{"title":"Physics of Die Attach Interfaces","authors":"D. R. Kitchen","doi":"10.1109/IRPS.1980.362958","DOIUrl":null,"url":null,"abstract":"The purpose of this investigation is to characterize the reactions occurring at the die/substrate interface and the causes of void formation and nonadherence observed in the die attach process. Using classical thermodynamics and a Differential Scanning Calorimeter, this study will clearly show that the gold-silicon eutectic is an undesirable alloy for brazing a circuit die to a gold-plated substrate. Furthermore, the study establishes a ternary gold-silicon-tin (¿) alloy as a more appropriate braze alloy. Experimental evidence is presented which demonstrates that the improved alloy has a melting point of 274°C, but can be varied from 196°C to 1410°C, and consistently wets the surface of a gold-plated substrate. This proposed ternary alloy was tested by brazing eleven separate dies to substrates at 300°C ± 10°C. A subsequent x-ray analysis of the brazed samples revealed no voids in the parting line between the die and substrate. Moreover, in order to test the strength of the bond between the die and substrate, the brazed combinations were subjected to the standard die-shear test performed according to MIL-STD-883B, 31 August 1977.5 In each case, the samples tested, exceeded a 1000 gram load, passing the MIL-SPEC for the given die area. Finally, testing the braze alloy on a production line indicates that lower brazing temperatures are used in the die attach process and that adherent bonds form between the die and substrate.","PeriodicalId":270567,"journal":{"name":"18th International Reliability Physics Symposium","volume":"91 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"1980-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"14","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"18th International Reliability Physics Symposium","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/IRPS.1980.362958","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 14
Abstract
The purpose of this investigation is to characterize the reactions occurring at the die/substrate interface and the causes of void formation and nonadherence observed in the die attach process. Using classical thermodynamics and a Differential Scanning Calorimeter, this study will clearly show that the gold-silicon eutectic is an undesirable alloy for brazing a circuit die to a gold-plated substrate. Furthermore, the study establishes a ternary gold-silicon-tin (¿) alloy as a more appropriate braze alloy. Experimental evidence is presented which demonstrates that the improved alloy has a melting point of 274°C, but can be varied from 196°C to 1410°C, and consistently wets the surface of a gold-plated substrate. This proposed ternary alloy was tested by brazing eleven separate dies to substrates at 300°C ± 10°C. A subsequent x-ray analysis of the brazed samples revealed no voids in the parting line between the die and substrate. Moreover, in order to test the strength of the bond between the die and substrate, the brazed combinations were subjected to the standard die-shear test performed according to MIL-STD-883B, 31 August 1977.5 In each case, the samples tested, exceeded a 1000 gram load, passing the MIL-SPEC for the given die area. Finally, testing the braze alloy on a production line indicates that lower brazing temperatures are used in the die attach process and that adherent bonds form between the die and substrate.