{"title":"Intermetallic growth kinetics of 2N gold wire ball bonds on aluminium bond pads annealed at 175 °C","authors":"C. D. Breach","doi":"10.1007/s13404-016-0184-6","DOIUrl":null,"url":null,"abstract":"<p>Gold ball bonding wires of 2N purity were bonded to Si devices with aluminium alloy bond pad metallisation. Specimens were subjected to isothermal annealing at 175?°C in a fan-circulated oven in an air environment for various times <i>t</i>. Samples of ball bonds from each specimen were cross-sectioned, and the total intermetallic thickness <i>x</i> at the centre of the ball bonds was measured. The kinetics of <i>x</i> versus <i>t</i> were modelled using a power law expression <span>\\( x(t)={x}_0+{\\theta}_1{t}^{\\theta_2} \\)</span> with growth parameter <i>θ</i>\n <sub>1</sub> and constrained parameter 0?<?<i>θ</i>\n <sub>2</sub>?≤?1. The parameter <i>θ</i>\n <sub>2</sub> is supposedly capable of differentiating between diffusion controlled intermetallic growth (grain boundary and/or volume diffusion) and interfacial reaction controlled growth with the presence/absence of grain growth solely based on its numerical value. The parameters <i>θ</i>\n <sub>1</sub>?and <i>θ</i>\n <sub>2</sub> and the regression function were obtained by log-transformation of both the response variable <i>x</i> and the explanatory variable <i>t</i> and data fitting using simple linear regression (SLR), taking care to perform diagnostic checks to confirm that transformed data satisfied the requirements of SLR. Simulated intermetallic thickness data obtained by sampling from lognormal distributions was observed to closely approximate the experimental observations. Values of the exponent <i>θ</i>\n <sub>2</sub> were observed in the range 0.2–0.4, which was interpreted as indicative of grain boundary dominated interdiffusion with grain growth.</p>","PeriodicalId":55086,"journal":{"name":"Gold Bulletin","volume":null,"pages":null},"PeriodicalIF":2.2000,"publicationDate":"2016-10-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1007/s13404-016-0184-6","citationCount":"5","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Gold Bulletin","FirstCategoryId":"5","ListUrlMain":"https://link.springer.com/article/10.1007/s13404-016-0184-6","RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"Chemistry","Score":null,"Total":0}
引用次数: 5
Abstract
Gold ball bonding wires of 2N purity were bonded to Si devices with aluminium alloy bond pad metallisation. Specimens were subjected to isothermal annealing at 175?°C in a fan-circulated oven in an air environment for various times t. Samples of ball bonds from each specimen were cross-sectioned, and the total intermetallic thickness x at the centre of the ball bonds was measured. The kinetics of x versus t were modelled using a power law expression \( x(t)={x}_0+{\theta}_1{t}^{\theta_2} \) with growth parameter θ1 and constrained parameter 0?<?θ2?≤?1. The parameter θ2 is supposedly capable of differentiating between diffusion controlled intermetallic growth (grain boundary and/or volume diffusion) and interfacial reaction controlled growth with the presence/absence of grain growth solely based on its numerical value. The parameters θ1?and θ2 and the regression function were obtained by log-transformation of both the response variable x and the explanatory variable t and data fitting using simple linear regression (SLR), taking care to perform diagnostic checks to confirm that transformed data satisfied the requirements of SLR. Simulated intermetallic thickness data obtained by sampling from lognormal distributions was observed to closely approximate the experimental observations. Values of the exponent θ2 were observed in the range 0.2–0.4, which was interpreted as indicative of grain boundary dominated interdiffusion with grain growth.
期刊介绍:
Gold Bulletin is the premier international peer reviewed journal on the latest science, technology and applications of gold. It includes papers on the latest research advances, state-of-the-art reviews, conference reports, book reviews and highlights of patents and scientific literature. Gold Bulletin does not publish manuscripts covering the snthesis of Gold nanoparticles in the presence of plant extracts or other nature-derived extracts. Gold Bulletin has been published over 40 years as a multidisciplinary journal read by chemists, physicists, engineers, metallurgists, materials scientists, biotechnologists, surface scientists, and nanotechnologists amongst others, both within industry and academia. Gold Bulletin is published in Association with the World Gold Council.