Library data peak searches

Modem pulse height analysers essentially contain dedicated digital computers which store and process data, as well as control the display and operation of the instrument. The computer will usually provide spectrum smoothing, peak search, peak identification, and peak integration routines. Peak identification may be made by reference to a spectrum library and radionuclide listing. Figure 10.15 summarizes such a pulse height analysis system. 

If the laboratory worker does not know of a reference to the preparation of a commercially available substance, he may be able to make a reasonable guess at the synthetic method used from published laboratory syntheses. This information, in turn, can simplify the necessary purification steps by suggesting probable contaminants. However, for other than macromolecules it is important that at least the NMR and IR spectra of the substance be measured. These measurements require no more than two to three milligrams (which are recoverable) of material and provides a considerable amount of information about the substance. Three volumes on the NMR spectra [C.J.Pouchert and J.Behnke, The Aldrich Library of C and FT-NMR Spectra, Vols 1—3, Aldrich Chemical Co., Inc, Milwaukee, Wl, 1993], and one on the infrared spectra [C.J.Pouchert, The Aldrich Library of FT-IR Spectra, 3nd ed, Aldrich Chemical Co., Milwaukee, Wl, 7959], as well as computer software [FT-IR Peak-search Data Base and Software, for Apple HE, IIC and II Plus computers and for IBM PC computers, Nicholet Instruments, Madison, Wl, 1984] contain data for all the compounds in the Aldrich catalogue and are extremely useful for identifying compounds and impurities. If the material appears to have several impurities these spectra should be followed by examination of their chromatographic properties and spot tests. Purification methods can then be devised to remove these impurities, and a monitoring method will have already been established. 

The third of the three adjacent peaks (29.207 minutes) is readily identified as 2-hydroxybenzaldehyde (salicylaldehyde) by comparison of the MS or IR spectrum with library data. It should be noted that the automated search routine in the MS software picked 3-and 4-hydroxybenzaldehyde as better matches than the 2-hydroxy compound, even though the 76 ion in the spectrum of the unknown is present only in the mass spectrum of the 2-hydroxy compound. However, the IR search routine correctly identified the 2-isomer. This illustrates that casual operators who rely on automated search routines for compound identification are much less likely to make errors when they have access to both IR and MS searching. 

The reproducibility and reliability of RIs makes it possible to create RI libraries and the identification can be achieved without authentic reference chemicals. The reliability and simplicity of RI monitoring is increased significantly by using a computer program that searches for the RI pattern, calculates the RIs for all peaks in the chromatogram, and then compares the indices with the library data. In addition to the identification of target chemicals, RIs can also be used to locate the interesting peaks between different kinds of GC-based analytical techniques (65). In this way, it is possible to ensure that all GC-based techniques used for identification focus on the same peaks even in samples with a complex mixture of chemicals. 

The forward search is the most rapid but demands unknown spectra of pure compounds to produce good results. If however the unknown spectrum includes peaks from unwanted impurities or of a mixture then it will not work correctly. The alternative reverse search strategy although slower is now required, whereby the resulting hit hst shows the best reference spectrum in the library as found in the unknown data. Peaks in the unknown which do not appear in the Hbrary... 

The position can be determined either from information derived from the peak location or by taking account of library data which indicate which peaks to expect within the particular multiplet. Both have their advantages and disadvantages. Obviously, unless a gamma-ray is in a library then it will not be taken account of and so a simple library directed approach cannot cope with the unexpected. On the other hand, small peaks within a multiplet and very close multiples may not be resolved by the peak search and incorrect peak areas may again result. 

Once the peaks have been collected and stored, the computer can be asked to work on the data to produce a mass spectrum and print it out, or it can be asked to carry out other operations such as library searching, producing a mass chromatogram, and making an accurate mass measurement on each peak. Many other examples of the use of computers to process mass data are presented in other chapters of this book. 

Most mass spectrometers for analytical work have access to a large library of mass spectra of known compounds. These libraries are in a form that can be read immediately by a computer viz., the data corresponding to each spectrum have been compressed into digital form and stored permanently in memory. Each spectrum is stored as a list of m/z values for all peaks that are at least 5% of the height of the largest peak. To speed the search process, a much shorter version of the spectrum is normally examined (e.g., only one peak in every fourteen mass units). 

Of course one may employ automated library searches ( library percent reports ) to check for compound identities, but algorithms for library matching are not infallible, and mass spectral libraries are not exhaustive, thus some compounds of interest will likely not be identified. Additional dilemmas are presented by mere reliance on retention times and library percent reports to ascertain the presence of common or unique peaks from among multiple mass spectral data files. As illustrated in Table 2.1, the TICs from the GC-MS of urine from four elephants evidence a peak at essentially the same retention time, but the library search results are inconclusive as to their common identity or lack thereof. As will be seen below, our novel macros can assist in making such decisions for a large number of peaks. 

As an example of the application of gas chromatography-mass spectrometry, Fig. 1.7 shows a reconstructed chromatograph obtained for an industrial sludge. The Finnigan MAT 1020 instrument was used in this work. Of the 27 compounds searched for, 15 were found. These data were automatically quantified. This portion of the report contains the date and time at which the run was made, the sample description, who submitted the sample and the analyst, followed by the names of the compounds. If no match for a library entry was found, the component was listed as not found . Also shown is the method of quantification and the area of the peak (height could also have been chosen). 

CH2N2 with 100% 3deld. Those samples were dissolved in CH2CI2 in concentrations 1.4mg/mL and injected for GC-FID and GC-MS analysis. The identification of methylated betulinic acid in extracts was done with use Wiley and NBS peak matching library search system. Authentic standard of the betuhnic acid and data reported in the literature were also used for further identification as described. 

Mass spectrometers provide computer output as bar graphs (Fig. 2.1) and as tabular data. Minor peaks, many of them resulting from possible impurities, occur at almost every mass unit. The minor peaks are frequently deleted in the bar graph (those < 0.5% have been omitted in Fig. 2.1). A search of the computer s library and a fit to these peaks may either identify the compound or suggest near structures. Peak heights are proportional to the number of ions of each mass. 

A computer file of about 19,000 peak wavenumbers and intensities, along with search software, is distributed by the Infrared Data Committee of Japan (IRDC). Donated spectra, which are evaluated by the Coblentz Society in collaboration with the Joint Committee on Atomic and Molecular Physical Data (JCAMP), are digitized and made available (64). Almost 25,000 ir spectra are available on the SDBS system developed by the NCLI as described. A project was initiated at the University of California, Riverside, in 1986 for the construction of a database of digitized ffir spectra. The team involved also developed algorithms for spectra evaluation (75). Other sources of spectral libraries include Sprouse Scientific, Aston Scientific, and the American Society for Testing and Materials (ASTM). 

These data for successive scans are then stored for subsequent manipulation. The reconstructed total ion current trace, equivalent to that obtained from a flame ionisation detector in gas chromatography, shows the variation of total ion current with time and allows spectra of interest to be identified. A typical example is shown in Fig. 6A. The background may be subtracted to give clean spectra, and their identification may be attempted using libraries of standard spectra. If a composite spectrum is obtained from two unresolved peaks, complex subtraction routines may be used to obtain a pure spectrum of each of the components. These may be separately submitted for library searching. The spectra may then be plotted or obtained as a mass versus intensity listing. 

---