Bond Dissociation Energies in the Gas Phase for Large Molecular Ions by Threshold Collision-Induced Dissociation Experiments: Stretching the Limits

Abstract

Accurate bond dissociation energies for large molecules are difficult to obtain by either experimental or computational methods. The former methods are hampered by a range of physical and practical limitations in gas-phase measurement techniques, while the latter require incorporation of multiple approximations whose impact on accuracy may not always be clear. When internal benchmarks are not available, one hopes that experiment and theory can mutually support each other. A recent report found, however, a large discrepancy between gas-phase bond dissociation energies, measured mass spectrometrically, and the corresponding quantities computed using density functional theory (DFT)-D3 and DLPNO-CCSD(T) methods. With the widespread application of these computational methods to large molecular systems, the discrepancy needs to be resolved. We report a series of experimental studies that validate the mass spectrometric methods from small to large ions and find that bond dissociation energies extracted from threshold collision-induced dissociation experiments on large ions do indeed behave correctly. The implications for the computational studies are discussed.

Surprising Homolytic Gas Phase Co−C Bond Dissociation Energies of Organometallic Aryl‐Cobinamides Reveal Notable Non‐Bonded Intramolecular Interactions

“…In this context, we have further investigated the possibility that the discrepancy could arise due to af ailure in either the threshold CID experiment, or the deconvolution of the experimental data by the L-CID program. To this end, we designed controle xperimentsw ith several models ystems to test the scalingoft he experimental methods with the size of the molecule, and also to isolatet he contribution of Londond ispersion, [52] the latter noncovalenti nteraction being the objecto ft he D3 correction applied to DFTc alculations. Thec ontrole xperiments revealed that the L-CID deconvolution of the T-CID experiment operates correctly for both small and large ions within the tested range and agreesw ith the calculated BDEs, with the fairly trivial proviso that the structures of the initial ion, as well as its dissociation products, have to be correct.T urning to potential issues with the computational results, the present results suggest a resolution of the previously observed discrepancy:good agreement with MeCbi + and poor agreement with AdoCbi + .A ni nsufficiently accurate consideration of the interactions among the side-chains themselves, and between the side-chains and the cleavable group on Co, most certainly impactst he result for the large adenosyl moiety,w ith its conformational flexibility,i ts hydrogen-bonding functionalities, and its extended psystem,a ll presenting possibilities for non-covalent interactions.…”

Modeling Gas-Phase Unimolecular Dissociation for Bond Dissociation Energies: Comparison of Statistical Rate Models within RRKM Theory

“…Consequently, we decided to evaluate the performance of the rate models on reproducing two kinds of rate data: microcanonical rates from Troe and Baer 46 and reaction cross-sections previously measured by us. 45 The former allows for a direct comparison of the rate models, and the latter is pertinent to our experiments.…”

“…In addition, we wanted well-behaving dissociation reactions to keep the comparisons free from complications, such as ambiguous reactant structures recently observed by us. 45 We chose a dataset of four cation-linked pyridine dimer dissociations (Scheme 1b), recently reported by us. 45 We selected the smallest and the largest well-behaving pyridine moieties from that study to sample different numbers of degrees of freedom, with both H + and Ag + linkers to sample two different E0 regions; in other words, we selected cases with varying kinetic shifts.…”

“…We recently investigated the reliability of the L-CID scheme with an experimental dataset of cationlinked pyridine dimer dissociations, designed to be structurally well-defined and well-behaved. 45 Herein, we report a theoretical assessment of the performance of the L-CID rate model by comparison to well-established statistical rate models within the RRKM framework. We compare the performance of the models on our recently published dataset of reaction cross-sections as well as a previously published dataset of microcanonical rates.…”

Application of continuous wave quantum cascade laser in combination with CIVP spectroscopy for investigation of large organic and organometallic ions

“…The precursor 3,5-bis((adamantan-1-yl)ethynyl)pyridine was synthesized and characterized previously. 20 Test compounds 3,5-di-tert-butyl-4-hydroxybenzaldehyde (2), 3,5di-tert-butylbenzaldehyde (3), and 2,6-di-tert-butylphenol are commercially available reagents, purchased from Alfa Aesar and Sigma Aldrich and used without additional purification. The coordination copper (I) tri(1H-indazol-1-yl)phosphine oxide acetonitrile complex (4•MeCN) was prepared in situ by mixing of tetrakis(acetonitrile)copper(I) hexafluorophosphate and tri(1Hindazol-1-yl)phosphine oxide (TIPO) in acetonitrile in the glovebox.…”