542939 A7 B7 五、發明説明（） 5_1發明領域： 本發明係有關於微影領域，特別關於近接效應及其 消除之方法。 5_2發明背景： 透過一微影製程由一玻璃光罩形成一光阻光罩，藉 由於一光罩下蝕刻產生一積集圖案，一積體電路中之最小 線見尺寸通ΐ稱做臨界尺寸（epical dimensi〇n, cd)，當 CD接近用於映像玻璃光罩及曝光光阻之燈源波長時（為一 深紫外線約1 83 0至365〇人之間），形成於光阻中之圖案 無法完整精確的複製玻璃光罩上之圖案，因圖案結構之周 遭情況對任何特定圖形結構有很大的影響，此現象一般稱 做“近接效應”。 經濟部中央標準局員工消費合作社印製 I—V-— - -- *: ί I HI . ϋϋ I (請先閲讀背面之注意事項再填寫本頁) 、?! 圖la-lc說明三種近接效應之不同證明，圖中線 2為分離的且無直接的鄰近物，而線3則擠在一起，線3 間之距離與其線寬度相當，雖然線3與線2於玻璃光罩上 有相同的寬度，由其映像於光阻上，近接效應造成線3較 線2窄。圖lb中，玻璃光罩上線4有一長度符合尺寸5, 但在光阻映像中則縮短’如圖所示。圖1 c中，原設計為 所示之正方形角落因圓形化效應，光阻在標示為6之區域 中將消失。 542939 A7 B7五、發明説明（） 經濟部中央標準局員工消費合作社印製 雖然近接效應之起因已了解，但欲計算任何圖案之 大小會非常複雜且耗時，然而，此為現今半導體工業之一 般習慣，執行一計算以產生一光學近接校正（Optical Proximity Correction，OPC)，應用於原玻璃光罩圖案補償 預期的光學近接效應。 玻璃光罩上之圖案轉移至光阻上之過程可概括總結 於以下四步驟： (1)光阻塗蓋，（2)曝光，（3)曝光後烘烤（PEB)，以及（4) 顯影。塗蓋光阻之表面可能有一反反射層（ARC)，近接效 應將被形成於光阻層中之駐波圖案程度所影響，然而，逐 出步驟2及3帶來近接效應之處，而且，小心地控制此兩 步驟之執行，近接效應可被消除，因此可除去 OPC之需 要及有關的代價高的計算。 在搜尋可能的習知技術過程中，發現幾篇重要的參 £ 考資料，包括 Itoo 等（USP 5,43 6，1 14 Jul 1 995)、Ootaka 等（USP 5,636,004 Jun 1 997)、以及 Gortych(USP 5,680,588 Oct 1 99 7)等人所提，他們全是討論數值孔隙及/或同調性 之重要性，Liu (USP 4,988,284 Jan 1991)等人主張需要曝 光後烘烤但用於電子束光阻，此外，溫度指定至少1⑽°c。 (請先閱讀背面之注意事項再填寫本百〇 ▼裝· 、^τ 本紙張尺度適用中國國家標準（CNS ) Α4規格（210X297公釐） 經濟部中央標準局員工消費合作社印製 542939 A7 B7 ------------------- 五、發明説明（） 5·3發明目的及概述： 本發明之目的為提供一種方法除去任何由於近接效 應造成之光阻圖案之扭曲。 本發明之另一目的為經由自行消除近接效應而除去 扭曲效應，而非只是補償它。 本發明之又一目的為該方法不會增加相對於微影製 程之成本且不允許近接效應產生。 藉由小心地控制伴卩近彳政影製程之二個獨立的變數可 以達成上述之目的，這些變數為進行曝光後烘烤之溫度、 曝光系統之數值孔隙以及部分同調參數，特別地，曝光後 烘烤之溫度應比製造商建議的低20至25 t：，數值孔隙 約0.5左右以及部分同調參數約0.8左右，若遵循這些規 則，證實降至線距/線寬比（dWy ratio)低於1且無近接效 應。 5-4圈式簡單說明： 圖1 a-1 c說明近接效應如何扭曲一光阻影像中線之形 狀及尺寸。 圖2描繪一系列不同PEB溫度之CD對線距/線寬比 本紙張尺度適用中國國家標準（CNS ) A4規格（210X：297公釐） (請先閱讀背面之注意事項再填寫本頁)542939 A7 B7 V. Description of the invention () 5_1 Field of invention: The present invention relates to the field of lithography, especially to the proximity effect and its elimination method. 5_2 Background of the Invention: A photoresist is formed from a glass photomask through a lithography process, and an accumulation pattern is generated by etching under a photomask. The minimum line size in an integrated circuit is generally referred to as the critical size. (Epical dimensioon, cd), when the CD approaches the wavelength of the lamp source used for the image glass mask and the exposure photoresist (a deep ultraviolet light between about 1,830 and 3,650 people), formed in the photoresist The pattern cannot completely and accurately reproduce the pattern on the glass mask. Because the surroundings of the pattern structure have a great influence on any specific pattern structure, this phenomenon is generally called "proximity effect". Printed by the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs I—V-—--*: ί I HI. Ϋϋ I (Please read the precautions on the back before filling out this page),?! Figure la-lc illustrates the three proximity effects The difference proves that line 2 in the figure is separated and has no direct neighbors, while line 3 is squeezed together. The distance between lines 3 is equivalent to its line width, although lines 3 and 2 have the same width on the glass mask. The width is reflected on the photoresistor. The proximity effect causes line 3 to be narrower than line 2. In FIG. 1b, the line 4 on the glass mask has a length corresponding to the size 5, but is shortened in the photoresist image as shown in the figure. In Figure 1c, due to the rounding effect of the square corners shown in the original design, the photoresist will disappear in the area labeled 6. 542939 A7 B7 V. Description of the Invention () Although printed by the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs, the cause of the proximity effect is known, it is very complicated and time-consuming to calculate the size of any pattern. However, this is the generality of the current semiconductor industry. Conventionally, a calculation is performed to generate an Optical Proximity Correction (OPC), which is applied to the original glass mask pattern to compensate for the expected optical proximity effect. The process of transferring the pattern on the glass mask to the photoresist can be summarized in the following four steps: (1) photoresist coating, (2) exposure, (3) post-exposure baking (PEB), and (4) development . The photoresist-coated surface may have an anti-reflective layer (ARC). The proximity effect will be affected by the degree of the standing wave pattern formed in the photoresist layer. However, the steps 2 and 3 bring out the proximity effect, and, By carefully controlling the execution of these two steps, the proximity effect can be eliminated, thus eliminating the need for OPC and the associated costly calculations. During the search for possible know-how, I found several important references, including Itoo et al. (USP 5,43 6, 1 14 Jul 1 995), Ootaka et al. (USP 5,636,004 Jun 1 997), and Gortych ( USP 5,680,588 Oct 1 99 7) and others mentioned that they were all discussing the importance of numerical porosity and / or homology. Liu (USP 4,988,284 Jan 1991) and others argued that post-exposure baking is required for electron beam photoresistance, In addition, the temperature is specified at least 1⑽ ° c. (Please read the precautions on the back before filling out this one hundred. ▼, ^ τ This paper size applies to China National Standard (CNS) A4 specifications (210X297 mm). Printed by the Consumers' Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs 542939 A7 B7- ------------------ V. Description of the invention (5.3) Purpose and summary of the invention: The purpose of the present invention is to provide a method to remove any photoresist pattern caused by the proximity effect. Another object of the present invention is to remove the distortion effect by eliminating the proximity effect by itself, instead of just compensating for it. Another object of the present invention is that the method does not increase the cost relative to the lithography process and does not allow the proximity effect. This can be achieved by carefully controlling the two independent variables of the Kanji Kinsei filming process, these variables are the temperature of the post-exposure bake, the numerical aperture of the exposure system, and some of the coherent parameters. In particular, Post-exposure baking temperature should be 20 to 25 t lower than the manufacturer's suggestion: the numerical aperture is about 0.5 and some coherent parameters are about 0.8. If these rules are followed, it is confirmed that the line spacing / The aspect ratio (dWy ratio) is lower than 1 and there is no proximity effect. The 5-4 circle simply explains: Figure 1 a-1 c illustrates how the proximity effect distort the shape and size of the center line of a photoresist image. Figure 2 depicts a series of different PEB temperature of CD to line pitch / line width ratio This paper size applies Chinese National Standard (CNS) A4 specification (210X: 297 mm) (Please read the precautions on the back before filling this page)
542939 A7 B7 五 、發明説明（ 圖。 圖3顯示一系列部分同調參數不同值之Cd對線距/ 線寬比之曲線圖。 圖4顯示使用製造商建議之曝光後烘烤溫度之部分 同調參數幾個不同值之CD對線距/線寬比之實驗$據。 圖5顯示使用依據本發明之主張所測定之曝光後烘 烤溫度’其部分同調參數幾個不同值之CD對線距/線寬比 之實驗數據。 5_5發明詳麵說明： 如稍早簡单敘述，微影中兩步驟曝光及曝光後烘烤 (PEB)決定任何近接效應之性質與範圍，導因於曝光之效 應為完全光學的起因以及導因於PEB之效應，來自光阻 之近接效應為主要因素。 (請先閱讀背面之注意事項再填寫本頁)542939 A7 B7 V. Description of the invention (Figure. Figure 3 shows a series of Cd vs. line pitch / line width ratios of different values of some coherence parameters. Figure 4 shows some coherence parameters of the baking temperature after exposure as recommended by the manufacturer Experimental data of several different values of CD-to-line spacing / line width ratio. Figure 5 shows the use of the post-exposure baking temperature measured in accordance with the claims of the present invention, with some different values of CD-to-line spacing / Experimental data of line width ratio. 5_5 Detailed description of the invention: As stated earlier, the two-step exposure and post-exposure baking (PEB) in lithography determine the nature and range of any proximity effect. The effect due to exposure is The cause of complete optics and the effect caused by PEB, the close effect from photoresistance is the main factor. (Please read the precautions on the back before filling this page)
裝· 玎 經濟、那中央標準局員工消費合作社印製 〜參考圖2，描繪一圖其中CD(單位為微米）為線距/線 見比之一函數，用於一系列不同的PEB溫度，CD值之範 圍介於約0.18至0·35微米之間，pEB之烘烤時間於所有 案例中均相同約等於2分鐘，但在一最小烘烤時間約丨分 鐘之後，烘烤時間不重要，此數據為使用K3〇G光阻層2 一底部抗反射層（BARC)存在下所得到。 曲線21之烘烤溫度為1 〇 〇 °c，由光阻製造商建議的 本紙張尺度適用中國國家標準（CNS ) A4規格（210X 297公釐） # 542939 A7 B7 i、發明説明（ 經濟部中央標準局員工消費合作社印製 值，顯示於曲線2 1上接續之每一曲線烘烤溫度為每欠減 少5 °C，最上端曲線22之烘烤溫度為75它，可見到cd 於線距/線寬比小於3時非常敏感，但此敏感度在較低pEB 溫度時明顯地降低，實際上，在p E B溫度7 5 °C時c D於 整個線距/線寬比範圍中只改變約4 %。 於圖3中，說明光學效應完全控制近接效應，在此 案例中PEB溫度假設具有上述之最佳條件並使用模擬技 術，兩主要光學參數之效應可有系統地研究。 兩用於模擬之光學參數為數值孔隙（NA)(可使用光圈 調整）及部分同調參數（σ)，後者定義為（光源NA)/(投射 ΝΑ)，可偵測一光學系統以測定此二數值，並可以修改的 aerial照明器（ASM步進器）或以放大過濾器尺寸（Nikon 步進器）作調整。 圖3中，如以下其他圖所示，να保持固定為0.55， s 如Ρ Ε Β案例中，此數值為所獲得之最高N A，我們已有許 多機會證實’ NA愈大近接效應愈小，如曲線31，同調參 數為“環狀2/3 ” ，因這些案例使用一環狀入射光分佈（環 狀半徑内徑··外徑=2/3)。 曲線32之σ為0.8(由Nikon X12B步進器可獲得之最 大值）而曲線33為0.4,數據顯示，當NA為0.55而σ為0·8 本紙張又度適用中國國家標準（CNS ) Α4規格（210Χ297公釐） (請先閱讀背面之注意事項再填寫本頁) 訂 542939 Α7 Β7 五、發明説明（ 結果在CD對線距/線寬比曲線幾乎為一平坦線’意味近接 效應已完全消除。 為確認這些結果是真實的而非由於不良的模擬，在 一真實的光阻上進行一實驗其舖設厚度在0.5與0.8微米 之間，圖4中，描繪使用不同σ值進行一系列曝光之結果， ΡΕΒ溫度在100 。(：（製造商之較佳值），σ值如下列所示： 曲線41為環狀2/3、曲線42為〇·8、曲線43為0.55、以 及曲線44為0.4，如前述，ΝΑ固定為0.55，則可見雖然 曲線43與曲線44最靠近，但無一曲線可視為平坦的，意 指近接效應仍然存在。 重覆上述實驗但這次使用农佳p E B溫度（此案例中8 Ο °C )，結杲如圖5，ΝΑ仍然固定為〇·55，σ值如下列所示： 曲線51為環狀2/3、曲線52為0.8、以及曲線53為0.4， 可見到於曲線52中，σ值為〇·8，CD不依賴線距/線寬比 降至線距/線寬比值小於1確認藉由適當的控制PEB溫度、 數值孔隙、以及部分同調參數，近接效應可被有效的消除， 亦應特別注意我們以t-BOC系列光阻得到此最佳結果。 (請先閱讀背面之注意事項再填寫本頁)Equipped with the economy, printed by the Consumer Standards Cooperative of the Central Standards Bureau ~ Refer to Figure 2, which depicts a picture in which CD (unit is micron) is a function of line pitch / line ratio, used for a series of different PEB temperatures, CD The value ranges from about 0.18 to 0.35 micron. The baking time of pEB is the same in all cases and is approximately equal to 2 minutes, but after a minimum baking time of about 丨 minutes, the baking time is not important. The data are obtained using K3OG photoresist layer 2 and a bottom anti-reflection layer (BARC). The baking temperature of curve 21 is 100 ° C. The paper size recommended by the photoresist manufacturer applies the Chinese National Standard (CNS) A4 specification (210X 297 mm) # 542939 A7 B7 i. Description of the invention (centrally by the Ministry of Economic Affairs) The printed value of the Consumer Cooperative of the Standard Bureau shows that the baking temperature of each curve continued on curve 21 is reduced by 5 ° C for each owe, and the baking temperature of the uppermost curve 22 is 75, which shows that cd is on the line spacing / Very sensitive when the line width ratio is less than 3, but this sensitivity is significantly reduced at lower pEB temperature. In fact, at a p EB temperature of 7 5 ° C, c D only changes about about the entire line pitch / line width ratio range. 4%. In Figure 3, it is shown that the optical effect completely controls the proximity effect. In this case, the PEB temperature is assumed to have the best conditions described above and the simulation technology is used. The effects of the two main optical parameters can be studied systematically. Two are used for simulation The optical parameters are numerical aperture (NA) (can be adjusted using the aperture) and some coherence parameters (σ). The latter is defined as (light source NA) / (projection NA). An optical system can be detected to determine these two values. Modified aerial illuminator (ASM Stepper) or to enlarge the filter size (Nikon stepper). In Figure 3, as shown in other figures below, να remains fixed at 0.55, s as in the case of Ρ Ε Β, this value is obtained The highest NA, we have many opportunities to confirm that 'the larger the NA, the smaller the proximity effect, such as curve 31, the coherence parameter is "ring 2/3", because these cases use a ring incident light distribution (ring radius inner diameter · · Outer diameter = 2/3). Σ of curve 32 is 0.8 (the maximum value obtained by Nikon X12B stepper) and curve 33 is 0.4. The data shows that when NA is 0.55 and σ is 0 · 8 Applicable to China National Standard (CNS) Α4 specification (210 × 297 mm) (Please read the precautions on the back before filling this page) Order 542939 Α7 Β7 V. Description of the invention A 'flat line' means that the proximity effect has been completely eliminated. To confirm that these results are real and not due to poor simulations, perform an experiment on a real photoresist with a laying thickness between 0.5 and 0.8 microns, as shown in Figure 4, Plot the results of a series of exposures using different σ values The PEB temperature is 100 ° C ((the manufacturer's preferred value), and the σ value is as follows: curve 41 is a ring 2/3, curve 42 is 0.8, curve 43 is 0.55, and curve 44 is 0.4 As mentioned above, NA is fixed at 0.55, so it can be seen that although curve 43 and curve 44 are closest, none of the curves can be regarded as flat, meaning that the proximity effect still exists. Repeat the above experiment but this time use Nongjia p EB temperature (this In the case 8 ° C), the result is shown in Fig. 5. NA is still fixed at 0.55, and the value of σ is as follows: curve 51 is a ring 2/3, curve 52 is 0.8, and curve 53 is 0.4. As for the curve 52, the value of σ is 0.8, and the CD does not depend on the line pitch / line width ratio to be reduced to a line pitch / line width ratio of less than 1. Confirm that by appropriately controlling the PEB temperature, numerical porosity, and some coherence parameters, close The effect can be effectively eliminated, and special attention should also be paid to the best results we obtain with t-BOC series photoresists. (Please read the notes on the back before filling this page)
訂 經濟部中央標準局員工消費合作社印製 本發明以較佳實施例說明如上，而熟悉此領域技藝 者，在不脫離習知技術之精神範圍内，當可作些許更動潤 飾，其專利保護範圍更當視後附之申請專利範圍及其等同 領域而定。 本紙張尺度適用中國國家標準（CNS ) Α4規格（210Χ297公釐） 542939 經濟部中央標率局員工消費合作社印袋 圖式符號說明 1，2,3,4 線 5 線長 6 光阻消失之區域 21 烘烤溫度100°C之曲 22 烘烤溫度75°C之曲彳 31 σ為2/3之曲線 32 σ為0.8之曲線 33 σ為0.4之曲線 41 σ為2/3之曲線 42 σ為0. 8之曲線 43 σ為0. 55之曲線 44 σ為0.4之曲線 51 σ為2/3之曲線 52 σ為0.8之曲線 53 σ為0.4之曲線Printed by the Consumers' Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs. The present invention is described in the preferred embodiment above. Those skilled in the art can make some modifications and modifications without departing from the spirit of the conventional technology. The scope of patent protection It should depend on the scope of the attached patent application and its equivalent fields. This paper scale applies Chinese National Standard (CNS) A4 specification (210 × 297 mm) 542939 Illustrated symbols for printed bags of the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs 1,2,3,4 Line 5 Line length 6 Area where the photoresist disappears 21 Bake temperature 100 ° C curve 22 Bake temperature 75 ° C curve 31 σ is 2/3 curve 32 σ is 0.8 curve 33 σ is 0.4 curve 41 σ is 2/3 curve 42 σ is 0.8 curve 43 σ is 0.5 55 curve 44 σ is 0.4 curve 51 σ is 2/3 curve 52 σ is 0.8 curve 53 σ is 0.4 curve
本紙張尺度適用中國國家標準（CNS ) Α4規格（210X2S (請先閲讀背面之注意事項再填寫本頁)This paper size applies to China National Standard (CNS) Α4 specification (210X2S (Please read the precautions on the back before filling this page)