G. S. Higashi
Bell Labs
Network
Latest external collaboration on country level. Dive into details by clicking on the dots.
Publication
Featured researches published by G. S. Higashi.
Applied Physics Letters | 1992
Zhifeng Sui; Patrick P. Leong; Irving P. Herman; G. S. Higashi; H. Temkin
Porous silicon that strongly emits in the visible was analyzed using Raman scattering. The spectrum peaks near 508 cm−1, has a width of ∼40 cm−1, and is very asymmetric. Using a model of phonon confinement, this suggests that the local structure of porous silicon is more like a sphere than a rod and has a characteristic diameter of 2.5–3.0 nm. Polarization Raman measurements suggest that the structure does not consist of a series of parallel columns.
Applied Physics Letters | 1989
G. S. Higashi; C. G. Fleming
Sequential surface reactions of trimethylaluminum and water vapor have been used to deposit Al2O3 on Si(100) surfaces. The self‐limiting nature of the surface reactions allows precise control of the thickness of the deposited layers and gives rise to films with highly conformal step coverage. High quality dielectrics have been deposited at temperatures as low as 100 °C. Resistivities of 1017 Ω cm, breakdown strengths of 8×106 V/cm, and interface‐state densities of 1011 states/eV cm2 have already been achieved and they suggest possible applications as a gate insulator or a dielectric passivation layer.
Applied Physics Letters | 1990
R. Hull; J. C. Bean; J. M. Bonar; G. S. Higashi; K. T. Short; H. Temkin; Alice E. White
It is shown that strain relaxation during annealing of Si/GexSi1−x/Si heterostructures is significantly enhanced if the strained GexSi1−x layers are implanted with p (B) or n (As) type dopants below the amorphization dose. Comparison of strain relaxation during in situ annealing studies in a transmission electron microscope between unimplanted and implanted structures reveals that the latter show residual strains substantialy below those for unimplanted structures. We propose that this enhanced relaxation is caused by increased dislocation nucleation probabilities due to the high point‐defect concentrations arising from implantation.
Applied Physics Letters | 1991
D. J. Eaglesham; G. S. Higashi; M. Cerullo
We describe a new low‐temperature clean for Si molecular beam epitaxy. Growth is carried out on Si wafers subjected to an ≊10–60 s clean in a buffered HF solution prior to insertion in the growth chamber. We demonstrate low defect densities (<105 cm−2) at 380 °C without either the conventional high‐temperature cleaning step to desorb a chemical oxide or the use of a glovebox for chemical cleaning and transfer to the vacuum chamber. Wafers are given an ≊200 °C prebake in situ to remove hydrocarbons, and then raised to the growth temperature prior to deposition. For (100) substrates the transition from amorphous deposition to crystalline growth occurs at ≊370 °C, below the temperature at which hydrogen should desorb. On (111) the minimum temperature for epitaxy is ≊500 °C, as expected. We attribute this difference to the large number of undercoordinated Si atoms present on (100), which allows growth to take place even on the hydrogen‐terminated surface. Secondary‐ion mass spectrometry suggests that contamin...
Journal of Applied Physics | 1991
M. L. Green; B. E. Weir; D. Brasen; Yong‐Fen Hsieh; G. S. Higashi; A. Feygenson; L. C. Feldman; Randall L. Headrick
Traditional techniques for growing Si‐Ge layers have centered around low‐temperature growth methods such as molecular‐beam epitaxy and ultrahigh vacuum chemical vapor deposition in order to achieve strain metastability and good growth control. Recognizing that metastable films are probably undesirable in state‐of‐the‐art devices on the basis of reliability considerations, and that in general, crystal perfection increases with increasing deposition temperatures, we have grown mechanically stable Si‐Ge films (i.e., films whose composition and thickness places them on or below the Matthews–Blakeslee mechanical equilibrium curve) at 900 °C by rapid thermal chemical vapor deposition. Although this limits the thickness and the Ge composition range, such films are exactly those required for high‐speed heterojunction bipolar transistors and Si/Si‐Ge superlattices, for example. The 900 °C films contain three orders of magnitude less oxygen than their limited reaction processing counterparts grown at 625 °C. The fi...
Applied Physics Letters | 1986
G. S. Higashi; C. G. Fleming
Excimer laser photolysis of organoaluminum adlayers has been used to catalytically activate the deposition of Al via thermal decomposition of triisobutylaluminum. The process exhibits good spatial selectivity and patterns with 4 μm resolution have been accurately reproduced. Patterned Al metallizations have been performed on Si, SiO2, Al2O3, and GaAs substrates and show promise for practical applications. Electrical measurements probing Al/substrate interface quality indicate that this technique may be suitable for the fabrication of rectifying contacts on GaAs.
Surface Science | 1995
G.J. Pietsch; Yves J. Chabal; G. S. Higashi
Abstract Chemo-mechanical polishing (CMP) of silicon with a colloidal suspension of silica (“Siton”) is the standard technology for the preparation of smooth, defect-free silicon starting surfaces for microelectronic device patterning. Despite its importance in device manufacturing, little is known about the microscopic removal mechanism during CMP that controls the resulting surface properties. With infrared spectroscopy we find that, after CMP, a surface termination by hydrogen predominates on Si(111) and Si(100). This H-termination is responsible for the observed strong hydrophobicity of the surface and its chemical stability (passivation) in air. Hydrophobicity (contact angle) and polishing removal rate strongly depend on the slurry pH and peak at pH ≈ 11. At this optimum pH a nearly “ideal” termination by monohydride is found on Si(111) which points to perfect atomic-scale surface planarity and chemical homogeneity. Si(100), after CMP, exhibits a more complex H-termination by mono-, di-, and trihydrides. At higher or lower pH, OH groups replace some of the hydride species both on CMP-Si(111) and CMP-Si(100). We present a microscopic removal mechanism which — on an atomic scale — is determined by an interplay of local oxidation by OH− and passivation by hydrogen.
Surface Science | 1991
Yves J. Chabal; Ph. Dumas; P. Guyot-Sionnest; G. S. Higashi
Picosecond sum-frequency generation measurements of the Si-H stretching vibration of an unreconstructed, ideally H-terminated Si(111) surface show that its lifetime is 0.8 ± 0.1 ns. High resolution infrared reflection-absorption measurements reveal a marked temperature dependence of the linewidth. frequency and intensity of the SiH stretching vibration. The width and frequency variations are completely accounted for by a weak coupling of this mode to a Si surface phonon centered at 210 ± 25 cm1. The loss in intensity, observed as the temperature is increased above 300 K, gives evidence for a strong coupling between the SiH stretching and bending modes.
Applied Physics Letters | 1985
G. S. Higashi; L. J. Rothberg
We report the results of ultraviolet irradiation of monolayers of organoaluminum complexes adsorbed to sapphire. Using surface vibrational spectroscopy, the compositional and structural changes caused by 248 and 193 nm surface illumination are documented. Electronic photochemistry of adsorbed trimethylaluminum is observed and can lead to nucleation of aluminum on the surface. A strong wavelength dependence to this process is observed which cannot be explained by assuming photochemical yields to be proportional to previously measured surface absorption cross sections. The implications for laser chemical vapor deposition of spatially localized metallic features are discussed.
Applied Physics Letters | 1993
Mau‐Tsu Tang; K. W. Evans‐Lutterodt; G. S. Higashi; T. Boone
We use synchrotron x‐ray diffraction to characterize the roughness of the buried Si(001)/SiO2 interface, for three types of oxide, without modification of the interface. We show that the thermal oxide interface is 0.5±0.1 times as rough as the native oxide interface, suggesting that the oxide growth decreases the roughness slightly. We also measure the roughness of a chemically grown oxide interface.