201137513 35147pif 六、發明説明： 【發明所屬之技術領域】 本發明是關於空白遮罩、光罩及其製造方法，其使得 可能在半導體微影製程（semiconductor lithography process)中達成高精度臨界尺寸（critical dimension，CD)， 且更特定而言，是關於空白遮罩、光罩及其製造方法，其 可應用於ArF (波長：i93nm)微影及ArF浸沒微影 (immersion lUhography)以達成小於65 nm之臨界尺寸。 【先前技術】 目前，高級半導體微處理技術對於滿足使高度整合半 導體積體電路所需之電路圖案之尺寸減小的需求變得日益 重要。對於高度積體電路’要求電路互連非常小以獲得低 電壓及高速操作’且對用於層間連接（interlayer connection)之接觸霍爾圖案（contact hall pattern)及對整 合所需之電路佈局做出之技術需求變得愈來愈高。因此， 為滿足此等需求，需要可使得能夠形成非常小的電路圖案 且減小製造形成有電路圖案之光罩的規模的技術。 特疋而5，為準確形成小於65 nm之超精度電路圖 案’目前正在開發空白遮罩結構，其中在不使用先前技術 抗姓劑膜作為侧遮罩的mu由料侧遮罩之無機 硬^膜_形成最終圖案之金屬膜。在用於硬遮罩用途 之空白遮罩t ’硬遮罩膜代替先前技術厚抗細膜，且金 屬膜之下β部分之縱橫比（aspeet rati。）及選擇性可由此 硬遮罩膜改良。因此，在金屬膜之乾式姓刻（吻⑽㈣） 4 201137513. 期間負載效應（loading effect)減小，使得可獲得優良的 CD平均值比目標值（mean t0 target，MTT)、CD線性度 和CD均^~性。 同時，在用於硬磁盤用途之先前技術中，其一般而言 具有一結構，其中在深度方向上具有均一組成比率的光屏 蔽膜、金屬膜、硬遮罩膜以及抗蝕劑膜依序形成於6〇25 尺寸合成石英玻璃基板上。此時，金屬膜可包含抗反射膜， 而硬遮罩膜用作蝕刻遮罩。在深度方向上薄膜之均一組成 比率意謂其在相同總體製程條件及製程環境下形成，諸如 薄膜形成期間的濺鍍靶（taget)、處理時間、功率、壓力、 反應氣體、惰性氣體類型以及輸入量。此外，薄膜之逐步 形成意謂其在用於控制每一薄膜之特性的不同製程條件下 沈積。即，其意謂薄膜相對於基板之減鑛製程條件是在不 連續步驟中實施，以便以逐步方式形成薄膜。 在其中組成比率在以逐步方式形成之薄膜的深度方 向上均一的薄膜製程中，每一層具有獨立執行其功能=優 點，因為製程改變在每一步驟中實施。舉例而言，為著重 於屏蔽具有用於光屏蔽膜之特定波長之暴露光的功能以及 減小抗反射膜之對暴露光之表面反射率的功能，僅需兩個 處理條件步驟。因此，存在形成具有兩層結構之金屬膜的 優點。 另一方面，由於嘗試更進一步減小規模以使得可達成 小於65 nm (尤其小於45 nm)之電路圖案，因此除針對 暴露光之光屏蔽功能及抗反射功能外，克服薄膜之均一 5 201137513 35147pif 性、化學耐久性特性、表面粗糙度、殘餘應力（阳4^ stress)、生長缺陷、CD均一性以及缺陷所面臨之問題的要 求亦變得更加嚴格。在大於65nm之圖案之達成期間，所 述效應在以上問題方面與待達成之臨界尺寸相比不明顯， 且不發生顯著問題，因為諸如365 nm i_線及248 之暴露光之波長為相對長的波長。然而，在使用諸如193 nm ArF之短波長以便達成小於65 nm之臨界尺寸的超精度 圖案形成期間’由藉由逐步方式形成之薄膜之現存殘餘應 力、薄膜之均一性以及薄膜密度產生的缺陷變為較顯著問 題，其隨著待達成之CD變得極小而需解決。 然而，難以解決以上問題，因為在藉由使用由逐步方 式形成之薄膜嘗試解決例如上述之複雜問題時，組成比率 之突然改變、薄膜密度之改變以及殘餘應力之改變經常發 生於薄膜之界面處。此外，為滿足光學密度之均一性、反 射率之均一性、化學耐久性、表面粗糙度及生長缺陷特性， 需要沈積另一薄膜之複雜製造製程。此複雜製程不僅致使 產生諸如粒子之缺陷，而且負載效應由於薄膜厚度之總體 增加而增加’使得減小解析度、CD、MTT以及CD均一 性的問題將發生。 【發明内容】 本發明提供一種二元空白遮罩、用於硬遮罩用途之空 白遮罩’以及用於ArF微影及ArF浸沒微影之相移空白遮 罩，其關於均一反射率、均一光學密度、薄膜之間的黏著、 殘餘應力、生長缺陷、化學耐久性、暴露耐久性以及表面 6 201137513 粗縫度均具有良好薄膜性質。 根據一例示性實施例，提供一種空白遮罩，其包含以 所陳述次序依序堆疊的透明基板、金屬膜、硬遮罩膜以及 抗蝕劑膜，其中所述空白遮罩具有在所述金屬膜及所述硬 遮罩膜中之至少一者中的至少一可變组成區，其中構成所 述膜之多個元素中之至少一元素的組成比率在所述膜之深 度方向上連續改變。 根據另一例示性實施例，提供一種光罩，藉由圖案化 及蝕刻空白遮罩獲得，所述空白遮罩包含：以所陳述次序 依序堆疊的透明基板、金屬膜、硬遮罩膜以及抗姓劑膜， 其中所述空白遮罩具有在所述金屬膜及所述硬遮罩膜中之 至少一者中的至少一可變組成區，其中構成所述膜之多個 元素中之至少一元素的組成比率在所述膜之深度方向上連 續改變。 根據另一例示性實施例，提供一種空白遮罩的製造方 法，藉由在透明基板上依序形成金屬膜、硬遮罩膜以及抗 蝕劑膜來製造，所述方法更包含：a)在維持第一沈積時間 中施加之相同製程條件的同時，在所述透明基板上沈積所 述金屬膜；以及b)形成可變組成區，其中藉由在反應氣 體仍接通的同時，以逐步或連續方式改變在第二沈積時間 中施加之許多製程條件中的至少一者，構成所述金屬膜之 多個元素中之至少一者的組成比率在所述金屬膜之深度方 向上連續改變，其中a)及b)實施之次序可改變，或a) 及b)可以交替方式實施。 201137513 35147pif 根據另一例示性實施例，提供一種空白遮罩的製造方 法’藉由在透明基板上依序形成金屬膜、硬遮罩膜以及抗 餘劑膜來製造，所述方法更包含：a)在所述透明基板上沈 積所述金屬膜；以及b)形成可變組成區，其中藉由在反 應氣體仍接通的同時，以逐步或連續方式改變在第二沈積 時間中施加之許多製程條件中的至少一者’構成所述金屬 膜之多個元素中之至少一者的組成比率在所述金屬膜之深 度方向上連續改變。 依據根據本發明之空白遮罩、光罩及其製造方洚，所 述空白遮罩可藉由在製造製程期間在薄膜之深度方向上選 擇性地且連續改變構成薄膜之至少多於一種的元素的含量 來改良薄膜之黏著、反射率均一性、化學耐久性以及殘餘 應力特性。另外’本發明使得具有優良性質之空白遮罩之 製造成為可能，且因此亦可達成具有優良性質之光罩的製 造。 【實施方式】 下文中將參考附圖詳細描述特定實施例。 圖1為根據例示性實施例之用於硬遮罩用途之空白遮 罩的截面圖。 參見圖1，根據本發明之用於硬遮罩用途之空白遮罩 100包含透明基板110、金屬膜12〇、硬遮罩膜130以及抗 钮劑膜140。透明基板11〇、金屬膜12〇、硬遮罩膜130以 及抗蝕劑膜140藉由依序堆疊而形成。此處，金屬膜及/ 或硬遮罩膜具有一區（下文稱為「可變組成區」），其中在 8 201137513 構成膜之元素中至少一元素之組成比率在膜之深度方向上 在自約1 at0/。至約50 at%之範圍内連續改變。 另外，其中在構成膜之元素中至少一元素之組成比率 在膜之深度方向上連續改變的可變組成區意謂用於控制每 薄膜之特性的製程條件是在不停止反應氣體是情況下實 施以包含以連續或逐步方式改變之多於一個的可變組成 區。亦即，為形成在薄膜之深度方向上改變的可變組成區， 在基板之薄膜之濺鑛期間以連續或逐步方式改變製程條 件0 “透明基板U0是藉由相對於基板母體材料執行多個拋 光（lapping)製程及多個研磨（p〇iishing)製程而形成。 用於透明基板110之材料可為選自合成石英、氟化鈣 (CaF2)及氟摻雜石英（F摻雜石英）中之一者。透明基板 110之尺寸為6025，且較佳193 nm下之雙折射 (birefringence)小於約2 nm/6.35 mm »此時，基板母體材 料由具有99.9999%以上純度之氧化矽（Si〇2)組成，且藉 由對合成石英錠進行切片（slicing)及邊緣碾磨（edge grinding)來製造以獲得152χ152±〇2賴之尺寸及約 mm以上之厚度。 透明基板110是藉由相對於基板母體材料執行多個拋 光製程及多個研磨製程而形成。首先，相對於具有 152χ152±0·2 mm之尺寸及約6·3 mm以上之厚度的基板母 體材料執行多個拋光製程。拋光製程是在執行一次拋光製 程的情況下鑒於製程效率使用相對大尺寸研磨粒在高壓下 201137513 35l47pif 執^。在此情況下，可容易達成減小厚度之目的，但包含 在冰度方向上自基板表面產生之裂縫的損壞發生。因此， 觀察到基板母騎料巾内部產生之裂縫作為後處理中之缺 陷。而且，由於_研練的作用，難叫成厚度準確性 目的。因此，較佳本發明中使用之透明基板11〇是藉 行多麵光餘來S造，以便相對絲板母歸料 陷減少及厚度準雜目的。在基板母體材料之拋光製程中 施加之研磨粒為選自碳化梦（sic)、金剛石（c)、氧化錯 (Zr02).和氧化|g (Al2〇3)之多於—個的研磨粒，且 粒子在多健紐料絲。此外，餘研錄之尺寸在 4μιη與20μιη之間。若研磨粒之尺寸小於約4μιη，則可容 易達成厚度準破性目#，但隨著處理時間增加，生產 小。若研磨粒之尺寸大於約2〇 μηι，則不可達成厚度 性目的，且缺陷水平減小。 隨後，相對於已被執行多個拋光製程之基板母體材料 執行多個研磨製程。在此多個研磨製程中使用之研漿 (slurry)包含氧化鈽（Ce〇2)、㈣二氧化$ (Si〇2)研磨 粒以及過氧化氫（H202)，且PH由硝酸（HN〇3)或氫氧 化斜（KOH)控制。較佳磨钱研漿之總體pH在研磨製程 中控制於6與η之間。特定而言，諸如氫氧化_ (κ 之無機鹼對基板母體材料具有蝕刻作用，進而具有展示综 效作用（synergy effect)之優點。同時，研磨粒起到在研 磨製程中實體移除基板母體材料的作用。此時，若研磨粒 之尺寸大於約5 μιη，則容易達成厚度減小目的，但可能難 201137513 JWpif 以獲得良好表面粗糙度。而且，由於在研磨粒之尺寸小於 約0.5 μιη的情況下經過拋光製程之基板母體材料之厚度 減小量較小，因此需要長處理時間，使得處理效率可降低t 因此’較佳氧化鈽研磨粒之尺寸在0.5 4„!與5 μιη之間。 此外，較佳在多個研磨製程中使用之膠態二氧化矽研磨粒 之尺寸在20 nm與200 nm之間。此時，若膠態二氧化矽 研磨粒之尺寸小於約20 nm，則研磨製程效率降低，且若 所述尺寸大於約200 nm,則在研磨製程中應達成之表面粗 糙度可能不令人滿意。 由於研磨製程是用於改良表面粗糙度之製程，因此較 佳在執行研磨製程時逐漸使用較小研磨粒。由於表面狀態 因拋光製程而相對粗糙，因此在第一研磨製程中與第二及 第三研磨製程相比應移除大量基板母體材料。因此，將具 有馬硬度及低可壓縮性之多孔錦塾（p0rous eerjum pad)用 作第一研磨製程中之研磨墊’且較佳在第二研磨製程中使 用SUBA#400至800作為軟墊。此時，研磨墊決定將移除 之基板母體材料的量’且表面狀態取決於所使用之墊的類 型。若使用低於#400之研磨墊，則處理時間由於相對高的 可壓縮性及彈性恢復率（elastic recovery rate )以及低硬度 特性而增加，使得研磨效率可降低。若使用高於#800之研 磨墊，則由於相對低的可壓縮性及彈性恢復率以及高硬度 特性而難以達成表面粗糙度目的。而且，較佳在第二研磨 製程中使用之軟墊之可壓縮性為約3%以上，且彈性恢復 率為約65%以上。在具有約3%以上可壓縮性之研磨墊中， 201137513 35l47pif 由於圍繞粒子之起毛層（nap layer)在研磨壓力施加於二 ，化矽粒子時彈性變形，因此研磨壓力經分散且吸收，使 知可在基板表面上發生之凹形缺陷的產生受到抑制。同 時，在具有超過65%之高彈性恢復率之研磨墊中，起毛層 可容易經壓縮及恢復，使得存在藉由在起毛層中不產生大 的剩餘二氧化梦粒子而抑制凹形缺陷的優點。且，在第三 研磨製程中將作為超軟墊之麂皮墊（suedepad)用作研磨 墊。由於第三研磨製程是用於最大地改良表面粗糙度及粒 子特性的製程，因此使用其中硬度較低且彈性恢復率及可 壓縮性相對大的超軟研磨墊。較佳在第三研磨製程中使用 之超軟研磨墊之可壓縮性大於約6%，且彈性恢復率大於 約72%。由於在第三研磨製程中應較嚴格控制缺陷，因此 應使用具有比第二研磨塾大的可壓縮性及彈性恢復率的研 磨墊。 同時’在研磨製程中使用之研磨墊中的凹槽可具有各 種形狀。作為一實例，可使用帶有具有25 mm間距、4 mm 寬度及0.5 mm深度之形狀之凹槽的研磨墊。研磨墊之凹 槽在研磨製程期間向透明基板110供應足夠量的研漿，使 得其起到增加透明基板110之研磨效率的作用。此時，可 能使用未形成有凹槽之研磨塾。凹槽之尺寸可取決於研磨 製程而改變，且是否使用形成有凹槽之研磨墊亦可取決於 研磨製程而選擇性地判定。此外，在研磨製程中使用之研 磨墊具有帶有兩個以上層之結構的情況下，較佳自精度表 面板方向研磨墊之對應於第二層之起毛層的厚度在約2〇〇 12 201137513 JDl^/pif μιη與約600 μπι之間。若起毛層之厚度小於約2〇〇 ，則 研磨墊之彈性恢復率減小，使得難以保證表面粗糙度，且 可具有相對於產生包含粒子之缺陷的不利效果。若起毛層 大於約600 μιη，則用於保證表面粗糙度之研磨效率減小。 金屬膜120形成於透明基板η。上且可作為單一芦曳 多層而形成。單-層或多層中之多於—層的層形成有二^ 變組成區，其中構成所述層之多於—個的元素之組成比率 在深度方向上連續改變。此時，在可變組成區中至少多 於一個的元素之組成比率在自約i at%至約5〇时％之範圍 内改變，且較佳可變組成區之厚度大於或等於5 nm。本發 明中揭露之金屬膜12〇為自透明基板11〇至形成有最終圖 案之表面形成之多層結構的薄膜，使得其藉由在最終圖案 以下之部分中包含所有薄膜（諸祕刻停止膜、應力減小 膜、光屏蔽膜以及抗反射膜）來指定，此外，相移膜（phase shifting film )可選擇性地形成於透明基板丨丨〇與金屬膜12〇 之間。因此，將其中形成相移膜之空白遮罩稱為半色調相 移（half-tone phase shift)空白遮罩。 而且’構成金屬膜120之元素之組成比率中氮及/或矽 之組成比率連續改變為必須的，且較佳組成比率改變之區 的厚度大於1 nm。氮起到不僅影響薄膜之諸如透射率及反 射率之光學性質而且影響化學耐久性及暴露耐久性的主要 因素的作用。舉例而言，若氮含量較高，則薄膜表面上之 反射率相對於薄膜之光學特性形成為較低。另一方面，若 氮含量較低’則薄膜表面上之反射率形成為較高。此外， 13 201137513 35147pif 響耐化學性㈣，謂面上對sc] 时化學性性質隨著I含量增加而劣化，且 ，之 ^哲1及硫酸之耐化學性性質隨著氮含量減小而展現優， f質。因此，触設計金屬層⑽赌得取決於 ^ =深度方向氮含量具有彼此不同的分佈圖以便考慮= =此時’由於在可變組成區之厚度小於約i邮的情^ 下難·得如絲提及之薄膜性f之有意義的改變/ ==具有多於一個的如下的區，其中氮含量連續。改變 的所迷區之厚度大於約1 nm。 變 同時，在本質上包含碎之金屬膜中，較佳具有 有矽兀素之可變組成區之厚度大於約5 nm的至少二區了 而且’在祕硬遮罩用途之空自遮罩中，重要的是金 厚度減小’且選擇具有優良魏學性性質之材料以 便減> 貞載效應。此時，相對於金屬膜12()之料 光學及耐化學性特性，且隨著她成比率之_增加^ 射率增加且耐化學性性質變得優良。此時，由於藉 金屬膜12G表面中之化學組成之性質而獲得耐化學性°，因 ^面t之石夕含量應較高。另一方面，若在比較具有相同 厚度之、賴_含4較高，财料增加，錄薄膜之厚 度將最終增加。因此，騎強以上條件，本質上需要其 發含量之組成㈣連續改變之可變組成區，且較佳此區 金屬膜120之深度方向上的厚度大於約5腿，更佳大 10 nm。 此外，組成比率之分佈在金屬膜12〇之寬度方向上均 201137513 且特定’較佳組献率之均-性在金屬膜之 Ϊ度方向上小於約跳，且在金屬膜12G中之薄膜深度方 向上之密度改變在G.2 g/em、2 G g/em3之間^此時，组 成比率之均-性可藉下等式計算，且測得位置在寬度 方向上為至少5個點以上^，薄膜之組成比率可藉由諸 如AES、XPS以及RBS之方法來分析。 [等式1] 組祕-种最大組成比率·最小組成比率Ϊ 一I '---X 100 (最大組成比率+最小組成比率丨 3同時，在薄膜之深度方向上之密度改變小於約 g/cm的情況下’效果不顯著，其類似於薄膜之密度不改變 的狀態’且存在的問題H薄動之應力在密度改變大於 約2.0g/cm3的情況下彼此顯著不同。而且，金屬膜12〇在 深度方向上具有不同的殘餘應力，且較佳包含其中殘餘應 力差大於約10 MPa的區。此時，金屬膜丨2〇包含其中多 於二個的元素之組成比率在其深度方向上連續改變的區， 特定而言較佳選擇性地包含碳以便減少膜中 此外，在金屬膜120之表面中形成可變組；=況 下，可將表面粗糙度控制為小於約l.〇nmRa。金屬膜12〇 之表面粗糙度由基板之表面粗糙度狀態、濺鍍之元素的種 類、反應氣體及製程條件決定。此時，可藉由以連續或逐 步方式改變反應氣體類型、沈積功率、腔室墨力以及在電 15 201137513 J3W/plf 浆仍接通時弓至腔室中以形成可變組成區（其中構 面或金屬膜120内之元素中的至少一元素之組成比率在膜 之深度方向上連續改變）之氣體之流動速率中的至少 f控制可變組成區之表面⑽度。此外，較佳氮原子之比 率與深度相比在金屬膜12〇之表面令相對高。此外，對於 ^屬膜U0而言較佳氮之組成比率高於在深度方向上在距 表面5 mn之厚度範圍内的氮之組成比率。在此情況下， 較佳193 nm下之反射率均一性將藉由特定而言相對於深 度方向使最外表面之氮含量分佈為相對高而在1%内。 一而且，在金屬膜120具有其中彼此不同之多於一個的 元素改變之多層結構的情況下，較佳在構成金屬膜之 相應薄膜中彼此鄰接之薄膜中相同元素中之多於—者大於 約5 nm之厚度連續形成，且在彼此鄰接之薄膜之界面處 組成比率在1 at%與50 at%之間改變。在用於硬遮罩用途 之ί白遮罩中’金屬膜12G為在最終®案化之後在表面上 暴露的膜。此時’由於在金屬膜具有均—之元素組成比率 的情況下至少包含屏絲之光屏蔽膜以及減少光反射之抗 反射膜’因此由於在沈積光屏蔽膜之後抗反射膜之沈積引 起的元素組成比率之突然改變，殘餘應力在界面處大量地 發生。此外，若逐步形成之薄膜界面（光屏蔽膜及抗反射 膜）之元素組成彼此不同，則薄膜密度之改變發生，且由 於姓刻比率在乾式蝕刻期間快速不同而將產生頸縮 (necking)問題。此頸縮不引起問題，因為由頸縮產生之 CD誤差在大於約65nm CD中在與待達成之圖案cd相比 201137513 時不顯著，但由於此缺陷的緣故而在達成小於約65nmCD 之圖案時可能發生。 在本發明中，由於可變組成區形成於表面上或膜内， 因此由組成比率及薄膜密度之突然改變引起之問題可減 少。此時，當與由逐步方式形成之習知薄膜比較時，薄膜 應力及乾式飯刻比率之改變在形成薄膜之元素之改變比率 在約1 at%内的情況下將較小。另一方面，若形成薄膜之 元素之改變比率大於約50 at%，則材料之間的殘餘應力快 速改變，使得在以逐步方式形成薄膜時所見之相同問題將 發生。因此，較佳形成至少一可變組成區，其中金屬膜12〇 内構成元素中多於一個的元素之改變比率在i故％與5〇 at%之間（更佳地，3 at%與30 at%之間）。 此外，隨著用於達成之圖案尺寸變得愈加小，所產生 之頸縮及應力將較大。因此，為解決此等問題，當形成金 屬膜120時形成可變組成區，使得由頸縮及應力引起之缺 陷可減少《此外，諸如耐化學性、應力及薄膜密度之性質 可改變，且因此可能製造在表面處具有優良薄膜性質之空 白遮罩。 同時，對於金屬膜121而言較佳在深度方向上蝕刻速 率之改變在至少多於一個的區段中在與表面部分比較時較 快。較佳蝕刻比率在薄膜之深度方向上增加，以便減少在 金屬膜120之乾式蝕刻期間產生的負載效應。此是由於自 由基離子（radical ion)與經蝕刻金屬膜之間的反應性在進 入深度方向時減小的事實（因為薄膜之厚度是在深度方向 17 201137513 35147pif 上）。因此，較佳蝕刻比率在進入深度方向時增加，以便形 成垂直圖案。可麟由在形成可變組成區雜職（n)、 氧（〇)以及碳（C)之組成比率來控制侧比率較快或較 慢。 下表中描述光學、物理及化學性質，諸如包含如上提 及之可變組之金屬膜12〇之厚度、表面處之光學密 度、耐化學性、殘餘應力以及平坦性。 [表1] 性質 -- 金屬膜之厚度 ~300 jT6O〇A 表面處之光學密度 針對具有193 nm波長之暴露光在約2.7與約 3.5之間 对化學性 (氨水：過氧化氫溶液：超純水= 1:1:5)中浸泡2小時之後，針對193 nm之 暴露光的反射率改變在約1%内》 殘餘應力 約100 MPa内 平坦性 士0.5 μιη 内 同時’根據本發明之空白遮罩1〇〇之金屬膜12〇可藉 由減鑛形成。此時，用於形成金屬膜12〇之濺鍍靶為選自 矽化钥（MoSi)、矽化銦鈕（M〇TaSi)、鉻（Cr)、鈕（Ta)、 鶴（W)、矽（Si)、钥（Mo)、鈦（Ti)以及釕（Ru)之 多於一個的元素。且在使用矽化鉬（M〇si)靶的情況下， 可使用具有諸如Mo:Si = 20 ato/o:80 at%以及Mo:Si = 10 at%:90 at%等組成比率的乾。而且，當逐步形成包含可變 組成區之金屬膜120時，1〇 at%與3〇 at%之間的Mo含於 MoSi滅鍍靶中用於沈積靠近基板之金屬膜12〇部分，且較 18 201137513 /pif 佳少於約10 at%的Mo含於MoSi濺鍍靶中用於形成金屬 膜120之表面。此外，金屬膜120可使用DC反應性磁控 減鍍（reactive magnetron sputtering )、RF 反應性磁控滅锻、 長拋減:鐘（long throw sputtering，LTS)以及離子束藏鍵方 法來沈積。在濺鍍期間，所使用之惰性氣體為選自由氬 (Ar)、氦（He)、氖（Ne)以及氙（Xe)或其混合物組成 之群組中的一者。且，所使用之反應氣體為選自由氧氣 (〇2)、氮氣（N2)、一氧化碳（CO)、二氧化碳（C02)、 二氧化氮（N02)、一氧化氮（NO)、一氧化二氮（N20)、 氨氣（NH3)以及甲烷（CH4)或其混合物組成之群組中的 一者。 此外，當藉由濺鍍製程形成金屬膜120時，較佳在自 約0.1 Pa至約0.15 Pa之壓力範圍下維持基板與靶之間的 約100 mm以上距離的狀態下進行製程。且，較佳濺鍍製 程之功率密度設定於〇.6W/mm與13 W/mm之間，且基板 加熱溫度設定於50。(：與300。(：之間。濺鍍製程期間之基 板溫度影響被濺鍍原子與基板之碰撞期間的黏合強度。因 此’可藉由在濺鍍之前加熱基板來增強基板與原子之間的 黏合強度’且薄膜之間的黏合強度在多層膜的情況下亦增 加。然而，基板與薄膜之間以及薄膜之間的黏合強度之效 果由於在加熱溫度在約50。(：以下的情況下的低溫而不顯 著，且其在加熱溫度大於約300 °C的情況下將因薄膜之沈 積期間以及沈積之後產生高殘餘應力而引起使平坦性劣化 的問題。因此’對於基板而言較佳在自約5〇。〇至約3〇〇ac 19 201137513 之範圍内，更佳在自約loot至約30(rC2範圍内進行加 熱。 硬遮罩膜130可形成為單一層或多層，且包含可變組 成區。此時，較佳硬遮罩膜130之厚度在5〇人與15〇人乏 間’且較佳片電阻（sheet resistance)小於約1 ki2/口。而且， 在硬遮罩膜130形成為包含彼此不同之多於一個的元素之 多層的情況下，較佳彼此鄰接之薄膜中之相同元素中之多 於者以約5 nm以上之厚度以及彼此鄰接之薄膜之界面 處1 at%與50 at%之間的組成比率改變而連續形成。且， 硬遮罩膜130包含具有在約5 nm以上厚度中連續改變之 氮含量的至少一區，且較佳在薄膜之深度方向上之密度改 變士 0.2g/cm3與2.0g/cm3之間。此外，對於硬遮罩膜13〇 而》較佳在深度方向上之姓刻速率改變在至少多於一個的 區中在與表Φ部分相比時較快。而且，較佳位於硬遮罩膜 13〇之下部部分處之金屬Μ 12〇㈣擇性在氣化物或氯化 乾式蝕刻期間大於約5。 同時，用於形成硬遮罩膜13〇之濺鍍靶為選自矽化鉬 (MoSi)、發化顧组（MoTaSi)、鉻（Cr)、組（Ta)、鎢（w)、 石=Ui)、鉬（Mo)、鈦（Ti)之一者。而且，硬遮罩膜13〇 疋、’星由DC反應性磁控滅鐘方法形成，且此時，可使用選 1由氬（Ar)、氦（He)、氖（Ne)以及氤（Xe)或其混 。物組成之群組的惰性氣體，且可使用選自由氧氣、 氮氣（n2)、一氧化碳（co)、二氧化碳（c〇2)、二氧化 氮（N〇2)、—氧化氮（N0)、—氧化二氮％。）、氨氣（νη〇 20 201137513 OJIH/pif 以及甲燒（CaO或其混合物組成之群組的反應氣體。 抗餘劑膜由包含強酸之抗蝕劑材料製成，且以1000 A 與2000人之間的厚度形成。且，較佳有機薄膜在抗蝕劑膜 140下方以小於約2⑻人之厚度形成且包含具有比抗蝕 劑膜4〇中之〉農度南的濃度的強酸。此時，位於抗钱劑膜 140下方之有機薄膜由顯影溶液顯影，無論是否應用暴露 製程。且’硬遮罩膜130之表面上的抗蝕劑材料塗層為化 學增幅型抗蝕劑（chemically amplified resist)。 圖2是根據本發明之用以製造用於硬遮罩用途之空白 遮罩之長抛錢鑛設備（long throw sputtering equipment)的 不意圖。 參見圖4，藉由使用由HIP方法製造之濺鍍靶，應用 如圖2說明之長拋濺鍍設備在具有6〇25尺寸之透明基板 11〇上形成金屬膜120，且M〇:Si之組成比率為1:9 (即， Mo:Si=l〇 at%:9〇 at〇/0) (s2〇〇)。在金屬膜 12〇 之沈積期間 的製程條件與下表中所示相同。 [表2] 製程條件 設定值 功率 0.7 kW 曼氣輸入 5 seem 製程壓力 0.05 Pa 藉由以如表2描述之製程條件在50秒期間針對金屬 膜120執行沈積製程之後的1〇秒期間緩慢地且連續地改變 氮氣輸入1.0—1.5 —2.0 (seem)而在金屬膜12〇中形成 201137513 35147pif 可變組成區。而且，在不斷開電漿的狀態下連續改變氮氣 輸入 5.0 — 9.0 — 13.0 — 9.0 — 5.0 (seem) ’ 且依序增加 功率0.7 — 1.01.2 — 1.5 (kW)以形成可變組成區。使 用分析儀器（n&k Technology 公司之 Analyzer 1512RT 儀 器）在148平方毫米之範圍内之49個點處量測如此製造之 金屬膜120之光學密度及反射率，且使用（^又尺儀器在中 心之一點處量測厚度。作為量測之結果，相對於193 nm 波長之暴露光之平均光學密度為3.0,光學密度之均一性 為0.02,平均反射率為18.3%，反射率之均一性為〇52%， 且厚度為484人，使得用於硬遮罩用途之空白遮罩之金屬 膜展現優良性質。 此外’在85%之硫酸及23%之SC-1(過氧化氫溶液：氨 水：超純水=1:1: 5)中浸泡2小時之後使用相同分析儀器 評估所製造金屬膜120上之反射率改變。作為結果，相對 於193 mn波長之暴露光之反射率改變對於硫酸為〇12%且 對於SC-1為0.42%，使得金屬膜展現優良性質。而且，由 於使用原子力光譜儀（Atomic Force Spectroscope，AFM ) 儀器量測所製造金屬膜120上之1μιηχ1μιη區域的表面粗 糙度，因此觀察到約0.53 mnRa的值，使得展現優良表面 粗糙度性質。 隨後，以Cr靶替換濺鍍靶，且藉由下表中描述之製 程條件在金屬膜120之上部部分上形成硬遮罩3〇 (S210) 〇 22 201137513 /pif [表3] 製程條件 設定值 功率 0.4 kW ~~ 製程壓力 0.5 Pa 氬氣輸入 3 seem ~ '~~· 甲烷氣體輸入 0.1 seem 二氧化氣氣體輸入 5 seem 在以如表3所描述之製程條件在10秒期間對硬遮罩 膜130執行沈積製程之後，在不斷開電漿的狀態下依序改 變二氧化氮輸入5.0 — 3.0 —> 1.0 (seem)，且依序增加功 率0.4 — ι.ο —ι·5 (k\V)以形成可變組成區。使用4點探 測設備量測如此製造之硬遮罩膜130之表面的片電阻，其 為324 Ω/口。因此，可確定硬遮罩13〇展現優良片電阻性 質。而且，藉由GXR儀器量測硬遮罩臈13〇之厚度，其 為106 A。使用歐傑電子光譜儀（a寧^咖 比:=後製造之用於 例=罩臈之組成比率分析 素之組成比率連續改變的區。硬遮罩膜包含其中每-元 隨後，在藉峻私硬遮I U)〇A厚度之有機_ 场成含有強酸且 ⑺正化學增_私 15⑽A厚度塗覆 遮罩（S220)。藉由先f來製造用於硬遮罩用途 製程將如此製造之用於硬 具有 FEP-171 之空白 23 201137513 35147pif ^用途之^白遮罩製造為光罩。此時，藉由5()]^電子 遮罩執行銘印（imprinting)，且在使用2遍之 顯影溶液執行顯影製程之後藉由氯化氣體制硬 遮罩膜130 »隨後，在再次移除⑽編14()之後藉由硬 遮罩膜130作為侧遮罩來侧金屬膜12〇。隨後，藉由 使用〇*姓刻溶液移除硬遮罩膜m來完成光罩的製造。隨 後，使用CD-SEM量測具有最終圖案之薄膜之cd，且確 定獲彳寸50 nm CD作為結果。 下文現將描述根據本發明之用於硬遮罩用途之空白 遮罩以及藉由習知逐步膜沈積技術製造之用於硬遮^途 之空白遮罩的評估結果。針對評估執行以下過程，使得藉 由習知逐步膜沈積技術製造空白遮罩。 册方法製造之雜乾之反應性Dc磁控賤鑛方法，在具 f 6025尺寸之透明基板上形成金屬膜，且其中·別之組 ，j @ ’ MG:Si=1()at%:9()at%)e 在沈積期間使 用之功率為約0.7 kw，Ar氣輸入為約5咖，製程壓力 為約0.05 Pa ’且製程時間為約55秒。隨後，在斷開電漿 =況下以L5kW之功率、5sccm之氬氣以及1〇靡之 氮氣形成抗反射膜。藉由相同量測儀器在8平方毫米之範 圍内之49個點處量測藉由此等製程條件逐步形成之膜的 光學密度及反射率。且，使用GXR儀器在中心處量測厚 度作為量測之結果，在藉由習知逐步膜沈積技術製造之 用於硬遮罩用途之空白遮罩上沈積的金屬膜相對於⑼ 腿波長之暴露光之平均光學密度為約3 〇，光學密度之均 24 201137513. 一性為約0.03 ,平均反射率為約19 8%，反射率均一性為 約1.12%，且厚度為約512人，使得用作用於硬遮罩用途 之空白遮罩之金屬膜不存在顯著問題。 、 隨後’在85%之硫酸及23%之SCM(過氧化氣溶液:氨 水：超純水=1:1: 5)中浸泡2小時之後在藉由習知逐步膜 沈積技術製造之用於硬遮罩用途之空白遮罩的金屬膜上量 測反射率改變。作為結果，反射率改變對於硫酸為〇 52% 且對於SC-1為0.83%，使得金屬膜展現優良性質，但可確 疋耐化學性性質與根據本發明之用於硬遮罩用途之空白遮 罩之包含可變組成區的金屬膜相比不良。而且，由於使用 AFM儀器量測藉由習知逐步膜沈積技術製造之用於硬遮 罩用途之空白遮罩的金屬膜上1μιη χ 區域之表面粗縫 度’因此觀察到約0.83 nmRa的值。因此可見，表面粗糙 度與藉由本發明之製造方法製造之硬遮罩膜13〇相比減小。 隨後’以Cr靶替換濺鍍靶。氬氣流動速率設定為3 seem’甲烷（CH4)氣體流動速率設定為〇1 sccm，且二 氧化氮氣體流動速率設定為5 seem。在藉由習知逐步膜沈 積技術製造之用於硬遮罩用途之空白遮罩之金屬膜的上部 部分上在0.4 kW功率及〇.5 Pa壓力下沈積硬遮罩膜歷時 12秒。使用4點探測設備在如此製造之用於硬遮罩用途之 I知空白遮罩之硬遮罩膜的表面上量測之薄膜的片電阻觀 察到是良好的，其量測值為約402 Ω/□，且由GXR量測之 硬遮罩膜之厚度為約116人。 雖然已參考特定實施例描述空白遮罩、光罩及其製造 25 201137513 35147pif 方法，但其不限於此。因此’熟習此項技術者將容易瞭解’ 在不脫離由所附申請專利範圍界定的本發明之精神和範圍 的情況下可對其做出各種修改和改變。 【圖式簡單說明】 自上文結合附圖做出之描述可更詳細理解例示性實 施例，附圖中： 圖1疋根據例示性實施例之用於硬遮罩用途之空白遮 罩的截面圖。 圖2是根據例示性實施例之用以製造用於硬遮罩用途 之空白遮罩之長拋濺鍍設備的示意圖。 圖3是繪示根據例示性實施例之用於硬遮罩用途之空 白遮罩的薄膜之組成比率分析之結果的曲線圖。 圖4是說明根據例示性實施例之實施製造用於硬遮罩 用途之空白遮罩之方法之過程的流程圖。 【主要元件符號說明】 100 :空白遮罩 110 :透明基板 120 :金屬膜/金屬層 130 :硬遮罩膜/硬遮罩 140 :抗飯劑膜 S200、S210、S220 :步驟標號 26201137513 35147pif VI. Description of the Invention: [Technical Field] The present invention relates to a blank mask, a photomask, and a method of fabricating the same, which makes it possible to achieve a high-precision critical dimension in a semiconductor lithography process (critical Dimensions, CD), and more specifically, for blank masks, reticle, and methods of making the same, which can be applied to ArF (wavelength: i93 nm) lithography and ArF immersion lUhography to achieve less than 65 nm The critical size. [Prior Art] At present, advanced semiconductor micro-processing technology is becoming increasingly important to meet the demand for reducing the size of circuit patterns required for highly integrated semiconductor body circuits. For highly integrated circuits 'requires very small circuit interconnections for low voltage and high speed operation' and for the contact hall pattern for interlayer connection and the circuit layout required for integration The technical needs are getting higher and higher. Therefore, in order to meet such demands, there is a need for a technique that can form a very small circuit pattern and reduce the scale of manufacturing a photomask in which a circuit pattern is formed. In particular, 5, in order to accurately form ultra-precision circuit patterns of less than 65 nm, a blank mask structure is currently being developed, in which the inorganic mask of the mat is not used without the use of the prior art anti-surname film as a side mask. Membrane_ forms a metal film of the final pattern. In the blank mask t 'hard mask film for hard mask use, it replaces the prior art thick anti-fine film, and the aspect ratio (aspeet rati) and selectivity of the β portion under the metal film can be improved by the hard mask film. . Therefore, the dry type of the metal film is engraved (kiss (10) (four)) 4 201137513. The loading effect is reduced during the period so that an excellent CD average value (mean t0 target, MTT), CD linearity, and CD uniformity can be obtained. Meanwhile, in the prior art for hard disk use, it generally has a structure in which a light shielding film, a metal film, a hard mask film, and a resist film having a uniform composition ratio in the depth direction are sequentially formed in 6〇25 size synthetic quartz glass substrate. At this time, the metal film may include an anti-reflection film, and the hard mask film is used as an etching mask. The uniform composition ratio of the film in the depth direction means that it is formed under the same overall process conditions and process environment, such as a taget during film formation, processing time, power, pressure, reactive gas, inert gas type, and input. the amount. In addition, the gradual formation of the film means that it is deposited under different process conditions for controlling the characteristics of each film. That is, it means that the demineralization process conditions of the film relative to the substrate are carried out in a discontinuous step to form the film in a stepwise manner. In a thin film process in which the composition ratio is uniform in the depth direction of the film formed in a stepwise manner, each layer has its function of performing independently = advantageous because process variation is carried out in each step. For example, to focus on shielding the function of exposed light for a particular wavelength of the light-shielding film and reducing the surface reflectance of the anti-reflective film to the exposed light, only two processing condition steps are required. Therefore, there is an advantage of forming a metal film having a two-layer structure. On the other hand, due to the attempt to further reduce the scale so that a circuit pattern of less than 65 nm (especially less than 45 nm) can be achieved, in addition to the light shielding function and anti-reflection function for exposed light, the uniformity of the film is overcome 5 201137513 35147pif The requirements for properties, chemical durability characteristics, surface roughness, residual stress, growth defects, CD uniformity, and defects are also becoming more stringent. During the achievement of a pattern greater than 65 nm, the effect is inconspicuous in terms of the above issues compared to the critical dimension to be achieved, and no significant problems occur because wavelengths of exposed light such as 365 nm i-line and 248 are relatively long. The wavelength. However, during the formation of a super-precision pattern such as a short wavelength of 193 nm ArF in order to achieve a critical dimension of less than 65 nm, the defects caused by the existing residual stress of the film formed by the stepwise manner, the uniformity of the film, and the film density are changed. For the more significant problem, it needs to be solved as the CD to be achieved becomes extremely small. However, it is difficult to solve the above problem because a sudden change in composition ratio, a change in film density, and a change in residual stress often occur at the interface of the film when attempting to solve a complicated problem such as the above by using a film formed by a stepwise manner. Further, in order to satisfy the uniformity of optical density, uniformity of reflectance, chemical durability, surface roughness, and growth defect characteristics, a complicated manufacturing process of depositing another film is required. This complex process not only causes defects such as particles, but also the loading effect increases due to the overall increase in film thickness', so that problems of reduced resolution, CD, MTT, and CD uniformity will occur. SUMMARY OF THE INVENTION The present invention provides a binary blank mask, a blank mask for hard mask use, and a phase shift blank mask for ArF lithography and ArF immersion lithography, with respect to uniform reflectivity, uniformity Optical density, adhesion between films, residual stress, growth defects, chemical durability, durability of exposure, and surface 6 201137513 have a good film property. According to an exemplary embodiment, there is provided a blank mask comprising a transparent substrate, a metal film, a hard mask film, and a resist film sequentially stacked in the stated order, wherein the blank mask has the metal At least one variable composition region of at least one of the film and the hard mask film, wherein a composition ratio of at least one of the plurality of elements constituting the film continuously changes in a depth direction of the film. According to another exemplary embodiment, there is provided a photomask obtained by patterning and etching a blank mask, the blank mask comprising: a transparent substrate, a metal film, a hard mask film, and sequentially stacked in the stated order An anti-surname film, wherein the blank mask has at least one variable composition region in at least one of the metal film and the hard mask film, wherein at least one of a plurality of elements constituting the film The composition ratio of an element continuously changes in the depth direction of the film. According to another exemplary embodiment, a method of manufacturing a blank mask is provided, which is manufactured by sequentially forming a metal film, a hard mask film, and a resist film on a transparent substrate, the method further comprising: a) Depositing the metal film on the transparent substrate while maintaining the same process conditions applied in the first deposition time; and b) forming a variable composition region, by gradually or while the reaction gas is still turned on Continuously changing at least one of a plurality of process conditions applied in the second deposition time, a composition ratio of at least one of a plurality of elements constituting the metal film continuously changing in a depth direction of the metal film, wherein The order of a) and b) implementation may vary, or a) and b) may be implemented in an alternating manner. 201137513 35147pif According to another exemplary embodiment, a method for manufacturing a blank mask is provided, which is manufactured by sequentially forming a metal film, a hard mask film, and an anti-surplus film on a transparent substrate, the method further comprising: a Depositing the metal film on the transparent substrate; and b) forming a variable composition region in which many processes applied in the second deposition time are changed in a stepwise or continuous manner while the reaction gas is still turned on At least one of the conditions 'the composition ratio of at least one of the plurality of elements constituting the metal film continuously changes in the depth direction of the metal film. According to the blank mask, the reticle, and the manufacturing method thereof according to the present invention, the blank mask can selectively and continuously change at least more than one element constituting the film in the depth direction of the film during the manufacturing process. The content of the film is improved by adhesion, reflectance uniformity, chemical durability and residual stress characteristics. Further, the present invention makes it possible to manufacture a blank mask having excellent properties, and thus it is also possible to manufacture a photomask having excellent properties. [Embodiment] Hereinafter, specific embodiments will be described in detail with reference to the accompanying drawings. 1 is a cross-sectional view of a blank mask for hard mask use, in accordance with an illustrative embodiment. Referring to Fig. 1, a blank mask 100 for hard mask use according to the present invention comprises a transparent substrate 110, a metal film 12, a hard mask film 130, and a resist film 140. The transparent substrate 11A, the metal film 12A, the hard mask film 130, and the resist film 140 are formed by sequentially stacking. Here, the metal film and/or the hard mask film has a region (hereinafter referred to as "variable composition region"), wherein the composition ratio of at least one of the elements constituting the film at 8 201137513 is in the depth direction of the film. About 1 at0/. Change continuously to approximately 50 at%. Further, the variable composition region in which the composition ratio of at least one of the elements constituting the film continuously changes in the depth direction of the film means that the process conditions for controlling the characteristics of each film are carried out without stopping the reaction gas. To include more than one variable composition region that changes in a continuous or stepwise manner. That is, in order to form a variable composition region which changes in the depth direction of the film, the process conditions are changed in a continuous or stepwise manner during the sputtering of the film of the substrate. "The transparent substrate U0 is performed by performing a plurality of substrates relative to the substrate matrix material. Formed by a lapping process and a plurality of polishing processes. The material for the transparent substrate 110 may be selected from the group consisting of synthetic quartz, calcium fluoride (CaF2), and fluorine-doped quartz (F-doped quartz). The transparent substrate 110 has a size of 6025, and preferably has a birefringence of less than about 2 nm/6 at 193 nm. 35 mm » At this time, the substrate precursor material has 99. 9999% or more of pure yttrium oxide (Si〇2) composition, and by slicing and edge grinding the synthetic quartz ingot to obtain a size of 152 χ 152 ± 〇 2 and about mm or more thickness. The transparent substrate 110 is formed by performing a plurality of polishing processes and a plurality of polishing processes with respect to the substrate precursor material. First, a plurality of polishing processes are performed with respect to a substrate mother material having a size of 152 χ 152 ± 0·2 mm and a thickness of about 6.3 mm or more. The polishing process is performed in the case of performing a polishing process in view of the process efficiency using relatively large-sized abrasive grains under high pressure 201137513 35l47pif. In this case, the purpose of reducing the thickness can be easily achieved, but the damage including the crack generated from the surface of the substrate in the ice direction occurs. Therefore, the crack generated inside the substrate mother-riding towel was observed as a defect in the post-treatment. Moreover, due to the role of _ training, it is difficult to call the purpose of thickness accuracy. Therefore, it is preferred that the transparent substrate 11 used in the present invention is formed by utilizing a multi-faceted light source so as to reduce the thickness of the wire plate and reduce the thickness. The abrasive particles applied in the polishing process of the substrate precursor material are selected from the group consisting of carbonized dream (sic), diamond (c), and oxidized fault (Zr02). And more than one oxidized |g (Al2〇3), and the particles are in the multi-strand filament. In addition, the size of the residue is between 4μηη and 20μιη. If the size of the abrasive particles is less than about 4 μm, the thickness quasi-breaking can be easily achieved, but as the processing time increases, the production is small. If the size of the abrasive particles is greater than about 2 〇 μηι, the thickness is not achieved and the defect level is reduced. Subsequently, a plurality of polishing processes are performed with respect to the substrate precursor material that has been subjected to a plurality of polishing processes. The slurry used in the plurality of grinding processes comprises cerium oxide (Ce〇2), (iv) oxidized $(Si〇2) abrasive particles, and hydrogen peroxide (H202), and the pH is composed of nitric acid (HN〇3). ) or osmium hydroxide (KOH) control. The overall pH of the preferred grinding slurry is controlled between 6 and η during the grinding process. In particular, an inorganic base such as KOH (kappa) has an etching effect on the substrate precursor material, thereby having the advantage of exhibiting a synergy effect. At the same time, the abrasive particles physically remove the substrate precursor material during the polishing process. At this time, if the size of the abrasive grains is larger than about 5 μm, it is easy to achieve the purpose of thickness reduction, but it may be difficult to obtain good surface roughness by 201137513 JWpif. Moreover, since the size of the abrasive grains is less than about 0. In the case of 5 μm, the thickness of the base material of the substrate subjected to the polishing process is small, so that a long processing time is required, so that the processing efficiency can be lowered. Therefore, the size of the preferred cerium abrasive grain is 0. 5 4 „! and 5 μηη. In addition, it is preferred that the size of the colloidal cerium oxide abrasive particles used in a plurality of grinding processes is between 20 nm and 200 nm. At this time, if the colloidal cerium oxide is ground If the size of the particles is less than about 20 nm, the polishing process efficiency is lowered, and if the size is greater than about 200 nm, the surface roughness to be achieved in the polishing process may be unsatisfactory. Since the polishing process is used to improve surface roughness Process, so it is better to gradually use smaller abrasive particles when performing the grinding process. Since the surface state is relatively rough due to the polishing process, a large amount should be removed in the first polishing process compared with the second and third polishing processes. The substrate precursor material. Therefore, a porous koi (p0rous eerjum pad) having horse hardness and low compressibility is used as the polishing pad in the first polishing process and it is preferred to use SUBA #400 to 800 in the second polishing process. As a cushion. At this time, the polishing pad determines the amount of substrate precursor material to be removed' and the surface state depends on the type of pad used. If a polishing pad lower than #400 is used, the processing time is relatively high. The compressibility and elastic recovery rate and low hardness characteristics increase, so that the polishing efficiency can be reduced. If a polishing pad higher than #800 is used, the compression rate and elastic recovery rate are high and high. It is difficult to achieve the purpose of surface roughness by hardness characteristics. Moreover, it is preferable that the compressibility of the cushion used in the second polishing process is about 3% or more, and the elastic recovery rate is about 65% or more. In the compressible polishing pad, 201137513 35l47pif Since the coating layer around the particle is applied to the polishing pressure at the polishing pressure, the polishing pressure is dispersed and absorbed, so that the polishing can occur on the surface of the substrate. The generation of concave defects is suppressed. Meanwhile, in the polishing pad having a high elastic recovery ratio of more than 65%, the raised layer can be easily compressed and recovered, so that there is no large residual oxidation in the raised layer. The advantage of dream particles suppressing concave defects. Moreover, a suedepad as a super soft pad is used as a polishing pad in the third polishing process. The grinding process is a process for maximally improving surface roughness and particle characteristics, so an ultra-soft polishing pad having a lower hardness and a relatively high elastic recovery rate and compressibility is used. It is preferably used in the third polishing process. The soft abrasive pad has a compressibility greater than about 6% and an elastic recovery rate greater than about 72%. Since the defects should be tightly controlled in the third polishing process, compressibility and elasticity greater than the second abrasive enthalpy should be used. Recovery rate of the polishing pad. Meanwhile, the groove in the polishing pad used in the polishing process can have various shapes. As an example, it can be used with a pitch of 25 mm, a width of 4 mm, and a width of 0. A polishing pad with a groove of 5 mm depth. The groove of the polishing pad supplies a sufficient amount of the slurry to the transparent substrate 110 during the polishing process, so that it serves to increase the polishing efficiency of the transparent substrate 110. At this time, it is possible to use a grinding burr which is not formed with a groove. The size of the groove may vary depending on the grinding process, and whether or not the polishing pad formed with the groove is used may be selectively determined depending on the polishing process. In addition, in the case where the polishing pad used in the polishing process has a structure with two or more layers, it is preferable that the thickness of the raised layer corresponding to the second layer of the polishing pad from the precision surface plate direction is about 2〇〇12 201137513 JDl^/pif μιη is between about 600 μπι. If the thickness of the raised layer is less than about 2 Å, the elastic recovery rate of the polishing pad is reduced, making it difficult to secure the surface roughness, and may have an adverse effect with respect to the generation of defects containing particles. If the raised layer is larger than about 600 μm, the grinding efficiency for ensuring surface roughness is reduced. The metal film 120 is formed on the transparent substrate η. It can be formed as a single reed multilayer. The layer of more than one of the single-layer or the plurality of layers is formed with a compositional region in which the composition ratio of more than one of the elements constituting the layer is continuously changed in the depth direction. At this time, the composition ratio of at least one element in the variable composition region is changed within a range from about i at % to about 5 Å, and the thickness of the preferred variable composition region is greater than or equal to 5 nm. The metal film 12〇 disclosed in the present invention is a film of a multilayer structure formed from the transparent substrate 11〇 to the surface on which the final pattern is formed, such that it includes all the films in the portion below the final pattern (the secret film, The stress reduction film, the light shielding film, and the antireflection film are specified, and in addition, a phase shifting film may be selectively formed between the transparent substrate 丨丨〇 and the metal film 12 。. Therefore, the blank mask in which the phase shift film is formed is referred to as a half-tone phase shift blank mask. Further, the composition ratio of nitrogen and/or ytterbium in the composition ratio of the elements constituting the metal film 120 is continuously changed, and the thickness of the region where the composition ratio is changed is more than 1 nm. Nitrogen acts as a major factor that not only affects the optical properties of the film such as transmittance and reflectance but also affects chemical durability and exposure durability. For example, if the nitrogen content is higher, the reflectance on the surface of the film is formed lower relative to the optical characteristics of the film. On the other hand, if the nitrogen content is lower, the reflectance on the surface of the film is formed to be higher. In addition, 13 201137513 35147pif resistance to chemical resistance (four), that is, the chemical properties of the surface on the sc] deteriorate with the increase of the I content, and the chemical resistance properties of the Zhe 1 and sulfuric acid show with the decrease of the nitrogen content. Excellent, f quality. Therefore, the touch design metal layer (10) is gambling depending on the fact that the nitrogen content in the depth direction has different distribution patterns from each other in order to consider == at this time, because the thickness of the variable composition region is less than about i. A meaningful change in the filminess f mentioned by the wire / == has more than one zone in which the nitrogen content is continuous. The thickness of the altered zone is greater than about 1 nm. In the meantime, in the case of a metal film which is essentially composed of a broken metal, it is preferred that the variable composition region having a halogen has a thickness of at least two regions of more than about 5 nm and that 'in a self-masking mask for the use of a hard mask, It is important to reduce the thickness of the gold 'and to select materials with excellent Wei Xue properties in order to reduce the 贞 load effect. At this time, with respect to the optical and chemical resistance characteristics of the metal film 12 (), as the ratio of her ratio increases, the chemical resistance property becomes excellent. At this time, since the chemical resistance is obtained by the nature of the chemical composition in the surface of the metal film 12G, the content of the surface t should be high. On the other hand, if the comparison has the same thickness, the y_4 is higher, the material increases, and the thickness of the recorded film will eventually increase. Therefore, the above conditions of the riding force essentially require a composition of the hair content (4) a continuously changing variable composition region, and it is preferable that the thickness of the metal film 120 in the depth direction is greater than about 5 legs, more preferably 10 nm. Further, the distribution of the composition ratio is 201137513 in the width direction of the metal film 12, and the uniformity of the specific 'best composition ratio is less than about hop in the twist direction of the metal film, and the film depth in the metal film 12G The density in the direction changes in G. 2 g / em, 2 G g / em3 ^ At this time, the uniformity of the composition ratio can be calculated by the following equation, and the measured position is at least 5 points in the width direction ^, the composition ratio of the film can be Analyzed by methods such as AES, XPS, and RBS. [Equation 1] Group secret - species maximum composition ratio · minimum composition ratio Ϊ I '---X 100 (maximum composition ratio + minimum composition ratio 丨 3 at the same time, the density change in the depth direction of the film is less than about g / In the case of cm, the effect is not significant, which is similar to the state in which the density of the film does not change, and there is a problem that the stress of the thin movement is greater than about 2. In the case of 0 g/cm3, they are significantly different from each other. Moreover, the metal film 12 has different residual stresses in the depth direction, and preferably includes a region in which the residual stress difference is more than about 10 MPa. At this time, the metal film 丨2〇 includes a region in which the composition ratio of more than two elements continuously changes in the depth direction thereof, and particularly preferably selectively includes carbon in order to reduce the film. Further, in the metal film 120 A variable group is formed in the surface; in which case, the surface roughness can be controlled to be less than about 1. 〇nmRa. The surface roughness of the metal film 12 is determined by the surface roughness state of the substrate, the type of the element to be sputtered, the reaction gas, and the process conditions. At this time, the variable composition region can be formed by changing the reaction gas type, the deposition power, the chamber ink force in a continuous or stepwise manner, and in the chamber when the electricity is still turned on. At least f of the flow rate of the gas in which the composition ratio of at least one of the elements in the face or the metal film 120 is continuously changed in the depth direction of the film controls the surface (10) of the variable composition region. Further, the ratio of the preferred nitrogen atom is relatively high on the surface of the metal film 12 相比 as compared with the depth. Further, the composition ratio of nitrogen which is preferable for the film U0 is higher than the composition ratio of nitrogen in the range of thickness of 5 mn from the surface in the depth direction. In this case, the reflectance uniformity at 193 nm is preferably such that the nitrogen content of the outermost surface is relatively high with respect to the depth direction within 1%. Further, in the case where the metal film 120 has a multilayer structure in which more than one element is changed different from each other, it is preferable that more than the same elements in the film adjacent to each other in the respective films constituting the metal film are larger than about A thickness of 5 nm is continuously formed, and the composition ratio changes between 1 at% and 50 at% at the interface of the films adjacent to each other. In the white mask for hard mask use, the metal film 12G is a film that is exposed on the surface after the finalization. At this time, 'the light-shielding film containing at least the screen and the anti-reflection film which reduces light reflection in the case where the metal film has a uniform element composition ratio' is therefore an element due to deposition of the anti-reflection film after depositing the light-shielding film A sudden change in the composition ratio, residual stress occurs in a large amount at the interface. Further, if the elemental compositions of the film interfaces (light-shielding film and anti-reflection film) which are formed stepwise are different from each other, a change in film density occurs, and a necking problem occurs due to a rapid difference in the ratio of the surnames during dry etching. . This necking does not cause problems because the CD error due to necking is not significant in the CD greater than about 65 nm compared to the pattern cd to be achieved 201137513, but due to this defect, a pattern of less than about 65 nm CD is achieved. May occur. In the present invention, since the variable composition region is formed on the surface or in the film, the problem caused by the sudden change in the composition ratio and the film density can be reduced. At this time, when compared with the conventional film formed by the stepwise manner, the change in the film stress and the dry rice ratio will be small in the case where the change ratio of the elements forming the film is within about 1 at%. On the other hand, if the ratio of the elements forming the film is more than about 50 at%, the residual stress between the materials changes rapidly, so that the same problem as seen when the film is formed in a stepwise manner will occur. Therefore, it is preferred to form at least one variable composition region in which the ratio of change of more than one of the constituent elements in the metal film 12 is between i% and 5 〇 at% (more preferably, 3 at% and 30) Between at%). In addition, as the size of the pattern used to achieve becomes smaller, the resulting necking and stress will be greater. Therefore, in order to solve such problems, a variable composition region is formed when the metal film 120 is formed, so that defects caused by necking and stress can be reduced. Further, properties such as chemical resistance, stress, and film density can be changed, and thus It is possible to manufacture a blank mask having excellent film properties at the surface. Meanwhile, it is preferable for the metal film 121 that the change in the etching rate in the depth direction is faster in at least more than one segment in comparison with the surface portion. The preferred etching ratio is increased in the depth direction of the film in order to reduce the load effect generated during the dry etching of the metal film 120. This is due to the fact that the reactivity between the radical ion and the etched metal film decreases as it enters the depth direction (because the thickness of the film is in the depth direction 17 201137513 35147pif). Therefore, the preferred etching ratio is increased as it enters the depth direction to form a vertical pattern. Kelin controls the side ratio faster or slower by forming a composition ratio of the variable composition area (n), oxygen (〇), and carbon (C). The optical, physical, and chemical properties are described in the following table, such as the thickness of the metal film 12A including the variable group as mentioned above, the optical density at the surface, the chemical resistance, the residual stress, and the flatness. [Table 1] Properties - Thickness of Metal Film ~300 jT6O〇A Optical Density at Surface The exposure light with a wavelength of 193 nm is about 2. 7 and about 3. After immersion for 2 hours in chemical (ammonia: hydrogen peroxide solution: ultrapure water = 1:1:5) for 5 hours, the reflectance change for exposed light at 193 nm is about 1%" Residual stress about 100 Flatness in MPa is 0. 5 μιη内内 The metal film 12 of the blank mask 1 according to the present invention can be formed by subtractive mineralization. At this time, the sputtering target for forming the metal film 12 is selected from the group consisting of a molybdenum (MoSi), an indium antimonide button (M〇TaSi), a chromium (Cr), a button (Ta), a crane (W), and a crucible (Si). ), more than one element of the key (Mo), titanium (Ti), and ruthenium (Ru). And in the case of using a molybdenum (M〇si) target, a dryness having a composition ratio such as Mo: Si = 20 ato/o: 80 at% and Mo: Si = 10 at%: 90 at% can be used. Moreover, when the metal film 120 including the variable composition region is formed stepwise, Mo between 1 〇 at% and 3 〇 at% is contained in the MoSi target to deposit a portion of the metal film 12 靠近 near the substrate, and 18 201137513 /pif Preferably less than about 10 at% of Mo is contained in the MoSi sputtering target for forming the surface of the metal film 120. Further, the metal film 120 can be deposited using DC reactive magnetron sputtering, RF reactive magnetron forging, long throw sputtering (LTS), and ion beam bonding. The inert gas used during sputtering is one selected from the group consisting of argon (Ar), helium (He), neon (Ne), and xenon (Xe) or a mixture thereof. Further, the reaction gas used is selected from the group consisting of oxygen (〇2), nitrogen (N2), carbon monoxide (CO), carbon dioxide (C02), nitrogen dioxide (N02), nitrogen monoxide (NO), and nitrous oxide ( One of a group consisting of N20), ammonia (NH3), and methane (CH4) or a mixture thereof. In addition, when the metal film 120 is formed by a sputtering process, it is preferably at about 0. 1 Pa to about 0. The process is carried out while maintaining a distance of about 100 mm between the substrate and the target under a pressure range of 15 Pa. Moreover, the power density of the preferred sputtering process is set at 〇. Between 6W/mm and 13 W/mm, and the substrate heating temperature is set at 50. (: and 300. (: between. The substrate temperature during the sputtering process affects the bonding strength during the collision between the sputtered atoms and the substrate. Therefore, it can be enhanced between the substrate and the atom by heating the substrate before sputtering. The bonding strength' and the bonding strength between the films also increase in the case of the multilayer film. However, the effect of the bonding strength between the substrate and the film and between the films is due to the heating temperature of about 50. (: The following case Low temperature is not significant, and it causes a problem of flatness deterioration due to high residual stress during deposition of the thin film and after deposition at a heating temperature of more than about 300 ° C. Therefore, it is preferable for the substrate. Approximately 5 〇. 〇 to about 3 〇〇 ac 19 201137513, more preferably from about loot to about 30 (rC2 heating). The hard mask film 130 can be formed as a single layer or multiple layers, and contains variable In this case, the thickness of the preferred hard mask film 130 is between 5 Å and 15 Å and the sheet resistance is less than about 1 ki 2 /port. Moreover, in the hard mask film 130 Formed to contain each other In the case of a plurality of layers of more than one element, it is preferred that more than one of the same elements in the film adjacent to each other has a thickness of about 5 nm or more and an interface between the films adjacent to each other at 1 at% and 50 at%. The composition ratio between the components is continuously changed. Further, the hard mask film 130 includes at least one region having a nitrogen content continuously changed in a thickness of about 5 nm or more, and preferably has a density change in the depth direction of the film. . 2g/cm3 and 2. Between 0g/cm3. Further, for the hard mask film 13', it is preferable that the rate change in the depth direction is faster in at least more than one region in comparison with the table Φ portion. Moreover, the metal Μ 12 〇 (4) preferably located at the lower portion of the hard mask film 13 择 is more than about 5 during vaporization or chlorination dry etching. Meanwhile, the sputtering target for forming the hard mask film 13 is selected from the group consisting of molybdenum molybdenum (MoSi), MoTaSi, chromium (Cr), group (Ta), tungsten (w), and stone = Ui. ), one of molybdenum (Mo) and titanium (Ti). Moreover, the hard mask film 13〇疋, 'star is formed by a DC reactive magnetron-off method, and at this time, it is possible to use argon (Ar), helium (He), neon (Ne), and xenon (Xe). ) or its mix. An inert gas of a group consisting of oxygen, nitrogen (n2), carbon monoxide (CO), carbon dioxide (c〇2), nitrogen dioxide (N〇2), nitrogen oxide (N0), % nitrous oxide. ), ammonia gas (νη〇20 201137513 OJIH/pif and toluene (a reaction gas of a group consisting of CaO or a mixture thereof. The anti-surplus film is made of a resist material containing a strong acid, and is 1000 A and 2000 people) The thickness between the layers is formed. Further, the preferred organic film is formed under the resist film 140 with a thickness of less than about 2 (8) people and contains a strong acid having a concentration higher than that of the resist film 4 。. The organic film under the anti-money film 140 is developed by the developing solution regardless of whether or not the exposure process is applied. And the coating of the resist material on the surface of the hard mask film 130 is a chemically amplified resist. Figure 2 is a schematic illustration of a long throw sputtering equipment for making a blank mask for hard mask use in accordance with the present invention. Referring to Figure 4, by using the HIP method The sputtering target is formed by forming a metal film 120 on a transparent substrate 11 having a size of 6 〇 25 using a long throwing sputtering apparatus as shown in FIG. 2, and the composition ratio of M 〇:Si is 1:9 (ie, Mo:Si) =l〇at%:9〇at〇/0) (s2〇〇). In gold The process conditions during the deposition of the film 12〇 are the same as those shown in the table below. [Table 2] Process conditions Setting value Power 0. 7 kW megawatt input 5 seem process pressure 0. 05 Pa slowly and continuously changes the nitrogen input during the 1 second period after performing the deposition process for the metal film 120 during the 50 second period with the process conditions as described in Table 2. 0-1. 5 — 2. 0 (seem) and a 201137513 35147pif variable composition region is formed in the metal film 12〇. Moreover, the nitrogen input is continuously changed without disconnecting the plasma. 0 — 9. 0 — 13. 0 — 9. 0 — 5. 0 (seem) ’ and increase the power by 0. 7 — 1. 01. twenty one. 5 (kW) to form a variable composition zone. The optical density and reflectance of the thus fabricated metal film 120 were measured at 49 points in the range of 148 mm 2 using an analytical instrument (N&k Technology's Analyzer 1512RT instrument), and used (^ and the instrument at the center) The thickness is measured at one point. As a result of the measurement, the average optical density of the exposed light with respect to the wavelength of 193 nm is 3. 0, the uniformity of optical density is 0. 02, the average reflectance is 18. At 3%, the uniformity of reflectance is 〇52%, and the thickness is 484, which makes the metal film of the blank mask for hard mask use exhibit excellent properties. In addition, the reflection on the produced metal film 120 was evaluated using the same analytical instrument after soaking for 2 hours in 85% sulfuric acid and 23% SC-1 (hydrogen peroxide solution: ammonia water: ultrapure water = 1:1: 5). The rate changes. As a result, the reflectance change of the exposed light with respect to the wavelength of 193 mn is 〇12% for sulfuric acid and 0% for SC-1. 42%, making the metal film exhibit excellent properties. Further, since the surface roughness of the 1 μm χ 1 μηη region on the produced metal film 120 was measured by an Atomic Force Spectroscope (AFM) instrument, about 0. The value of 53 mnRa makes it exhibit excellent surface roughness properties. Subsequently, the sputtering target is replaced with a Cr target, and a hard mask 3 〇 is formed on the upper portion of the metal film 120 by the process conditions described in the following table (S210) 〇22 201137513 /pif [Table 3] Process condition setting value Power 0. 4 kW ~~ process pressure 0. 5 Pa argon input 3 seem ~ '~~· methane gas input 0. 1 seem Dioxide gas input 5 seem After performing a deposition process on the hard mask film 130 during 10 seconds in the process conditions as described in Table 3, the nitrogen dioxide input is sequentially changed without disconnecting the plasma. . 0 — 3. 0 —> 1. 0 (seem), and increase the power by 0. 4 — ι. ο —ι·5 (k\V) to form a variable composition area. The sheet resistance of the surface of the thus fabricated hard mask film 130 was measured using a 4-point detecting device, which was 324 Ω/□. Therefore, it can be confirmed that the hard mask 13 〇 exhibits excellent sheet resistance properties. Moreover, the thickness of the hard mask 臈13〇 was measured by a GXR instrument, which was 106 A. Using the Oujie electronic spectrometer (a 宁^咖 ratio: = after manufacturing the composition ratio of the = 臈 分析 分析 分析 分析 分析 分析 。 。 。 。 。 硬 硬 硬 硬 硬 硬 硬 硬 硬 硬 硬 硬 硬 硬 硬 硬 硬 硬 硬 硬 硬 硬 硬 硬Hard cover IU) 〇A thickness of organic _ field containing strong acid and (7) positive chemical increase _ private 15 (10) A thickness coated mask (S220). Fabrication for hard masking by f first The process will be manufactured for hard use. Blanks with FEP-171 23 201137513 35147pif ^The use of white masks is made into a reticle. At this time, imprinting is performed by the 5()]^ electronic mask, and the hard mask film 130 is made of chlorinated gas after the development process is performed using the development solution of 2 times. (10) After the pattern 14 (), the side metal film 12 is covered by the hard mask film 130 as a side mask. Subsequently, the manufacture of the photomask is completed by removing the hard mask film m using a 姓* surname solution. Subsequently, the cd of the film having the final pattern was measured using a CD-SEM, and the obtained 50 nm CD was determined as a result. The evaluation results of the blank mask for hard mask use and the blank mask for hard mask manufactured by the conventional step-by-step film deposition technique according to the present invention will now be described. The following procedure was performed for the evaluation to make a blank mask by conventional progressive film deposition techniques. The reactive Dc magnetron-based antimony ore method produced by the method of the invention forms a metal film on a transparent substrate having a size of 6060, and among them, j @ ' MG:Si=1()at%:9 ()at%)e The power used during deposition is about 0. 7 kw, Ar gas input is about 5 coffee, the process pressure is about 0. 05 Pa ' and the process time is about 55 seconds. Subsequently, an anti-reflection film was formed at a power of L5 kW, an argon gas of 5 sccm, and a nitrogen gas of 1 Torr under the disconnection of the plasma. The optical density and reflectance of the film which was gradually formed by such process conditions were measured at 49 points in the range of 8 mm 2 by the same measuring instrument. And, using a GXR instrument to measure the thickness at the center as a result of the measurement, the metal film deposited on the blank mask for hard mask use manufactured by the conventional step-by-step film deposition technique is exposed to the (9) leg wavelength. The average optical density of light is about 3 〇, and the optical density is 24 201137513. One sex is about 0. 03, the average reflectance is about 19 8%, and the reflectance uniformity is about 1. 12%, and a thickness of about 512 people, makes the metal film used as a blank mask for hard mask use without significant problems. Subsequently, after immersing for 2 hours in 85% sulfuric acid and 23% SCM (peroxide gas solution: ammonia water: ultrapure water = 1:1: 5), it is made of hard by conventional progressive film deposition technology. The reflectance is measured on the metal film of the blank mask for mask use. As a result, the reflectance change is 〇 52% for sulfuric acid and 0 for SC-1. 83% makes the metal film exhibit excellent properties, but it is confirmed that the chemical resistance property is inferior to the metal film containing the variable composition region of the blank mask for hard mask use according to the present invention. Moreover, since the surface roughness of the 1 μιη 区域 region on the metal film of the blank mask for hard mask use manufactured by the conventional step-by-step film deposition technique was measured using an AFM apparatus, it was observed that about 0. The value of 83 nmRa. Therefore, it can be seen that the surface roughness is reduced as compared with the hard mask film 13 制造 manufactured by the manufacturing method of the present invention. The sputtering target is then replaced with a Cr target. The argon flow rate was set to 3 seem' methane (CH4) gas flow rate was set to 〇1 sccm, and the nitrogen dioxide gas flow rate was set to 5 seem. On the upper portion of the metal film of the blank mask for hard mask use, which is manufactured by conventional progressive film deposition techniques, is at 0. 4 kW power and 〇. A hard mask film was deposited under a pressure of 5 Pa for 12 seconds. The sheet resistance of the film measured on the surface of the hard mask film of the blank mask thus used for hard mask use was observed to be good using a 4-point detecting device, and the measured value was about 402 Ω. /□, and the thickness of the hard mask film measured by GXR is about 116 people. Although the blank mask, the reticle, and the method of manufacturing the same have been described with reference to the specific embodiments, it is not limited thereto. It will be apparent to those skilled in the art that various modifications and changes can be made without departing from the spirit and scope of the invention as defined by the appended claims. BRIEF DESCRIPTION OF THE DRAWINGS Exemplary embodiments can be understood in more detail from the above description in conjunction with the accompanying drawings in which: FIG. 1 is a cross section of a blank mask for hard mask use according to an exemplary embodiment. Figure. 2 is a schematic illustration of a long throw sputter apparatus for fabricating a blank mask for hard mask use, in accordance with an illustrative embodiment. 3 is a graph showing the results of composition ratio analysis of a film for a blank mask for hard mask use, according to an exemplary embodiment. 4 is a flow chart illustrating a process of a method of fabricating a blank mask for hard mask applications, in accordance with an illustrative embodiment. [Main component symbol description] 100: blank mask 110: transparent substrate 120: metal film/metal layer 130: hard mask film/hard mask 140: anti-rice film S200, S210, S220: step number 26