Bo Nordén

Four levels of pattern recognition

Abstract Problems of pattern recognition in chemistry and other subjects can be divided conveniently into four different types depending on the level of scope of the problem. (1) Classification into one of a number of defined classes. As an example blood samples taken from persons known to be either controls or welders are considered. The problem is whether trace element concentrations in these samples contain information on whether or not a person is a welder. (2) Level 1 plus the possibility that an object is an outlier, i.e. does not belong to any of the defined classes. As an example, the use of 13C-n.m.r. data to decide whether 2-substituted norbornanes have the exo or endo structure is discussed. (2A) Level 2, asymmetric. This situation occurs when one class does not have a systematic structure, but another class is homogeneous and can be described by a level 2 model. This occurs in the classification of materials or compounds as good or bad, active or inactive, and in binary classifications. As an example the use of trace element data to classify steel samples as having good or poor properties of strength is discussed. (3) Level 2 plus the ability to relate the variables measured to external properties of continuous character. As an example, the classification of a series of chemical compounds as β -receptor blockers, β -receptor stimulants, or neither, on the basis of their structural variables is discussed. In addition, relations between these structural variables and the measured biological activity are sought within each of the two classes. (4) Level 3 with the difference that several external property variables in the objects are measured. It may be desirable to use variables of the objects both for classification and for relations to several property variables: such examples are numerous in analytical chemistry.

Characterization of Peat Samples by Diffuse Reflectance FT-IR Spectroscopy

A Fourier transform infrared spectroscopic study of Sphagnum peat, Carex peat, and mixtures of these peat types is reported. Intensity values from selected wavenumber regions of the infrared spectra were correlated with calorific value, degree of decomposition, carbon, and nitrogen content, and content of amino acids and amino sugars, with the use of the method of partial least-squares modelling with latent variables (PLS). The application of this method is shown to give valuable information about properties of peat which have no easily measured quantitative relationship to the infrared spectrum.

Multivariate Data Analysis of In Situ Pulp Kinetics Using 13 C CP/MAS NMR

Abstract 13C CP/MAS NMR spectra of birch pulp samples were measured at regular intervals during the kraft pulping process. General multivariate data analysis methods based on principal components were used to extract the spectral information content and to relate these components to chemical descriptors such as lignin content. It was shown that this approach has an improved predictive ability relative to traditional methods. Model parameters could also be used to construct subspectra with independent information.

Principal components analysis of complex n.m.r. spectra from heterogeneous material

Abstract Principal components (PC) analysis is applied to raw data profiles of solid state 13 C CP/MAS n.m.r. spectra of heterogeneous organic material. Information about various peat classes is obtained from component scores plots, where similar types of samples ( Carex or Sphagnum peat) group together. Substructures in these groups that correspond to directions of increased decomposition are found. Component loadings are used to estimate spectra from peat constituents e.g. bitumen/humic substances and carbohydrates are uncovered in the first two loadings. This may be a good method to use in combination with other techniques, such as deconvolution and synthesis of spectra.

Spectroscopic investigations of peat-water interactions: an ESR, FT-IR, and NMR study.

Despite the existence of naturally occurring paramagnetic species in peat, it is difficult to use them unambiguously for determination of the water-binding properties of peat. In the present investigation we have therefore used extrinsic probe molecules, viz stable nitroxide radicals. By means of these reporter molecules, it is possible to follow the rate of disappearance of water in peat at various temperatures. Samples from different sorts of peat (Sphagnum and Carex) with various degrees of humification have been investigated. The binding between water and peat have been estimated. The DRIFT technique was used to investigate the interaction between water and homogenized batches of Sphagnum and Carex peat of both high and low degrees of decomposition. As an aid to spectral interpretations, the hydrogen bonding (3700 to 3000 cm-1) region of the spectra was analyzed using principal component analysis and partial least squares modeling with latent variables. A very good correlation is obtained between DRIFT data and the water content of the peat samples. Water molecules are more strongly bound to the surface of a high-humified peat sample than to a surface of low-humified peat. The hydrogen bonding interaction between water and functional groups in the peat structure is also reflected by the vibrational spectrum of peat. We used 13C CP/MAS nuclear magnetic resonance to monitor changes in chemical composition, revealing a decrease of carbohydrates and a relative increase in aromatic and aliphatic structures as humification proceeds. 1H spin-spin relaxation measurements are used to detect different fractions of water in peat. The data indicate the existence of a multitude of fractions, which can be interpreted as hard bound water, capillary water, and bulk water.

CHARACTERIZATION AND CLASSIFICATION BASED ON MULTIVARIATE DATA ANALYSIS

Abstract The characterization of chemical or biological samples by multidimensional data followed by an appropriate data analysis provides a means for the quantitative classification of new samples. Data analytic problems and ways to solve them are discussed using two examples. The first concerns the classification of micro-organisms (fungi) on the basis of pyrolysis-gaschromatography data. The second concerns the determination of the chemical structure of organic compounds on the basis of C-13 NMR data.

Spin and electron distributions in heme-cyanide models and hemeproteins

Proton NMR spectra of low-spin Fe(III) cyanoprotoheme as prosthetic group in a number of proteins are presented. The diagonally positioned 1-, 5- and 3-, 8-methyl groups obey shifts proportional to the Fe(III)/(II) reduction potential Em7, which indicates a pseudo-contact interaction. The correlation with Em7 is understandable if one postulates an enhanced rhombic distortion, dominating the Fe-methyl dipolar interactions. Hartree-Fock-Slater quantum chemical calculations show no significant changes of spin density as a function of the Fe-L5 distance, except at the iron atom and predominantly in the 3dxz and 3dyz orbitals. 4p orbitals, on the other hand, uphold most of the changes of electron density. We also observe a principal difference in the amino acid sequences in the heme-accommodating pocket of oxygen carriers and two-electron transmitters.

A 13C-NMR study of mutant hemoglobins with altered oxygen affinity

The 13CO‐NMR spectra of carbonylhemoglobins Saint Mandé (β 102Asn → Tyr), Malmö (β 97His → Gln), Hôtel Dieu (β 99Asp → Gly) and Ao have been determined. The positions of the 13CO resonances for hemoglobins Ao, Malmö and Hôtel Dieu were similar indicating similar ligand environments for all. The 13CO resonance for the β‐subunit of Saint Mandé was upfield‐shifted compared to the others. This is evidence that structural changes at the β 102 position directly affect iron‐ligand bonding as well as quaternary structure.

SPECTROSCOPIC PEAT CLASSIFICATION AND CALIBRATION USING ELECTRON SPIN RESONANCE AND MULTIVARIATE DATA ANALYSIS

Electron Spin Resonance (ESR) measurements were performed on particle size fractions of high and low humified Sphagnum fuscum and Carex rostrata. Each peat sample was divided into seven particle size groups, where the size varies between >2.0 and <0.045 mm. Each peat sample was also subjected to traditional chemical analysis, where the amounts of some chemical constituents have been determined: ash, carbon, hydrogen, nitrogen, sulphur, rhamnose, fucose, arabinose, xylose, mannose, galactose, glucose, and Klason lignin (including the bitumen fraction). Two respiration rate data, the carbon dioxide emission after 2 days and 25 days, have also been included. In order to handle all data in a rational manner, multivariate data analysis was used. According to principal component analysis of the ESR data, each peat type forms a well defined class, which implies that a calibration model has to be created for each peat class. The main differences between the different peat classes were the amount of stable organic radicals, but two peat classes could only be separated based on differences in type of organic radicals. The partial least squares modeling, i.e., modeling of the correlation between the ESR measurements and the chemical constituents and the respiration data, works well for the low humified Carex peat type class and the low humified Sphagnum peat, slightly worse with the high humified Carex peat and the high humified Sphagnum peat types. Many of the dependent variables were well modeled, and unknown test samples were correctly predicted