@misc{oai:ir.soken.ac.jp:00000189,
 author = {平野, 真也 and ヒラノ, シンヤ and HIRANO, Shinnya},
 month = {2016-02-17, 2016-02-17},
 note = {Synchrotron-radiation (SR) stimulated processes have attracted much attention in semiconductor technology for fabricating submicron- and nano-scale devices, because these are low-temperature and low-damage processes with high selectivity in the excitation of chemical reactions.  From this viewpoint, the understanding and the control of the chemical nature of chemisorbed hydrogen on Si(100) surfaces (H(D)-Si(100)) are important research subjects in SR processes as they are commonly occurring processes.  The purpose of his thesis is clarifying the annealing and SR irradiation effects on the structure of the H(D)-Si(100) surfaces.
  It is well known that the structure of H(D)-Si(100) surfaces is strongly dependent on the condition of the atomic hydrogen exposure.  The monohydride surface which is prepared by a saturation exposure of atomic hydrogen at high temperatures of around 600 K yields the ordered 2x1 structure.  The saturation exposure of atomic hydrogen on a 2x1 surface at 380±20 K yields an ordered 3x1 structure which consists of alternating coupled monohydride (H-Si-Si-H) and dihydride (H-Si-H) units.  It is also reported that saturation exposure below 400 K produces a 1x1 structure which consists mainly of mono- and dihydride and of some trihydride species.  Concerning the 1x1 surface obtained by the hydrogen saturation exposure (H(D)-Si(100) 1x1), it is reported that the structure
is a disordered 3x1 consisting of H-Si-Si-H and H-Si-H when formed at 300 K.
  High-resolution infrared (IR) absorption spectroscopy is a powerful tool for the detection of adsorbed hydrogen on silicon surfaces and for the investigation of the associated structures and reaction mechanisms.  However, the SR irradiation and annealing effects on the hydrogen terminated systems, have not yet been investigated in detail using this technique.
  In this work, the structure of the H(D)-Si(100) 1x1 surface formed at 400 K and its change by annealing and SR irradiation have been investigated by BML-IRRAS method using a CoSi 2  BML Si(100) substrate, and reflective high-energy electron diffraction (RHEED).　 Here, the initial substrate temperature of 400 K was selected so that the existence of trihydride, which complicates the analysis, can be ignored.　 The SR irradiation experiments were carried out using the SR beam without monochromatization from the beam line 4B of the 0.75 GeV storage ring (UVSOR) at the Institute for Molecular Science.
  From measurements of IRRAS of silicon hydrides on the Si(100) BML substrate surfaces as a function of hydrogen exposure at 400 K.  Two peaks, which are assigned to the 818 symmetric stretching and Sin. symmetric bending vibration bands, respectively, are observed.  At around the saturation exposure (1000 L), the peak height of the SiH symmetric stretching vibration band decreases to about 273 of that in 100 L, accompanied by the appearance of a SiH 2  bending vibration band and the change of RHEED pattern from 2x1 to 1x1.
  Therefore, it shows that only monohydrides and dihydrides exist as dominant species at the saturation exposure region.  Furthermore, since the integrated absorbance (IA) of the Si-D stretchinw vibration band increases about 1.5 times, as described later, upon 650 K annealing, which decomposes dideuteride to monodeuteride, it is considered that the distribution of Si atoms between dihydride and monohydride is 1:2 on the H-Si(100) 1x1 surface.  Since, 3x1 unit cell consisting of iterating H-Si-H and coupled monohydride (H-Si-Si-H) is a stable structure at 400 K, and H-Si(100) 1x1 surface formed at 300 K is made up of disordered 3x1 units, it is quite possible from the above result that the structure of the H-Si(100) 1x1 surface at 400 K is also a disordered 3x1 structure consisting of H-Si-H and H-Si-Si-H units.
  Next, the effects of the annealing and SR irradiation on the D-Si(100) 1x1 surface were investigated.  From the annealing study, the observed IRRAS spectral change due to annealing is explained from the results of the temperature programmed desorption (TPD) experiments for the hydrogen (deuterium) adsorbed Si(100) surface.  It is known that the β1  peak of TPD, which starts to rise at about 670 K and peaks at around 760 K (dT/ds=1.7 K), indicates the hydrogen desorption from monohydride phase, and the β2 peak which starts to rise at around 550 K and peaks at 630 K, indicates the hydrogen desorption from dihydride phase.  Therefore, the rapid decrease of the IA at around 700 K is ascribed to the β1 desorption and the corresponding peak shift to the lower frequency side is explained mainly by the decrease of the dipole-dipole coupling interaction.  Concerning the β1 desorption, the preparing mechanism is supported by the scanning tunneling microscopy (STM) experiments, i.e., monohydride (monodeuteride) desorbs through the precursor form of H-Si-Si-H (D-Si-Si-D).  Therefore, it is concluded that the surface after the annealing at 670 K or higher is covered only by D-Si-Si-D, if there exist no defects on the surface, and the sharp SiD stretching vibration band observed at around 1525 cm -1 in the case of annealing study is assigned to the D-Si-Si-D symmetric stretching vibration.
  On the other hand, concerning this β2 decomposition, the IA of the SiD stretching vibration band increases by about 1.5 times with annealing ternperature increase from 400 to 650 K.  This supports the fact that the distribution of Si atoms between dideuteride and monodeuteride is about 1:2 on the D-Si(100) 1x1 surface and β2 desorption is explained by the thermal decomposition of 2SiD 2 to D-Si-Si-D + D 2 (gas).
  Next, the change of the SiD stretching vibration band due to the annealing plus SR irradiation is observed.  Here, the SR was irradiated on the surface during 15 min annealing at each annealing temperature.  It is reported that the SR irradiation decomposes di- and trihydrides but not the monohydride.  In the present case, a slight increase of the IA of the SiD stretching vibration band is observed upon 570 K annealing plus SR irradiation.  This increase is explained by the decomposition of dideuteride to monodeuteride by the SR irradiation, since no increase is observed only upon 570 K annealing.  The decomposition of dideuteride to monodeuteride by the SR irradiation is more clearly shown by taking the difference spectrum between the spectra before and after the SR irradiation.  It has also been found that the RHEED pattern did not change from 1x1 by the SR irradiation on the D-Si(100) 1x1 at 400 K.  It is assumed that the decomposition of dideuteride by SR at low temperature does not mainly result in the formation of D-SI-Si-D.  The main process is probably the decomposition of single D-Si-D.  Concerning this dideuteride decomposition by SR irradiation, not only the desorption of deuterium atom by breaking the Si-D bond, but also the possibility of the etching (desorption of SiD or SiD2) must be considered.
  By comparing both cases, they discussed this etching effects from a different viewpoint.  By annealing at 650 K or higher, if there is no defect on the Si(100) surface, the existing surface deuteride species should become to be only D-Si-Si-D, and therefore, the SiD stretching vibration band shape should become sharp and symmetric as observed in annealing study.  However, the observed results are quite different in the case of SR irradiation plus annealing.  The SR irradiation effefcts clearly appear in the shape of the IRRAS SiD stretching vibration band i.e., at the annealing temperature higher than 650 K, where only monodeuteride exists, the band shape becomes broad and asymmetric.  In the case of annealing (670 K) + SR irradiation.  This asymmetric shape can be explained by the etching of the Si(100) surface by the SR irradiation.  If there are no defects on the Si(100) surface, the surface hydride species should be, as already mentioned, only D-Si-Si-D, which are thermodynamically stable in the high temperature region where the surface migration of deuterium atoms frequently occurs.  But, the observed SiD stretching vibration band shape shows that the surface species are not only D-Si-Si-D.  Therefore, it is concluded that some defects are generated by SR irradiation on D-Si(100) 1x1 surface which consists of D-Si-Si-D and D-Si-D, and it is Si-D, because monodeuteride is not decomposed by SR irradiation.  Since only monodeuteride exists on the surface after 650 K annealing, the component of the low frequency side tail of the SiD stretching vibration band in the high temperature region above 650 K is assigned to the stretching vibration of Si-D at the defect site generated by the etching.  This means that the SR irradiation induces not only the desorption of D by breaking the Si-D bond, but also the desorption of SiD and/or SiD2  by breaking the backbonds of D-Si-D., application/pdf, 総研大甲第373号},
 title = {赤外反射吸収分光法によるSi(100)表面に化学吸着した 水素(重水素)の表面反応に関する研究},
 year = {}
}