Spectral Databases—Identifying Unknowns

Variations on the spectral peaks from different species of the same genus were also observed. Three species of Pseudomonas produced the spectra shown in Figure 14.2. These spectra are clearly unique and were used to correctly identify unknown samples. Because of peak ratio reproducibility issues in bacterial protein profiles obtained by MALDI MS,11 a fingerprint approach that had been used for other mass spectrometry approaches has not been used. The profile reproducibility problem was first recognized by Reilly et al.12,13 and later researched by others in the field.14,15 As a later alternative, a direct comparison of the mass-to-charge ratio (m/z) of the unknown mass spectral peaks with a database of known protein masses has been used to identify unknown samples.14... 

The confidence level in the identificadon capabiUty for gc/ir/ms is enhanced by the ability to perform computer library searching of large spectral databases. Unknown spectra are searched against reference databases and a hit quality number, indicating how well the unknown spectrum matches the library spectra, is generated. For the very small peak at 19 min in Figure 4, peak 6 in the TIC, the library search identified the component as a-phellandrene [99-83-2]y for both the ic and ms data. Because these data are complementary and generated from two completely independent principles of... 

NMR spectra can be used to identify unknown compounds through spectral pattern matching. A number of companies, instrument manufacturers, government agencies, and other sources pubhsh collections of reference spectra in electronic format and in hardcopy. These spectral databases may contain spectra of more than 200,000 compounds. The... 

There are two ways to interpret such spectra. The first is to compare the spectrum you have with those in a searchable computerized mass spectral database. The second is to evaluate the spectrum using the interpretation procedure described subsequently. In either case, once an unknown compound has been identified from its mass spectrum, the pure compound should be obtained and analyzed under the same conditions as the sample for conhrmation. Over 10 milhon chemical compounds have been identified. No mass spectral database contains spectra for every possible compound, although mass spectral databases of over 400,000 spectra are available. The mass spectral database from the US National Institute for Standards and Technology (NIST) may be purchased from a number of licensed vendors. Limited mass spectra from NIST are available online in the NIST Chemistry WebBook (http //webbook.nist.gov). Commercial vendors and publishers offer specialized mass spectral hbraries of compounds, such as environmental compounds, pharmaceuticals, natural products, oil industry compounds, and the like. 

In this article we have introduced the Bruker SGF Profiling method for the authentication, verification and quality control of fiuit juices. In addition to tihe quantification of a large array of characteristic compounds, this fully automated NMR screening technique uses statistical models for the estimation of fhiit content or the origin of the juice. This analysis tool can show known and unknown deviations from normality. Currently, routines are under development to identify unknown deviations by constructing spectral patterns which can be compared to an existing reference compound database. ... 

Nuclear Magnetic Resonance Spectroscopy (NMR) for the verification of the Chemical Weapons Convention (CWC) has the same basic aim as conventional NMR Unknown substances should be identified by recording spectra with high resolution and sensitivity and subsequent comparison to a spectral database or the expert s experience. 

Searching for a mass spectrum of an unknown substance in a mass spectral database often very quickly identifies the unknown, provided the corresponding spectrum is already available in the database. However, even if this is not case, the search might deliver similar spectra of closely related compounds, thus facilitating structure elucidation [81]. 

Although the coverage of these databases is enormous, and an easy-to-use interface is provided, one should be aware of potential pitfalls. The high score of a hit can be wrong, simply because the spectrum of the actual compound is still not included in database. A spectrum may be flawed by superimposition with some impurity peaks, which can in turn mislead the search. As a result, the unknown can generally not be perfectly identified by simple conparison with a library spectrum [83]. Nonetheless, mass spectral databases are highly useful, because even under unfavorable circumstances, they deliver spectra of similar compounds or isomers... 

Various infrared spectral databases or libraries are available, which contain collections of the infrared spectra of a number of specific chemical species. Spectral search or data retrieval is a technique enabling one to identify a material of unknown origin by comparing its spectrum with library spectra, or to make a guess at the chemical structure of the unknown material from the similarity of its spectmm to some library spectra. 

These are very simple examples, but provide a plan of attack for the problems at the end of the chapter. It is highly unlikely that an analyst can identify a complete unknown by its IR spectrum alone (especially without the help of a spectral library database and computerized search). For most molecules, not only the molecular weight, but also the elemental composition (empirical formula) from combustion analysis and other classical analysis methods, the mass spectrum, proton and C NMR spectra, possibly heteroatom NMR spectra (P, Si, and F), the UV spectrum, and other pieces of information may be required for identification. From this data and calculations such as the unsaturation index, likely possible structures can be worked out. 

Access to a database of chemical structures and associated spectral parameters can accelerate the process of identifying an unknown. This is especially true when the data are utiHzed to produce NMR prediction algorithms and these are now available for a number of nuclei, most commonly and but also including P and F. Electronic content databases of N data are available from a number of sources [26—28], and the largest and most up-to-date source of data is associated with the ACD/NNMR predictor program, presently used in the laboratory of one of the authors (G. E. M.). 

---