WEKO3
アイテム
Studies of Chemical Reactions with Atomic Hydrogen on Si (100) Surfaces by Infrared Reflection Absorption Spectroscopy
https://ir.soken.ac.jp/records/219
https://ir.soken.ac.jp/records/21984f6fc11-bbc1-4fcf-984a-d69d394985eb
| 名前 / ファイル | ライセンス | アクション |
|---|---|---|
要旨・審査要旨 / Abstract, Screening Result (577.3 kB)
|
||
|
本文 / Thesis (5.5 MB)
|
| Item type | 学位論文 / Thesis or Dissertation(1) | |||||
|---|---|---|---|---|---|---|
| 公開日 | 2010-02-22 | |||||
| タイトル | ||||||
| タイトル | Studies of Chemical Reactions with Atomic Hydrogen on Si (100) Surfaces by Infrared Reflection Absorption Spectroscopy | |||||
| タイトル | ||||||
| タイトル | Studies of Chemical Reactions with Atomic Hydrogen on Si (100) Surfaces by Infrared Reflection Absorption Spectroscopy | |||||
| 言語 | en | |||||
| 言語 | ||||||
| 言語 | eng | |||||
| 資源タイプ | ||||||
| 資源タイプ識別子 | http://purl.org/coar/resource_type/c_46ec | |||||
| 資源タイプ | thesis | |||||
| 著者名 |
王, 志宏
× 王, 志宏 |
|||||
| フリガナ |
ワン, ジーホン
× ワン, ジーホン |
|||||
| 著者 |
WANG, Zhihong
× WANG, Zhihong |
|||||
| 学位授与機関 | ||||||
| 学位授与機関名 | 総合研究大学院大学 | |||||
| 学位名 | ||||||
| 学位名 | 博士（理学） | |||||
| 学位記番号 | ||||||
| 内容記述タイプ | Other | |||||
| 内容記述 | 総研大甲第657号 | |||||
| 研究科 | ||||||
| 値 | 数物科学研究科 | |||||
| 専攻 | ||||||
| 値 | 07 構造分子科学専攻 | |||||
| 学位授与年月日 | ||||||
| 学位授与年月日 | 2003-03-24 | |||||
| 学位授与年度 | ||||||
| 値 | 2002 | |||||
| 要旨 | ||||||
| 内容記述タイプ | Other | |||||
| 内容記述 | Chemical reactions on silicon surfaces are very important and interesting from the viewpoint of both semiconductor technology and surface science. In this thesis they shall focus on (I) the hydrogen diffusion into silicon bulk causing infrared peak width broadening, (II) the reaction of water with ideally hydrogen terminated Si(100)-(2x1) surfaces and (III) the reaction of atomic hydrogen with water covered Si(100)-(2x1) surfaces. Hydrogen reaction with Si(100) surface is an important model system in surface science studies. This system has been used extensively in semiconductor fabrication technology preventing the Si(100) surfaces from contamination and also employed as a precursor in several chemical vapor deposition reactions. Interaction of hydrogen with silicon surfaces has been extensively studied by several surface techniques such as temperature programmed desorption(TPD), scanning tunneling microscopy (STM), electron energy loss spectroscopy (EELS) as well as Fourier-transform infrared spectroscopy (FTIR). These investigations have thrown light on several aspects of hydrogen reaction with Si(100) surfaces. When silicon surface is exposed to atomic hydrogen, it generally forms (3x1), (2x1) and (1x1) surface structures depending on the adsorption conditions. However, very few reports have appeared on hydrogen diffusion into silicon bulk on nearly ideally hydrogen terminated silicon surfaces. Most of them have focused on theoretical and TPD studies. They have investigated the dependence of the line width of the coupled monohydride symmetric stretching vibration on the H-terminated Si(100)-(2x1) surface as a function of temperature and hydrogen exposure by infrared reflection absorption spectroscopy (IRRAS) using CoSi2 buried metal layer substrate (BML-IRRAS) (Fig. 1). They find that even for nearly ideally H-terminated Si(100) surface, the line width changes significantly depending on the hydrogen exposure and exposure temperatures. The minimum line width observed on the nearlyideally H-terminated surface at around 670K and 500L hydrogen exposure agrees or even better than the reported homogeneous line width determined by the dephasing effects. The dependence of line width broadening on hydrogen exposure and temperature on nearly ideal H-terminated regions can not be explained by either dephasing effects or inhomogeneities due to the coexistence of higher hydrides, or contamination by the residual water. The line width broadening is also not due to the surface roughness induced by the hydrogen etching. They suggest that the line width broadening is essentially caused by subsurface hydrogen. They have carried out a number of experiments to investigate the presence of subsurface hydrogen and its effect on the infrared line width broadening. The Si(100) surface was initially cleaned by flashing it to high temperature (～1150K) and exposed to 5000L atomic deuterium at about 670K. Then the surface silicon deuterides were completely replaced by exposing 500L atomic hydrogen at 620K. The sample was then annealed for one minute at fixed higher temperatures in order to check the reappearance of surface deuteruum from sxlicon bulk by IR at 1525 cm-1. They find that with increasing annealing temperatures, the decomposition rate of coupled monohydride (νSi-H CM) characterized by 2098 cm-1 becomes faster. This is simultaneously followed by the appearance of coupled monodeuteride peak at 1525 cm-1 which can only be explained by the diffusion of deuterium atoms from silicon bulk to the surface at higher annealing temperatures. They have also measured the amount of deuterium incorporated into the Si bulk by conventional TPD experiment for the samples made by the similar process of deuterium exposure and replacement by hydrogen as described above. The νSi-H CM IR line width, separately measured for 500L hydrogen exposure, is found to increase from 2.1cm-1 at 673K to 3.2cm-1 at 598K which is roughly in proportion to the amount of deuterium atoms incorporated in the Si bulk as determined by TPD measurements. Similarly, by keeping the exposure temperature constant, for example at 673K, the IR peak width broadening occurs from 2.1cm-1 at 500L to 2.5cm-1 at 1000L of hydrogen exposure. They have carried out several measurements to confirm these trends. The IR and TPD experiments clearly demonstrate that the diffusion of hydrogen atoms does occur into the Si(100) bulk and causes inhomogeneous broadening of the IR line width. The interaction of water with Hydrogen-terminated Si(100) surface is another important topic for investigation by BML-IRRAS technique. For this purpose, nearly ideally H-terminated Si(100) surfaces were prepared by exposing the clean Si(100) surface to 500L hydrogen at 650K. These surfaces were exposed to controlled amount of water and the changes occurring on the surfaces were monitored by BML-IRRAS as a function of the water exposure. The surface oxidation of Si has been observed due to water adsorption and it is concluded that the nearly ideally H-terminated Si(100) surface is still quite reactive with water. However, it has also been noted that the reactivity of H-terminated surface has diminished due to the passivation effect of the surface hydride layer on Si against the background water in the ultrahigh vacuum. The water-adsorbed Si(100) systems have received considerable attention due to its apparent simplicity and the widespread use of H2O in industrial oxidation processes. In spite of many excellent scientific and technological studies, much remains to be understood about this system at microscopic level. Vibrational spectroscopy and ab initio quantum chemical cluster calculations have established that there is no barrier to dissociative chemisorption of H2O on Si(100)-(2x1) surfaces forming stable Si-H and Si-OH bonds. Annealing of water exposed Si surfaces induces the insertion of oxygen atoms into Si back bonds or the formation of epoxides through dehydrogenation. It has also been reported that the initial Si(100) surface is comprised of an array of isolated and intra-row coupled dimers, which are coupled by a hydrogen bonding between OH groups that reside on the same end of adjacent dimers in a dimer row. It is proposed that such infer-dimer coupling facilitates the subsequent transfer of oxygen between dimers forming oxygen agglomeration, so that the initial oxidized surface is comprised of an inhomogeneous array of zero-, one-, and two-oxygen containing dimers. Recently the water-covered Si(100) surface exposed to atomic hydrogen at 220K has been studied using external transmission (ET) infrared geometry method (Weldon et al. J. Chew. Phys. 113, 2440, (2000)). These experiments conclusively show that the atomic hydrogen exposure induces oxygen insertion which is usually achieved by annealing the water-covered Si(100) surface. However, in this case, single O incorporation predominates which can be explained by the following mechanism. The first step is the elimination of Hz into gas phase by the abstraction of hydrogen atom from the surface Si-OH group forming the SiO･ radical species. The oxygen radical is preferentially inserted into the Si-Si dimer bond, thereby forming the single oxygen-inserted species. In spite of its scientific and technological importance, the investigation on atomic hydrogen induced-oxidation has only been performed over a limited range of temperature and hydrogen exposures. In the present work, the BML-IRRAS has been used to investigate the atomic hydrogen-induced oxidation on the water-adsorbed Si(100)-(2x1) surface over a wide range of temperatures (268～373K) and exposures of atomic hydrogen (Fig.2). The BML-IRRAS has high sensitivity for the perpendicular component over a wide frequency range including less than 1000 cm-1 (the frequency cut off region of a multiple internal reflection (MIR) geometry). Owing to this unique characteristic of BML-IRRAS, a series of oxidized and nonoxidized SiH2, which are not observed or very week by the ET geometry because of its low sensitivity to the perpendicular component, are observed clearly for the first time, and the new oxidation mechanism is proposed. Using the well-documented B3LYP gradient corrected density functional method with the polarized 6-31G** basis set for all atoms, They calculated the IR vibrational frequencies. The calculated harmonic frequencies are obtained by multiplying uniform frequency shift factors for each type of vibration, which were determined by using the assignment-established modes: νSiH CM = 2099cm-1, νSiH CM(M)=993cm-1, and νSiO2CM(O,M)= 1042cm-1 (Weldon at al.Phys.Rev.Lett. 79,2851,(1997)), νSiH ID = 2090cm-1, νSiH AD = 2107cm-1, δSiH ID = 902cm-1, and δSiH AD = 913cm-1 (Noda et al. Chem. Phys. Left. 326, 163 (2000)). By comparing the observed peaks with calculations many important peaks are uniquely assigned. (The definitions of these symbols are given in the captions of Fig. 3 ). The most interesting observations are three pairs of doublet peaks, 901 and 916cm-1, 926 and 938cm-1, and 963 and 982cm-1. These are assigned to δSiH ID and δSiH AD with zero, one and two inserted oxygen atoms at Si back bonds, respectively. The perpendicular dynamic dipole moments of these modes make these peaks insensitive for the ET geometry (Even at 60 degree incidence, the δSiH ID peak intensity observed on the H:Si(100)-(3x1) surface by ET geometry (Weldon et al. Phys. Rev. Left. 79, 2851,(1997)) is about 1/5 of the BML-IRRAS method (Noda et al. Chem. Phys. Left. 326, 163 (2000)). The small peaks observed at 990～1050cm-1 range are assigned to the SiO stretching mode of coupled monohydrides with one to three oxygen atoms at the Si back bonds and/or the Si-Si dimer bond. The strong peak at 2114cm-1 is assigned to the overlapping of the perpendicular components of νSiH CM(M) and νSiH CM(O,M). The peak at 2141cm-1 observed in the high atomic dose region is assigned to the overlapping of the perpendicular components of νSiH CM(oo,oo) and νSiH CM(oo,o). The peaks at 2198 and 2185cm-1 are assigned to the overlapping of the perpendicular components of νSiH2AD(oo,oo) and νSiH3AD(oo,o), and νSiH2AD(oo), respectively. The 2108cm-1 peak may be assigned to the overlapping of νSiH CM(o,oo)(calc. = 2108cm-1) and νSiH CM(o,o)(calc. = 2106cm-1). The band observed at ～1107cm-1 is assigned to the overlapping of νSiO ID and νSiO AD with one to four oxygen atoms inserted into Si back bonds and νSiO CM with three and four oxygen atoms inserted into Si-Si dimer and/or Si back bonds. From the studies, they have found that the atomic hydrogen-induced oxidation on water covered Si(100)-(2x1) surface has different reaction routes depending on the atomic hydrogen dose. At low atomic hydrogen dose region (< 100L), the mechanism is well explained by the reaction H-Si-Si-OH + 2H → H-Si-0-Si-H or H-Si-Si(0)-H + H2 proposed by Weldon et al. It is observed that the vibrational band at 982 cm-1 appears in IRRAS spectra only on Si(100) surface exposed to atomic hydrogen but not on the Si surface thermally annealed after water adsorption. The appearance of this peak, even in the lower hydrogen exposure region (< 100L), suggests the existence of another hydrogen atom-induced oxidation channel, H-Si-Si-OH + 2H → SiH2 + Si(0)H2 which is followed by the formation of two oxygen atoms inserted SiH2 through the infer-dimer oxygen atom migration. The increase of 982cm-1 peak intensity with decreasing 2114cm-1 peak intensity at hydrogen exposures greater than 100L suggests that the O-inserted dihydride species are also formed by the hydrogen atom-induced reaction of O-inserted dimer species such as H-Si-O-Si-H + 2H → SiH2 + Si(O)H2. The observed IRRAS spectra show that double oxygen insertion is clearly favored over single oxygen insertion. This is also reasonable from the thermo-dynamical point of view. The calculated energy of AD(00) is 0.449eV and 0.438eV lower than those of AD(0,0) and AD(0,0'), respectively. Also, CM(00) is 0.561eV and 0.509eV more stable than the respective energies of CM(0,0) and CM(0,0'). |
|||||
| 所蔵 | ||||||
| 値 | 有 | |||||
| フォーマット | ||||||
| 内容記述タイプ | Other | |||||
| 内容記述 | application/pdf | |||||
| 著者版フラグ | ||||||
| 出版タイプ | AM | |||||
| 出版タイプResource | http://purl.org/coar/version/c_ab4af688f83e57aa | |||||
