China
Africa
Explore chapters and articles related to this topic
Detector Fabrication
Published in Alan Owens, Semiconductor Radiation Detectors, 2019
For epitaxial produced material, the end product is a wafer of semiconducting material between 2 inches and 8 inches in diameter, although for most compound semiconductors, wafer sizes are limited to 2 inches. Once processed and patterned with particular device structures, on-wafer testing of each chip is carried out and the wafer is separated into individual chips, commonly referred to as wafer dicing. Depending on the wafer material and its thickness, dicing is achieved by (i) scribing along selected crystallographic planes and breaking, (ii) cutting with a high precision diamond blade or (iii) laser cutting. For melt grown crystals, the ingot or boule is first sliced into wafers using a diamond tip or wire saw and the surfaces lapped and polished prior to patterning and dicing. All other operations beyond this point are identical.
Recent Advances in Floorplanning
Published in Charles J. Alpert, Dinesh P. Mehta, Sachin S. Sapatnekar, Handbook of Algorithms for Physical Design Automation, 2008
However, this approach presents some new challenges. These different chips have to be extracted from the wafer by cutting (dicing) the wafer. Existing wafer dicing technologies are somewhat restrictive, making chip locations on the reticle vital to optimizing the chip yield. For example, the side-to-side wafer dicing technology cuts the wafer using horizontal and vertical cutlines that traverse the entire length of the wafer (Figure 12.8). Within a given reticle, these lines may either cut through dies rendering them useless or might leave large margins making the dies unacceptably large.
Forward stepwise random forest analysis for experimental designs
Published in Journal of Quality Technology, 2021
Machine learning is an artificial intelligence technology that applies computational algorithm to create systems that can learn from data and build analytical models for accurate prediction. It has been widely applied in many fields for analyzing various types of data that have complicated model structures. In experimental designs, the cases discussed in the literature or the examples introduced in the textbooks usually involve numerical data assumed to be linear and follow a normal distribution. The analysis of variance (ANOVA) method based on the F-test and the regression method based on the t-test are often employed for data analysis and model fitting. In practice, however, experiments may have different types of responses and distributions of the data may not be always normal. For instance, responses in wafer-dicing experiments can be categorical: NG (crack) or OK (no crack); advance rates in the drill experiment (Daniel 1976) are not normal. In some cases, the relationships between the responses and factors are more complicated and the structures of the underlying true models are usually uncertain. Although transformation can make the data approximately linear and normal, it is sometimes difficult to find a suitable one. If the data disobey these assumptions, analysis results based on the F-test and t-test can lead to incorrect decisions.
Whereabouts of missing atoms in a laser-injected Si: Part 1
Published in Philosophical Magazine, 2019
Daisuke Kawaguchi, Hiroyuki Iwata, Hiroyasu Saka
Laser-matter interaction is of practical and academic interest. Practically it has been successfully applied to dry and nearly debris-free wafer dicing of Si wafers, i.e. the separation of Si wafers into individual dices [1]. Of academic interest is the mechanism by which a laser-injected Si is fractured. Microstructures of a laser-induced modified volume (hereinafter denoted by LIMV) have been studied comprehensively by the present authors [2–4]. Figure 1 illustrates schematically an overall picture of LIMV. It can be summarised as follows from the focal region toward the entrance-surface of a laser beam: ① Voids accompanied with heavily compressed regions of diamond Si ④ and nano-cracks (①’).② Voids virtually free from defects.③ A dislocated region with a high density of glide-set dislocations.⑤ Chimney-like region accompanied with a few amount of sessile dislocations.⑥ A large crack is observed to emanate from near the heavily dislocated regions but only very rarely.