Structural identification by differential mass spectrometry as a criterion for selecting the best quantum chemical calculation of formation enthalpy for tetrachlorinated biphenyls

Abstract

RATIONALE The assignment of correct structures for isomers with similar mass spectra (e.g. polyhalogenated aromatic compounds) is not always successful when spectral libraries alone are employed or, even worse, when the compounds are not present in commercial spectral libraries. METHODS We present a computational method based on differential mass spectrometry (Diff‐MS) for the validation of formation enthalpy (ΔfH) series calculated using quantum chemistry for the fragments produced in electron ionization (EI)‐MS. The method simulates the chemical structure identification (CSI) of isomers with similar mass spectra using differential mass spectra and ΔfH series. The best ΔfH values were those from which the correct structures could be derived. RESULTS We have used six tetrachlorinated biphenyl isomers (TeCBs 44, 46, 52, 66, 74, 77). Their EI mass spectra were acquired at 70 eV and, for the principal ions, five series of ΔfH values were computed by the semi‐empirical methods, AM1, MINDO3, MNDO, PM3, and RM1. The generation of differential mass spectra and the correlation with the ΔfH series for the calculation of probabilities from the list of structural assignments were carried out with the ordering algorithm (ORD) of the CSI‐Diff‐MS Data Analysis 3.1.1 program. CONCLUSIONS Intelligent software, used for structural elucidation based on MS and QCC, was employed to select the best values of the formation enthalpies of TeCBs. The advantages and disadvantages of the semi‐empirical methods for the calculation of ΔfH values for different TeCB ions are critically presented. The best semi‐empirical methods were RM1, AM1 and MINDO3, which can be used to calculate the ΔfH database necessary to identify TeCB isomers. This approach allowed the correct assignment of structures for isomers with very similar mass spectra and demonstrated the reliability of the correlation between differential mass spectra and the formation enthalpies of the fragment ions. Copyright © 2012 John Wiley & Sons, Ltd.

New Quantitative Structure-Fragmentation Relationship Strategy for Chemical Structure Identification Using the Calculated Enthalpy of Formation as a Descriptor for the Fragments Produced in Electron Ionization Mass Spectrometry: A Case Study with Tetrachlorinated Biphenyls

“…Recently, we reported a method of determining the best series of Δ f H values obtained with semiempirical QCC methods, which used the simulation of CSI by the ORD algorithm of six chemical standards (tetrachlorinated biphenyls, TeCBs 44, 46, 52, 66, 74, and 77). 10 The lists of probabilities of possible structural assignments were obtained by the specialized software package, Chemical Structure Identification by Differential Mass Spectra (CSI-Diff-MS Analysis 3.1.1; BET2 Software, Konigsbrunn, Germany). 9 The best results were obtained using Δ f H calculated by semiempirical methods, AM1, MINDO3, and MNDO for M +• ions; RM1 for [M − Cl] + and [M − 2Cl] +• ions; PM3 for [M − 3Cl] + ; and AM1 for [M − 4Cl] +• .…”

“…9 The best results were obtained using Δ f H calculated by semiempirical methods, AM1, MINDO3, and MNDO for M +• ions; RM1 for [M − Cl] + and [M − 2Cl] +• ions; PM3 for [M − 3Cl] + ; and AM1 for [M − 4Cl] +• . 10 Although not considered the most efficient, these calculation methods are faster than other methods, but even so, we performed over 33 000 clicks for each of them, because QCC software is not yet adapted for the rapid creation of thermochemical databases for a large number of fragments. Yet, this is worth the effort, because these databases do not have to be recalculated for each analysis.…”

“…These cases often do not correspond with actual experimental situations in which only a few standards from a large group of isomers are available, and the analyte can have any of the structures proposed for the unknown isomers. We recently built such an analytical situation using as an example the TeCBs group for which we already knew several good QCC methods 10 and for which there was practical importance of identifying the various isomers. PCBs are a group of 209 isomers with varying degree of chlorination which are important environmental 11 and food 12 contaminants.…”

“…When several isomeric ionic structures result for a fragmentation, Δ f H minimum is used, because the corresponding ion is the most stable and has the essential contribution to the ionic current (IC). 4,10 From here onward, we have used only the relative minimum of the enthalpy of formation, simply given as Δ f H. The values calculated in this way were imported into the Δ f H library of the CSI-Diff-MS Analysis 3.1.1 software (Figure S-2 and Tables S-1 and S-2, Supporting Information). LCOR Algorithm.…”

“…We can distinguish two groups of high similarity: group 1 (PCB 44, 46, and 52) and group 2 (PCB 66, 74, and 77) with values ranging between 92% and 96%. 10 CSI Probabilities List. We have calculated the probabilities of the CSI list using the CSI-Diff-MS 3.1.1 software, the database of enthalpies of formation (Δ f H) calculated by HyperChem 8.0.10, and the mass spectra generated by EI-MS (see above).…”