Skip to main content
main-content

Über dieses Buch

June 1986 brought together some of the world's leaders in computer­ enhanced analytical spectroscopy at Snowbird, Utah, for what the attendees decided to call "The First Hidden Peak Symposium." With the remarkable advances in both computer hardware and software, it is interesting to observe that, while many computational aspects of spectroscopic analysis have become routine, some of the more fundamental problems remain unsolved. The group that assembled included many of those who started trying to interpret chemical spectroscopy when computers were ponderous, slow, and not very accessible, as well as newcomers who never knew the day that spectrometers were delivered without attached computers. The synergism was excellent. Many new ideas, as well as this volume, resulted from interactions among the participants. The conclusion was that progress would be made on more fundamen­ tal problems now that hardware, software, and mathematics were coming together on a more sophisticated level. The feeling was that the level of sophistication is now adequate and that it is only a matter of time before automated spectral interpretation surpasses all but the most advanced human experts.

Inhaltsverzeichnis

Frontmatter

Optimization and Exploratory Data Analysis

Frontmatter

Chapter 1. Development of an AI-Based Optimization System for Tandem Mass Spectrometry

Abstract
Artificial intelligence (AI) is that branch of computer science that attempts to understand and model intelligent behavior with the aid of computers. In general these attempts to have machines emulate intelligent behavior fall far short of the competence of humans. However, in the area of expert systems, computer programs have been developed that can achieve human performance, and in limited aspects even exceed it.
Carla M. Wong, Hal R. Brand

Chapter 2. Curve Fitting and Fourier Self-Deconvolution for the Quantitative Representation of Complex Spectra

Abstract
Most techniques for multicomponent analysis require spectral features due to each component to be distinguished in some manner. Many of these methods are based on minimizing the standard deviation between the measured spectrum and a synthetic spectrum. The latter spectrum may be synthesized from a linear combination of spectra of mixtures of known composition, as it is in whole spectrum curve-fitting techniques(1,2) or the K-matrix method,(3) or it may be completely synthetic, i.e., generated from the combination of N bands of known shape with different center wave numbers, v̄ i 0 , peak absorbances, A i , and widths, γ i .(4,5) (In this chapter γ will be defined as the full width at half-height.) The greater the dissimilarity of the spectra of each component, the more accurately may least-squares techniques be applied. Conversely, when absorption bands are located so close together that all spectral features are completely unresolved, least-squares methods tend to break down.
Peter R. Griffiths, John A. Pierce, Gao Hongjin

Chapter 3. Evolutionary Factor Analysis in Analytical Spectroscopy

Abstract
Factor analysis (FA) is a computational tool for solving multidimensional problems which has found increasing importance in analytical spectroscopy.(1) Abstract factor analysis (AFA) reveals the number of spectroscopically visible components in a series of related mixtures containing the same components but in varying proportions.(1–3) Target factor analysis (TFA) verifies the presence or absence of suspected components.(1,4,5) Rank annihilation factor analysis (RAFA)(6,7) allows quantification of a particular component without recourse to any knowledge of the other components.
Edmund R. Malinowski

Chapter 4. Numerical Extraction of Components from Mixture Spectra by Multivariate Data Analysis

Abstract
The authors know by experience that, if presented in purely mathematical terms, the multivariate data analysis techniques which will be discussed in this chapter can be difficult to understand for nonmathematicians. Therefore we will explain techniques such as factor and discriminant analysis, as well as subsequent procedures aimed at extracting chemical information, primarily in geometrical terms. After the basic principles have been visualized, these principles will then be rationalized mathematically so that a link can be established with the more mathematically oriented literature. The paragraphs containing the mathematical rationalizations can be skipped, if desired.
Willem Windig, Henk L. C. Meuzelaar

Chapter 5. Simultaneous Multivariate Analysis of Multiple Data Matrices

Abstract
The application of pattern recognition methods to two-way tables of data emanating from mass spectrometers, chromatographs, and a variety of other complex analytical instruments is now widespread. In particular principal components analysis (Pearson,(1) but see Gnanadesikan(2) for a more recent introduction) is used to obtain low-dimensional representations of the relationships between rows and columns of two-way tables. In pyrolysis mass spectrometry this technique, usually called “factor analysis,” has been particularly powerful.(3,4)
Halliday J. H. MacFie

Chapter 6. Multivariate Calibration: Quantification of Harmonies and Disharmonies in Analytical Data

Abstract
The simplest string instruments have only one string. It is possible to play nicely on such primitive instruments, because our memory can recognize a series of successive sounds in terms of melody and rhythm.
Tormod Næs, Harald Martens

Spectral Interpretation and Library Search

Frontmatter

Chapter 7. Automated Spectra Interpretation and Library Search Systems

Abstract
Identification and structure elucidation of organic compounds is today mostly done with spectroscopic methods; the most popular methods include infrared spectroscopy (IR), mass spectrometry (MS), nuclear magnetic resonance spectroscopy (NMR), and spectroscopy in the ultraviolet and visible region of the electromagnetic spectrum (UV/vis).
J. T. Clerc

Chapter 8. Carbon-13 Nuclear Magnetic Resonance Spectrum Simulation

Abstract
Carbon-13 nuclear magnetic resonance (CNMR) spectroscopy is a powerful tool for organic structure elucidation because the signals observed are directly related to the immediate surroundings of the skeletal carbon atoms. Therefore, a great deal of information directly relevant to the skeletal arrangement of the structure is accessible. Modern NMR spectrometers generate huge quantities of data rapidly, increasing the demand for tools to aid the spectroscopist in the analysis of NMR data.
Peter C. Jurs, Debra S. Egolf

Chapter 9. The Evolution of an Automated IR Spectra Interpretation System

Abstract
Commercial spectral identification and search systems are intended to be used by persons who are not expert spectroscopists. Most pure search systems focus directly on the best matches to spectra that are contained in the library. However, an expert narrows the scope of a search by first identifying patterns in the spectrum assigned to its functional groups or structural units. Then, when searching the library for matching spectra, attention is focused on compounds containing these structural units. The effect of the structural unit prefilter on the library search is to eliminate materials that coincidently match the peaks of the unknown but are chemically unrelated. A significant factor in this search mode is the guidance the searcher receives when the spectrum is not contained in the search library. The evolution of a spectrum recognition system that follows these procedures is described as one of the earliest examples of a commercially successful expert system.
Abraham Savitzky

Chapter 10. Novel Advances in Pattern Recognition and Knowledge-Based Methods in Infrared Spectroscopy

Abstract
Theoreticians can calculate exactly the frequencies at which infrared radiation will be absorbed by a given molecule, assuming the spatial arrangements of the atoms and the strengths of the bonds are known. This assumption is realistic for very small or highly symmetrical larger molecules. For those types of molecules, excellent theoretical treatments have been made. However, the vast majority of molecules have vibrational characteristics and interatomic interactions that are too complex for adequate theoretical treatment, so empirical treatment of the data becomes necessary.
Hugh B. Woodruff

Chapter 11. Library Storage and Retrieval Methods in infrared Spectroscopy

Abstract
The use of infrared (IR) spectroscopy as the method of choice for qualitative organic analysis is well established, from the days of the Infracord through to the Fourier transform (FT) instruments of today. IR predates most other forms of molecular spectroscopy as a useful tool for the analytical organic chemist. The unique fingerprinting and identification ability provided by an IR spectrum result from the fact that the peaks in the spectrum correspond to vibrational modes that are characteristic of the complete molecule and to other modes that are directly related to specific functional groups in the molecule. This combination of group frequencies and the well-known “fingerprint” region (1400–400 cm-1) in IR spectra has made comparison of an unknown spectrum to a standard spectrum from a reference material or a reference library a commonly accepted method for compound confirmation.
Stephen R. Heller, Stephen R. Lowry

Chapter 12. Synergistic Use of Infrared, 13C Nuclear Magnetic Resonance, and Mass Spectral Data in Analysis Schemes for the Identification of Organic Mixture Components

Abstract
The implementation of analysis schemes that incorporate spectrometric information from multiple sources has been demonstrated to enhance the reliability of identification procedures for unknown compounds. Most notably, infrared spectrometry (IR) has been utilized in combined(1,2) and independent(3) measurements with mass spectrometry (MS) to minimize false positive identifications endemic to single-source detection schemes. This chapter will be devoted to a survey of computer-assisted algorithms generated in our laboratory which exploit the synergistic potential of multiple detectors for the analysis of unknown organic mixtures. The triad of analytical techniques that are generally recognized to yield the greatest source of complementary information for organic structure elucidation-mass spectrometry, infrared spectrometry, and 13C nuclear magnetic resonance spectrometry (NMR)-provide the data for the various algorithms that are constructed.
David A. Laude, Charles L. Wilkins

Backmatter

Weitere Informationen