TWI243200B - Chemical/mechanical polishing slurry and chemical mechanical polishing method using the same - Google Patents
Chemical/mechanical polishing slurry and chemical mechanical polishing method using the same Download PDFInfo
- Publication number
- TWI243200B TWI243200B TW91133022A TW91133022A TWI243200B TW I243200 B TWI243200 B TW I243200B TW 91133022 A TW91133022 A TW 91133022A TW 91133022 A TW91133022 A TW 91133022A TW I243200 B TWI243200 B TW I243200B
- Authority
- TW
- Taiwan
- Prior art keywords
- slurry
- patent application
- scope
- layer
- oxide
- Prior art date
Links
Landscapes
- Mechanical Treatment Of Semiconductor (AREA)
- Finish Polishing, Edge Sharpening, And Grinding By Specific Grinding Devices (AREA)
Abstract
Description
1243200 ⑴ 疚、發鴨說明 (發明說明應敘明:發明所屬之技術頜域、先前技術、内容、實施方式及圖式簡單說明) 相關申j青案之交又參考 本申請案主張8/ 1 6/2 0 0 2申請之韓國專利申請案第 2002-48403之優先權,該申請案在此提出供參考。 發明技術領域 本發明係關於製造微電子裝置中所用之CMP(化學/機 械研除)研漿。尤其,本發明係關於可快速移除標地層且 使研除終止層有效的純化之具有南選擇性之C Μ P研漿。 發明背景 半導體裝置因多層連結而持續變得更整合且更小’且 因此需增加加工步驟之量以便在晶圓上形成多重導電層 或絕緣層。為減少半導體裝置製造中之步驟,因此通常使 用C Μ Ρ (其為化學及機械製程之結合)。C Μ Ρ係在1 9 8 0年後 其 由 IBM 公 司(international Business Machines Corporation)發展出。因為其發展使CMP在製造64 Mbit或 更大記憶體及非記憶體裝置之幾乎所有製造階段時用作 核心微處理技術。目前’ C M P對於製造下一代g丨g a b i t等級 DRAM記憶體或相等等級之非技藝裝置之應用有進一步 考量。 CMP為平整化製程之一類型。CMP中’係將不規則之晶 圓表面壓向相對旋轉之研除墊’同時使研磨研槳流到晶圓 及研除墊之接觸區。C M P經由化學及物理反應使不規則晶 圓表面平整。C Μ P之性能係由諸因素如C MP裝置之操作條 1243200 ⑺ 發日臟5¾驗' 該等因素中,C Μ P所用研漿之種類為影響研除性能之重 要因素。然而,對於傳統C Μ Ρ研漿,其係用於形成淺溝槽 分離(STI)或形成層間電介質(ILD)層,於其中形成使 D R A Μ之源/排區暴露之自動排列接觸孔,欲研除之目標 層(例如氧化物層)相對於研除終止層(例如氮化矽層)之 選擇性比為4 : 1,其選擇性並不良。因此,會使研除終止 層過度研除。為避免終止層之過度研除,因此需增加研除 終止層之厚度。而且,針對一般研漿,C Μ Ρ後留下之研除 終止層厚度會改變,導致不平整之晶圓表面且降低後續裝 置製造製程之寬容度。因此,會損及半導體裝置之特性。 因此,需要一種新穎、高選擇性之CMP研漿。 發明概要 本發明係針對一種製造具有高選擇性CMP研漿之方法 。本發明之C ΜΡ研漿具有高選擇性,可使目標層快速移除 ,且使研除終止層有效的鈍化。本發明之C Μ Ρ研漿之目標 層對研除終止層之選擇性為約30 : 1至約5 0 : 1。 本發明之一目的係提供一種研漿,包括金屬氧化物研 磨顆粒、移除加速劑、分子量約1,0 0 0至約1 0 0,0 0 0之陰離 子聚合物鈍化劑、C 1 - C 1 2陰離子鈍化劑及水。 較好,金屬氧化物研磨顆粒包括氧化鈽、氧化矽、氧〜 化链、氧化鈦、氧化錯及氧化鍺,或前述材料之任何結合 物。較好,移除加速劑包括鱗酸鹽化合物或亞鱗酸鹽化合 物。較好,陰離子性聚合物鈍化劑包括聚(丙烯酸)、聚 (甲基丙烯酸)、聚(丙烯酸-馬來酸)、聚(甲基丙烯酸-馬來 1243200 綱 0) 酸)、聚(丙烯酸-丙烯醯胺)、聚(丙烯腈-丁二烯-丙烯酸) 、聚(丙烯腈-丁二烯-甲基丙烯酸)或前述材料之衍生物或 鹽,或前述材料之任何結合物。 本發明之研漿尚可包括p Η控制劑。研漿之pH較好在約3 至約6之間。 依另一目的,本發明之C Μ P研漿係用於化學/機械研除 製程。例如,研漿可用於化學/機械研除其上沉積研除終 止層及欲研除之目標層。欲研除之目標層可為氧化物層, 且研除終止層可為氮化矽層。CMP製程包括將晶圓加於 CMP裝置上,且對晶圓表面上之目標層進行CMP,同時在 晶圓上之目標層表面與研除墊間施加’本發明之C ΜΡ研漿。 本發明之此等及其他目的、目標、特徵及優點由下列 詳細敘述及較佳具體例將變得顯而易見,本文係伴隨附圖 研讀。 附圖簡要敘述 圖1為說明製備本發明化學/機械研除(C ΜΡ)研漿之方 法之流程圖。 圖2為可使用本發明之CMP研漿進行CMP方法之CMP裝 置簡圖。 圖3為說明本發明之一目的之C Μ Ρ方法之流程圖。 圖4為說明移除速率變化作為磷酸氫胺量之函數之氧 化物(PE-TEOS)之圖。 圖5、6及7為說明氮化矽(3丨>〇與?£-丁£〇父移除速率及選 擇性變化作為使用本發明各種研漿之不同添加劑添加量 1243200 1_ (4) f日辨鮮 函數之圖。 圖8為說明SiN及PE-TEOS移除速率之變化作為研漿pH 函數之圖。 圖9為說明SiN及PE-TEOS之Γ電位作為研漿pH函數之 圖。 .', 圖1 0、1 1及1 2為說明以本發明之C MP研漿製備欲研除之 樣品晶圓之方法之剖面圖。 圖1 3為使用本發明之研漿進行CMP後之樣品晶圓之剖 面圖。 圖14為氧化物移除速率作為CMP研漿選擇性之函數之 圖。 圖1 5為說明溶解陰離子鈍化劑產生陰離子聚合物之方 法之反應圖之列舉圖示。 較佳具體例之詳細敘述 依據本發明之化學/機械研除(CMP)研漿及使用該研漿 之C MP法將參照附圖更進一步敘述。應了解本發明可包含 許多不同形式,且本文所述之具體例應不對本發明之範圍 造成任何限制。事實上,列舉之具體例係提供使該揭示更 徹底且完全,且充分涵蓋熟習本技藝者了解之本發明觀念 。需進一步了解附圖中,層及區域之厚度係經簡化且放大 使之更清晰,且所用相同之參考數目係指相同或類似之元 件。 通常,本發明之高選擇性C Μ P研漿較好包括研磨顆粒及 第一、第二及第三種添加劑之混合物,以調整研漿之特性 1243200 發:明就日月顧; <% ' " k VSW "7、、\' 、 \>、/ (5)1243200 说明 Guilt and explanation (Invention description should state: technical jaw field, prior technology, content, implementation mode, and drawings of the invention are briefly explained.) Refer to the claim of this application for the submission of the relevant application. 8/1 The priority of Korean Patent Application No. 2002-48403 filed on 6/2 0 0 2 is hereby incorporated by reference. FIELD OF THE INVENTION The present invention relates to a CMP (chemical / mechanical removal) slurry used in the manufacture of microelectronic devices. In particular, the present invention relates to a CMP slurry having a Southern selectivity that can be quickly removed from the target formation and effectively removed the terminating layer. BACKGROUND OF THE INVENTION Semiconductor devices continue to become more integrated and smaller due to multilayer connections, and therefore the number of processing steps needs to be increased in order to form multiple conductive or insulating layers on a wafer. In order to reduce the number of steps in the fabrication of semiconductor devices, C MP (which is a combination of chemical and mechanical processes) is often used. CMP was developed after 1980 by the International Business Machines Corporation. Because of its development, CMP is used as the core micro-processing technology in almost all manufacturing stages of manufacturing 64 Mbit or larger memory and non-memory devices. At present, CMP has further consideration for the application of manufacturing next-generation g a g a b i t level DRAM memory or equivalent non-technical devices. CMP is one type of planarization process. In the CMP, “the irregular surface of the wafer is pressed against the relatively rotating grinding pad” and the polishing paddle is flowed to the contact area between the wafer and the grinding pad. C M P smoothes the surface of irregular crystals through chemical and physical reactions. The performance of C MP is determined by various factors, such as the operation strip of the C MP device 1243200 ⑺ hair date dirty 5¾ '. Among these factors, the type of slurry used by C MP is an important factor affecting the performance of the removal. However, for conventional CMP slurry, it is used to form a shallow trench separation (STI) or an interlayer dielectric (ILD) layer in which an automatic alignment contact hole is formed that exposes the source / drain region of the DRA M. The selectivity ratio of the target layer (for example, the oxide layer) to the stop layer (for example, the silicon nitride layer) is 4: 1, and the selectivity is not good. Therefore, the research termination layer will be over-engineered. In order to avoid excessive removal of the termination layer, it is necessary to increase the thickness of the termination layer. Moreover, for general lapping, the thickness of the lapping stop layer left after CMP will change, resulting in uneven wafer surfaces and reducing the latitude of subsequent device manufacturing processes. Therefore, the characteristics of the semiconductor device are impaired. Therefore, a new and highly selective CMP slurry is needed. SUMMARY OF THE INVENTION The present invention is directed to a method for manufacturing a CMP slurry with high selectivity. The C MP slurry of the present invention has high selectivity, can quickly remove the target layer, and effectively passivate the removal termination layer. The selectivity of the target layer of the CMP slurry of the present invention to the removal of the termination layer is about 30: 1 to about 50: 1. An object of the present invention is to provide a slurry, including metal oxide abrasive particles, removal accelerator, anionic polymer passivating agent with a molecular weight of about 1,000 to about 100, 0 0, and C 1-C. 1 2 Anionic passivator and water. Preferably, the metal oxide abrasive particles include hafnium oxide, silicon oxide, oxygen chain, titanium oxide, oxide and germanium oxide, or any combination of the foregoing materials. Preferably, the removal accelerator includes a phosphonate compound or a phosphonate compound. Preferably, the anionic polymer passivation agent includes poly (acrylic acid), poly (methacrylic acid), poly (acrylic acid-maleic acid), poly (methacrylic acid-maleic 1243200 class 0) acid), poly (acrylic acid- Acrylamide), poly (acrylonitrile-butadiene-acrylic acid), poly (acrylonitrile-butadiene-methacrylic acid) or derivatives or salts of the foregoing materials, or any combination of the foregoing materials. The slurry of the present invention may further include a p Η control agent. The pH of the slurry is preferably between about 3 and about 6. According to another object, the CMP slurry of the present invention is used in a chemical / mechanical removal process. For example, the slurry can be used for chemical / mechanical removal of the deposition termination layer and the target layer to be removed. The target layer to be removed may be an oxide layer, and the stop layer may be a silicon nitride layer. The CMP process includes adding a wafer to a CMP device, and performing CMP on a target layer on the surface of the wafer, and applying 'the C MP slurry of the present invention between the surface of the target layer on the wafer and the pad. These and other objects, objects, features and advantages of the present invention will become apparent from the following detailed description and preferred specific examples, which are read in conjunction with the accompanying drawings. Brief Description of the Drawings Figure 1 is a flow chart illustrating a method for preparing a chemical / mechanical removal (CMP) slurry of the present invention. Fig. 2 is a schematic diagram of a CMP apparatus capable of performing a CMP method using the CMP slurry of the present invention. FIG. 3 is a flowchart illustrating a CMP method which is an object of the present invention. Figure 4 is a graph illustrating the change in removal rate of oxide (PE-TEOS) as a function of the amount of ammonium phosphate. Figures 5, 6 and 7 are illustrations of silicon nitride (3 丨 > 〇 and? £-££) parental removal rate and selectivity as different additives used in the various slurries of the invention. Addition amount 1243200 1_ (4) A graph of the daily discriminant function. Figure 8 is a graph illustrating the change in the removal rate of SiN and PE-TEOS as a function of the pH of the slurry. Figure 9 is a graph illustrating the Γ potential of SiN and PE-TEOS as a function of the pH of the slurry. ', Figures 10, 11 and 12 are cross-sectional views illustrating a method for preparing a sample wafer to be removed using the CMP slurry of the present invention. Figure 13 is a sample after CMP using the slurry of the present invention. A cross-sectional view of the wafer. Figure 14 is a graph of oxide removal rate as a function of CMP slurry selectivity. Figure 15 is a list of reaction diagrams illustrating the method of dissolving an anionic passivating agent to produce an anionic polymer. Detailed description of specific examples The chemical / mechanical milling (CMP) slurry according to the present invention and the CMP method using the slurry will be further described with reference to the drawings. It should be understood that the present invention may include many different forms and is described herein The specific examples should not limit the scope of the present invention in any way. Specific examples are provided to make the disclosure more thorough and complete, and fully cover the concept of the invention understood by those skilled in the art. It is necessary to further understand that in the drawings, the thickness of layers and regions has been simplified and enlarged to make it clearer and used The same reference number refers to the same or similar components. Generally, the highly selective CMP slurry of the present invention preferably includes a mixture of abrasive particles and first, second, and third additives to adjust the characteristics of the slurry. 1243200 Issue: Tomorrow and Sun; <% '" k VSW " 7 ,, \', \ >, / (5)
。尤其,本發明之C Μ P研漿包括金屬氧化物作為研磨顆粒 。適用之金屬氧化物包括氧化筛、氧化石夕、氧化紹、氧化 鈦、氧化锆、氧化鍺、或類似材料或前述材料之混合物。 較好,金屬氧化物研磨顆粒係與去離子水混合,使濃度以 研漿之總重為準約0.5%至約25% (wt)(此後寫成wt%)。若 金屬氧化物研磨顆粒之量低於0.5 w t %,則研除速率變得 太低。若金屬氧化物研磨顆粒之量大於2 5 w t %,則無法 確定研漿分散液之安定性。基於此等理由,上述濃度範圍 為研磨顆粒之較佳範圍。 第一種添加劑較好為移除加速劑。較好,移除加速劑 包括可增加移除欲研除之目標層之速率之材料。適用之移 除加速劑包含例如填酸鹽化合物或亞碌酸鹽化合物。較佳 之鱗酸鹽化合物包含例如填酸氫銨、墙酸二氫敍、鱗酸二 鼠舒、雙(2 -乙基己基)鱗酸鹽' 2 -胺基乙基二鼠填酸鹽、 4 -氣笨二偶氮六氟磷酸鹽、硝基苯二偶氮六氟磷酸鹽、六 *%·. In particular, the CMP slurry of the present invention includes metal oxides as abrasive particles. Suitable metal oxides include sieves, stone oxides, oxides, titanium oxides, zirconias, germanium oxides, or similar materials or mixtures of the foregoing materials. Preferably, the metal oxide abrasive particles are mixed with deionized water so that the concentration is about 0.5% to about 25% (wt) based on the total weight of the slurry (hereinafter referred to as wt%). If the amount of the metal oxide abrasive particles is less than 0.5 wt%, the removal rate becomes too low. If the amount of the metal oxide abrasive particles is more than 25 wt%, the stability of the slurry dispersion cannot be determined. For these reasons, the above-mentioned concentration range is a preferable range of the abrasive particles. The first additive is preferably a removal accelerator. Preferably, the removal accelerator includes materials that increase the rate at which the target layer to be removed is removed. Suitable removal accelerators include, for example, salt filling compounds or linoleates. Preferred phosphonium salt compounds include, for example, ammonium bisulfate, dihydroxanthate, diratharyl linoleate, bis (2-ethylhexyl) phosphonate '2-aminoethyl dirrolates, 4 -Airborne diazo hexafluorophosphate, nitrobenzene diazo hexafluorophosphate, six *% ·
氟磷酸銨、雙(2,4-二氣苯基)氣磷酸鹽、雙(2 -乙基己基) 氫磷酸鹽、氟磷酸鈣、氯磷酸二乙酯、氯硫代磷酸二乙酯 、六氟磷酸鉀、焦磷酸鹽、六氟磷酸四丁基銨、六氟磷酸 四乙基敍、或前述材料之任何結合物。較佳之亞填酸鹽化 合物包含例如雙(2 -乙基己基)亞磷酸鹽。 依其一具體例,移除加速劑之添加濃度以研漿之總重 為準,較好為約0.0 0 1 w t %至約1 0 w t %。若移除加速劑之 量低於0.0 0 1 w t %,則研除速率變得太低。若移除加速劑 之量超過1 0 w t %,則研漿分散液之安定性變得不良,且 -10 - 1243200 (6) 研除速率無法控制。基於此等理由,上述濃度 加速劑之較佳範圍。依另一具體例,移除加速 較好在約0.0 1 w t %至約1 w t %之間。 針對第二種添加劑,較好使用研除終止層 其亦可提升研磨顆粒之分散安定性,且調整研 。當研除終止層為氮化矽層,且目標層為氧化 好使用陰離子聚合物鈍化劑作為第二種添加裔 合物鈍化劑之分子量基於研聚之流動性,較4 約1 0 0,0 0 0。適用之陰離子聚合物鈍化劑包含 (例如聚(丙烯酸)及聚(曱基丙烯酸)),共聚物 烯酸-馬來酸)、聚(曱基丙烯酸-馬來酸)、及聚 烯醯胺)),三聚物(例如聚(丙烯腈-丁二烯-择 (丙烯腈-丁二烯-曱基丙烯酸)),或前述材料之 ,或前述材料之任何結合物。 依其一具體例,陰離+聚合物鈍化劑之添 之總重為準,較好約0.0 0 1 w t %至約1 0 w t %。 合物鈍化劑之量低於0.0 0 1 w t %,則鈍化劑無 子聚合物鈍化劑之量超過約1 0 w t %,則研除 低。依另一具體例,陰離子聚合物鈍化劑之添 之總重為準約0.0 1 w t %至約1 w t %。 較好使用鈍化劑作為提昇研除終止層之鈍 三種添加劑。當研除終止劑為氮化矽層時 C 1 - C 1 2陰離子鈍化劑。適用之C 1 - C 1 2陰離子 例如苯二酸S旨化合物如苯二酸氫舞。 mmm :ί: 範圍為移除 劑之添加量 之對話劑, 漿之流動性 物層時,較 4。陰離子聚 r子約1,0 0 0至 例如均聚物 (例如聚(丙 (丙稀酸-丙 3烤酸)、聚 衍生物或鹽 加量以研漿 若陰離子聚 效,若陰離 速率變得太 加量以研漿 化作用之第 ,較好使用 鈍化劑包含 1243200 _ ⑺ 杳明顯職 依其一具體例,C 1 - C 1 2陰離子鈍化劑之添加量以研漿 之總重為準,較好約0 . 0 0 1 w t %至約1 0 w t %。若C 1 - C 1 2陰 離子鈍化劑之量低於約0.0 1 w t %,則無法確保高的研除選 擇性。若C 1 - C 1 2陰離子鈍化劑之量超過約1 0 w t %,則研 除速率變得太低。基於此等理由,上述濃度範圍為移除加 速劑之較佳範圍。依另一具體例,鈍化劑(第三種添加劑) 之添加量以研漿之總重為準較好約0.0 1 w t %至約1 w t %。 本發明另一具體例之CMP研漿尚包括pH控制劑,以提 供研漿最佳之p Η。研漿之p Η隨著目標層及研除終止層之 成份而變。當欲研除之目標層為氧化物層,且研除終止層 為氮化矽層,則較好將研漿之ρ Η調整在約3至約6之間。 適用之pH控制劑包含鹼如氫氧化鉀(ΚΟΗ)、氫氧化銨 (NH4〇H)、氫氧化鈉(Na〇H)、四曱基銨氫氧化物(TMAH) 或膽驗。另外,適用之ρ Η控制劑包含如琉酸(Η 2 S〇4 )、鹽 s复(hci)、磷酸(η3ρ〇4)、硝酸(ην〇4)、或乙酸(CH3C〇〇H)。 較好,本發明之CMP研漿水溶液中,由於第一、第二及 第三種添加劑之作用,因此會使目標層之移除速率增加, 且可使研除終止層更有效的鈍化,因此獲得高的研除選擇 性。例如,藉甴使用移除加速劑作為第一種添加劑,欲研 除之目標層(例如氧化物層)之移除速率可調整在約2,0 0 0 至約6,0 0 0埃/分鐘之間。 其次,藉由使用陰離子性聚合物鈍化劑(較好分子量在 約1 , 0 0 0至約1 0 0,0 0 0之間)作為第二種添加劑,將陰離子 聚合物溶於水性C Μ P研漿中作為鈍化劑,且亦限制研磨顆 1243200 ⑻ 粒在整體相中之移動性,影響研漿之流動性。Ammonium fluorophosphate, bis (2,4-diphenylphenyl) aerophosphate, bis (2-ethylhexyl) hydrogen phosphate, calcium fluorophosphate, diethyl chlorophosphate, diethyl chlorothiophosphate, six Potassium fluorophosphate, pyrophosphate, tetrabutylammonium hexafluorophosphate, tetraethylammonium hexafluorophosphate, or any combination of the foregoing. Preferred underfilling compounds include, for example, bis (2-ethylhexyl) phosphite. According to a specific example, the additive concentration for removing the accelerator is based on the total weight of the slurry, and is preferably about 0.01 to 10 wt%. If the amount of the accelerator removed is less than 0.001 wt%, the removal rate becomes too low. If the amount of accelerator removal is more than 10 wt%, the stability of the slurry dispersion becomes poor, and -10-1243200 (6) the removal rate cannot be controlled. For these reasons, a preferable range of the above-mentioned concentration accelerator. According to another specific example, the removal acceleration is preferably between about 0.01 wt% and about 1 wt%. For the second kind of additive, it is better to use a research removal layer, which can also improve the dispersion stability of the abrasive particles, and adjust the research. When the stop layer is a silicon nitride layer and the target layer is oxidized, the molecular weight of the anionic polymer passivating agent used as the second additive passivating agent is based on the mobility of the research polymer, which is about 1 0 0,0 compared to 4 0 0. Suitable anionic polymer passivating agents include (such as poly (acrylic acid) and poly (fluorenyl acrylic acid)), copolymers of enoic acid-maleic acid), poly (fluorenyl acrylic acid-maleic acid), and polyoxamine) ), Terpolymer (such as poly (acrylonitrile-butadiene-selective (acrylonitrile-butadiene-fluorenyl acrylic acid)), or one of the foregoing materials, or any combination of the foregoing materials. According to one specific example, The total weight of the anion + polymer passivating agent is based on the total weight, preferably about 0.01 to 1 wt%. If the amount of the compound passivating agent is less than 0.0 0 1 wt%, the passivating agent is polymerized without particles. If the amount of chemical passivation agent exceeds about 10 wt%, the research will be low. According to another specific example, the total weight of the anionic polymer passivation agent is about 0.0 1 wt% to about 1 wt%. Passivation is preferably used As an inactive three additives to improve the removal of the terminating layer. When the terminating agent is a silicon nitride layer, C 1-C 1 2 anion passivation agent. Suitable C 1-C 1 2 anions such as phthalic acid S purpose compounds such as Hydrogen phthalate. Mmm: ί: Dialogue agent in the range of the amount of remover added, compared with 4 when the slurry is a fluid layer. Anionic poly-r is about 10,000 to, for example, a homopolymer (such as poly (propyl (propyl-acrylic acid) -propionic acid), a poly derivative or a salt added to grind the slurry. If the anion polymerizes, if the anion rate Become too large to rank first for slurrying, it is better to use a passivator containing 1243200 _ 杳 杳 Obviously according to one specific example, the amount of C 1-C 1 2 anionic passivating agent is based on the total weight of the slurry It is preferably about 0.01 wt% to about 10 wt%. If the amount of the C 1 -C 1 2 anionic passivator is less than about 0.01 wt%, a high selectivity cannot be ensured. The amount of C 1-C 1 2 anionic deactivator exceeds about 10 wt%, and the removal rate becomes too low. For these reasons, the above-mentioned concentration range is a preferred range for removing the accelerator. According to another specific example The addition amount of the passivating agent (the third additive) is preferably about 0.01 wt% to about 1 wt% based on the total weight of the slurry. The CMP slurry of another specific example of the present invention further includes a pH control agent. Provide the best p 研 of the slurry. The p Η of the slurry varies with the composition of the target layer and the removal termination layer. When the target layer to be removed is an oxide layer, and The removal termination layer is a silicon nitride layer, and the ρ Η of the slurry is preferably adjusted between about 3 and about 6. A suitable pH control agent includes a base such as potassium hydroxide (K0Η), ammonium hydroxide (NH4). H), sodium hydroxide (NaOH), tetramethylammonium hydroxide (TMAH) or biliary test. In addition, suitable ρ Η control agents include, for example, sulphuric acid (Η 2 S〇 4), salt s complex ( hci), phosphoric acid (η3ρ〇4), nitric acid (ην〇4), or acetic acid (CH3COOH). Preferably, in the CMP slurry aqueous solution of the present invention, due to the effects of the first, second, and third additives, the removal rate of the target layer can be increased, and the termination layer can be more effectively passivated. Obtain high selectivity. For example, by using a removal accelerator as the first additive, the removal rate of the target layer (such as an oxide layer) to be removed can be adjusted from about 2,000 to about 6,000 angstroms / minute. between. Secondly, by using an anionic polymer passivating agent (preferably having a molecular weight between about 1,000 and about 100,000) as the second additive, the anionic polymer is dissolved in the aqueous CMP As a passivating agent in the slurry, it also restricts the mobility of the abrasive particles 1243200 3200 particles in the overall phase, affecting the flowability of the slurry.
之陰 子聚 .〇相 面排 正電 時可 中之 體相 同步 子聚 離子 鈍化 中, 氧化 化物 中之 點’ 薄膜 之pH 溶於水性研漿中之陰離子聚合物鈍化劑產生大量 離子聚合物,如圖1 5中所述之反應圖中之說明。陰離 合物藉由靜電吸附力,強力的吸收於正電荷表面之J: 反的,陰離子性聚合物由於靜電排斥力,受負電荷表 斥。因此,C Μ P研漿中之陰離子聚合物吸收於鈍化之 荷表面上。因此,晶圓之負電荷區可選擇性研除,同 保護正電荷區免於研除。·如上述,溶於水性C Μ Ρ研漿 陰離子聚合物可作為鈍化劑,亦可限制研磨顆粒在整 中之移動,影響研聚之流動性。當晶圓之表面經歷不 驟時,可快速移除晶圓之大步驟區,同時可藉由陰離 合物鈍化劑緩慢移除小步驟區。較好使用C 1 - C 1 2陰 鈍化劑作為第三種添加劑,以提昇上述陰離子聚合物 劑之純化作用。 事實上,C 1 - C 1 2陰離子鈍化計較好溶於水性研漿 且帶負電,例如當使用氫苯二酸鉀時為C 8 Η 4〇4 2 -。 為增加使用鈍化劑功能之研除選擇性,欲研除之 物目標層在水性研漿中應帶負電,且水性研漿中之氮 研除終止層應帶正電。水性研漿卡之薄膜表面在研漿 pH低於等電點時,為帶正電,當研漿之pH大於等電 則帶負電。因此,藉由控制研漿之ρ Η,可使二不同 充電成相反極性。因此,依本發明之一具體例,研漿 較好使用ρ Η控制劑調整至最佳水準。 為增加加工寬容度且降低碟化現象之發生,較好調整 1243200 翻響獅 (9) 本發明之C Μ P研漿,使目標層對研除終止層之選擇性在約 3 0 : 1至約5 0 : 1之間。 此後,將參考圖1敘述製備本發明具體例之高選擇性 C Μ P研漿之方法。 首先,將研磨顆粒置於含去離子水(步驟S 1 0)之高剪力 混練機中。使去離子水及研磨顆粒預先混合(步驟S 1 2 )。 預先混合(步驟S 1 2)中,研磨顆粒之量以最終研漿之總重 為準,係調整在約0.5 wt%至約0.2 5 wt%之間,且研漿之 p Η係調整在弱酸範圍及弱驗範圍中。 使用泵浦,將研聚混合物移入適當之分散裝置(例如介 質研磨機或超高壓分散裝置)中,接著在高壓下充分的分 散(步驟S 1 4)。任何分散裝置均可使用。然而,考量各種 因素,如分散液形成之再製性,降低分散過程中之污染, 及分散後之平均粒徑與分散性,較好使用超高壓分散裝置 。當考量耐久性時,較好超硬分散室(其為超高壓分散裝 置之核心部分)係由鑽石形成。高壓分散製程之過程中, 係之加約1 0,0 0 0 p s i至約2 0,0 0 0 p s i之適當壓力。若壓力低 於該範圍,則分散效力變得太低。若壓力大於該範圍,則 會對系統效力及室之耐久性產生負面影響。 經由超高壓分散控制平均研漿粒徑後,可進一步添加 去離子水,以適當的控制研漿中研磨顆粒之濃度。 接著,添加第一、第二及第三種研磨料,以達到所需 之研漿特性(步驟S 1 6)尤其,移除加速劑(其協助移除欲研 除之目標層)之添加量約為0.0 0 1 w t %至約1 0 w t %。陰離子 14 1243200 _ no) 丨 fans 聚合物鈍化劑及C 1 - C 1 2陰離子鈍化劑之添加量各在約 0.0 1 w t % 至約 1 0 w t % 之間。 當添加第一種、第二種及第三種添加劑無法使研漿到 最佳之p Η時,則可視情況添加p Η控制劑,以獲得最佳p Η。 添加以製備本發明C Μ Ρ研漿之添加劑量及/或ρ Η控制劑 量較好藉由水性C Μ Ρ研漿間之作用決定,且係針對最佳之 研除作用最佳化。 當添加劑混合且調整研漿之ρ Η後,使研漿通過過濾器 ,快速移除相對大之顆粒(步驟S 1 8)。該過濾步驟降低研 除過程中目標層表面上發生之刮痕。研漿過濾製成後,分 析研漿以設定其一般物理特性及性能。 隨後將參考例如圖2及3敘述本發明各具體例之CMP方 法。 參考圖2,在CMP裝置中,使其上具有研除墊21之研除 板2 0與以馬達(未顯示)旋轉之第一旋轉軸2 2相連且繞著 其旋轉。將研除頭3 0置於研除墊2 1之上,且附接於以馬達 (未顯示)旋轉之第二旋轉軸3 2之上。該研除頭係以與研除 板20相反之旋轉方向旋轉。晶圓36(欲進行CMP)係藉由裝 置在研除頭30表面上之夾具34可分離的固定於研除頭30 上。由研漿供給管線4 0將本發明之研漿4 2供給於研除板2 0 之一邊。 依本發明之CMP製程(參考圖2及3),製備欲進行CMP之 晶圓及C Μ Ρ研漿4 2 (步驟S 5 0 )。接著,將晶圚3 6固定於研 除頭3 0上,且將C ΜΡ研漿4 2供給於晶圓3 6 (步驟S 5 2 )上。 1243200 (ii) 晶圓3 6可包含欲研除之目標層之氧化物層及作為研除終 止層之氮化石夕層。本發明之高選擇性C Μ P研漿可施加於將 進行淺溝分離(S ΤI)及層間電介質(丨L D)膜形成之晶圓上 。依現行s TI及IL D模型成製程,氧化物層(如以高密度電 漿化學蒸氣沉積(H D P C V D)形成之高密度電漿氧化物層 ,或以電漿提昇之化學蒸氣沉積(PECVD)形成之電漿提昇 之四乙基原矽酸鹽(PE-TEOS)層))為欲研除之層,且氮化 矽層(其係由壓力為數百mTorr或更低之低壓化學蒸氣沉 積(LPCVD)形成,或.在約5〇〇°C至約600 °C之高溫下以 PEC VD形成,功能如同研除終止層。 當本發明之高選擇性研漿供給至晶圓(步驟S 5 2)後,研 除溝槽充填層或IL D層(步驟S 5 4 )。 本發明將參考下列實驗例更詳細敘述。熟習本技藝者 慣闬且可輕易推斷之技術資料並未充分敘述於下列實驗 例中。下列實驗例係說明用’並非企圖限制本發明之範 圍。 實驗例1 該實驗中’以圖1中所述之方法為準,各種研漿係藉由 將不同量之移除加速劑磷酸氫銨(A Η P )添加於含1 w t %第 一種添加劑氧化鈽(金屬氧化物研磨顆粒)之水溶液中製 備。 PE-丁EOS空白晶圓(各具有8 0 0 0埃厚度之氧化物 (P E - T E〇S )層)係如樣品晶圓般製備。 接著在樣品晶圓上執行,同時將含不同量A Η p (如表1 1243200The anion polymer in the positive phase. When the surface is positively charged, the phase synchronizer in the polyion passivation, the point in the oxide's film pH. The anionic polymer passivating agent dissolved in the aqueous slurry produces a large number of ionic polymers. This is illustrated in the reaction diagrams described in FIG. 15. Anionic compounds are strongly absorbed on positively charged surfaces by electrostatic adsorption. J: Inversely, anionic polymers are repelled by negative charges due to electrostatic repulsion. Therefore, the anionic polymer in the CMP slurry is absorbed on the passive surface. Therefore, the negatively-charged regions of the wafer can be selectively removed, while the positively-charged regions are protected from being removed. · As mentioned above, the anionic polymer soluble in aqueous CMP slurry can be used as a passivating agent, and it can also limit the movement of the abrasive particles in the whole, affecting the flowability of the polymerization. When the surface of the wafer is inexperienced, the large step area of the wafer can be quickly removed, and the small step area can be slowly removed by the anion passivation agent. C 1-C 1 2 anion passivation agent is preferably used as the third additive to improve the purification effect of the above anionic polymer agent. In fact, the C 1 -C 1 2 anionic passivation meter is better soluble and negatively charged in aqueous slurry, for example, C 8 Η 40 4 2-when using potassium hydrophthalate. In order to increase the selectivity of the passivation function, the target layer to be removed should be negatively charged in the aqueous slurry, and the nitrogen removal termination layer in the aqueous slurry should be positively charged. The surface of the water-based slurry card is positively charged when the pH of the slurry is lower than the isoelectric point, and it is negatively charged when the pH of the slurry is greater than the isoelectric point. Therefore, by controlling the ρ 研 of the slurry, two different charges can be made to have opposite polarities. Therefore, according to a specific example of the present invention, it is preferable to adjust the pulverizing agent to an optimal level using a ρρ control agent. In order to increase the processing latitude and reduce the occurrence of dishing, it is better to adjust 1243200 Fansang Lion (9) The CMP slurry of the present invention, so that the selectivity of the target layer to remove the termination layer is about 30: 1 to Between about 50: 1. Hereinafter, a method for preparing a highly selective CMP slurry of a specific example of the present invention will be described with reference to FIG. First, the abrasive particles are placed in a high-shear kneader containing deionized water (step S 10). The deionized water and the abrasive particles are mixed in advance (step S 1 2). In the pre-mixing (step S 1 2), the amount of the abrasive particles is based on the total weight of the final slurry, which is adjusted between about 0.5 wt% and about 0.2 5 wt%, and the p system of the slurry is adjusted to a weak acid. Range and weak test range. Using a pump, move the ground polymer mixture into a suitable dispersing device (such as a media grinder or ultra-high pressure dispersing device), followed by sufficient dispersing under high pressure (step S 1 4). Any dispersion device can be used. However, considering various factors, such as the reproducibility of the dispersion, reducing the pollution during the dispersion process, and the average particle size and dispersibility after dispersion, it is better to use an ultra-high pressure dispersion device. When durability is considered, it is preferable that the ultra-hard dispersion chamber (which is the core part of the ultra-high-pressure dispersion apparatus) is formed of diamond. During the high-pressure dispersion process, an appropriate pressure of about 10, 0 0 0 p s i to about 20, 0 0 0 p s i is added. If the pressure is lower than this range, the dispersion efficiency becomes too low. If the pressure is greater than this range, it will negatively affect the system efficiency and the durability of the chamber. After the average slurry size is controlled by ultra-high pressure dispersion, deionized water can be further added to appropriately control the concentration of abrasive particles in the slurry. Next, add the first, second, and third abrasives to achieve the required slurry properties (step S 1 6). In particular, remove the amount of accelerator (which helps remove the target layer to be removed). From about 0.01 wt% to about 10 wt%. Anion 14 1243200 _ no) 丨 fans The amount of polymer passivating agent and C 1-C 1 2 anionic passivating agent are each about 0.0 1 w t% to about 10 w t%. When the first, second, and third additives are not added to the optimum p Η, then the p Η control agent may be added as appropriate to obtain the optimal p 可视. The amount of addition and / or ρ Η controlling agent added to prepare the CMP slurry of the present invention is preferably determined by the effect of the aqueous CMP slurry, and is optimized for the best removal effect. After the additives are mixed and the pH of the slurry is adjusted, the slurry is passed through a filter to quickly remove relatively large particles (step S 1 8). This filtering step reduces scratches that occur on the surface of the target layer during the study. After the slurry has been filtered, the slurry is analyzed to set its general physical properties and properties. The CMP method of each specific example of the present invention will be described later with reference to Figs. 2 and 3, for example. Referring to Fig. 2, in a CMP apparatus, a grinding plate 20 having a grinding pad 21 is connected to a first rotation shaft 22 rotated by a motor (not shown) and rotated therearound. The grinding head 30 is placed on the grinding pad 21 and attached to a second rotation shaft 32 which is rotated by a motor (not shown). The grinding head is rotated in a direction opposite to that of the grinding plate 20. The wafer 36 (to be subjected to CMP) is detachably fixed to the polishing head 30 by a jig 34 provided on the surface of the polishing head 30. The slurry 42 of the present invention is supplied to one side of the slurry removing plate 20 through the slurry supply line 40. According to the CMP process of the present invention (refer to FIGS. 2 and 3), a wafer to be subjected to CMP and a CMP slurry 4 2 are prepared (step S 50). Next, the wafer 36 is fixed to the research head 30, and the CMP slurry 42 is supplied to the wafer 36 (step S52). 1243200 (ii) The wafer 36 may include the oxide layer of the target layer to be removed and the nitride nitride layer as the stop layer for removal. The highly selective CMP slurry of the present invention can be applied to a wafer that will be formed with shallow trench separation (STI) and interlayer dielectric (LD) films. According to the current s TI and ILD model formation process, oxide layers (such as high-density plasma chemical vapor deposition (HDPCVD) formed high-density plasma oxide layer, or plasma-lifted chemical vapor deposition (PECVD) formed The plasma-lifted tetraethylorthosilicate (PE-TEOS) layer) is the layer to be removed, and the silicon nitride layer (which is deposited by low-pressure chemical vapor with a pressure of several hundred mTorr or lower ( LPCVD) formation, or. PEC VD formation at a high temperature of about 500 ° C to about 600 ° C, which functions like removing a termination layer. When the highly selective slurry of the present invention is supplied to a wafer (step S 5 2) After that, the trench filling layer or the IL D layer is removed (step S 5 4). The present invention will be described in more detail with reference to the following experimental examples. Technical information that is familiar to the skilled person and can be easily inferred is not fully described in In the following experimental examples, the following experimental examples are intended to illustrate the use of 'not intended to limit the scope of the present invention. Experimental Example 1 In this experiment', the method described in FIG. 1 shall prevail. Accelerator ammonium hydrogen phosphate (A Η P) added to 1 wt% of the first additive Prepared in an aqueous solution of thallium (metal oxide abrasive particles). PE-butyl EOS blank wafers (each oxide (PE-TEOS) layer with a thickness of 8000 angstroms) are prepared as sample wafers. Then execute it on the sample wafer, and at the same time will contain different amounts of A Η p (see Table 1 1243200
中所示)之研漿加於其上’且測量氧化物移除速率。c M P 係使用Μ I R R A - Μ E S Α裝置(R〇D E L· C 〇 .,L t d ),在下列條件 下進行:I C 1 0 0 0堆積墊,薄膜力為3 . 5 p s i,維持之環力為 4.5 psi,且内管力為3.5 psi,轉檯速度為HO rpm’頭速 度為1 0 4 r p m,且研漿流速為3 0 0 0毫升/分鐘。C MP結杲列 於表1中。圖4為結果之圖示。 表1 AHP, wt% 氧化物(PE-TEOS)移除速率,埃/分鐘 pH 〇 1655 5 0.05 4602 5 0.1 14499 5 0.5 2875 5 由表1及圖4可清楚看出’相較於未添加A Η P,當添加 A Η Ρ時,會增加氧化物移除速率。然而,當A Η Ρ之添力σ量 進一步增加時或使氧化物移除速率下降。該結杲相信係由 於過量之磷酸鹽基溶於研漿中,且吸附在研磨顆粒上阻礙 C Μ Ρ所致,此可由當添加過量A Η Ρ時’會使研磨顆粒沉澱 推斷。考量適當之氧化物移除速率及預防研磨顆粒沉澱, A Η Ρ之較佳添加量經測定約1 〇 w t %或更低,且較好約1 w t % 〇 實.驗例2 該實驗中,藉由添加〇 . 2 5 w t %之A Η P及不同量之聚(曱 基丙烯酸)(Ρ Μ A )作為陰離子聚合物鈍化劑於含1 w t %氧 化鈽之水溶液中製備各種研漿。 製備PE-TEOS及氮化矽(Si3N4)對照晶圓作為樣品晶圓 ,其中氧化物(Ρ E - T E〇S )層及氮化矽層之沉積厚度分別為 1243200 ι_ι (13) | ·獄明:瘦頁 8,000 埃及 2,000 埃。 在樣品晶圓上進行C Μ Ρ,同時使用與實驗例1相同C Μ Ρ 裝置及在相同操作條件下,供給以表2組合物製備之研漿 。C Μ Ρ過程中測量氧化物及氮化矽移除速率。結果列於下 表2中。圖5為結果之圖示說明。 表2 AHP, wt% PM A, wt% pH 移除速率,埃/分鐘 -選擇性 氧化物 (PE-TEOS) 氮化矽 (Si3N4) 0.25 0.05 5 3993 1078 3.7:1 0.25 0.1 5 3706 975 3.8:1 0.25 0.5 5 2448 679 3.6:1 0.25 1 5 1617 478 3.4:1(Shown in) on top of it 'and the oxide removal rate was measured. c MP is performed using Μ IRRA-Μ ES Α device (RODEL.CO., L td) under the following conditions: IC 1000 stacking pad, film force is 3.5 psi, the ring force is maintained It is 4.5 psi, the inner tube force is 3.5 psi, the turntable speed is HO rpm, the head speed is 104 rpm, and the slurry flow rate is 300 ml / min. C MP results are listed in Table 1. Figure 4 is a graphical representation of the results. Table 1 AHP, wt% oxide (PE-TEOS) removal rate, Angstroms / minute pH 〇1655 5 0.05 4602 5 0.1 14499 5 0.5 2875 5 As can be clearly seen from Table 1 and Figure 4, compared to the absence of A Y P, when A Y P is added, it will increase the oxide removal rate. However, when the amount of additive force σ of A Y P is further increased or the oxide removal rate is decreased. This result is believed to be caused by the excess phosphate group dissolving in the slurry and being adsorbed on the abrasive particles to hinder C MP, which can be inferred from the fact that the abrasive particles are precipitated when an excessive amount of A PP is added. Taking into account the appropriate oxide removal rate and the prevention of the precipitation of abrasive particles, the preferred addition amount of AΗP is determined to be about 10 wt% or less, and preferably about 1 wt%. Test Example 2 In this experiment, Various slurries were prepared by adding 0.2 wt% AAP and different amounts of poly (fluorenyl acrylic acid) (PMA) as anionic polymer passivating agents in an aqueous solution containing 1 wt% hafnium oxide. Preparation of PE-TEOS and silicon nitride (Si3N4) control wafers as sample wafers, in which the thickness of the oxide (PE-TE0S) layer and the silicon nitride layer were 1243200 ι_ι (13) | : Thin page 8,000 Egypt 2,000 Angstroms. C MP was performed on the sample wafer, while using the same C MP device as in Experimental Example 1 and under the same operating conditions, a slurry prepared with the composition of Table 2 was supplied. Oxide and silicon nitride removal rates were measured during the CMP process. The results are listed in Table 2 below. Figure 5 is a graphical illustration of the results. Table 2 AHP, wt% PM A, wt% pH removal rate, Angstroms / minute-selective oxide (PE-TEOS) silicon nitride (Si3N4) 0.25 0.05 5 3993 1078 3.7: 1 0.25 0.1 5 3706 975 3.8: 1 0.25 0.5 5 2448 679 3.6: 1 0.25 1 5 1617 478 3.4: 1
由表2及圖5可清楚看出,當作為陰離子聚合物純化劑 之Ρ Μ Α量增加時,氮化矽移除速率會下降。證明陰離子聚 合物鈍化劑之作用。然而,如上述,增加陰離子聚合物鈍 化劑之量會使氧化物移除速率直線下降。對僅使用PM A 作為鈍化劑之所有樣品晶圓,選擇性相對的較低。 實驗例3 該實驗中,係添加0.2 wt %之ΑΗΡ及不同量之苯二酸氫 甲(Ρ Η Ρ )作為C卜C 1 2陰離子鈍化劑至1 w t %氧化鈽之水溶 液中製備各種研漿。 製備如實驗例2相同之樣品晶圓,且在如實驗例2之相 同C Μ P裝置及相同操作條件下進行C Μ P,C Μ P過程中測量 氧化物及氮化物移除速率及選擇性。結果列於下表3中。 圖6為結果之圖示說明。 -18- 1243200 難:葬:猶、; > ϊ-t V ·ν ν>«· V ν > (14)It is clear from Table 2 and FIG. 5 that as the amount of PMA as an anionic polymer purification agent increases, the silicon nitride removal rate decreases. Prove the effect of anionic polymer deactivator. However, as mentioned above, increasing the amount of anionic polymer passivating agent causes the oxide removal rate to decrease linearly. For all sample wafers using only PMA as a passivator, the selectivity is relatively low. Experimental Example 3 In this experiment, 0.2 wt% AΗP and different amounts of hydrogen phthalate (P Η P) were added as Cb C 1 2 anionic passivating agent to a 1 wt% hafnium oxide aqueous solution to prepare various slurries. . The same sample wafer as in Experimental Example 2 was prepared, and the same MP device and the same operating conditions as those in Experimental Example 2 were used to perform C MP. The oxide and nitride removal rate and selectivity were measured during the C MP process. . The results are listed in Table 3 below. Figure 6 is a graphical illustration of the results. -18- 1243200 Difficult: Funeral: Still, > ϊ-t V · ν ν > «· V ν > (14)
AHP, wt% PHP, wt% pH 移除速幸,埃/分鐘 選擇性 氧化物 (PE-TEOS) 氮化矽 (Si3N4) 〇 0 5 1242 542 2.3:1 0.2 〇 5 2668 1206 1 2.2:1 0.2 0.01 5 2615 1155 2.3:1 0.2 0.1 5 2668 899 3.0:1 0.2 0.25 5 2744 772 3.6:1 由表3及圖6可清楚看出,用作C 1 - C 1 2陰離子鈍化劑之 Ρ Η P量增加時,會使氮化矽移除速率下降。證明陰離子鈍 化劑之鈍化作用。而且,、相較於為增加陰離子鈍化劑之個 案,增加陰離子鈍化劑之量會使氧化物移除速率稍增加。 如實驗例2(其中僅使用一類鈍化劑(ΡΜΑ))般,當僅添加 Ρ Η Ρ作為鈍化劑時,所有試驗樣品中之選擇性相對較低。 實驗例4 該實驗中,以實驗例1、2及3之結果為準,藉由添加0.0 5 w t %之A Η Ρ、Ρ Μ Α及Ρ Η Ρ於含1 w t %氧化铞之水溶液中製備 各種研漿。PMA之量固定在0.05wt%,且改變PHP之量製 備不同研漿。 製備如實驗例2之相同樣品晶圓,且使用如實驗例2相 同之CMP裝置且在相同操作條件下進行CMP。CMP過程中 測量氧化物及氮化矽移除速率及選擇性。結果列於下表4 中。圖7為結果之圖示說明。 -19 - 1243200 (15) 表4 ΑΗΡ, wt% PMA, wt% PHP, wt% pH 移除速率,埃/分鐘 選擇性 氧化物 (PE-TEOS) 氮化矽 (Si3N4) 0.05 0.05 0.01 5 4641 82 57:1 0.05 0.05 0.05 5 4890 75 65:1 0.05 0.05 0.1 5 4960 62 80:1 0.05 0.05 0.5 5 3902 41 95:1 由表4及圖7可清楚發現,當二類陰離子鈍化劑(Ρ Μ A及 Ρ Η P ) —起使用時,會獲得高的氧化物對氮化矽之選擇 性。 實驗例5 該實驗例中,藉由添加0.0 5 w t % A Η Ρ、0.0 5 w t % Ρ Μ A 及0.0 5 w t % Ρ Η P於含1 w t %氧化筛之水溶液中,接著以 1 Μ-H2S〇4溶液及氫氧化鉀標準溶液調整pH,製備不同pH 等級之各種研漿。 如實驗例2般製備相同之樣品晶圓,且使用如實驗例2 般相同之CMP裝置且在相同操作條件下進行CMP。CMP 過程中測量氧化物及氮化矽移除速率。結果列於下表5中 。圖8為結果之圖示說明。 表5 pH 移除速率,埃/分鐘 選擇性 氧化物 (PE-TE0S) 氮化矽 (Si3N4) 4 3544 40 89:1 4.5 3904 44 89:1 5 4214 61 69:1 5.5 4458 82 54:1 6 5190 129 40:1 7 4750 1197 4:1 -20 - 1243200 一一…… (16) 袭明麗南:嫌義, 由表5及圖8可清楚發現,當研漿之ρ Η增加時,氧化物 (PE-TEOS)及氮化矽之移除速率會增加,且當pH到達且超 過約6時,氮化矽之移除速率會明顯增加。 另外,當ρ Η為約6或更高時,選擇性會突然的下降。顯 然的,ρ Η為6或更低之C Μ Ρ研紧對於增加氧化物對氮化石夕 之選擇性有效。在較高ρ Η下選擇性下降係由於稀負載氣 化矽層上之鈍化劑較少所致。基本上,此係由於鈍化劑之 分解度及氮化矽層之表面電荷隨著ρ Η而變所致。 針對分解度,ΡΗΡ之官能基-COOH及ΡΜΑ之分解常數約 為4.2,因此在pH為5或更低下可完全分解。然而,當pH 值為中性化鹼性ρ Η時,分解度下降。因此,鈍化中所含 之較少負電荷會由鈍化劑產生,使選擇性下降。 依氮化矽表面電荷之觀點,選擇性之下降可解釋如下 。層之固有表面電荷可以以r(Q電位,mV)表示。層中C 電位之符號改變時之溶液ρ Η稱之為等電點。如圖9中所示 ,HDPC VD或PEC VD氧化物層之等電點約為pH 3,且 LPCVD或高溫PECVD氮化矽層約為pH 6。當研漿之pH低 於層之等電點時,層之表面在研漿中帶正電。當研漿之pH 大於層之等電點時,層之表面在研漿中帶負電。因此,當 研漿中之pH約為3至6時,氮化矽層之表面帶正電,且氧 化物層之表面帶負電。因此,當陰離子聚合物鈍化劑及 C 1- C 2 0陰離子鈍化劑分解且帶負電時,鈍化劑會烯負載 氮化矽層之表面且使其有效的鈍化。 1243200 /Ί、 雜禪涵 (1/) j <、二、 , 實驗例6 該實驗中,藉由添加0 0 5 w t % A Η P、0,0 5 w t % P Η P及 陰離子聚合物純化劑於含1 w t %氧化錦之水溶液中製備各 種研漿。使用聚不同分子量之陰離子聚合物鈍化劑製備不 同研漿。 製備如實驗例2相同之樣品晶圓’且使用與實驗例2相 同之C Μ P裝置及在相同操作條件下進行c M p。C M P過程中 測量氧化物及氮化矽移除速率及選擇性。結果列於了表6 中 〇 盘 ,添加劑 移除速率,埃/分鐘 選擇性 氧化物 (PE-TEOS) 氮化碎 (Si3N4) ΑΗΡ+ΡΗΡ 十 PAA(Mw:l,200) 5567 140 40:1 AHP+PHP+PAA(Mw;5,000) 6414 139 46:1 AHP-fPHP+PAA(Mw;8,000) 5179 123 42:1 AHP+PHP+PAA-co'MA (Mw; 3’000). 4241 94 45:1 AHP+PHP+PAA-co-MA (Mw; 50,000) 3982 117 34:1 表6中,ΜΡΑΑΠ為聚(丙烯酸)之簡寫’ ’’PAA-co-MA”為聚 (丙烯酸-馬來酸)之簡寫,且,,M w ”為分子量。由表ό可清楚 看出,當陰離子聚合物鈍化劑之分子量增加時’氧化物對 氮化矽之選擇性會下降。對加工寬容度而言較佳之選擇性 範圍為約3 0 : 1至約5 0 ·· 1。因此’較好使用分子量約1,0 0 0 至約1 0 0,0 0 0之陰離子聚合物鈍化劑。 f驗例7 該實驗中,樣品晶圓(各包括全部3 9個晶片)係經由0 . 1 2 微米D R A Μ之S T I製程製備。參考圖1 0 ’係經氧化物層1 0 2 1243200AHP, wt% PHP, wt% pH Quick removal, Angstrom / minute selective oxide (PE-TEOS) Silicon nitride (Si3N4) 〇 0 5 1242 542 2.3: 1 0.2 〇5 2668 1206 1 2.2: 1 0.2 0.01 5 2615 1155 2.3: 1 0.2 0.1 5 2668 899 3.0: 1 0.2 0.25 5 2744 772 3.6: 1 It is clear from Table 3 and Figure 6 that the amount of P Η P used as C 1-C 1 2 anionic passivating agent Increasing it will decrease the silicon nitride removal rate. Demonstrate the passivation of anionic passivation agents. Moreover, compared to the case of increasing the anionic deactivator, increasing the amount of the anionic deactivator slightly increases the removal rate of the oxide. As in Experimental Example 2 (in which only one type of passivation agent (PMA) is used), when only PΗP is added as a passivation agent, the selectivity in all test samples is relatively low. Experimental Example 4 In this experiment, based on the results of Experimental Examples 1, 2 and 3, it was prepared by adding 0.0 5 wt% of AΗP, PM A and PΗP in an aqueous solution containing 1 wt% scandium oxide. Various mortars. The amount of PMA was fixed at 0.05 wt%, and the amount of PHP was changed to prepare different mortars. The same sample wafer as in Experimental Example 2 was prepared, and the same CMP apparatus as in Experimental Example 2 was used and CMP was performed under the same operating conditions. During CMP, oxide and silicon nitride removal rates and selectivity were measured. The results are listed in Table 4 below. Figure 7 is a graphical illustration of the results. -19-1243200 (15) Table 4 ΑΗΡ, wt% PMA, wt% PHP, wt% pH removal rate, Angstroms / minute Selective Oxide (PE-TEOS) Silicon Nitride (Si3N4) 0.05 0.05 0.01 5 4641 82 57: 1 0.05 0.05 0.05 5 4890 75 65: 1 0.05 0.05 0.1 5 4960 62 80: 1 0.05 0.05 0.5 5902902 95: 1 It can be clearly found from Table 4 and Figure 7 that when the second type of anionic passivating agent (P M A And P Η P) when used together, high oxide selectivity to silicon nitride will be obtained. Experimental Example 5 In this experimental example, by adding 0.0 5 wt% AΗP, 0.0 5 wt% P M A, and 0.0 5 wt% PΡ P in an aqueous solution containing 1 wt% oxidation sieve, followed by 1 M- H2S04 solution and potassium hydroxide standard solution were used to adjust the pH to prepare various mortars with different pH levels. The same sample wafer was prepared as in Experimental Example 2, and the same CMP apparatus was used as in Experimental Example 2 and CMP was performed under the same operating conditions. Measure oxide and silicon nitride removal rates during CMP. The results are listed in Table 5 below. Figure 8 is a graphical illustration of the results. Table 5 pH removal rate, Angstroms / minute Selective Oxide (PE-TE0S) Silicon Nitride (Si3N4) 4 3544 40 89: 1 4.5 3904 44 89: 1 5 4214 61 69: 1 5.5 4458 82 54: 1 6 5190 129 40: 1 7 4750 1197 4: 1 -20-1243200 One by one ... (16) Xi Minglinan: Suspect, it can be clearly found from Table 5 and Figure 8 that when the ρ 研 of the mortar increases, the oxide The removal rate of (PE-TEOS) and silicon nitride will increase, and when the pH reaches and exceeds about 6, the removal rate of silicon nitride will increase significantly. In addition, when ρ Η is about 6 or higher, the selectivity may suddenly decrease. Obviously, CMP with ρ 6 of 6 or less is effective for increasing the selectivity of oxides to nitrides. The decrease in selectivity at higher ρ 系 is due to less passivation agent on the thinly-loaded silicon dioxide layer. Basically, this is due to the degree of decomposition of the passivation agent and the surface charge of the silicon nitride layer as ρ ρ changes. Regarding the degree of decomposition, the decomposition constants of the functional groups-COOH and PMA of PP are about 4.2, so they can be completely decomposed at a pH of 5 or lower. However, when the pH value is neutralized basic ρ Η, the degree of decomposition decreases. Therefore, less negative charges contained in the passivation are generated by the passivation agent, which reduces the selectivity. From the viewpoint of silicon nitride surface charge, the decrease in selectivity can be explained as follows. The intrinsic surface charge of a layer can be expressed in r (Q potential, mV). The solution ρ 时 when the sign of the C potential in the layer changes is called the isoelectric point. As shown in FIG. 9, the isoelectric point of the HDPC VD or PEC VD oxide layer is about pH 3, and the LPCVD or high-temperature PECVD silicon nitride layer is about pH 6. When the pH of the slurry is lower than the isoelectric point of the layer, the surface of the layer is positively charged in the slurry. When the pH of the slurry is greater than the isoelectric point of the layer, the surface of the layer is negatively charged in the slurry. Therefore, when the pH in the slurry is about 3 to 6, the surface of the silicon nitride layer is positively charged and the surface of the oxide layer is negatively charged. Therefore, when the anionic polymer passivating agent and the C 1-C 2 0 anionic passivating agent are decomposed and negatively charged, the passivating agent will load the surface of the silicon nitride layer and passivate it effectively. 1243200 / Ί, Miscellaneous Zen (1 /) j <, 2 ,, Experimental Example 6 In this experiment, by adding 0 0 5 wt% A Η P, 0, 0 5 wt% P Η P and anionic polymer The purifying agent was used to prepare various slurries in an aqueous solution containing 1 wt% oxidized brocade. Different anionic polymer deactivators with different molecular weights are used to prepare different slurries. The same sample wafer 'as in Experimental Example 2 was prepared, and using the same CMP device as in Experimental Example 2 and performing c M p under the same operating conditions. The C M P process measures oxide and silicon nitride removal rates and selectivity. The results are shown in Table 6. Additive removal rate, Angstroms / minute Selective Oxide (PE-TEOS) Nitriding (Si3N4) ΑΗΡ + ΡΗΡ Ten PAA (Mw: 1, 200) 5567 140 40: 1 AHP + PHP + PAA (Mw; 5,000) 6414 139 46: 1 AHP-fPHP + PAA (Mw; 8,000) 5179 123 42: 1 AHP + PHP + PAA-co'MA (Mw; 3'000). 4241 94 45 : 1 AHP + PHP + PAA-co-MA (Mw; 50,000) 3982 117 34: 1 In Table 6, MPAΑΠ is shorthand for poly (acrylic acid) '' 'PAA-co-MA "is poly (acrylic acid-maleic acid) ), And M w ”is a molecular weight. It is clear from Table 6 that as the molecular weight of the anionic polymer passivator increases, the selectivity of the 'oxide to silicon nitride decreases. A better selectivity range for processing latitude is from about 30: 1 to about 50 ·· 1. Therefore, it is preferred to use an anionic polymer passivating agent having a molecular weight of about 1,000 to about 1,000. f Examination Example 7 In this experiment, the sample wafers (each including all 39 wafers) were prepared by the ST I process of 0.1 μm DRMA. Referring to FIG. 10 ′ system via oxide layer 1 0 2 1243200
發.日月說明淹頁1 及氮化矽層1 0 4依序沉積在晶圓上。氮化矽層1 0 4係以 LPCVD,在壓力數百mToΓΓ或更低下,或以PECVD在約5 00 至約6 0 0 °C之高溫範圍下,使用二氣矽烷及氨(N Η 3)作為反 應物氣體,沉積至厚度5 5 0埃。形成光阻圖案1 0 6以定義溝 槽區後,使用光阻劑圖案1 〇 6作為蝕刻保護蝕刻,形成氮 化矽層圖案1 0 4,其作為硬質保護及研除終止層。隨後, 如圖1 1中所示,移除光阻劑圖案1 0 6,且使用氮化石夕層圖 案1 0 4作為蝕刻保護部分蝕刻氧化物層1 0 2及晶圓1 0 0,因 此形成許多淺溝槽1 0 7。接著,如圖1 2所示,沉積氧化物 層1 0 8至厚度為5 5 0 0埃,以充填溝槽1 0 7,且覆蓋氮化矽層 圖案104之表面。此處,氧化矽層108係使用PECVD形成 PE-TEOS層。 本發明之高選擇性研漿係藉由添加0.05 wt% ΑΗΡ、0.05 w t % Ρ Μ Α及0.0 5 w t % Ρ Η Ρ於含1 w t %氧化鈽之水溶液中, 接著藉由調整pH至5製備。 使用4玄研聚在樣品晶圓上進1'亍C ΪΜ P ’且便用樣品晶圓上 之氮化矽層圖案104作為研除終止層。使用6ED裝置 (StrasbaughCo.Ltd.),且使用 IC1000 上墊及 Suba4 次要墊 進行CMP。所得最終結構敘述於圖I 3中。 針對所得結構,測量作為研除終止層之氮化矽層圖案 I 0 4之剩餘厚度。計算剩餘厚度及其偏差之平均,且測量 I 0 0微米X I 0 0微米面積中之碟化深度。結果列於下表7 中 〇 1243200The sun and the moon explained that the flooded page 1 and the silicon nitride layer 104 were sequentially deposited on the wafer. The silicon nitride layer 104 is LPCVD at a pressure of several hundred mToΓΓ or lower, or PECVD at a high temperature range of about 500 to about 600 ° C, using digas silane and ammonia (N Η 3) As a reactant gas, it was deposited to a thickness of 5 50 angstroms. After the photoresist pattern 106 is formed to define the trench area, the photoresist pattern 106 is used as an etch protection etch to form a silicon nitride layer pattern 104 as a hard protection and removal stop layer. Subsequently, as shown in FIG. 11, the photoresist pattern 106 is removed, and the oxide layer 10 and the wafer 100 are etched using the nitride stone layer pattern 104 as an etch protection part, so that Many shallow trenches 1 0 7. Next, as shown in FIG. 12, an oxide layer 108 is deposited to a thickness of 5500 angstroms to fill the trench 107 and cover the surface of the silicon nitride layer pattern 104. Here, the silicon oxide layer 108 is a PE-TEOS layer formed using PECVD. The highly selective slurry of the present invention is prepared by adding 0.05 wt% AHP, 0.05 wt% PM, and 0.05 wt% PΗP in an aqueous solution containing 1 wt% ytterbium oxide, and then prepared by adjusting the pH to 5. . 4'Yanken Polymer was used to advance 1 '亍 C ΪM P' on the sample wafer and the silicon nitride layer pattern 104 on the sample wafer was used as the removal stop layer. A 6ED device (Strasbaugh Co. Ltd.) was used, and CMP was performed using an IC1000 pad and a Suba4 secondary pad. The resulting final structure is described in Figure I3. With respect to the obtained structure, the remaining thickness of the silicon nitride layer pattern I 0 4 as the removal stop layer was measured. Calculate the average of the remaining thickness and its deviation, and measure the depth of dishing in an area of 100 micrometers by 100 micrometers. The results are listed in Table 7 below. 1243200
樣品 晶圓 編號 CMP條件 (下壓板速度) 剩餘氮化物層 圖案之厚度偏 差,埃 剩餘氮化物層 圖案之平均厚 度,_埃 卜,7匕 深度 ,埃 1 3 psi - 93 rpm 53 512 233 2 3.3 psi - 93 ipm 55 502 278 3 1 3.5 psi - 103 rpm 59 510 1 261 4 3.5 psi — 110 rpm 54 505 250 由表7可清楚發現,當加上本發明研漿時,移除之氮化 物研除終止層為3 0 - 5 0埃。而且,剩餘氮化物層之厚度偏 差低至5 0 - 6 0埃。顯然的,本發明之研漿可提供高氧化物 對氮化物之選擇性。由表7可明顯發現,依C Μ P條件,本 發明研漿可明顯降低碟化現象。 實驗例8 該實驗中,使用具有不同氧化物對氮化物選擇性之研 漿測量氧化物(PE-TEOS)移除速率。樣品晶圓之製造及 CMP均如實驗例7般進行,且測量氧化物(PE-TEOS)移除 速率。結果如表8及圖14所示。Sample wafer number CMP conditions (lower platen speed) Deviation of the thickness of the remaining nitride layer pattern, the average thickness of the remaining nitride layer pattern, _Eb, 7 depth, Angstrom 1 3 psi-93 rpm 53 512 233 2 3.3 psi-93 ipm 55 502 278 3 1 3.5 psi-103 rpm 59 510 1 261 4 3.5 psi — 110 rpm 54 505 250 It is clear from Table 7 that when the slurry of the present invention is added, the nitrides removed are removed. The termination layer is 30-50 angstroms. Moreover, the thickness deviation of the remaining nitride layer is as low as 50 to 60 angstroms. Obviously, the slurry of the present invention can provide high oxide-to-nitride selectivity. It can be clearly found from Table 7 that, according to the condition of CMP, the slurry of the present invention can significantly reduce the dishing phenomenon. Experimental Example 8 In this experiment, a slurry having different oxide-to-nitride selectivity was used to measure the oxide (PE-TEOS) removal rate. The sample wafer manufacturing and CMP were performed as in Experimental Example 7, and the oxide (PE-TEOS) removal rate was measured. The results are shown in Table 8 and FIG. 14.
A 選擇性 氧化物移除速率,埃/分鐘 12:1 3030 19:1 3237 20:1 3327 26:1 2784 33:1 1200 36:1 1233 40:1 1080 45:1 1347 50:1 1396 79:1 1880 104:1 1822 108:1 1911 24 1243200(20) 由表8及圖1 4可清楚發現,當研漿之選擇] 至約5 0 : 1時,氧化物移除速率在約1,0 0 0至約 鐘間。考量低碟化現象之氮化物移除速率(圖 及氧化物移除速率二者,較佳之氧化物移P 1,0 0 0至約1,5 0 0埃/分鐘之間。因此,較好使用 :1至約5 0 : 1之本發明研漿,以降低碟化現| 如上述,當使用本發明之高選擇性研漿時 層之厚度變化極小,且晶圓表面及均勻且沒有 據此,可增加後續製程之寬容度。另外,相較 CMP研漿,可降低研除終止層及欲研除之層之 雖然本發明亦經參考其較佳具體例特殊的 ,但熟習本技藝者應了解可進行型態及細節上 ,但均不離附屬申請專利範圍所定義之本發 圍。 圖式代表符號說明 l在約3 0 : 1 1,5 0 0埃/分 7中之結果) 余速率在約 選擇性約3 0 L 。 ,研除終止 碟化現象。 於使用傳統 起始厚度。 顯示及敘述 之各種改變 明精神及範 20 研 除 板 2 1 研 除 墊 22 轉 軸 32 轉 車由 30 研 除 頭 34 夾 具 3 6 晶 圓 40 研 漿 供給管線 42 研 漿A Selective oxide removal rate, Angstroms / minute 12: 1 3030 19: 1 3237 20: 1 3327 26: 1 2784 33: 1 1200 36: 1 1233 40: 1 1080 45: 1 1347 50: 1 1396 79: 1 1880 104: 1 1822 108: 1 1911 24 1243200 (20) It can be clearly found from Table 8 and Figure 14 that when the slurry is selected] to about 50: 1, the oxide removal rate is about 1.0. 0 to about 0 minutes. Considering the low removal rate of nitride (both the figure and the oxide removal rate), the preferred oxide shift is between P 1, 0 to about 1,500 angstroms / minute. Therefore, it is better Use: 1 to about 50: 1 of the slurry of the present invention to reduce dishing | As mentioned above, when the highly selective slurry of the present invention is used, the thickness variation of the layer is extremely small, and the surface of the wafer is uniform and without evidence. Therefore, the latitude of subsequent processes can be increased. In addition, compared with the CMP slurry, the removal termination layer and the layer to be removed can be reduced. Although the present invention is also referred to its specific examples, it is familiar to those skilled in the art It should be understood that the types and details can be carried out, but they do not deviate from the original hairpin defined by the scope of the patent application for the subsidiary application. The representation of the symbols on the diagram indicates the result of l in about 30: 1 1, 50 angstroms / min. 7) The rate is about 30 L at a selectivity of about. To study the termination of discontinuation. For use with traditional starting thickness. Various changes in display and narrative Ming spirit and norms 20 Research and removal of boards 2 1 Research and removal of pads 22 Rotary shafts 32 Turns of cars by 30 research and removal of heads 34 Clamps 3 6 Crystal circles 40 Research slurry Supply lines 42 Research slurry
1243200 (21) 100 晶 圓 102 墊 氧 化 物層 104 氮 化 矽 層 106 光 阻 圖 案 108 氧 化 物 層1243200 (21) 100 crystal circle 102 pad oxide layer 104 silicon nitride layer 106 photoresist pattern 108 oxide layer
-26 --26-
Claims (1)
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
KR10-2001-0025873A KR100459696B1 (en) | 2001-04-05 | 2001-05-11 | Chemical mechanical polishing slurry, and chemical mechanical polishing process and shallow trench isolation process employing the same |
KR10-2002-0048403A KR100464429B1 (en) | 2002-08-16 | 2002-08-16 | Chemical mechanical polishing slurry and chemical mechanical polishing method using the same |
Publications (2)
Publication Number | Publication Date |
---|---|
TW200403329A TW200403329A (en) | 2004-03-01 |
TWI243200B true TWI243200B (en) | 2005-11-11 |
Family
ID=37025326
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
TW91133022A TWI243200B (en) | 2001-05-11 | 2002-11-11 | Chemical/mechanical polishing slurry and chemical mechanical polishing method using the same |
Country Status (1)
Country | Link |
---|---|
TW (1) | TWI243200B (en) |
-
2002
- 2002-11-11 TW TW91133022A patent/TWI243200B/en not_active IP Right Cessation
Also Published As
Publication number | Publication date |
---|---|
TW200403329A (en) | 2004-03-01 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP4272409B2 (en) | Chemical mechanical polishing slurry and chemical mechanical polishing method using the same | |
JP5290769B2 (en) | CMP slurry and semiconductor wafer polishing method using the same | |
JP4537010B2 (en) | Chemical mechanical polishing slurry and chemical mechanical polishing method using the same | |
JP6989493B2 (en) | Methods and Compositions for Machining Dielectric Substrates | |
TWI450952B (en) | Cmp slurry and a polishing method using the same | |
JP5326492B2 (en) | Polishing liquid for CMP, polishing method for substrate, and electronic component | |
JP5552475B2 (en) | Cerium oxide production method, cerium oxide obtained therefrom and CMP slurry containing the same | |
TWI413678B (en) | Polishing liquid | |
CN105814163A (en) | Mixed abrasive polishing compositions | |
TW201435070A (en) | Aqueous dispersion for chemical mechanical polishing, and chemical mechanical polishing method | |
WO2017114309A1 (en) | Chemical mechanical polishing slurry and application thereof | |
TW201345988A (en) | Chemical mechanical polishing composition having chemical additives and methods for using same | |
JP2022060218A (en) | Tungsten chemical mechanical planarization (cmp) with low dishing and low erosion topography | |
US11066575B2 (en) | Chemical mechanical planarization for tungsten-containing substrates | |
TWI512809B (en) | Method for forming through-base wafer vias for fabrication of stacked devices | |
JP4851535B2 (en) | Polishing selectivity adjusting auxiliary agent and CMP slurry containing the same | |
TWI243200B (en) | Chemical/mechanical polishing slurry and chemical mechanical polishing method using the same | |
TWI662096B (en) | Cmp compositions selective for oxide and nitride with improved dishing and pattern selectivity | |
CN112552824B (en) | Polishing composition and polishing method | |
TW202116968A (en) | High oxide removal rates shallow trench isolation chemical mechanical planarization compositions | |
CN114787304B (en) | Shallow trench isolation chemical mechanical planarization polishing of low oxide trench recesses | |
TWI744696B (en) | High oxide vs nitride selectivity, low and uniform oxide trench dishing in shallow trench isolation (sti) chemical mechanical planarization polishing (cmp) | |
TWI790509B (en) | High oxide film removal rate shallow trench isolation (sti) chemical mechanical planarization (cmp) polishing | |
WO2024142722A1 (en) | Polishing material, polishing method, method for producing semiconductor component, and additive liquid for polishing | |
KR20200073479A (en) | Chemical mechanical polishing slurry composition |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
MK4A | Expiration of patent term of an invention patent |