Elsevier

Computers & Chemistry

Volume 24, Issue 5, July 2000, Pages 585-594
Computers & Chemistry

Dehydrogenation processes and molecular clusters in mass spectra of organometallic and coordination compounds

https://doi.org/10.1016/S0097-8485(00)00057-7Get rights and content

Abstract

A method for the calculation of components from the complex molecular pattern is proposed. The modelling of molecular ion region in mass spectrum is applied to cases where for detection of the dehydrogenation processes effects such as losses of protons, hydrogen radicals or hydrogen molecules may occur. The parts of (MH) and (M2H) bands are determined as components of the picture observed in the molecular ion range of mass spectrum. Positive results of the modelling show, the hydrogen losses should be considered in resulted spectrum interpretation. The components contributions were computed by the least squares method, in which optimisation is based on Hooke–Jeeves procedure. Such an approach resulted in model fits within 1% precision for the cluster containing five or more peaks. Applications of the method are presented for 2-methyl-selenolo[2,3-b]-pyridine C8H7NSe, ethyl-digermane C2H10Ge2 and methyl-mercuric-dicyandiamide C3H10HgN4.

Introduction

Each mass spectrum consists of three groups of peaks with different importance for interpretation.

  • The fragmentation peaks. The signals are associated with fragmentation ions, essential for monoisotopic methods of spectrum interpretation for organic compounds as well as coordination or organometallic ones.

  • The isotopic peaks. The signals are poorly applicable for interpretation except those contained in molecular cluster1, particularly for organic compounds (Silverstein and Bassler, 1970).

  • Background. These peaks, usually small, are inapplicable for mass spectrum interpretation. Commonly, the first step of investigation depends on background elimination using mathematical extraction (Bieler and Biemann, 1974, Gorączko, 1998).

The first two peak groups always occur together and build the fragmentation ion cluster. Usually the pattern consists of few or several peaks but only 2–3 of the most intensive are applied for interpretation.

A significant part of mass spectrum is the molecular ion region. In electron impact mass spectrometry, the molecular ion originates from the reaction of an electron with a molecule of the investigated compound:M+ē→M++2ē

Molecular ions carry information about the investigated compound, since they contain all the structural fragments and have an element composition characteristic for the compound (Płaziak, 1997). The intensities and locations of peaks belonging to molecular clusters allow the determination of molecular mass as well as molecular formula of a compound. The relations between intensities of the molecular m+ as well as isotopic peaks2 (m+1)+ and (m+2)+ are based on molecular formula investigations in a classical method for organic compounds containing C, H, N, O, F, I, and P (Beynon and Williams, 1964, Ege, 1970). Determination of molecular formulas is more difficult for organic compounds with Cl, Br, Si and S, but previous theoretical calculation of molecular band (Beynon et al., 1968) solved this problem for compounds containing up to ten atoms of S and Si, up to eight atoms of Br and Cl, alone or in combination. Such procedure did simplify mass spectra interpretation of organic compounds with more complex elemental contents. For that reason, the molecular formulas of such compounds can be easily determined from isotopic peak intensities.

However, the presence of multiisotopic elements, typical for many organometallic and coordination compounds, makes the method of Beynon and Williams inapplicable. Since the molecular patterns of these compounds are usually very complex, the method of calculation must be completely different from monoisotopic interpretation.

Knowledge of the compound origin, some physical or chemical properties and results collected from other instrumental methods can be a source of additional information about the investigated compound. These data help to formulate hypothesis about the element formula, which is a starting point for theoretical predictions of molecular cluster (Hsu, 1984, Gorączko and Szymura, 1999) based on natural isotopic abundance (IUPAC, 1991). An agreement of experimental results with calculated ones points to compliance of the theoretical model with experimental.

Sometimes, significant discrepancies between experimental molecular band and predicted cluster may be observed, even for correctly made calculations and precisely performed experiments. One can consider three possible sources, which may be responsible for such a situation.

  • Differences between published natural isotopic abundances (DeLaerter et al., 1991) and those occurring locally. Usually, this error affects only one element, and thus the differences are not significant.

  • Discrepancies arising from a strong background. Sufficient purification of investigated compounds and proper operation of the mass spectrometer should solve this problem completely.

  • Overlapping of molecular cluster and patterns associated with hydrogen loss processes. Since these reactions generate bands located practically within the same range as common molecular patterns, the molecular cluster can be seriously disturbed after mass spectrum normalisation. This problem will be discussed below.

Section snippets

Decomposition of the molecular ion by dehydrogenation processes

Some compounds, e.g. alcohols, amines, phenols, easily lose hydrogen under mass spectrometry conditions. The meaning of the term ‘hydrogen loss’ suggests something well defined, i.e. hydrogen, however, the reaction may also be concerned with the detachment of a radical, a proton or an hydrogen molecule (, ). It appears that the term ‘dehydrogenation’ would be better for such a loss of hydrogen species, because it does not define the form of the eliminated hydrogen. The hydrogen elimination

Modelling of quasi-molecular pattern structure

The experimental data are compared with predictions (Gorączko and Szymura, 1999) based on the molecular formula of the investigated compound and the corresponding natural isotopes abundance (DeLaerter et al., 1991). A resulting molecular cluster contains signals with intensities of Pcalc.i which are located within the range from m1 to mn. The peaks of the experimental molecular cluster M with intensities Pexp.i are positioned within the region from m1 to mk (Scheme A).m1m2m3mk−2mk−1mkPexp.1Pexp

Examples of experimental molecular cluster reconstruction

The application of the procedure is presented for experimental mass spectra of three selected compounds. All of them contain a multiisotopic element in molecule, i.e. selenium, germanium or mercury. Molecular regions of their experimental spectra differ from the molecular clusters calculated theoretically.

Conclusions

The modelling of the molecular cluster region of mass spectra is needed for the interpretation of patterns disturbed by dehydrogenation processes, i.e. loss of a proton, hydrogen radical or hydrogen molecule.

The proposed method applies a simple idea, sometimes realised in the ‘hand made’ mass spectra interpretation and utilises additivity rules commonly used in instrumental analysis. The modelling effects are obtained very rapidly by the least squares method, in which optimisation is based on

Acknowledgements

The authors are grateful to Professor S.E. Wanke from University of Alberta, Canada for suggesting some useful modifications to the manuscript.

References (16)

  • A.J. Gorączko et al.

    Comput. Chem.

    (1999)
  • J.H. Beynon et al.

    Mass and Abundance Tables for Use in Mass Spectrometry

    (1964)
  • J.H. Beynon et al.

    The Mass Spectra of Organic Molecules

    (1968)
  • J.E. Bieler et al.

    Anal. Lett.

    (1974)
  • Croisy, J.C., 1974. Org. Mass Spectrom, 10 970 ; CAS#: 55108-57-1; NIST#:...
  • J.R. DeLaerter et al.

    J. Phys. Chem. Ref. Data

    (1991)
  • G. Ege

    Zahlentafeln zur Massenspektrometrie und Elementaranalyse

    (1970)
  • Freudenthal, J., Gramberg, L.G., 1975. Catalogue of Mass Spectra of Pesticides, National Institute of Public Health,...
There are more references available in the full text version of this article.

Cited by (8)

View all citing articles on Scopus
View full text