Tadao Tamura

Neural network system for the identification of infrared spectra

A neural network system has been developed on a personal computer to identify 1129 infrared spectra. The system is composed of two steps of networks. The first step classifies 1129 spectra into 40 categories, and each unit of the output layer is connected to one of the 40 networks in the second step, which identify each spectrum. Each network is composed of three layers. The input, intermediate, and output layers are composed of 250, 40, and 40 units, respectively. Intensity data at 250 wavenumber points between 1800 and 550 cm−1 of the infrared spectra are entered into the input layer of each network. The training of the networks was carried out with the spectral data of 1129 compounds stored in the SDBS system, and thus the networks were successfully constructed. On the basis of the results, the system has been developed by preparing pre- and post-processing programs. The system can identify each unknown spectrum within 0.1 s, and is quite efficient for identifying infrared spectra on a personal computer.

Identification of Chemical Structures from Infrared Spectra by Using Neural Networks

Structure identification of chemical substances from infrared spectra can be done with various approaches: a theoretical method using quantum chemistry calculations, an inductive method using standard spectral databases of known chemical substances, and an empirical method using rules between spectra and structures. For various reasons, it is difficult to definitively identify structures with these methods. The relationship between structures and infrared spectra is complicated and nonlinear, and for problems with such nonlinear relationships, neural networks are the most powerful tools. In this study, we have evaluated the performance of a neural network system that mimics the methods used by specialists to identify chemical structures from infrared spectra. Neural networks for identifying over 100 functional groups have been trained by using over 10 000 infrared spectral data compiled in the integrated spectral database system (SDBS) constructed in our laboratory. Network structures and training methods have been optimized for a wide range of conditions. It has been demonstrated that with neural networks, various types of functional groups can be identified, but only with an average accuracy of about 80%. The reason that 100% identification accuracy has not been achieved is discussed.

Infrared Qualitative Analysis Automated by Use of an Electronic Computer and a Spectral Data File with Band Intensity Information

An infrared date file with intensity information has been prepared, based on the first 5000 spectra of the data cards of the Infrared Data Committee of Japan (IRDC). A computer program has been developed for the qualitative analysis of binary mixtures using this data file. The program has been applied to 11 binary mixtures; for ten of them it has provided correct answers among the first three compounds outputted by the computer. The computer time necessary for the analysis of a sample mixture was approximately 1 min in TSS mode, excluding the time for the input and output of data. The one unsuccessful case has been proved to be due to the unusual character of the infrared spectrum of one of the components (C2Cl6). Hence this failure does not indicate any serious defect in the program; the developed program has proved to be of practical use.

NEURAL NETWORK SYSTEM FOR THE IDENTIFICATION OF MATERIALS FROM INFRARED SPECTRA

A neural network system has been developed on a personal computer to identify 1129 infrared spectra of chemical substances. The system is composed of two steps of networks. The first step classifies 1129 spectra into 40 categories, and each unit of the output layer is connected to one of the 40 networks in the second step, which identify each spectrum. Each network is composed of three layers. The input, intermediate, and output layers are composed of 250, 40 and 40 units, respectively. Intensity data at 250 wavenumber points between 1800 and 550 cm −1 of infrared spectra are entered into the input layer of each network. The training of the networks was carried out with spectral data of 1129 compounds stored in the SDBS system, and thus the networks were successfully constructed. On the basis of the results, the system has been developed by preparing pre- and post-processing programs. The system can identify each unknown compounds from infrared spectra within 0.1 s, and it is quite efficient for identifying compounds on a personal computer.

Estimation of substitution form of benzene ring from IR spectra by pattern recognition.

A pattern recognition technique for estimating four types of substituted benzene from IR spectra has been studied using the spectral data of 675 compounds. The accuracy of estimation was evaluated by using a linear learning machine method. It was found that, when there are 250 data points in a computer memory, the wavenumber range of 1650650 cm -1 and the wavenumber interval of 4 cm-1 are sufficient, and that the four types of benzene can be estimated with above 90 % accuracy.

Computer elucidation of functional groups from infrared spectra.

スペクトルデータベースシステムSDBSに集積されている約9500種の化合物の赤外吸収スペクトルのデータを用いて,スペクトルから種々の官能基の有無を推定する方法について検討した.それぞれの官能基を含むグループと含まないグループについてコンピューターを用いて平均スペクトルを算出し,その差スペクトルを識別関数とした.その関数と検体のスペクトルとの間の相関係数を算出し,その値の大小によって官能基の有無を推定する方法を採用した.その結果,C=OやNO2などでは90%近い確度で推定可能であるが,C=CやC≡Nなどでは推定の誤差がかなり高くなることが分かった.

Determination of equilibrium constant between nitrogen dioxide and dinitrogen tetraoxide by means of infrared absorption spectroscopy

赤外用気体セルを空気恒温そう中に入れて種々の温度に保ち,その中に二酸化窒素を種々の圧力に満たし,四酸化窒素との平衡のもとに赤外線吸収スペクトルを測定した.二酸化窒素の1617cm-1の吸収帯と四酸化窒素の1735cm-1または1255cm-1の吸収帯とをkcy bandに用い,その面積強度から各分子種の分圧を算出した.K=P2NO2/PN2O4(PNO2,PN2O4はそれぞれ二酸化窒素と四酸化窒素との分圧)という式に従って,種々の温度での平衡定数を求めた.このようにして得られた平衡定数の値にはかなりのバラツキがみられたが,その温度変化から2NO2→N2O4反応のΔHとして約17.5kcal/molという値を得た.