TW400599B - Method of confirming the reasons of the defects in the IC process - Google Patents

Method of confirming the reasons of the defects in the IC process Download PDF

Info

Publication number
TW400599B
TW400599B TW87116161A TW87116161A TW400599B TW 400599 B TW400599 B TW 400599B TW 87116161 A TW87116161 A TW 87116161A TW 87116161 A TW87116161 A TW 87116161A TW 400599 B TW400599 B TW 400599B
Authority
TW
Taiwan
Prior art keywords
wafer
electrostatic discharge
same
metal layer
defects
Prior art date
Application number
TW87116161A
Other languages
Chinese (zh)
Inventor
Ren-Shian Ye
De-Fang Huang
Jau-Shing Jang
Jr-Jian Hung
Original Assignee
Taiwan Semiconductor Mfg
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Taiwan Semiconductor Mfg filed Critical Taiwan Semiconductor Mfg
Priority to TW87116161A priority Critical patent/TW400599B/en
Application granted granted Critical
Publication of TW400599B publication Critical patent/TW400599B/en

Links

Landscapes

  • Testing Or Measuring Of Semiconductors Or The Like (AREA)

Abstract

This invention provides a method of confirming whether the reason of defects is caused by the electrostatic discharge (ESD) in the IC process. Such a method includes the steps of: firstly, providing a wafer formed thereon a semiconductor device, and the wafer comprising a metal layer; next, providing another wafer formed thereon the same semiconductor device with the first wafer, but replacing the metal layer with an insulation layer; applying the two wafers to any unit of the same process and allowing the first wafer to have the defects caused by the approximate electrostatic discharge (ESD); and finally, detecting whether the same defects appear in the second wafer. If yes, it is known that the defects are not caused by ESD; otherwise, the defects are caused by ESD.

Description

五、發明說明(1) 本發明是有關於半導體製程技術,且特別是有關於一 種在I C製程中用來確認缺陷成因的方法。 在某些特用IC (AS 1C)的製程中,會發現其内金屬介 電層(IMD)在經過洗務器(scrubber)的清洗後,表面上會 有介層破裂(via-crack)的缺陷產生。掃瞄式電子顯微鏡 (SEM)顯示出這些介電層上有凹陷(cave)或爆裂(crack)情 形,同時在能量散佈光譜(EDS ; Energy Dispeaicm Spectrum)中顯示出該些缺陷中含有A1、F、ci等元素,這 表示底下的金屬層也因此而露出,並有遭到侵兹的情形。 由於via-crack總是伴隨在金屬圖案之上發生,而且 SEM也顯示出底下的金屬有爆開的情形,因此靜電放電 (ESD ;electrostatic discharge)首先被視為主要的肇 因;另一個同樣可用來解釋via-crack總是發生在金屬圖 案之上的假没’則疋4曰向内金屬介電層間應力的(stress) 影響。儘管如此’在習知技術中並沒有一個簡單而有效的 測試方法可以讓吾人得知,究竟哪一項才是造成前述 via-crack的真正原因。 為了找出造成via-crack的根本原因,本案發明人進 行了一系列的實驗’在實驗的過程中發現:在兩組不同的 洗滌器中,其中via-crack出現比例較高(70〜80%)的洗滌 器’在清洗過程時所產生的靜電場總是比正常值高(約 -3 — 5 KV/inch ’正常約-1〜-2 KV/inch)。這樣的結果顯 示出via-crack與ESD似乎具有高度的相關性。在另外的實 驗中’本案發明人則研究洗滌器的參數與via_crack的關V. Description of the invention (1) The present invention relates to semiconductor process technology, and in particular, to a method for confirming the cause of defects in an IC process. In some special IC (AS 1C) manufacturing processes, it will be found that after the metal dielectric layer (IMD) is cleaned by a scrubber, the surface will have a via-crack. Defects occur. Scanning electron microscopy (SEM) revealed that these dielectric layers have caves or cracks. At the same time, the energy dispersive spectrum (EDS; Energy Dispeaicm Spectrum) showed that these defects contained A1 and F. , Ci and other elements, which means that the underlying metal layer is also exposed as a result, and may be invaded. Because via-crack always occurs on top of the metal pattern, and the SEM also shows that the underlying metal is bursting, electrostatic discharge (ESD; electrostatic discharge) is first considered the main cause; another is also available To explain that via-crack always occurs on the metal pattern, then the effect of inward metal dielectric interlayer stress is described. However, there is no simple and effective test method in the conventional technology that can let us know which one is the real cause of the aforementioned via-crack. In order to find out the root cause of via-crack, the inventor of this case conducted a series of experiments. 'During the experiment, it was found that in two different sets of scrubbers, the proportion of via-crack was higher (70 ~ 80%). The electrostatic field generated by the scrubber 'during the cleaning process is always higher than the normal value (about -3-5 KV / inch' normally about -1 ~ -2 KV / inch). Such results show that via-crack and ESD seem to be highly correlated. In another experiment, the inventor of this case studied the relationship between the parameters of the scrubber and via_crack.

C:\Program Files\Patent\0503-3793-E.ptd第 4 頁 五、發明說明(2). 係,包括喷 果發現旋轉 關連性,但 加。然而, 例之洗滌器 產生。 為了進 吾人將原本 chuck)更換 程序。新的 因此更換過 5-0.8 KV/i 水柱的壓力、旋轉速度、以及洗滌的時間,社 ,^與洗滌時間與vi a —crack的產生沒有什麼^ 7的壓力升向則會使via-crack的數目增 後的實驗中,即使吾人將高via_crack比 、、水柱壓力降低,仍無法避免via_crack的 、乂瞭解靜電放電(ESD)與via-crack的關係, 氟龍(Teflon)材質的旋轉夾頭(叩4 ^ π電材質的旋轉夾頭,然後進行同樣的洗滌 導體失頭是用來降低晶圓上所累積的靜電荷, 後在洗滌進行時於晶圓中央的靜電場僅有〇 nch ’請參照表1 : 表 1 鐵氟龍夾頭 靜電場 (KV/inch) Via-crack 數目 Via-crack 晶圓數 總晶圓數 -4 〜-5 2 2 4 _導體夾頭 0.5 〜0.8 6 5 8 # φ 表1可知將鐵氟龍夾頭更換為導體夾頭後,儘管 分。葱已經大幅地降低,但是via_crack的數目依然沒有 減少的跡象。 雖然上述降低靜電場的實驗無法使via-crack的數目 減少,吾人仍然不能排除靜電放電為via crack肇因的可C: \ Program Files \ Patent \ 0503-3793-E.ptd page 4 V. Description of the invention (2). The system, including spraying, finds rotation relevance, but adds. However, for example, a scrubber was produced. In order to enter, I will chuck) the replacement procedure. The new one has changed the pressure, rotation speed, and washing time of 5-0.8 KV / i water column. There is nothing to do with the washing time and vi a —crack ^ The pressure rise of 7 will make via-crack After increasing the number of experiments, even if I reduce the via_crack ratio and the pressure of the water column, we still cannot avoid via_crack. I understand the relationship between electrostatic discharge (ESD) and via-crack. Teflon rotating chuck (叩 4 ^ π rotating chuck made of electrical material and then performing the same washing. The conductor loss is used to reduce the electrostatic charge accumulated on the wafer. After the washing process, the electrostatic field in the center of the wafer is only 0nch ' Please refer to Table 1: Table 1 Teflon chuck electrostatic field (KV / inch) Via-crack number Via-crack wafer number total wafer number -4 to -5 2 2 4 _conductor chuck 0.5 to 0.8 6 5 8 # φ Table 1 shows that after replacing the Teflon chuck with the conductor chuck, although the points have been reduced. The onion has been greatly reduced, but the number of via_crack has not decreased. Although the above experiments to reduce the electrostatic field cannot make via-crack Decrease in number, I still cannot rule out Electrical discharge is via crack at cause may

五、發明說明(3) 月匕性’因為靜電放電有可能/ + 生,而如果電荷流向導體央贺Y柱接觸晶圓的一瞬間發 可能無法感測到ESD的快速辦的速度太快,靜電偵測器有 實際上晶圓已經因此而受到文二而仍顯示出低靜電值’但 由上述可知’在習知枯化丄 法可以用來確認沒有**種簡單有效的方 或是另有其他的原因。疋否為靜電放電所造成的結果V. Description of the invention (3) Due to the possibility of electrostatic discharge / +, if the charge flow to the Y-pillar of the Yoshioka ’s Y-pillar is in contact with the wafer, it may not be able to detect the ESD fast. The electrostatic detector has the fact that the wafer has been subjected to the second article and still shows a low electrostatic value. However, it can be known from the above that in the conventional drying method, it can be used to confirm that there are no simple and effective methods or other methods. There are other reasons.疋 Is it a result of electrostatic discharge?

有鑑於此,本發明的主I 中用來確認缺陷成因的方法,a : ”共一種在IC製程 成因的情形時,經由本發明的^到有關懷疑ESD為缺陷 澄清。 Θ的方法’可以快速獲得確認或 述目❸’本發明提供—種在1c製程中用來確認 静電放電(ESD)疋否為缺陷成因的方法,包括下列步驟: (a)提供一第一晶圓,其上形成有半導體元件,其中至少 —金屬層,(b)提供一第二晶圓,在其上形成與第一 晶圓相同之半導體元件,但其中以一絕緣層取代上述金屬 層’(c )將上述兩晶圓施以任意相同之處理單元,並藉使 第二晶圓上出現疑似靜電放電所造成之缺陷;以及(d)比 較第二晶圓是否出現與第一晶圓相同的缺陷,如果是,則 表示上述缺陷並非靜電放電所造成;反之,則表示靜電放 電為缺陷之成因所在。 =在上述方法中,第二晶圓之絕緣層與第一晶圓之金屬 層最好具有相同之圖案,而其中絕緣層之材質,例如是以 電漿加強化學氣相沈積法(PE-CVD)所形成之氮化矽。In view of this, the method for confirming the cause of the defect in the main I of the present invention, a: "A total of one method is used to clarify the suspected ESD as a defect when the IC process is caused. The method of Θ can quickly Obtaining confirmation or description: 'The present invention provides a method for confirming whether electrostatic discharge (ESD) is the cause of a defect in the 1c process, including the following steps: (a) providing a first wafer on which There are semiconductor elements, at least-a metal layer, (b) a second wafer is provided on which the same semiconductor element as the first wafer is formed, but an insulating layer is used instead of the metal layer '(c) The two wafers are subjected to any identical processing unit and a defect caused by a suspected electrostatic discharge appears on the second wafer; and (d) comparing whether the second wafer has the same defects as the first wafer, if it is , It means that the above defect is not caused by electrostatic discharge; otherwise, it means that electrostatic discharge is the cause of the defect. = In the above method, the insulating layer of the second wafer and the metal layer of the first wafer preferably have the same Pattern, and the material of the insulating layer is, for example, silicon nitride formed by plasma enhanced chemical vapor deposition (PE-CVD).

C:\ProgramFiles\Patent\0503-3793-E.ptd第 6 頁 五 、發明說明(4) 為讓本發明之上述和其他目的、特徵、和優點能更明 顯易懂’下文特舉一較佳實施例,作詳細說明如下: 實施例 由於vi a-crack總是伴隨在金屬圖案之上發生,因此 除了靜電放電(ESD)為可能的肇因外,另一個可能的原因 是内金屬介電層(IMD)的應力影響所致。為了找出真正的 原因’本實施例準備了兩個具有相同元件的晶圓,並將其 中一者的金屬層,以相同圖案的絕緣層取代之: ★準備一半導體晶圓,按習知方式在其基底上製作M〇s 等半導體元件,並在其A lCu金屬層上’形成内金屬介電層 (IMD)。在另一方面,製作另一個具有相同元件的半導體曰 晶Ϊ,.但其中以一層具有相同圖案的氮化矽絕緣層,以取 代前述之AlCu金屬層,該氮化矽絕緣層係以pE_CVD法形 成’厚度約7000埃,與金屬層厚度相當。 在本實施例中之I層是以下述方法製作:首先以 PE CVD法沈積一層厚約2〇〇〇埃的第一介電層;然後以 SA-CVD法(次壓化學氣相沈積法)形成厚約5〇〇〇埃的氧化 層,再以PE-CVD法沈積厚約2 0 0 0埃的覆蓋氧化層。接 先以厚度約6 00 0埃的旋塗式玻璃層(s〇G)進行平H坦化,’ ,之完全地回蝕刻後,再以PE-CVD法形成厚度約M〇〇埃的 弟二介電層,完成内金屬介電層的製作。 、的 將上述兩晶圓施以同樣的洗膝程序:在停止旋轉 (spin off)的條件下,以口徑0.2践的喷嘴頭進行清洗 (請參照本案發明人另一件共同審理中的案件:「^ I c製C: \ ProgramFiles \ Patent \ 0503-3793-E.ptd page 6 5. Description of the invention (4) In order to make the above and other objects, features, and advantages of the present invention more comprehensible, the following is a better example. The embodiment is described in detail as follows: In the embodiment, since vi a-crack always occurs on the metal pattern, in addition to electrostatic discharge (ESD) as a possible cause, another possible cause is the inner metal dielectric layer. (IMD). In order to find out the real reason, this embodiment prepared two wafers with the same elements, and replaced the metal layer of one of them with the insulating layer of the same pattern: ★ Prepare a semiconductor wafer according to the conventional method Semiconductor devices such as Mos are fabricated on the substrate, and an inner metal dielectric layer (IMD) is formed on the Al Cu metal layer. On the other hand, another semiconductor with the same elements is made, but a silicon nitride insulating layer with the same pattern is used to replace the aforementioned AlCu metal layer. The silicon nitride insulating layer is made by pE_CVD method. The thickness is about 7000 Angstroms, which is equivalent to the thickness of the metal layer. The I layer in this embodiment is made by the following method: firstly, a first dielectric layer having a thickness of about 2000 angstroms is deposited by a PE CVD method; and then a SA-CVD method (sub-pressure chemical vapor deposition method) is used An oxide layer with a thickness of about 5000 angstroms is formed, and then a cover oxide layer with a thickness of about 2000 angstroms is deposited by a PE-CVD method. Subsequently, the spin-coated glass layer (s0G) with a thickness of about 600 angstroms was first flattened, and after being completely etched back, a PE-CVD method was used to form a brother with a thickness of about OO angstroms. Two dielectric layers complete the fabrication of the inner metal dielectric layer. The above two wafers are subjected to the same knee washing procedure: under the condition of spin off, cleaning with a nozzle diameter of 0.2 mm (please refer to another case in the joint trial by the inventors of this case: "^ I c

五 '發明說明(5) 程中用來檢測洗蘇器清洗時所造成之應力影響的方法」 )。清洗過後之晶圓以K L A掃瞄其v i a _ c r a c k缺陷的數量 其結果列於表2 : 表 2 製程條件 缺陷數目 AlCu+IMD >100 S13N4+IMD >100 由上表可知,即使以相同圖案的絕緣層取代原先的 AlCu金屬層,Via-crack依然會出現,由此可見via_crack 的發生確實與靜電放電無關,應為應力所造成之結果。 經由上述可知,本發明的確提供了一個簡單而有效的 方法’可以快速確認ESD是否究竟為缺陷成因。 雖然本發明已以一較佳實施例揭露如上,然其並 以限定本發明,任何熟習此技藝者’在不脫離:^明用 神和範圍内’當可作各種之更動與潤飾,因此本^日 精 護範圍當視後附之申請專利範圍所界定者為準。x明之保Five 'Invention description (5) method used to detect the impact of stress caused by the scrubber cleaning process'). After cleaning the wafer, the number of via _ crack defects was scanned by KLA. The results are shown in Table 2: Table 2 Number of defects under process conditions AlCu + IMD > 100 S13N4 + IMD > 100 As can be seen from the above table, even with the same pattern Via-crack will still appear if the insulating layer replaces the original AlCu metal layer. It can be seen that the occurrence of via_crack is indeed not related to electrostatic discharge and should be the result of stress. It can be known from the foregoing that the present invention does provide a simple and effective method 'to quickly confirm whether ESD is the cause of the defect. Although the present invention has been disclosed as above with a preferred embodiment, but it is also used to limit the present invention. Anyone skilled in this art can not make any changes but can use various modifications and decorations. ^ The scope of Japanese precision protection shall be determined by the scope of the attached patent application. x 明 之 保

附頁 靜電場 (KV/inch) Via-crack 數目 Via-crack 晶圓數 總晶圓數 鐵氟龍夾頭 -4*— 5 2 2 4 導體夾頭 0.5 〜0.8 6 5 8 製程條件 缺陷數目 AlCu+IMD >100 S13N4+IMD >100 C:\Program Files\Patent\0503-3793-E.ptd第 頁Attached sheet Electrostatic field (KV / inch) Via-crack number Via-crack wafer number total wafer number Teflon chuck-4 * — 5 2 2 4 conductor chuck 0.5 to 0.8 6 5 8 process condition defect number AlCu + IMD > 100 S13N4 + IMD > 100 C: \ Program Files \ Patent \ 0503-3793-E.ptd page

Claims (1)

六 '申請專利範圍 h丨· 一種在IC製程中用來確認靜電放電(ESD)是否為缺 陷成因的方法,該方法包括下列步驟: Ca)提供一第一晶圓,其上形成有半導體元件,其中 至少包含一金屬層; (b )提供一第二晶圓,在其上形成與第一晶圓相同之 、導體元件,但其中以一絕緣層取代上述金屬層; :〜(c)將上述兩晶圓施以任意相同之處理單元,並 該第一晶圓上出現疑似靜電放電所造成之缺陷;以曰 (d)比較該第二晶圓是否出現與該第—晶圓及 ^ ’如果是’則表示上述缺陷並非靜電放電所&目同的缺 之’則表示靜電放電為缺陷之成因所在。 ^成;反 2.如申請專利範圍第1項所述之方法,其中 之絕緣層與第一晶圓之金屬層具有相同之圖案。第二晶圓 3 ·如申請專利範圍第2項所述之確認方法, 讀绝 緣層的材質為氮化矽。 ’其中 4. 如申請專利範圍第1項所述之方法,其中a 元包括:將上述兩晶圓經過相同的洗滌器洗膝遠處理單 似靜電放電所造成之缺陷之其他製程。 5相同有疑 5. 如申請專利範圍第1項所述之方法,其中= 之上更包括.一内金屬介電層。 該金屬層6 'Patent Application Scope h 丨 · A method for confirming whether electrostatic discharge (ESD) is the cause of a defect in an IC process, the method includes the following steps: Ca) providing a first wafer on which a semiconductor element is formed, It includes at least one metal layer; (b) providing a second wafer on which the same conductive element as the first wafer is formed, but in which the metal layer is replaced by an insulating layer; ~ (c) the above The two wafers are subjected to any identical processing unit, and a defect caused by a suspected electrostatic discharge appears on the first wafer; and (d) comparing whether the second wafer appears with the first wafer and ^ 'if If yes, it means that the above defect is not the same as that of electrostatic discharge. It means that electrostatic discharge is the cause of the defect. 2. The method described in item 1 of the scope of patent application, wherein the insulating layer and the metal layer of the first wafer have the same pattern. Second wafer 3 • The confirmation method described in item 2 of the scope of patent application, the material of the read insulation layer is silicon nitride. ′ Among them 4. The method described in item 1 of the scope of patent application, wherein a yuan includes: other processes of subjecting the above-mentioned two wafers to the same scrubber and treating the defects caused by electrostatic discharge alone. 5 Same doubts 5. The method as described in item 1 of the scope of patent application, wherein the above includes an inner metal dielectric layer. The metal layer C:\ProgramFiles\Patent\0503-3793-E.ptd第 9 頁C: \ ProgramFiles \ Patent \ 0503-3793-E.ptd page 9
TW87116161A 1998-09-29 1998-09-29 Method of confirming the reasons of the defects in the IC process TW400599B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
TW87116161A TW400599B (en) 1998-09-29 1998-09-29 Method of confirming the reasons of the defects in the IC process

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
TW87116161A TW400599B (en) 1998-09-29 1998-09-29 Method of confirming the reasons of the defects in the IC process

Publications (1)

Publication Number Publication Date
TW400599B true TW400599B (en) 2000-08-01

Family

ID=21631499

Family Applications (1)

Application Number Title Priority Date Filing Date
TW87116161A TW400599B (en) 1998-09-29 1998-09-29 Method of confirming the reasons of the defects in the IC process

Country Status (1)

Country Link
TW (1) TW400599B (en)

Similar Documents

Publication Publication Date Title
JP4659751B2 (en) Method for forming a low-K dielectric in a semiconductor manufacturing process
US7341943B2 (en) Post etch copper cleaning using dry plasma
US6872322B1 (en) Multiple stage process for cleaning process chambers
JP3185150B2 (en) Method for manufacturing semiconductor device
US6872652B2 (en) Method of cleaning an inter-level dielectric interconnect
US7166534B2 (en) Method of dry cleaning photoresist strips after via contact etching
US6660624B2 (en) Method for reducing fluorine induced defects on a bonding pad surface
US6547647B2 (en) Method of wafer reclaim
US6251806B1 (en) Method to improve the roughness of metal deposition on low-k material
US7892984B2 (en) Reduction of defects formed on the surface of a silicon oxynitride film
TW400599B (en) Method of confirming the reasons of the defects in the IC process
CN1222018C (en) Mfg. method of semiconductor and treatment liquid
US6890864B2 (en) Semiconductor device fabricating method and treating liquid
Lim et al. Comparison of in situ and ex situ plasma-treated metalorganic chemical vapor deposition titanium nitride thin films
JP2003124316A (en) Method for manufacturing semiconductor device and treatment liquid
US6926590B1 (en) Method of improving device performance
US6153497A (en) Method for determining a cause for defects in a film deposited on a wafer
KR970006937B1 (en) Metal wiring method in semiconductor
KR100698742B1 (en) Fabricating method of semiconductor device
KR100254620B1 (en) Method for forming tungsten-plug of semiconductor device
KR20010065863A (en) A method for forming metal film pattern for eliminating polymer
US6806103B1 (en) Method for fabricating semiconductor devices that uses efficient plasmas
TW575911B (en) Method of rendering good material property and reducing resistivity of titanium nitride
Gabriel et al. Transient fuse structures: The role of metal etching vs. dielectric deposition
KR19990009344A (en) Cleaning solution and cleaning method using the same

Legal Events

Date Code Title Description
GD4A Issue of patent certificate for granted invention patent
MM4A Annulment or lapse of patent due to non-payment of fees