Experimental and Computational Study of the Group 1 Metal Cation Chelates with Lysine: Bond Dissociation Energies, Structures, and Structural Trends

Abstract

The kinetic energy dependence of the collision-induced dissociation (CID) of Group 1 metal cations (M+ = Li+, Na+, K+, Rb+, and Cs+) chelated to the amino acid lysine (Lys) was measured by threshold CID using a guided ion beam tandem mass spectrometer. The simple loss of neutral lysine is the only dissociation channel observed with the heavier alkali metal cations, whereas CID of Li+(Lys) yields other competing channels including loss of NH3 (the dominant channel at low energy) and eight other reactions. Analysis of the kinetic energy-dependent cross sections yields experimental M+(Lys) bond dissociation energies (BDEs) of 376 ± 30, 219 ± 13, 160 ± 10, 141 ± 6, and 128 ± 4 kJ/mol for Li+, Na+, K+, Rb+, and Cs+, respectively. Computational searches yielded 18 distinct, low-energy structural families related to sites of M+ binding in M+(Lys) complexes and 10 distinct, low-energy structural families for neutral lysine. Among the four levels of theory and three basis sets used, four different ground conformers of M+(Lys) and four different ground conformers of lysine were found, including a ground conformer of K+(Lys) and Cs+(Lys), [Nε,CO­(OH)], and its higher energy zwitterionic analogue, [Nε,CO2 –], that better explains recent infrared multiple photon dissociation action spectroscopy results. Computational results for predicted ground structures of M+(Lys) complexes yielded computed BDEs in reasonable agreement with experiment.

Benchmarking Cation−π Interactions: Assessment of Density Functional Theory and Möller–Plesset Second-Order Perturbation Theory Calculations with Optimized Basis Sets (mp2mod) for Complexes of Benzene, Phenol, and Catechol with Na+, K+, Rb+, and Cs+

“…Concretely, the popular B3LYP functional, employed in early studies of cation−π binding energies, , is here tested with the 6-31G­(d) (model A) and 6-311+G­(d,p) (model B) basis sets, employing the Hay–Wadt effective core potential (ECP) for the larger cations. Next, the M06 functional, extensively employed by Armentrout and co-workers, ,, is used in this work in combination with three different basis sets, namely, 6-31G­(d), 6-311+G­(d,p), and the large def2-TZVPPD (model C) basis set, adopting two different ECPs for the larger cations: the Hay–Wadt, with the Pople 6-31G basis sets, and def2-TZVPPD ECP, consistent with the third, larger, basis set. Finally, three additional functionals are considered in the present work: M06-2X, recently applied to the study of cations interacting with curved aromatic species, PBE0, , also recently employed in the study of neutral Na + – and K + –benzene complexes, and ωB97XD, applied to arene–Cs + calculations .…”

“…As detailed in the original papers, , these basis sets were derived from the original 6-31G­(d,p) basis set, by tuning the α d exponents of the d polarization functions on heavy atoms (separately for carbon and oxygen) and the α p exponents of the p functions on hydrogen, to reproduce the aromatic homodimer interaction energies of a reference CCSD­(T) training set, yielding a standard deviation with respect to the CCSD­(T) validation set lower than 2 kJ/mol, for both benzene and catechol . As far as the metal ions are concerned, following the suggestion of Armentrout and co-workers, ,, we choose to employ the large def2-TZVPPD basis set, again considering explicitly all core electrons for the two smaller ions and resorting to ECPs for the other two.…”

“…An important consequence of this finding is that many conclusions drawn from investigations carried out on interactions between metal ions and aromatic molecules in vacuo could be in first approximation extended to the condensed phase . In this framework, the extensive and long lasting work of the Armentrout’s group, based on both experimental measures and theoretical calculations of binding energies (Δ E ’s) in complexes formed by M + ions and organic molecules, − furnishes a sound reference database, against which computational tools can be validated or tuned.…”

“…The aim of the present work is therefore 3-fold. The first goal is to validate MP2 mod applicability and performances with respect to higher level methods, in particular CCSD­(T), and to a number of lower level DFT approaches, previously employed in the field of cation−π interactions. − ,,,, To this end, also in view of future applications to trimers or larger cation–aromatic clusters, we will employ the MP2 mod approach, coupling it with the modified basis sets recently developed by us to treat π–π complexes. , Next, to demonstrate MP2 mod computational convenience, the method will be applied to compute and analyze several two-dimensional IPESs of M + –aromatic unit pairs, for different metals (Na + –Cs + ) and aromatic moieties (benzene, phenol, and catechol). To our knowledge, an accurate yet systematic investigation of IPES cross section in cation−π interactions has been seldom reported, often limited to few one-dimensional curves, due to the computational cost of the accurate method employed.…”

“…Notwithstanding the aforementioned very good agreement with the experimental measures and the deep insight gained in the basic physics underlying cation−π interactions, the routine application of such high level methods appears to be still out of reach, due to their demanding computational cost. Hence, given the importance of cation−π interactions in biosystems ,,,, a lot of effort has been devoted to develop and/or validate accurate yet feasible methods, − ,,,− capable to reliably explore more extended regions of the interaction potential energy surfaces (IPESs), larger monomers, or complexes consisting of more than two units. The latter aim appears to be particularly challenging, as a simple benzene–M + –benzene trimer already requires not only a reliable estimate of the cation−π interactions but also of their interplay with the π–π interactions, established between the two aromatic units.…”