Molecular structure from mass spectra

Intensive applications of pattern recognition methods in chemistry were started with pioneering works about spectral interpretations by Isenhour, Jurs, and Kowalski (1969 - 1971). Numerous papers deal with the automatic prediction of molecular structures from mass spectra, infrared spectra and nuclear magnetic resonance spectra (Chapter 13). Predictive abilities of 80 to 95 % are typical of these applications. 

The identification of a molecular structure from a mass spectrum requires good chemical detective work. Let s see how that is done by trying to identify a simple compound,... 

Mass spectrometry (MS) provides the molecular weight and valuable information about the molecular formula, using a very small sample. High-resolution mass spectrometry (HRMS) can provide an accurate molecular formula. The mass spectrum also provides structural information that can confirm a structure derived from NMR and IR spectroscopy. 

We note that the earliest computational efforts in this field were the identification of molecular structures from spectra, and they followed the exhaustive, deterministic mode (see Spectroscopic Databases). In the DENDRAL project molecular features were determined from spectroscopic data and then assembled into molecular structures consistent with the observed mass spectra. This problem is quite analogous to chemical compound design, with the spectrum as the target property set. 

Multivariate data analysis usually starts with generating a set of spectra and the corresponding chemical structures as a result of a spectrum similarity search in a spectrum database. The peak data are transformed into a set of spectral features and the chemical structures are encoded into molecular descriptors [80]. A spectral feature is a property that can be automatically computed from a mass spectrum. Typical spectral features are the peak intensity at a particular mass/charge value, or logarithmic intensity ratios. The goal of transformation of peak data into spectral features is to obtain descriptors of spectral properties that are more suitable than the original peak list data. 

Typical MS/MS configuration. Ions produced from a source (e.g., dynamic FAB) are analyzed by MS(1). Molecular ions (M or [M + H]+ or [M - H]", etc.) are selected in MS(1) and passed through a collision cell (CC), where they are activated by collision with a neutral gas. The activation causes some of the molecular ions to break up, and the resulting fragment ions provide evidence of the original molecular structure. The spectrum of fragment ions is mass analyzed in the second mass spectrometer, MS(2). 

By measuring a mass spectrum of normal ions and then finding the links between ions from the metastable ions, it becomes easier to deduce the molecular structure of the substance that was ionized originally. 

Fullerenes are described in detail in Chapter 2 and therefore only a brief outline of their structure is presented here to provide a comparison with the other forms of carbon. The C o molecule, Buckminsterfullerene, was discovered in the mass spectrum of laser-ablated graphite in 1985 [37] and crystals of C o were fust isolated from soot formed from graphite arc electrodes in 1990 [38]. Although these events are relatively recent, the C o molecule has become one of the most widely-recognised molecular structures in science and in 1996 the codiscoverers Curl, Kroto and Smalley were awarded the Nobel prize for chemistry. Part of the appeal of this molecule lies in its beautiful icosahedral symmetry - a truncated icosahedron, or a molecular soccer ball, Fig. 4A. 

The mass spectrum of 2-methylbenzaldehyde suggests an aromatic compound because of the intensity of the molecular ion and peaks at m/z 39, 51, and 65 (see Figure 6.2). The loss of hydrogen atoms and loss of 29 Daltons from the molecular ion indicate that this is an aromatic aldehyde. Looking up m/z 91 in Part III suggests the following structure ... 

MS-MS is a term that covers a number of techniques in which two stages of mass spectrometry are used to investigate the relationship between ions found in a mass spectrum. In particular, the product-ion scan is used to derive structural information from a molecular ion generated by a soft ionization technique such as electrospray and, as such, is an alternative to CVF. The advantage of the product-ion scan over CVF is that it allows a specific ion to be selected and its fragmentation to be studied in isolation, while CVF bring about the fragmentation of all species in the ion source and this may hinder interpretation of the data obtained. 

The MS-MS spectrum of the (M + H)+ ion from the parent drug contains an ion at m/z 465, the structure of which is indicated in Figure 5.40. The mass spectrum of metabolite 1 indicates that it has a molecular weight of 482 Da, while the MS-MS spectrum from its MH+ ion contains both an ion at m/z 466 and aXm/z 364, also present in that from the MH+ of the parent drug. It is not unreasonable, although not necessarily always correct, to assume that the ion of... 

The NMR spectrum of 4 showed signals at 7-9 (m,9H,ArH) and 10.5 (s,lH,NH). Data from the elemental analyses have been found to be in conformity with the assigned structures. Furthermore, the molecular ion recorder in the mass spectrum is also in agreement with the molecular weight of the compound. 

A new amorphous alkaloid has been recently isolated from the Chinese plant T. bufalina (Ervatamia hainanensis) collected on Hainan Island (53). Its mass spectrum showed a molecular ion at m/z 382, corresponding to C23H30N2O3. From the fragmentation pattern, this compound would appear to be a coronaridine derivative in which a C2H50 unit is attached to the aliphatic moiety of the molecule. The structure 111 with (S) configuration at C-3 was determined by a detailed analysis of its H-NMR spectrum (Table IV) in comparison with the data of other ibogan alkaloids. 

---