1) These findings indicate that "juvenile acute nonherpetic encephalitis" or a subset of this disorder is mediated by an antibody-associated immune response against NR1/NR2 heteromers of the NMDA receptor (NMDAR). 2) Our patients' clinical features emphasize that anti-NMDAR encephalitis is severe but potentially reversible and may precede by years the detection of an ovarian teratoma. 3) Although recovery may occur without tumor removal, the severity and extended duration of symptoms support tumor removal.

A Si epitaxial layer was selectively grown on Si02-patterned Si(100) with no miscut and on 1', 3', and 4' miscut vicinal surfaces by ultrahigh-vacuum chemical-vapor deposition using disilane. On the patterned Si(100) surfaces with and without miscuts, faceted (100) layers grew. Although [ 111 I is energetically the most stable surface, the resulting facets had [311I orientation. This means that the [311I faceting is related to the growth kinetics rather than the energetics. Macroscopically, the [311 j faceting is caused by the slower growth rate of the (311)surface than that of the (100) surface. On the vicinal surfaces, the top surface of the selectively grown layer was not vicinal but had exactly (100) orientation. This indicates the ready incorporation of adatoms into the step edges of (100) terraces and the subsequent step-flow growth mode. When the flowing step reaches the end of the top surface, it is swallowed into the facet and disappears. This decreases the step density and makes the vicinal surface flat. With respect to the swallowing of the growing steps at the facet corner, we propose a microscopic model for the [311) faceting.

Using a scanning tunneling microscope, we examine the evolution of the surface con6guration of GaAs(001) upon annealing after molecular-beam-epitaxy growth interruption.Soon after growth interruption, the surface exhibits many two-dimensional islands elongated along the [110] direction and a ragged step con6guration.Continuous annealing after growth interruption causes changes in the surface topography. Annealing for 2 min makes the surface very smooth by decreasing the number of islands and smoothing the step shapes. Further annealing for about 20 min causes step bunching. The decrease in the number of islands and smoothing of step shapes are explained by shortening of the total step length.Control of the step structure of GaAs(001) surfaces grown by molecular-beam epitaxy (MBE) is of technological importance for developing superlattice structures.Step configurations on the surface or at interfaces have been intensively studied by reflection high-energy electron diffraction (RHEED) and cross-sectional transmission electron microscopy (TEM) observations. Both methods provide one-dimensional information about the surface or interface structures.However, MBE growth on GaAs surfaces oriented exactly to (001) is generally in a twodimensional island (2D island) growth mode, because of the large terrace width compared to the diffusion length of the growth-determining species on the surface. Photoluminescence (PL) studies of quantum wells' suggest that after growth interruption these 2D islands disappear by atomic diffusion on the surface. Cross-sectional highresolution TEM observations show a very smooth heterointerface for A1As films on GaAs(001) grown by MBE with growth interruption. To study this growthinterruption efl'ect by direct two-dimensional observations, we examined a GaAs(001) surface grown by MBE using a scanning tunneling microscope (STM). STM observations of various reconstructed GaAs(001) structures have been reported by Pashley et al. and Biegelsen et al. In this report, we observe by STM evolution of the GaAs(001)2x4 surface configurations due to growth interruption, and discuss the anisotropy of island and step shapes. Our STM is the same type as the field ion-scanning tunneling microscope apparatus developed by Sakurai et al. , but equipped with an MBE chamber to observe MBE grown surfaces without exposing them to air. Substrates were exactly oriented N+-type GaAs(001) wafers, prepared by chemical etching using H2SO4 and H202 solutions. The surface oxide layers on the substrates were removed by heating above 600 C in an As flux in the MBE chamber. The GaAs epitaxial layer growth was started by opening a Ga cell shutter and was stopped by closing it. This was the growth-interruption step. (In MBE growth interruption, the Ga shutter is closed but the As shutter remains open. ) During the MBE growth and after growth interruption the substrate was kept at 590'C while being exposed to the As flux. The Ga and As flux were 3X10 s and 1 x10 Torr, respectively. The fluxes were measured by a flux gauge at th...

No abstract

No abstract