The interactions of ethanol with a Brønsted acid site (a hydroxy group) and with a Lewis acid site (a metal ion) on surfaces of seven metal oxides (SiO2-Al2O3, SiO2, TiO2, ZnO, MgO, MnO and CdO) have been investigated using the DV-Xα method. The oxides were selected from a series of catalysts, whose selectivities for the dehydration and the dehydrogenation of ethanol had been experimentally determined by P. Sabatier and A. Mailhe. By comparing the results of the DV-Xα calculations from several models of ethanol interacting with the Brønsted or Lewis acid sites with the experimental selectivities, it was concluded that a proton shift from the Brønsted site to the ethanol oxygen occurs when ethanol strongly interacts with the site and that this shift changed the electronic state of ethanol for the dehydration to be favorable. These results indicate that the dehydration of ethanol on oxide catalysts proceeds by the E1 mechanism.
We have developed multimedia type chemistry CAL software by Netscape Navigator browser.We have converted CAL software, made by ExpandBook form, into HTML form, whose theme is for physical chemistry experiments. Multimedia type chemistry CAL material with video movie really helps students to visualize chemistry operations in practical experiments. Anyone can make a teaching document by HTML language easily if one memorizes a procedure. When we tried the intranet use of CAL software in a remote place (separate room) and video movies of CAL, it produced smooth movement and was useful for the learning student.
An artificial neural network simulation was applied to the recognition and reproduction of time series data whose amplitudes and frequencies simultaneously change with time. The model is composed of two neural networks respectively predicting the change of amplitudes and frequencies. The results of our model were compared with those obtained by the least squares method using four kinds of model functions. Our model gives higher quality results than the least squares method especially in the prediction of amplitude change.
We have performed parallel processing on the ab initio crystal orbital calculations of one-dimensional polymers using the TCGMSG message passenger. Cray T932 (32 CPU), IBM RS6000 cluster (4 CPU), and personal computer cluster with Intel Pentium and Pentium-Pro (4 CPU) were used for the computational environment. The two electron integrals and the diagonalization of the different k-points were computed in parallel. We found that the parallel processing on the diagonalisation of the different k-points were totally effective for the energy band calculations. The problem solved here does not require massive matrix calculations and we can obtain the speed of the supercomputer with the personal computer cluster.