Ceilidh ... RedOx reaction in skewed geometry environment (modeling via wstate variation).

Firefly and PC GAMESS-related discussion club

Learn how to ask questions correctly

We are NATO-free zone

RedOx reaction in skewed geometry environment (modeling via wstate variation).

Hello open-minded Qchemists,
I am investigating a redox reaction which may be facilitated by skewing of the reagent’s geometry. As a 3D gripping device a protein is used. The reaction is activation of a triplet O2 molecule by oxidation of indirectly participating co-factor. As the O2 is too small molecule to skew with a molecular device, the geometry a cofactor would be an option which allow channeling of protein structural features (like normal modes for example) to facilitate otherwise tough reaction. The question is the viability of the hypothesis (with further experimental proofing with synthetic co-factor candidates if the answer is so-so positive) and the best way to test the hypothesis with computations.
As a preliminary research I’ve calculated multitude of reactions routes that are assuming a geometrical skews. Firstly the geometry of my system of co-factor, several residues of the protein active site, a O2 molecule and a natural co-factor was geometrically optimized with 2-states CASSCF(6,9) at sampled but frozen values of the reaction coordinate (O-O bond distance, r={1.415, 1.5, 1.6, 1.7, 1.8, 1.9A}). This provided a basic route of the reaction. As a source of alternative but either “optimal” reaction routes, the variation of two state contribution was used (sampled at each reaction coordinate value by series {0.1, 0.3, 0.5, 0.7, 0.9} excited state contribution, ground state is =1.0-the_contribution). Geometry of the system was then optimized and the energy is measured at each node of the mesh.
The PES of such wstate variation is show on the attached fig. It looks supportive for the model of skewed mechanism as activation barrier can be passed by significantly lowered route in wstates space. Better yet this is true for the beginning of the reaction only and the slope is flat as O-O distance shakes out of the equilibrium (i.e. one specific skew is needed and sufficient). As a criticism, a bit artificial mechanism of optimal geometries generation can be mentioned.
I am going to measure distribution of reagent geometries induced by the protein with long molecular dynamic of the protein bound to intermediate products which are carefully parameterized from QM data. Particularly slow growing method should be fine to interpolate between sampled geometries of the intermediates above at ns time-scale. The geometries statistics is expected to be a measure of the reaction route actually chosen (by the protein) but it takes a while and for novices like me it’s nice to know an expert opinions about investing ones time into things that are not gona work. :)
So, many thanks for your comments about the hypothesis and the approach!
Alex

This message contains the 24 kb attachment
[ wstates.pdf ] PES of wstate variation

Thu Oct 9 '14 1:55am

This message read 707 times