Jose Arvin M. Plomantes, Ruby Ann D. Mamangun, Armando T. Clarina, R. Guevara
{"title":"加强凹凸厚阻光刻:建立过程控制,以消除铜柱的基础","authors":"Jose Arvin M. Plomantes, Ruby Ann D. Mamangun, Armando T. Clarina, R. Guevara","doi":"10.1109/EPTC.2018.8654440","DOIUrl":null,"url":null,"abstract":"In order to fabricate tall post technologies, chemically amplified positive resist with high viscosity is usually used. Thick and uniformly coated resist is patterned during lithography to act as mold prior copper (Cu) electroplating. An unoptimized photolithography process can consequently result to defects in the plated Cu pillar – among which is Cu footing. This defect poses electrical and reliability risks such as shorting and Cu migration.In this study, Cu post footing is resolved by enhancing the thick resist lithography process. The baseline recipe was initially optimized in order to set the parameters for the design of experiment (DOE). Among the settings checked include soft bake and post-exposure bake time and z-axis settings to optimize the heat transfer process, and the develop spin direction to improve the developer coverage. The confirmation run yielded a 20.6% improvement in resist undercut measurements.Using the baseline split as comparison, a 24-split multiple-facet four-variable full-factorial design of experiment was executed by taking into consideration the soft bake (125C, 80140–125C, 80–140C), expose (1800, 2000 mJ/cm2), postexposure bake (100C, 105C) and puddle develop settings (8x, 11x).The best DOE split (140C, 1800 mJ/cm2, 100C, 8x puddle) resulted to 58.8% reduction in the actual resist undercut which translates to Cu foot elimination. This is backed up by improvement in the undercut uniformity across the wafer as the standard deviation was also reduced by 63.9%. Considering a theoretical resist undercut to ensure optimization without jeopardizing the critical dimension, an 81.4% and 81.9% improvement was observed on the undercut readings and standard deviation, accordingly.Among the parameters involved, post-exposure bake, soft bake, and develop settings are shown to have significant effects in the improvement. The optimized heat transfer during the bake process helped in the uniform solvent dissipation and enhanced resist-to-substrate adhesion. Also, the stabilization of the photoactive compound (PAC) to carboxylic acid conversion during post-exposure bake at lower temperature ensured avoidance of over production of soluble acid. Finally, development at less puddle intervals caused controlled resist dissolution and removal.","PeriodicalId":360239,"journal":{"name":"2018 IEEE 20th Electronics Packaging Technology Conference (EPTC)","volume":"33 1 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2018-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":"{\"title\":\"Enhancing Bump Thick Resist Lithography: Establishing Process Controls to Eliminate Copper Pillar Footing\",\"authors\":\"Jose Arvin M. Plomantes, Ruby Ann D. Mamangun, Armando T. Clarina, R. Guevara\",\"doi\":\"10.1109/EPTC.2018.8654440\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"In order to fabricate tall post technologies, chemically amplified positive resist with high viscosity is usually used. Thick and uniformly coated resist is patterned during lithography to act as mold prior copper (Cu) electroplating. An unoptimized photolithography process can consequently result to defects in the plated Cu pillar – among which is Cu footing. This defect poses electrical and reliability risks such as shorting and Cu migration.In this study, Cu post footing is resolved by enhancing the thick resist lithography process. The baseline recipe was initially optimized in order to set the parameters for the design of experiment (DOE). Among the settings checked include soft bake and post-exposure bake time and z-axis settings to optimize the heat transfer process, and the develop spin direction to improve the developer coverage. The confirmation run yielded a 20.6% improvement in resist undercut measurements.Using the baseline split as comparison, a 24-split multiple-facet four-variable full-factorial design of experiment was executed by taking into consideration the soft bake (125C, 80140–125C, 80–140C), expose (1800, 2000 mJ/cm2), postexposure bake (100C, 105C) and puddle develop settings (8x, 11x).The best DOE split (140C, 1800 mJ/cm2, 100C, 8x puddle) resulted to 58.8% reduction in the actual resist undercut which translates to Cu foot elimination. This is backed up by improvement in the undercut uniformity across the wafer as the standard deviation was also reduced by 63.9%. Considering a theoretical resist undercut to ensure optimization without jeopardizing the critical dimension, an 81.4% and 81.9% improvement was observed on the undercut readings and standard deviation, accordingly.Among the parameters involved, post-exposure bake, soft bake, and develop settings are shown to have significant effects in the improvement. The optimized heat transfer during the bake process helped in the uniform solvent dissipation and enhanced resist-to-substrate adhesion. Also, the stabilization of the photoactive compound (PAC) to carboxylic acid conversion during post-exposure bake at lower temperature ensured avoidance of over production of soluble acid. Finally, development at less puddle intervals caused controlled resist dissolution and removal.\",\"PeriodicalId\":360239,\"journal\":{\"name\":\"2018 IEEE 20th Electronics Packaging Technology Conference (EPTC)\",\"volume\":\"33 1 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2018-12-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"1\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2018 IEEE 20th Electronics Packaging Technology Conference (EPTC)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/EPTC.2018.8654440\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2018 IEEE 20th Electronics Packaging Technology Conference (EPTC)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/EPTC.2018.8654440","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Enhancing Bump Thick Resist Lithography: Establishing Process Controls to Eliminate Copper Pillar Footing
In order to fabricate tall post technologies, chemically amplified positive resist with high viscosity is usually used. Thick and uniformly coated resist is patterned during lithography to act as mold prior copper (Cu) electroplating. An unoptimized photolithography process can consequently result to defects in the plated Cu pillar – among which is Cu footing. This defect poses electrical and reliability risks such as shorting and Cu migration.In this study, Cu post footing is resolved by enhancing the thick resist lithography process. The baseline recipe was initially optimized in order to set the parameters for the design of experiment (DOE). Among the settings checked include soft bake and post-exposure bake time and z-axis settings to optimize the heat transfer process, and the develop spin direction to improve the developer coverage. The confirmation run yielded a 20.6% improvement in resist undercut measurements.Using the baseline split as comparison, a 24-split multiple-facet four-variable full-factorial design of experiment was executed by taking into consideration the soft bake (125C, 80140–125C, 80–140C), expose (1800, 2000 mJ/cm2), postexposure bake (100C, 105C) and puddle develop settings (8x, 11x).The best DOE split (140C, 1800 mJ/cm2, 100C, 8x puddle) resulted to 58.8% reduction in the actual resist undercut which translates to Cu foot elimination. This is backed up by improvement in the undercut uniformity across the wafer as the standard deviation was also reduced by 63.9%. Considering a theoretical resist undercut to ensure optimization without jeopardizing the critical dimension, an 81.4% and 81.9% improvement was observed on the undercut readings and standard deviation, accordingly.Among the parameters involved, post-exposure bake, soft bake, and develop settings are shown to have significant effects in the improvement. The optimized heat transfer during the bake process helped in the uniform solvent dissipation and enhanced resist-to-substrate adhesion. Also, the stabilization of the photoactive compound (PAC) to carboxylic acid conversion during post-exposure bake at lower temperature ensured avoidance of over production of soluble acid. Finally, development at less puddle intervals caused controlled resist dissolution and removal.