根據Global Information針對絕緣閘雙極電晶體(Insulated Gate Bipolar Transistors, IGBT)的市場佔有率分析、產業趨勢與統計、成長預測，預測期內絕緣閘雙極電晶體市場複合年成長率為7.45%[1]，所以代表這個市場淺力無窮，值得投入開發研究。
在絕緣閘雙極電晶體製造裡，在完成邏輯元件如閘極（Gate）、射極(emitter)、集極(collector)等後，需有與焊墊(bonding pad)作連接的裝置，這個裝置就叫做接觸窗(Contact）。再來要將導電材料填入，一般是使用金屬材料鎢(W)，利用鎢金屬連接邏輯元件與焊墊。其用鎢金屬化學氣相沉積(W CVD Deposition)將鎢金屬沉積到接觸窗通道，因無法很準確的把鎢金屬只沉積到接觸窗通道頂就停止，故都會沉積超過接觸窗通道頂，後再利用化學研磨機台或蝕刻回蝕將多餘的鎢金屬去除，而完成接觸窗。
蝕刻分為濕蝕刻又稱為等向性蝕刻，及乾蝕刻又稱為異向性蝕刻。本實驗機台為反應離子蝕刻類型，有著較高的選擇比的化學性蝕刻，輔以用電場控制電漿離子轟擊晶圓，藉以提高整個蝕刻反應速率。本實驗所要介紹的乾蝕刻機台，是科林研發的Lam 2300機型，主要由供電模組、大氣傳送模組、真空傳送模組及生產模組等四大部分組成，生產模組主要是蝕刻反應腔體。
本研究內容利用田口方法進行製程實驗，應用直交表來進行實驗得到實驗數據，根據實驗結果來計算S/N比，再根據S/N比與重要控制因子蝕刻率最大化，故選用望大特性，故選定單一沉積機台、單一蝕刻機台、單一量測機台及單一操作人員為本實驗的固定因子。關鍵因子的選擇在於蝕刻機台本身，那蝕刻機台本身會影響蝕刻率的因素有解離電場功率(TCCT power）、電漿來源的化學氣體流量大小，包括六氟化硫(SF₆)流量、四氟化碳(CF₄)流量、氧氣(O₂)流量等。故取4因子、3水準，選用田口L_9 (3^4)直交表來進行實驗配置，因子包含解離電場功率(TCCT power）、六氟化硫（SF₆）、四氟化碳(CF₄）、氧氣（O₂）等3種氣體的流量。
經由實驗結果得知，本實驗利用田口方法，選擇關鍵因子，經由S/N比分析或是蝕刻率分析，皆得到一樣的最佳化配置驗證參數，即為解離電場功率(TCCT power)為1000瓦，六氟化硫(SF₆)流量為80毫升/分鐘，四氟化碳(CF₄)流量為50毫升/分鐘，氧氣(O₂)流量為85毫升/分鐘。將最佳化配置驗證的參數，實驗30組控片蝕刻率，得到這30組平均蝕刻率為7726Å/分鐘，未實驗前的蝕刻率為3126Å/分鐘，增幅149%。故以本實驗最佳化配置驗證參數得到驗證，確實可以得到最大蝕刻率。

In IGBT manufacturing, after completing the logic components such as the gate, emitter, and collector, a structure is required to connect these elements to the bonding pad. This structure is called a contact window. Next, a conductive material must be filled into the contact window, typically using tungsten (W) metal. Tungsten is used to electrically connect the logic components to the bonding pad. Tungsten metal is deposited into the contact window channel via tungsten chemical vapor deposition (W CVD). Since it is difficult to precisely stop the deposition right at the top of the contact window channel, tungsten is usually deposited beyond the top of the channel. The excess tungsten is then removed using chemical mechanical polishing (CMP) or etch-back processes to complete the contact window structure.
This study uses the Taguchi method to conduct process experiments. An orthogonal array was applied to obtain experimental data. Based on the results, the signal-to-noise (S/N) ratio was calculated, and the etch rate was used as the primary indicator for optimization. Since the objective is to maximize the etch rate, the "larger-the-better" characteristic was selected. To reduce variation, fixed factors were set, including using the same deposition tool, same etching tool, same measurement tool, and same operator throughout the experiment.
The key factors affecting the etch rate lie in the etching system itself. These include the TCCT power (transformer-coupled capacitively-tuned power), and the flow rates of the plasma source gases: sulfur hexafluoride (SF₆), carbon tetrafluoride (CF₄), and oxygen (O₂). Therefore, four factors with three levels each were chosen, and the Taguchi L₉ (3⁴) orthogonal array was adopted for the experimental design. The factors include: TCCT power, SF₆ flow rate, CF₄ flow rate, O₂ flow rate.
Based on the experimental results, using the Taguchi method and selecting the key factors, both S/N ratio analysis and etch rate analysis yielded the same optimal parameter configuration:
TCCT power: 1000 W, SF₆ flow rate: 80 mL/min, CF₄ flow rate: 50 mL/min, O₂ flow rate: 85 mL/min.
Using this optimized parameter set, 30 control wafers were etched, and the average etch rate was measured to be 7726 Å/min. Before optimization, the etch rate was 3126 Å/min, showing a 149% increase. Therefore, the optimization configuration obtained through this experiment is verified to effectively maximize the etch rate.

[1]Global Information網址:https://www.gii.tw/report/moi1644876-insulated-gate-bipolar-transistors-market-share.html, 2025/01/05

[2]NEIL H.E. WESTE and DAVID MONEY HARRIS, CMOS VLSI Design: A Circuits and Systems Perspective., Addison Wesley, pp.19-29, 2010

[3]Yu-Chong Tai, Yong Xu and Fukang Jiang, SURFACE-MICROMACHINED PRESSURE SENSOR AND HIGH PRESSURE APPLICATION, United States Patent, US6740595B2（FIG.9–10）, 2004

[4]林美慧, "淺溝槽隔離製程的缺陷研究", 國立交通大學工學院專班半導體材料與製程設備學程, pp. 6-8, 2007

[5]Keren J. Kanarik, Samantha Tan, and Richard A. Gottscho, " Atomic Layer Etching: Rethinking the Art of Etch", Lam Research Corporation, Fremont, California 94538, United States, pp. 4814-4821, 2018

[6]Michael Quirk and Julian Serda, Semiconductor Manufacturing Technology, Prentice Hall, pp.545, 2006

[7]Michael Quirk and Julian Serda, Semiconductor Manufacturing Technology, Prentice Hall, pp. 519-546, 2006.

[8]M. R. Neisser, " INTERNATIONAL ROADMAP FOR DEVICES AND SYSTEMS", IEEE IEDM , 2020 EDITION

[9]林庚民, "鎢化學機械研磨-研磨速率與流量行為分析", 元智大學化學工程與材料科學學系, pp. 6-17, 2022

[10]李輝煌,田口方法-品質設計的原理與實務, 高立圖書有限公司, 2010

[11]Mahato G.C., Niranjan and Waquar Aarif Abu, "Analysis on IGBT Development", RVS College of Engineering and Technology, Jamshedpur India, pp.2, 2016

[12]黃景裕, "針對半導體製程金屬層良率提昇研究", 國立交通大學工學院半導體材料與製程設備學程, pp.9, 2010

[13]英飛凌公司網址: https://www.infineon.com/cms/en/ , 2025

[14]R.Manikandan^1 and R.Raja Singh^2, "IoT based IGBT Power Module Health Monitoring System with RUL Estimation for Electric Drive Applications ", 1_(School of Electrical and Electronics Engineering.Vellore Institute of Technology,Vellore,Tamil Nadu,India), 2_(Automotive Research Center,Vellore Institute of Technology,Vellore,Tamil Nadu,India), pp.2, 2017

[15]戴光助, "GaN HEMT 打線接合構裝技術之研究", 國立交通大學工學院半導體材料與製程設備學程, pp.33, 2012

[16]Blacke, Douglas Otto, "Evaluation of CVD Tungsten Metallization for Integrated Circuits application", The University of Arizona., pp.1-100, 1980

[17]G. Srinivasan, M.F. Bain, S. Bhattacharyya, P. Baine, B.M. Armstrong, H.S. Gamble, D.W. McNeill，"Tungsten silicide contacts to polycrystalline silicon and silicon–germanium alloys", Materials Science and Engineering: B, Volumes 114-115, pp.223-227, 2004

[18]Michael Quirk and Julian Serda, Semiconductor Manufacturing Technology, Prentice Hall, pp.196, 2006

[19]廖木生, "變壓耦合式電漿製程設備之蝕刻率批片控制", 國立交通大學機械工程系所, pp.8-9, 2005

[20]S.M. Sze, VLSI Technology, 2nd ed., McGraw-Hill, pp. 339–342, 1988.

[21]Campbell, S.A., The Science and Engineering of Microelectronic Fabrication, 2nd ed., Oxford Univ. Press, pp.307–315, 2001.

[22]Lieberman, M. A., and Lichtenberg, A. J., Principles of Plasma Discharges and Materials Processing, 2nd ed., Wiley-Interscience, pp.417-430, 2005.

[23]LONG MAOLIN & MARSH RICKY&PATERSON ALEX, 用於電漿蝕刻腔室之變壓器耦合電容調諧電路，中華民國106130537號專利，2013

[24]Hong Xiao, 羅正忠 and 張鼎張 " Introduction to Semiconductor Manufacturing Technology", Prentice Hall, p. 324, 2003

[25]科林研發, Lam 2300 Versys Kiyo45 Process Module Spare Parts Identifier技術手冊, pp.1-78, 2008

[26]愛德華edwardsvacuum公司網址: https://www.edwardsvacuum.com/zh-tw, 2025

[27]Lieberman, M. A., and Lichtenberg, A. J., Principles of Plasma Discharges and Materials Processing (2nd ed.), wiley online library, pp.592-595, 2005

[28]菲希爾測試儀器有限公司網址: https://www.helmut-fischer.com/zh/ , 2025