈性模數。 在描述了本發明之後，咸信參照以下試驗將會使同樣内 容變得更加顯而易見。應瞭解到，該等試驗僅為說明目的 而k供且不應被理解為限制了本發明。例如，儘管可在試 驗室環境（laboratory setting)下進行下文所描述之試驗，但 是熟習此項技術者可將特定數目、尺寸與數量調節至適當 值以用於全面的工廠環境（plant setting)。 試驗 進行試驗以·· 1)表徵塗覆有研磨劑之熱塑性泡沫基板之 作為塗覆時間之函數的化學組份；2)表徵塗覆有研磨劑之 96108.doc 200525017 泡沫基板的化學特性；及3)量測塗覆有研磨劑之研磨墊之 作為塗覆時間之函數的研磨特性。 將一熱塑性泡沫基板形成至直徑為大約12〇 、厚度為 約〇.3cm之環形研磨墊。商業上獲得之表示為，,J-6〇SE，f的熱 塑性泡沫基板（J_f0am，來自JMS piastics，Neptune NJ)包括 存在於商業上提供之基板中之約18% EVA、約16至約20% /月石粉及其餘為聚乙烯與其它添加物（如矽酸鹽）之摻合 物。以一商業切割刀片（型號D51〇〇 K1，來自Fecken_Kirfel， Aachen，Germany)削切j_60薄片。接著，以一水性/異丙醇 溶液手動地清潔該等薄片。 接著，藉由將所削切之基板置放於一習知商業射頻輝光 放電（RFGD)電漿反應器之反應腔室中，以一研磨劑塗覆 J 60SE基板，該電漿反應器具有一溫控電極組態（pE_2s ; 面級能量系統，Medford，NY)。藉由在維持於350 mTorr之 反應腔室内持續引入3〇秒之一級電漿反應物氬而開始基板200525017 IX. Description of the invention: [Technical field to which the invention belongs] The present invention is directed to the creation of items such as glass, semiconductors, dielectrics, metals and their composites, magnetic storage media and integrated circuits Chemical mechanical grinding of smooth, ultra-flat surfaces. [Previous technology] Chemical mechanical polishing (CMP) has been successfully used to planarize metal and dielectric films. In the seemingly possible planarization mechanism, I believe that the polishing process involves the pad material and the wafer surface in the presence of a polishing slurry. Tight contact between high points. In the remote case, the corrosive material generated from the reaction between the wafer surface being polished and the polishing slurry is removed by shearing at the pad_wafer interface. The elastic properties of the pad material can significantly affect the final flatness and polishing rate. The elastic properties are a function of the intrinsic polymer and its foam structure. Historically, polyurethane-based mats have been used in CMPs because of their high strength, hardness, modulus, and high elongation at break. Although they can achieve good uniformity and effective configuration reduction, the force that removes surface materials quickly and uniformly decreases as a function of use. The decrease in material removal rate as a function of time observed for polyurethane-based mats has been attributed to changes in the mechanical response of these mills under critical shear conditions. In general, the functional exhaustion of polyurethane-based CMP pads is attributed to the breakdown of the pads due to the interaction between the polishing slurry and the grate used in polishing. In addition, in order to solve the problem in the case of a polyurethane pad, a surface modification within the surface and the surface of 96108.doc 200525017 itself is generated, which can be harmful to uniformity. Alternatively, the surface modification of the material used for the CMP polishing pad in a certain shape can improve application efficiency. "These modifications may be temporary and therefore require frequent replacement or revocation of CM—Generally, urethane also requires a break-in period before grinding, except after In addition to the repair disc revocation. When the grinding equipment is in the idle mode, it is often necessary to keep the traditional pads wet. When using polyurethane or similar conventional pads, these characteristics undesirably reduce the overall CMP Efficiency. Therefore, what I need is an improved CMP pad that can provide a highly flat surface and have an improved lifetime during CMP without experiencing the above problems. [Summary of the Invention] To address the aforementioned shortcomings of the prior art In one embodiment of the present invention, a polishing pad for chemical mechanical polishing is provided. The polishing pad includes an abrasive body including a thermoplastic foam substrate. The thermoplastic foam substrate has a surface including concave cells. The abrasive on the inner surface of one of the concave cells includes an inorganic metal oxide containing a carbide or a nitride. Another embodiment of the present invention is directed to a A method for preparing a polishing pad. The method includes exposing closed cells in a thermoplastic foam substrate to provide a substrate surface including concave cells. The method further includes coating the substrates with an abrasive including an inorganic metal oxide. An inner surface of one of the concave cells' in which carbides and nitrides are incorporated into the inorganic metal oxide during coating. [Embodiment] 96108.doc 200525017 The present invention benefits from previously unrecognized advantages, its use The thermoplastic polymer is used as a substrate for depositing a uniform cold layer of abrasive on the concave cells. The inner surface of the concave cells has been found to form an excellent container for containing a uniform layer of abrasive. Assume that the concave cells The center serves as an excellent nucleation point for coating 'because the surface energy of the cell at the center is the lowest. Xianxin's initial coating at this position facilitates the application of abrasives to the concave cells. The uniform coverage of the inner surface thereby promotes the abrasive performance of the pad having the surface. The abrasive of the present invention includes an inorganic compound containing a nitride or a carbide. It is an oxide. Using the special inorganic metal oxide to coat the surface of the pad can also advantageously make the polishing pad surface permanent hydrophilic. Preferably, the inorganic metal oxide has the ability to incorporate rat or slave compounds into the crystal lattice. Atomic lattice. The use of these abrasive surface coatings enhances polishing by modifying the mechanical properties of the surface of the polishing pad. For example, changing the nitride or carbide content of a particular inorganic metal oxide will allow fine adjustments Mechanical characteristics of the polishing pad surface to match the mechanical characteristics of the surface being polished. Also, matching the mechanical characteristics of the polishing pad surface to the mechanical characteristics of the surface being polished will improve polishing rate selectivity and reduce processing-induced defects Such as scratching. The adjustment of the mechanical properties of the surface is achieved by utilizing changes in the thermoplastic polymer substrate produced by the plasma enhanced chemical vapor deposition (PECVD) surface coating and a first-order thermally induced reaction in the thermoplastic substrate surface. An embodiment of the present invention is a polishing pad for chemical mechanical polishing of semiconductor devices. FIG. 1 presents an exemplary polishing pad of the present invention. The polishing pad 丨 96108.doc 200525017 includes an abrasive body 110. The abrasive body 110 includes an abrasive 13G having a thermoplastic foam substrate U5e with a surface including concave cells 125 and an inner surface 135 of the concave cells 125, including an inorganic metal oxide including a carbide or a nitride. The abrasive 130 includes a ceramic compound composed of one or more inorganic metal oxides, or a plurality of inorganic metal oxides, on the surface of a thermoplastic bead substrate by a plasma enhanced chemical vapor deposition (PECVD) process. Formed by grafting first-order reactants on 120. As explained further below, iPECVD can be modified to promote the inclusion of carbides or nitrides in inorganic metal oxides in the current month. The combination of carbides or nitrides, or both, in the crystal lattice of an inorganic metal oxide is preferred. For example, when the inorganic metal oxide includes dioxide, the Ba lattice may include a petrified salt of a structured si_〇_si. These structures have a tetrahedral and distorted tetrahedral configuration. Nitrides may be nitrides that are incorporated into such lattices. Alternatively, when the inorganic metal oxide includes titanium oxide, the nitride may include nitride incorporated in a titanium oxide crystal lattice. Similarly, carbon can be cut with titanium carbide and incorporated in abrasive 130, and its inorganic oxides include sulphur dioxide and zinc oxide, respectively. In some preferred examples, the nitrides such as rat fossils include about 10 mole percent of abrasive 1 3 0, and in the case of pots and slabs, in the embodiment, such as carbonization Silicon carbide includes about ⑺ mole percentage of abrasive. In some preferred embodiments, the milling " 30 may include both nitrides and carbides at these concentrations. > When the field ECVD process is extended for a longer period of time, the decrease in the degree of spread of the stone in the polishing agent ’leads to a decrease in the ratio of oxygen to silicon. In some advantageous embodiments of the polishing pad 96108.doc 200525017, wherein the abrasive 130 includes silicon dioxide, the O: Si ratio is at least about 8: 1, and in some cases, it is at least about 9.9: 1. The inclusion of nitrides and carbides in the abrasive 130 provides an additional previously unknown method of altering the mechanical characteristics of the polishing pad 100 and thereby alters the polishing characteristics of the pad. In some preferred embodiments, the polishing pad 100 has a hardness greater than 60 KPa, and more preferably greater than 70 KPa. In other preferred embodiments, the polishing pad 100 has a hardness between about 62 KPa and about 70 KPa. In some preferred embodiments, the polishing pad 100 has a modulus of elasticity greater than about 3 MPa, and more preferably 4 MPa or greater. In other preferred embodiments, the polishing pad has a modulus of elasticity between about 4 MPa and 8.2 MPa, respectively. In yet other embodiments, the polishing pad 100 has a loss modulus ' of about 0.4 MPa or greater and more preferably at least about 0.6 MPa. As disclosed in U.S. Patent 6,579,604 and U.S. Application No. 10 / 241,074 (which is incorporated herein by reference), various oxygen-containing organometallic compounds that can be used as secondary reactants in processing An inorganic metal oxide of abrasive 130 is produced in the process. For example, the secondary plasma mixture may include 1 teramanganese or a button transition metal. However, any metal element capable of forming a volatile organic subordinate compound (such as a metal ester containing one or more oxygen atoms) and capable of being grafted to a polymer surface is suitable. Silicon can also be used as the metal part of organometallic secondary plasma mixtures. In these embodiments, the organic portion of the organometallic reagent may be acetic acid, acetic acid, or oxy-oxygen fragment. In the Chevrolet embodiment, the inorganic metal oxides of the abrasive agent 13 include: Shixiu = or titanium oxide ', such as sulphur dioxide or dioxin, respectively; tetraethoxylan polymer; or titane oxidation物 polymer. 96108.doc 200525017 Other secondary plasma reactants include ozone, alkoxysilanes, water, ammonia, alcohols, mineral spirits, or hydrogen peroxide. In some preferred embodiments, the primary plasma reactant includes: a titanium ester; a button alkoxide, including a button alkoxide in which the oxide portion has 1-5 carbon atoms; an aqueous solution of manganese acetate; Fierce burning oxides in olein; acetic acid; acetamidine g isoacids; Ming oxides; alkoxy aluminates; alumina; oxane oxides, among which the burning oxides have 1 to 5 Atom breaking; alkoxy zirconate; magnesium acetate; Other examples of secondary plasma reactants include, for example, oxysilanes and ozone, alkoxysilanes and ammonia, titanium esters and water, titanium oxides and alcohols, or titanium vinegar and ozone. Some preferred embodiments of the thermoplastic foam substrate 115 include cross-linked polyolefins such as polyethylene, polypropylene, and combinations thereof. In certain preferred embodiments, the thermoplastic foam substrate 115 includes closed cell foam beads of a crosslinked homopolymer or copolymer. Examples of closed cell foam crosslinked homopolymers including polyethylene (PE) include: VolaraTM and VolextraTM from Voltek (Lawrence, MA); Aliplast ™ from JMS Plastics Supply, Inc. (Neptune, NJ); or Senflex T-Cell ™ (Rogers Corp., Rogers, CT). Examples of closed-cell foams including cross-linked copolymers of polyethylene and vinyl acetate (EVA) include: VolextraTM (from Voltek Corp.); Senflex EVAtm (from Rogers Corp ·); and J-foamTM (from jms Plastics JMS Plastics Supply, Inc.). In other preferred embodiments, the closed cell thermoplastic foam substrate 115 includes a crosslinked ethylene vinyl acetate copolymer and a low density polyethylene copolymer (ie, preferably between about 0.1 and about 0.3 gm / cc) Of the blend. In yet other advantageous 96108.doc • 10-200525017 embodiments, the blend has ethylene vinyl acetate between about 1: 9 disks about @ ^ w 〃, αν · 1. Ethylene weight ratio. In certain preferred embodiments, the blend includes EVA in a range from about 5 to about 45 Wt%, preferably from about 6 to about 24%, and more preferably from about 12 to about 24 wt%. We believe that these blends have the benefit of producing the desired closed cells 14 of a thermoplastic foam substrate! 15 with a small size (e.g., between about 10 and about 500 microns in diameter, and more Preferably between about 5G and 15G microns). In another aspect, it is better to implement an exchange, with a ratio of ethylene vinyl acetate to polyethylene acetate between about 0.6 · 9 · 4 and about u ·· 8.2. In an even more preferred embodiment, the blend has an ethylene vinyl acetate: polyethylene weight ratio between about 0.6 ·· 9 · 4 and about 1.2: 8.8. As further illustrated in Figure 1, #plastic bubble bead substrate u5 includes closed cells. The term closed cells as used herein 4 refers to a substrate gas or other gas (such as nitrogen or Defined by the film occupied by He), He Guji's closed cells 140 form a generally concave cell 135 formed when the substrate is cut. However, the concave cells 135 need not be smooth or bowed2. Conversely, the concave cells 35 may have irregular shapes and sizes. Several factors such as the components of the thermoplastic foam substrate 115 and the procedure used to prepare the thermoplastic foam substrate 115 may affect the closed cells 14 and the concave cells. Shape and size of 135. As further illustrated in FIG. 1, the thermoplastic foam substrate 115 may be coupled to an optional backsheet material 145. In some preferred embodiments, the backsheet material 145 is produced from a door. The rigid backing plate advantageously limits the compressible 11 and elongation of the foam during grinding, which in turn reduces the "scumming" effect during metal grinding by CMP. The surface sag effect. In some cases, the backing material 45 includes Gaomi 96108.doc 200525017 is also "diluted (ie, 'greater than about 0.98 gm / cc), and more preferably concentrated high density polyethylene.纟 In some cases, the chemical bonding of 150, which is a conventional adhesive such as epoxy resin or other materials known to those skilled in the art, is used to achieve the light-weight adhesion of thermoplastic substrate 115. In some preferred implementations, coupling is achieved by extrusion coating molten backsheet material 145 onto thermoplastic 115. In still other preferred embodiments, the backsheet material 145 is thermally welded to the thermoplastic foam substrate 115. Another aspect of the present invention is a method for preparing a polishing pad for chemical mechanical polishing. Figures 2 to 4 present selected steps in an exemplary method of preparing a polishing pad 2000. Any embodiment of the polishing pad and its component parts (including the primary and secondary plasma reactants described above) can be incorporated into the method of preparing the polishing pad 2000. Turning now to FIG. 2, which shows the thermoplastic foam exposed to the polishing pad The closed cells 2 in the bead substrate 2 20 are provided to provide a polishing pad 200 having a part of the structure after the substrate surface 230 including the concave cells 2 4. The concave cells 24 are formed on the surface 230 of the substrate by cutting. The term chipping as used herein means any treatment used to cut off a thin layer of a surface of the substrate 220 to expose the concave cells 240 in the thermoplastic foam substrate 22. Cutting can be achieved by using any conventional technique known to those of ordinary skill. 3 and 4 illustrate selected stages in the surface coating process. Turning first to FIG. 3, it illustrates a partially completed polishing pad 2000 after exposing the substrate surface 230 to the initial plasma reactant and then exposing it to the secondary plasma reactant in a PECVD process. Exposure to the initial plasma reactants formed a thermoplastic foam substrate 22 repair 96108.doc -12-200525017 modified surface 310. It is important to carefully control the conditions and duration of plasma withdrawal to avoid excessive damage to the thermoplastic foam substrate 220. For example, an excessively high or uncontrolled radio flow discharge electrode temperature may cause the thermoplastic foam substrate 220 to melt, bend, or break. In some preferred embodiments of the method, the temperature of the radiation current discharge electrode is maintained at about 20 °. (: And 100 ° C 'and more preferably between about 30 ° C and 5CTC. In some cases, radio frequency (RF) operating power between about 250 and about 1000 Watts is used, and better Between about 300 Watts and 400 Watts. In some preferred embodiments, the initial plasma reactant includes an emotional gas such as neon, and more preferably argon or helium. In some cases, exposure to the initial plasma For between about 1 second and about 60 seconds, and more preferably about 30 seconds. In some embodiments, the pECvD reaction chamber is maintained between about 300 mToir and about 400 mTorr, and more preferably about 35 〇mT〇rr. Figure 3 also shows a partially completed polishing pad 200 after exposing the modified surface 3 10 to a secondary plasma reactant. In some preferred embodiments, the secondary plasma reactant includes tetraethyl Oxysilane (TEOS) or titanyl oxide (TYZOR). In some cases, the secondary plasma reactant also includes a first plasma reactant, such as TEOS or TYZOR gas phase mixed with helium or argon gas. Exposure The secondary plasma reactant will cause the secondary plasma reactant to graft to the modified surface 3 1 0 to form an inorganic gold Oxide abrasive 320. The abrasive 320 coats the inner surface 330 of the concave cells 240. In addition, carefully control the conditions and duration of exposure to the secondary plasma reactants to avoid damaging the thermoplastic bead substrate 220 or Abrasive 320, and coated to achieve long-lasting abrasive 320. In some embodiments, the pecvd reaction chamber is maintained between about 300 mTorr and about 400 mTorr, and more preferably 96108.doc -13- 200525017 about 350 mToir. In some cases, the radiant flow discharge electrode temperature is maintained between about 20 ° C and 100t, and more preferably between about 30 ° C and 50. (:. In this case, it is used at about 50 and 111? Operating power between about 500 watts, and more preferably between about 250 watts and about 350 watts. Turning now to FIG. 4, which illustrates a period of at least about 30 minutes after exposure to a secondary plasma reactant Partially finished polishing pad 200. In some preferred embodiments, exposure to the secondary plasma reactants is between about 30 minutes and about 60 minutes. In other preferred embodiments, exposure to Secondary plasma reactants last between about 30 minutes and about 45 minutes These periods of exposure advantageously enhance the nitride or slave compound, or both of these incorporated into the inorganic metal oxide of abrasive 3 2 0. In some embodiments of the method, the closure of the thermoplastic foam substrate 22 0 The interior of the cells 410 includes a nitrogen gas. The nitrogen gas may react with a secondary electroplating reactant to form a nitride. In other embodiments of the method, at least a portion of the 'thermoplastic foam substrate 220 and the secondary plasma reactant The reaction is performed to form carbides. For example, in some embodiments, a carbon-based substance within a depth of about 1 micron from the self-modified surface 310 of the thermoplastic foam substrate 220 may react with a secondary plasma reactant. As discussed above and further explained in the Examples section that follows, the deposition of abrasive 320 by PECVD modifies the surface characteristics of certain thermoplastic foams 23 (such as poly-fe foams). Surface coating of thermoplastic foam surface 3 10 by abrasive 320 for up to about 30 minutes occurs by a mechanism. However, after this period, surface coating occurs by a different mechanism. This in turn results in a polishing pad surface with a unique difference in surface micromechanical and chemical reactions (depending on the coating time). 96108.doc -14- 200525017 In these 6 cases, as the coating time increases, the temperature of the temperature thermoplastic foam substrate 220 increases. This in turn leads to a degassing of nitrogen gas for foaming the substrate 22 and located in the closed cells 410 of the substrate 240. In some embodiments, for example, where the abrasive 32o includes silicon dioxide, the degassed nitrogen reacts with the silicon species on the surface of the polishing pad, which results in the formation of a chemical transformation from 0 to 3 S "N4 substance. Of course, a similar reaction can occur in embodiments where the abrasive includes other inorganic metal oxides such as titanium oxide. Similarly, during a long coating time, 3iq ion bombardment of the surface of the thermoplastic foam substrate 24 Perceived carbon-based amount will be generated on the surface 31 of the pad 200. In some embodiments, for example, where the abrasive 32 () includes silicon dioxide, these groups react with the silicon substance to form silicon carbide ( sic), which is then incorporated into the abrasive 320 of the coating pad 200. Compared to the original thermoplastic foam substrate or a substrate subjected to a short coating period, substances such as S "N4 and SiC are incorporated into the abrasive 32. Modified The characteristics of the polishing pad 2000 are improved, such as enhancing its rigidity and hardness and changing its elastic modulus. After describing the present invention, the same will become more apparent with reference to the following experiments. It should be understood that these tests are provided for illustrative purposes only and should not be construed as limiting the invention. For example, although the tests described below can be performed in a laboratory setting, those skilled in the art can adjust specific numbers, sizes, and quantities to appropriate values for use in a comprehensive plant setting. The test was conducted to: 1) characterize the chemical composition of the abrasive coated thermoplastic foam substrate as a function of coating time; 2) characterize the chemical characteristics of the abrasive coated 96108.doc 200525017 foam substrate; and 3) Measure the abrasive characteristics of the abrasive pad coated abrasive as a function of coating time. A thermoplastic foam substrate was formed to a circular polishing pad having a diameter of about 120 and a thickness of about 0.3 cm. Commercially available as, J-6〇SE, f thermoplastic foam substrate (J_f0am, from JMS piastics, Neptune NJ) including about 18% EVA, about 16 to about 20% present in commercially available substrates / Moonstone powder and the rest are blends of polyethylene and other additives (such as silicate). A commercial cutting blade (model D5100 K1 from Fecken_Kirfel, Aachen, Germany) was used to cut the j_60 sheet. The sheets were then manually cleaned with an aqueous / isopropanol solution. Then, by placing the cut substrate in the reaction chamber of a conventional commercial RFGD plasma reactor, and coating the J 60SE substrate with an abrasive, the plasma reactor has a temperature Control electrode configuration (pE_2s; Surface Energy System, Medford, NY). The substrate was started by continuously introducing 30 seconds of first-order plasma reactant argon in a reaction chamber maintained at 350 mTorr
之電聚處理。電極溫度維持在3〇〇c，並使用3〇〇瓦特之RF 操作功率。隨後，於〇.1〇 SLM下自約0至約45分鐘範圍内之 日可期内’引入二級反應物，並包括混合有He或Ar氣體之 TE〇s °藉由在單體儲集層溫度（MRT ; 5〇 土 1〇。〇下之二級 反應物單體的氣相背壓（Bp)來支配氣體流中之二級反應物 白勺量。 u 藉由研磨晶圓來檢查j6〇se研磨墊之研磨特性，該等晶圓 ' 具有一約4000埃厚之鎢表面與一下層之約250埃厚之钽障 壁層。藉由使用一商業研磨器（產品號EP〇222，來自Ebara 96108.doc 200525017Its electro-polymerization processing. The electrode temperature was maintained at 300c and an RF operating power of 300 watts was used. Subsequently, a secondary reactant is introduced within a period of time ranging from about 0 to about 45 minutes under 0.10 SLM, and includes TE0s mixed with He or Ar gas by storing in the monomer. Layer temperature (MRT; 50 ° C to 10.0 ° C of the secondary reactant monomer vapor phase back pressure (Bp) dominates the amount of secondary reactants in the gas stream. U Check by grinding the wafer The polishing characteristics of j6〇se polishing pads, these wafers' have a tungsten surface with a thickness of about 4000 Angstroms and a tantalum barrier layer with a thickness of about 250 Angstroms. By using a commercial abrasive (product number EP〇222, From Ebara 96108.doc 200525017
Technologies, Sacramento，CA)來分析鶴研磨特性。除非另 有說明，否則藉由使用基板之約25 kPa之下壓力（d〇Wn force)、約100至約250 rpm之台速（產品號MSW2000，來自 Rodel，Newark DE)來分析鎢研磨之移除率。使用一被調節 至pH值約為2之習知研磨漿（產品號MSW2000,來自R0del， Newark DE) 〇 圖5說明在不同時期後之塗覆有TEOS之基板之表面的 FTIR光譜。在一FTIR光譜儀（FTIR 1727, Perkin_Elmer系統 偵測器，其裝備有系列-I FTIR顯微鏡（MCT偵測器）並具有 自10,000至370 cm·1之光譜範圍）上獲得光譜。在約1〇1〇與 約950 cm·1處之訊號分別係基於矽石之不對稱的Si_〇_Si伸 展與石夕酸鹽之Si-Ο-Χ(其中X係指不為四面體組態之聚合的 -(Si-O-Si)n-結構）伸展。在85〇 cm·1處之訊號係歸因於自由 及締合的矽烷醇（Si-Ο-Η)。透過氫鍵結之矽烷醇締合物的締 合程度隨矽烷醇之表面濃度的增加而增加。 如圖6所說明，當塗覆時間增加至高達約30分鐘時，該等 訊號兩者皆單調地下降，此係歸因於Si_〇之表面濃度之淨 下降。其後，在沈積動力學與機制中存在一變化，其指示 增加之表面Si-Ο濃度。後者觀察與TE〇s沈積機制與動力學 之通常所接受之觀念係不一致的，且促進了對塗覆處理之 進一步研究，尤其係在30分鐘後之塗覆時期時。 奈米壓痕（nanoindentation)測試用於分析表面塗覆塗層 之機械特性’且更具體言之’用於量測彈性模數及硬度。 在不同時期内，於塗覆有研磨劑之熱塑性泡沫基板上進行 96108.doc 200525017 壓痕。NAN〇TEST 6_，位於高級材料與特徵工心Technologies, Sacramento, CA) to analyze crane grinding characteristics. Unless otherwise stated, the migration of tungsten grinding is analyzed by using a substrate pressure of about 25 kPa (d0Wn force) and a table speed of about 100 to about 250 rpm (product number MSW2000 from Rodel, Newark DE). Divide the rate. A conventional polishing slurry (product number MSW2000, from Rodel, Newark DE) adjusted to a pH of about 2 was used. Figure 5 illustrates the FTIR spectra of the surface of a TEOS-coated substrate after different periods. Spectra were obtained on a FTIR spectrometer (FTIR 1727, Perkin_Elmer system detector, equipped with a Series-I FTIR microscope (MCT detector) and having a spectral range from 10,000 to 370 cm · 1). The signals at about 1010 and about 950 cm · 1 are based on the asymmetric Si_〇_Si extension of silica and Si-O-X of oxalate (where X is not a tetrahedron). The configured polymeric- (Si-O-Si) n-structure) stretches. The signal at 85 cm · 1 is due to the free and associative silanol (Si-O-Η). The degree of association of the hydrogen-bonded silanol association increases with the increase of the surface concentration of silanol. As illustrated in FIG. 6, when the coating time is increased up to about 30 minutes, both of these signals decrease monotonically, which is due to the net decrease in the surface concentration of Si_0. Thereafter, there was a change in sedimentation kinetics and mechanism, which indicated an increased surface Si-O concentration. The latter observation is inconsistent with the generally accepted notion of TE0s deposition mechanism and kinetics, and has facilitated further research into coating processes, especially during the coating period after 30 minutes. The nanoindentation test is used to analyze the mechanical characteristics of the surface coating coating 'and more specifically to measure the elastic modulus and hardness. 96108.doc 200525017 indentation was performed on a thermoplastic foam substrate coated with an abrasive at different times. NAN〇TEST 6_, located in advanced materials and features
㈤and0，FL)之奈米M痕器（腿。lndenter)用於所有量測。機’ 器依靠於-振動隔絕臺上並被封閉於—溫控箱内。置放於 該箱前任一側上的兩個獨立加熱器提供一熱障壁。温控号 被設定至高於室溫約2或之值，具有± 〇rc之所預Z 穩定性。麼痕器被允許安置持續至少半小時以在開始試驗 前獲得熱穩定性。 值’藉由執行所塗覆之研磨塾的預備測試而建立壓痕參 數’如壓痕器之類型、最大深度及負載/卸载率。吾人^現 研磨塾表面含有約幾微米之不同尺寸的㈣（叫仙似微 孔。基於該等觀察，選擇尖端直徑為〜lmm之球體壓痕器， 使得該壓痕器將取樣足夠的墊材料。為了相同的原因，選 擇大於500微米之空間解析度。在初始負載為〇i mN(其= 一機器參數）之極低負載範圍下執行壓痕。控制參數被設定 至深度控制的，且各個墊被壓痕為不同深度，其最大深产 為10,000奈米。結果通常被檢查為1〇個壓痕之平均值。 Oliver-Pharr方法與Hertz方法兩者皆用於評估與驗證該等 結果。藉由使用Oliver-Pharr方法來分析自奈米壓痕器中所 獲得之負載-深度（p-h)曲線。 在壓痕試驗期間，研磨劑塗覆之研磨墊展示了非均勻之 穿透。視覺檢查壓痕（p-h)曲線顯示了 一些獨特特徵。如圖7 所示，自曲線中可辨別被標記為”pop-in”（Xb)、 或”kink-back”（Xc)之獨特事件。舉例而言，當壓痕器尖端 突然穿透至樣本中時，npop-in，，在壓縮週期期間發生。該等 96108.doc -18- 200525017 事件與諸如塗覆時間、貞載/卸鮮及壓痕深度之若干試驗 參數相關聯。由所塗覆之研磨墊所展示之混合回應展現 了 . pop-in事件更經常發生於壓痕器穿透深度為丨〇⑼奈米 左右時；”p〇P-out"似乎不受壓痕深度影響；且負載/卸載率 受该專事件兩者影響。 對各種深度及不同塗覆時間之p-h曲線的進一步分析展 現了負載曲線陡然地上升並下降，該χ座標或該過渡點之深 度被表示為Xa。類似地，乂!3與以係對應的乂座標或奈米深 度。壓痕器尖端至塗層中之該非均句穿透可能係由發生塑 料變形所引起的。吾人認為塑料變形係CMp墊之關鍵屬 性，其影響了 CMP處理之效率。因此，負載_深度曲線中之 上述事件被假設為墊效能之預測者。吾人發現初始表面穿 透事件（Xa)為PECVD塗覆時間、最大穿透深度與負載率之 函數。例如，圖8顯示Xa與TEOS塗覆時間之代表性相互關 係。數據暗示Xa與已藉由介電塗層而修改之表面泡沫的厚 度有關。 圖9與圖1〇分別說明研磨墊在不同TE〇s塗覆時間之藉由 使用Ohver-Pharr方法所計算的硬度與彈性模數。吾人發現 有效表面模數與硬度隨著塗覆時間的增加而增加。對於3〇 刀名里、40分|里與45分鐘之塗覆時間，研磨墊分別具有約65 KPa、62 KPa與70 Kpa之硬度值。對於較短之塗覆時間，硬 度值為60 KPa或更小。對於30分鐘、40分鐘與45分鐘之塗 覆時間，研磨墊分別具有約4MPa、5.5?^3與8.2]\^8之彈 性模數。對於較短之塗覆時間，彈性模數值為3 MPa或更小。 96108.doc -19- 200525017 墊表面之機械特性的變化歸因於沈積於泡沫基板上之塗 層的效應。該等數據指示了先前所說明之FTIR數據的不連 續性指示塾表面化學反應之變化，而非TE〇s衍生之塗層之 淨移除。 藉由使用商業設備對所塗覆之研磨墊樣本執行動力機械 刀析（DMA)，該商業設備在以下情況下操作：張力模式為 -125至200°C，頻率為1 Hz，振幅為10微米且程序化加熱率 為5°C/min。使用液體氮來達成次周圍溫度。在做出量測之 前，使樣本在預定之初始溫度下平衡1〇分鐘。將所有研磨 墊樣本製備成具有15 cm x5 cm之相同尺寸，且在DMA量測 之剷將其真空乾燥（3〇°C，在〜1 X 1〇_2 Torr下）24小時，以便 避免必須考慮濕氣效應。 如圖11中所說明’ DMA研究指示了在長pECVD塗覆時間 之損耗模數的突然變化。與對於1〇分鐘至4〇分鐘之塗覆時 間之自0.37至0.23範圍的值相比，在45分鐘之塗覆時間處， 損耗模數增加至約0.6 MPa。 此與PECVD塗覆僅修改基板表面之通常所接受之觀念相 反。孩令人吃驚之結果暗示了會發生在表面塗覆期間改變 泡沬基板之表體機械特性之其它處理。其假設熱塑性泡沫 基板中之剩餘反應物在PECVD塗覆期間以依賴時間之方式 反應。 藉由使用X-射線光電子光譜術（XPS)進一步表徵了經受 不同塗覆時間之研磨墊的表面修改。在1〇·1() 之基礎壓 力下刼作一商業χ_射線光電子光譜儀，且藉由使用金屬金 96108.doc -20- 200525017 標準（Au (4 π/2): 84.0 ± 0.1 eV)來校正該光譜儀。將能量為 1253 eV、功率為250 W之非單色Mg K ocX-射線源用於該分 析。藉由使用相對於285.0 eV處之外來碳線之氫部分之鍵結 能量而參照的鍵結能量尺度（energy scale)，移除由研磨塾 樣本所產生之充電移位（charging shift)。藉由使用商業軟體 執行峰值去卷積。 XPS分析提供了關於由TEOS吸附與***所引起之構形之 化學性質之若干見識。碳（Is)訊號被解析為三個主峰值：在 〜285.0 eV處之兩個峰值，其對應於C-C與C-H鍵結；及與在 〜286.5 eV處之所觀察之峰值’其對應於C-0鍵結。一居中 於〜289至〜289.3 eV之峰值歸因於尿素[-〇_C(NH2) =〇]官能 基’其來自用於熱塑性泡泳基板製造處理中之剩餘嘴吹 劑。對於分別被塗覆40與50分鐘之樣本而言，觀察到在 〜283.6 eV附近之另一峰值，且該峰值被試驗性地分配至 C-Si鍵結。 圖12呈現適合在以下TEOS塗覆時間之後自墊所獲得之㈤and0, FL) Nano M mark device (leg. Indenter) is used for all measurements. The machine's device relies on a vibration isolation table and is enclosed in a temperature control box. Two independent heaters placed on either side of the front of the box provide a thermal barrier. The temperature control number is set to a value of about 2 or higher than room temperature, and has a pre-Z stability of ± 〇rc. Modifiers were allowed to be installed for at least half an hour to obtain thermal stability before starting the test. The value ' establishes indentation parameters ' such as the type of indenter, maximum depth, and load / unload rate by performing a preliminary test of the coated abrasive. I am now grinding the surface of the millet with a few micrometers of millet (called immortal-like micropores. Based on these observations, a ball indenter with a tip diameter of ~ 1mm is selected so that the indenter will sample enough pad material For the same reason, choose a spatial resolution greater than 500 microns. Perform indentation in an extremely low load range with an initial load of 0 mN (which = a machine parameter). The control parameters are set to depth-controlled and each The pads are indented at different depths, with a maximum deep production of 10,000 nm. The results are usually checked as the average of 10 indentations. Both the Oliver-Pharr method and the Hertz method are used to evaluate and verify these results. The load-depth (ph) curve obtained from the nanoindenter was analyzed by using the Oliver-Pharr method. During the indentation test, the abrasive-coated abrasive pad showed non-uniform penetration. Visual inspection The indentation (ph) curve shows some unique characteristics. As shown in Figure 7, unique events marked as "pop-in" (Xb), or "kink-back" (Xc) can be discerned from the curve. For example, Say, when the creaser tip The npop-in, which suddenly penetrates into the sample, occurs during the compression cycle. These 96108.doc -18- 200525017 events are associated with several test parameters such as coating time, load / unload, and indentation depth The hybrid response demonstrated by the coated abrasive pads demonstrates. Pop-in events occur more often when the indenter penetration depth is around 丨 〇⑼nm; "p〇P-out " seems to be uncompressed The depth of the trace is affected; and the load / unload rate is affected by both the special event. Further analysis of the ph curves of various depths and different coating times shows that the load curve rises and falls abruptly, the depth of the χ coordinate or the transition point It is expressed as Xa. Similarly, 乂! 3 corresponds to the 乂 coordinate or nanometer depth. The penetration of the uneven sentence from the tip of the indenter to the coating may be caused by plastic deformation. I think plastic Deformation is a key attribute of CMP pads, which affects the efficiency of CMP processing. Therefore, the above events in the load-depth curve are assumed to be predictors of pad performance. We found that the initial surface penetration event (Xa) was when PECVD was applied , Maximum penetration depth, and load factor. For example, Figure 8 shows the representative correlation between Xa and TEOS coating time. The data suggests that Xa is related to the thickness of the surface foam that has been modified by the dielectric coating. Figure 9 The hardness and elastic modulus calculated by using the Ohver-Pharr method at different TE0s coating time of the polishing pad at different coating times are illustrated separately from FIG. 10. We have found that the effective surface modulus and hardness increase with the increase of the coating time. For a coating thickness of 30 kPa, 40 minutes, and 45 minutes, the polishing pad has hardness values of about 65 KPa, 62 KPa, and 70 Kpa, respectively. For shorter coating times, the hardness value is 60 KPa or less. For coating times of 30 minutes, 40 minutes, and 45 minutes, the polishing pads have elastic modulus of about 4 MPa, 5.5? ^ 3, and 8.2] \ ^ 8, respectively. For shorter coating times, the modulus of elasticity is 3 MPa or less. 96108.doc -19- 200525017 The change in the mechanical properties of the pad surface is due to the effect of the coating deposited on the foam substrate. These data indicate the discontinuity of the previously described FTIR data indicating changes in surface chemical reactions, not the net removal of TE0s-derived coatings. Powered mechanical sample analysis (DMA) was performed on the coated abrasive pad samples using a commercial device that operated under a tension mode of -125 to 200 ° C, a frequency of 1 Hz, and an amplitude of 10 microns And the programmed heating rate is 5 ° C / min. Use liquid nitrogen to achieve sub-ambient temperatures. Before making the measurement, the sample is allowed to equilibrate for 10 minutes at a predetermined initial temperature. All abrasive pad samples were prepared to have the same size of 15 cm x 5 cm and vacuum dried (30 ° C, ~ 1 X 1〇_2 Torr) on a DMA measuring shovel for 24 hours to avoid the need to Consider moisture effects. As illustrated in Figure 11, the 'DMA study indicates a sudden change in the loss modulus at long pECVD coating times. Compared to a value ranging from 0.37 to 0.23 for a coating time of 10 minutes to 40 minutes, the loss modulus increased to about 0.6 MPa at a coating time of 45 minutes. This is contrary to the commonly accepted concept that PECVD coating only modifies the substrate surface. The surprising results suggest that other treatments that alter the mechanical properties of the surface of the foam substrate during surface coating can occur. It assumes that the remaining reactants in the thermoplastic foam substrate react in a time-dependent manner during PECVD coating. Surface modification of abrasive pads subjected to different coating times was further characterized by the use of X-ray photoelectron spectroscopy (XPS). A commercial X-ray photoelectron spectrometer was fabricated at a base pressure of 10.1 (), and by using metal gold 96108.doc -20- 200525017 standard (Au (4 π / 2): 84.0 ± 0.1 eV) Calibrate the spectrometer. A non-monochromatic Mg K ocX-ray source with an energy of 1253 eV and a power of 250 W was used for the analysis. By using a bonding energy scale that is referenced to the bonding energy of the hydrogen portion of the extraneous carbon wire at 285.0 eV, the charging shift generated by the milled radon sample is removed. Perform peak deconvolution by using commercial software. XPS analysis provides several insights into the chemical properties of the configuration caused by TEOS adsorption and fragmentation. The carbon (Is) signal is interpreted as three main peaks: two peaks at ~ 285.0 eV, which correspond to CC and CH bonding; and the peaks observed at ~ 286.5 eV, which correspond to C- 0 bond. The one-centered peak at ~ 289 to ~ 289.3 eV is attributed to the urea [-〇_C (NH2) = 〇] functional group 'which is derived from the remaining mouth blowing agent used in the manufacturing process of the thermoplastic bubble substrate. For the samples coated for 40 and 50 minutes respectively, another peak was observed around ~ 283.6 eV, and this peak was experimentally assigned to the C-Si bond. Figure 12 presents suitable results obtained from the pad after the following TEOS coating times
Si(2p)包裝之XPS目fl號的例不性峰值：（a)l〇 min、（b)20 min、（c)30 min、（d)40 min 及（e)45 min。峰值被識別為： (l)Si-O、（2)矽酸鹽、（3)Si-N及（4)Si-C鍵結。各光譜被去卷 積成在10 2.3與〜1 〇 3 · 4 e V處之兩個主峰值，其分別對靡、於 矽酸鹽與Si-Ο物質中之鍵結。 該等數據指示了對於短塗覆時間（如，小於〜30分鐘）而 言，墊表面富含矽烧醇，與在低處理溫度下沈積之TE0S薄 膜一致。如圖13中之進一步說明，自XPS數據計算之氧-矽 96108.doc -21- 200525017 強度比率早期在塗覆期間為高的，其指示了在沈積塗層中 之矽烷醇的濃度為高的。矽烷醇濃度隨著塗覆時間達到 分鐘而降低，接著開始增加。舉例而言，如圖9所示，在分 別為10 mm、20 min、3〇 min、4〇 _與45論之塗覆時間 後 ’ 〇:Si比率等於約 7.4、4.8、3.6、7·1 與 9·9。 再次轉向圖^’對於儿〜扣與“㈤沁之塗覆時間’於〜丨⑽」 ev處觀察到小峰值，其對應於Si_N鍵結。在此同一時期内， 在Si-N比率中存在突然減少，其指示了表面上之氮物質的 增加。 該等觀察暗示了基於PECVD之塗覆涉及了若干處理間之 競爭。基於PECVD之塗覆於泡沫表面上之（31(^與以〇2)上產 生了矽石與矽酸鹽兩者。此外，基板之表面化學反應作為 塗覆時間之函數而改變。對於在3〇分鐘以下之塗覆時間， 存在來自對表面之Ar離子轟擊之沈積物的淨蝕刻。樣本亦 自電漿加熱，因此熱處理亦會發生。當塗覆時間增加，基 板表面上之矽酸鹽含量開始降低，且墊變得更密集，以便 增加該墊之硬度。 對於30分鐘或更長之塗覆時間，基板溫度為足夠高以導 致用於使基板發泡或分解任何留在泡沫中之剩餘噴吹劑之 氮氣體之除氣以產生氮氣體。氮在塾表面上之氣相或Si物 質中與含Si中間體進行反應，以在費用si〇2的情況下形成 諸如SisNU之氮化物。以高達1〇莫耳％之濃度將該等氮化物 併入研磨劑中。此外，對於該等塗覆時間而言，對泡來表 面之離子轟擊在墊表面上產生了可感知之碳基量。該等基 96108.doc -22- 200525017 與石夕物質進行反應以形成碳化物，如碳化石夕（Sic)，以高達 1 0莫耳％之濃度將其併入研磨劑中。 圖14比較經受不同塗覆時間時期之具有TEOS之熱塑性 泡沫基板之相對毯覆性鎢移除率（W-RR)與靜態摩擦係數 (COF)。W-RR與COF兩者都隨著塗覆時間增加至3〇分鐘而 增加，其表示研磨劑之厚度的增加。對於在3〇與6〇分鐘之 間的塗覆時間而言，W-RR與COF都下降，並接著增加。該 等結果暗示了墊藉由一用於塗覆了高達3〇分鐘之表面之機 制及藉由一用於塗覆了大於30分鐘之表面之不同機制而出 現研磨，此係歸因於表面微機械與化學反應中之差異。 【圖式簡單說明】 圖1說明本發明之研磨墊的橫截面圖； 圖2-4說明本發明之用於製備研磨墊之方法中所選步驟 的橫截面圖； 圖5呈現在各種時期後之塗覆有包括四乙氧基矽烷 (TEOS)之研磨劑前驅物之熱塑性泡沫研磨墊之樣本的代表 性近紅外光譜； 圖6說明曝露於不同塗覆時期之具有te〇s之代表性熱塑 性泡殊研磨塾之近紅外訊號的變化； 圖7說明熱塑性泡沫研磨墊在被塗覆有TE〇s後之例示性 鋸齒狀曲線； 圖8說明具有TEOS之熱塑性泡沫研磨墊之（Xa)”p〇p-in，， 作為塗覆時間之函數的代表性變化； 圖9說明具有TEOS之熱塑性泡沫研磨墊之硬度作為塗覆 96108.doc -23- 200525017 時間之函數的代表性變化； 、圖10 #明具有TEOS之熱塑性泡沫研磨墊之彈性模數作 為塗覆時間之函數的代表性變化； 圖11說明具有TEOS之熱塑性泡沫研磨墊之儲存與損耗 模數作為塗覆時間之函數的代表性變化； 圖12呈現在各種塗覆時間時期後之具有TE〇s之研磨墊 的代表性XPS光譜； 圖1 3呈現自在各種塗覆時間時期後之具有TE〇s之研磨 墊之XPS光譜中計算之氧-矽強度比值的變化；且 圖14呈現具有te〇S之熱塑性泡沫墊研磨墊之相對毯覆 性鶴移除率（WRR)與靜態摩擦係數（c〇F)作為塗覆時間之 函數的變化。 【主要元件符號說明】 100 研磨塾 110 研磨體 115 熱塑性泡沫基板 120 表面 125 凹面泡孔 130 研磨劑 135 内表面/凹面泡孔 140 封閉泡孔 145 背板材料 150 黏著劑 200 研磨墊 96108.doc -24- 200525017 210 220 230 240 310 320 330 410 封閉泡孔 熱塑性泡沫基板 基板表面/熱塑性泡沫 凹面泡孔/熱塑性泡沫基板 ' 修改的表面 . 研磨劑 内表面 凹面泡孔/研磨墊表面Examples of peaks in the case of Si (2p) packaging XPS mesh fl: (a) 10 min, (b) 20 min, (c) 30 min, (d) 40 min, and (e) 45 min. The peaks were identified as: (1) Si-O, (2) silicate, (3) Si-N, and (4) Si-C bonding. Each spectrum is deconvolved into two main peaks at 10 2.3 and ~ 10 3 · 4 e V, which are respectively bonded to the silicate and the Si-O substance. These data indicate that for short coating times (eg, less than ~ 30 minutes), the surface of the pad is rich in silanol, which is consistent with TEOS films deposited at low processing temperatures. As further illustrated in FIG. 13, the oxygen-silicon 96108.doc -21- 200525017 strength ratio calculated from the XPS data was early during the coating period, which indicates that the silanol concentration in the deposition coating was high. . The silanol concentration decreases as the coating time reaches minutes, and then begins to increase. For example, as shown in FIG. 9, after the coating times of 10 mm, 20 min, 30 min, 40 °, and 45 °, respectively, the ratio of 〇: Si is equal to about 7.4, 4.8, 3.6, and 7.1. With 9 · 9. Turning again to the figure ^ 'For the child ~ button and the "coating time of" ㈤ 沁 "at ~ 丨 ⑽" a small peak is observed at ev, which corresponds to the Si_N bond. During this same period, there was a sudden decrease in the Si-N ratio, which was indicative of an increase in nitrogen species on the surface. These observations suggest that PECVD-based coating involves competition between several processes. PECVD-based coatings on the foam surface (31 (^ and 〇2) produced both silica and silicate. In addition, the chemical reaction of the substrate surface changes as a function of coating time. For coating times below 0 minutes, there is a net etch from deposits bombarded with Ar ions on the surface. The sample is also heated from the plasma, so heat treatment will also occur. When the coating time increases, the silicate content on the substrate surface Begins to decrease, and the pad becomes denser in order to increase the hardness of the pad. For coating times of 30 minutes or more, the substrate temperature is high enough to cause the substrate to foam or decompose any remaining left in the foam The degassing of the nitrogen gas of the blowing agent produces nitrogen gas. The nitrogen reacts with the Si-containing intermediate in a gas phase or a Si material on the surface of the rhenium to form a nitride such as SisNU at a cost of SiO2. These nitrides are incorporated into the abrasive at a concentration of up to 10 mole%. In addition, for these coating times, ion bombardment of the bubbled surface produces a appreciable amount of carbon based on the surface of the pad .Base 961 08.doc -22- 200525017 reacts with Shi Xi substances to form carbides, such as Carbide Si (Sic), which is incorporated into the abrasive at a concentration of up to 10 mole%. Figure 14 compares different coatings Relative blanket removal of tungsten (W-RR) and static coefficient of friction (COF) of a thermoplastic foam substrate with TEOS over time. Both W-RR and COF increase with coating time to 30 minutes. Increasing, which indicates an increase in the thickness of the abrasive. For coating times between 30 and 60 minutes, both W-RR and COF decrease, and then increase. These results suggest that the pads are used by Grinding occurs on a mechanism that coats surfaces for up to 30 minutes and by a different mechanism for coating surfaces that are longer than 30 minutes, due to differences in surface micromechanical and chemical reactions. [Figure] Brief description of the formula] Figure 1 illustrates a cross-sectional view of a polishing pad of the present invention; Figures 2-4 illustrate cross-sectional views of selected steps in the method for preparing a polishing pad of the present invention; Figure 5 presents coating after various periods Thermoplastic coated with abrasive precursors including tetraethoxysilane (TEOS) Representative near-infrared spectra of samples of foamed foam polishing pads; Figure 6 illustrates changes in the near-infrared signal of a representative thermoplastic foam polishing pad with te0s exposed to different coating periods; Figure 7 illustrates the thermoplastic foam polishing pads in Exemplary zigzag curve after being coated with TE0s; Figure 8 illustrates (Xa) "p0p-in, a thermoplastic foam abrasive pad with TEOS, as a representative change as a function of coating time; Figure 9 Explain the representative change in hardness of a thermoplastic foam abrasive pad with TEOS as a function of coating time 96108.doc -23- 200525017; Figure 10 #Ming out the elastic modulus of a thermoplastic foam abrasive pad with TEOS as a function of coating time Figure 11 illustrates representative changes in storage and loss modulus of a thermoplastic foam abrasive pad with TEOS as a function of coating time. Figure 12 presents polishing pads with TE0 after various coating time periods. Representative XPS spectrum of FIG .; FIG. 13 shows the change of the oxygen-silicon intensity ratio calculated from the XPS spectrum of a polishing pad with TE0s after various coating time periods; and FIG. 14 presents a graph with Changes in Relative Blanket Crane Removal Rate (WRR) and Static Coefficient of Friction (coF) of a thermoplastic foam pad abrasive pad with te0S as a function of coating time. [Description of main component symbols] 100 grinding 110 grinding body 115 thermoplastic foam substrate 120 surface 125 concave cell 130 abrasive 135 inner surface / concave cell 140 closed cell 145 back sheet material 150 adhesive 200 polishing pad 96108.doc- 24- 200525017 210 220 230 240 310 320 330 410 Surface of closed cell thermoplastic foam substrate substrate / thermoplastic foam concave cells / thermoplastic foam substrate 'modified surface. Abrasive inner surface concave cells / pad surface
96108.doc -25-96108.doc -25-