No abstract

IRMPD spectroscopy and computational chemistry techniques have been used to determine that the proton- and sodium-bound dimers of proline exist as a mixture of a number of different structures. Simulated annealing computations were found to be helpful in determining the unique structures of the protonated and sodiated dimers, augmenting chemical intuition. The experimental and computational results are consistent with the proton-bound dimer of N-protonated proline bound to zwitterionic proline. There was no spectroscopic evidence in the 3200-3800 cm(-1) region for a canonical structure which is predicted to have a weak N-H stretch at about 3440 cm(-1). A well resolved band at 1733 cm(-1) from a previous spectroscopic study (DOI: 10.1021/ja068715a ) was reassigned from a high energy canonical isomer to the C=O stretch of a lower energy zwitterionic structure. This band is a free carboxylate C=O stretch where protonated proline is hydrogen bonded to the other carboxylate oxygen which is also involved in an intramolecular hydrogen bond. Fifteen structures of the sodium bound proline dimer were computed to be within 10 kJ mol(-1) of Gibbs energy and eight structures were within 5 kJ mol(-1). None of these structures can be ruled out based on the experimental IRMPD spectrum. They all have an N-H stretching band predicted in a position that agrees with the experimental spectrum. However, only structures where one of the proline monomers is in the canonical form and having a free O-H bond can produce the band at ∼3600 cm(-1).

A combination of IRMPD spectroscopy, collision-induced dissociation, deuterium isotopic substitution, and computational chemistry was used to determine the structure and unimolecular chemistry of [Zn(Pro-H)](+) and the singly hydrated complex in the gas phase. Five competing dissociation channels were observed: loss of H2O, CO, CO2, and HCOOH and the main fragmentation pathway, loss of neutral Zn. By comparing the IRMPD spectrum with the predicted IR spectra of the lowest energy structures, it was confirmed that [Zn(Pro-H)](+) complex is deprotonated at the amine moiety, and a hydrogen from either C2 or C5 migrated to Zn(2+). In this H-type complex, ZnH(+) was chelated between the amine nitrogen and the carbonyl oxygen. Calculations of the potential energy surface revealed that the loss of neutral zinc is energetically more favorable than the loss of dehydrogenated proline leading to ZnH(+) product. Furthermore, calculations on all five primary decomposition routes, all beginning with the lowest energy structure, revealed that loss of Zn has the lowest energy requirement, consistent with it being the most abundant product of unimolecular dissociation following collisional or IR multiphoton activation. For the singly hydrated complex, [Zn(Pro-H)(H2O)](+), IRMPD spectroscopy confirms a structure with water added to the H-type structure and intramolecularly hydrogen bonded to the deprotonated amine site. This structure is not the lowest-energy [Zn(Pro-H)(H2O)](+) isomer, but it is the one where water is added to the lowest energy [Zn(Pro-H)](+) isomer.

The unimolecular chemistry of the [Zn(Pro-H)(Pro)](+) complex following collisional or infrared multiple photon activation was studied, and interestingly was found to lose H2 as one of the main dissociation pathways. Furthermore a second dehydrogenation step, forming [Zn(Pro-H)(Pro)-2H2](+), was also observed. When proline was substituted for sarcosine, also a secondary amine, a single dehydrogenation was observed. In contrast, [Zn(Gly-H)(Gly)](+) and [Zn(Ala-H)(Ala)](+) were found to lose H2O as their primary fragmentation route with no dehydrogenation observed. Tandem mass spectrometry, deuterium substitution, and infrared spectroscopy were used to determine the origin of the H atoms in the losses of H2, as well as for other fragmentation routes, including the loss of H2O. The hydrogen atoms for H2 loss from [Zn(Pro-H)(Pro)](+) was found to originate on the amine group and primarily from C5 on the non-deprotonated proline, with a smaller contribution from the C2 hydrogen. Both hydrogens for H2O loss were determined to be from labile hydrogens. Potential energy surfaces were computed for the H2 loss and H2O loss routes for both [Zn(Pro-H)(Pro)](+) and [Zn(Gly-H)(Gly)](+) and were compared. For [Zn(Pro-H)(Pro)](+), H2 loss was found to be the pathway with the lower energy requirement than for H2O loss, and the opposite was found for [Zn(Gly-H)(Gly)](+). The greater basicities of proline and sarcosine are most likely responsible for stabilizing the 3 coordinate Zn(2+) transition states en route to H2 loss, compared to those complexes formed with the much less basic glycine or alanine.

No abstract