Chemical mechanical polishing (CMP) 1 is the final and thus the most important step of a Si wafer shaping process (after cutting, lapping, and chemical etching) to produce an atomically flat and defectfree surface for further electronic device manufacturing. The process of CMP consists of rotation of a soft polishing pad on a Si wafer with alkaline silica (SiO 2 ) slurry. It has been known that CMP is a complicated process, where mechanical wear by the SiO 2 particles and chemical corrosion by the slurry solution occur simultaneously. This complexity leads to the difficulty of analyzing a microscopic Si removal process that controls the polished surface morphology on an atomic scale. Schnegg et al. have proposed the removal process in which OH Ϫ adsorption on the topmost Si atoms catalyze the corrosive reaction by H 2 O, resulting in cleavage of Si back bonds. 2 Pietsch et al. have verified this mechanism by Fourier transform infrared (FTIR) spectroscopy of the Si surface immediately after CMP. They presented a removal mechanism, in which a silicic acid species, i.e., Si(OH) 4 , (OH) 3 Si-O-Si(OH) 3 ,... is formed by local oxidation by OH Ϫ and frictionally removed by the mechanical action of the SiO 2 particles. 3 Although the formation of silicic acid species has been identified, the frictional interaction between the SiO 2 particle and those species has not fully been understood.In this study, we examine the origin of the frictional interaction by considering a chemical reaction between Si and SiO 2 during wear. We use an atomic force microscope (AFM), which is a powerful tool for atomic scale wear analysis. 4-6 The wear test is performed by scratching a thermally oxidized SiO 2 film on a Si wafer with a Si AFM tip in KOH solution at various pH values. The chemical reaction between Si and SiO 2 is investigated by comparing the Si removal volume in moles with that of SiO 2 by the wear test, in order to clarify the Si removal mechanism on an atomic scale. ExperimentalA commercial-type AFM with a closed glass fluid cell was used in this experiment (Digital Instruments, Santa Barbara, CA). 7 A four-sided pyramidal single-crystal Si tip, mounted on a microfabricated Si cantilever (Digital Instruments, Santa Barbara, CA) was used for both the wear test and AFM observation. The cantilever is 125 m long and has a spring constant of 20-50 N/m. A SiO 2 film was grown on a nondoping Si(100) wafer by thermally oxidizing (1000CЊ, 3 h) in a dry O 2 atmosphere. Prior to the thermal growth, the wafer was cleaned by using the RCA procedure. 8 The thickness of the oxide film was approximately 100 nm. KOH solutions were prepared by mixing reagent grade chemicals and deionized water. The pH value of the solution was adjusted to add buffered solutions, H 3 BO 3 (0.30 kmol/m 3 ) and Na 2 B 4 O 7 (0.075 kmol/m 3 ).After the solution was injected into the cell, the Si tip was brought to the SiO 2 film and scratched the film surface for the wear test. An area of 10 ϫ 10 m was scratched with a tip scan speed of 20 m/s at normal loads r...

No abstract

The removal of the carbon from Nd magnet scraps is indispensable for high-quality recycling by the induction melting method as a preliminary process. The Nd magnet scraps can be decarburized to a level of less than 0.03 mass% by using an oxygen source at high temperatures, as reported in Part 1. The decarburized Nd magnet scraps can then be deoxidized by using the Ca reduction to a level that allows commercial melting in an induction furnace, as reported in Part II. However, the undesirable iron oxide (Fe 2 O 3 ) which causes a disadvantage for Ca reduction is inevitably generated by using an oxygen source at high temperatures. The aim of this work is to investigate an economical decarburization method in which only the carbon sources in Nd magnet scraps are decarburized, without generating iron oxide. The grinding sludge as Nd magnet scraps is effectively decarburized to a level of less than 0.03 mass% without generating any iron oxide by heating at above 1073 K under a pressure of less than 5.32 × 10 −2 Pa. The amount of oxygen in the decarburized powder is about 8 mass%, which is lower in comparison with its value in Part 1. In this report, the decarburization mechanism under reduced pressure using the grinding sludge, and its economic significance prior to the decarburizing method described in Part 1, are discussed.

No abstract

No abstract