Previous Page  |  Table of Contents  |  Next Page

Christopher J. Cramer, Fellow

Development and Application of Computational Chemistry for Purposes of Investigating Phenomena of Biological and Environmental Interest

These researchers stress the limit of present day computational chemistry tools in order to examine large systems of relevance to one or more areas in chemistry. In general, they focus on systems containing unpaired electrons, multiple metal atoms, or systems of a very large organic nature. Furthermore, they try to include any condensed-phase effects into the calculations in order to make them more relevant to experimental observations. This work is ongoing in several major areas.

99/35

99/48

99/66

99/99

99/111

99/112

99/120

99/124

99/128

99/139

99/147

99/161

99/174

99/183

99/217

99/218

The first area focuses on RNA dynamics. Nuclear magnetic resonance has proven to be a powerful technique for determining the time-averaged structure of polynucleic acids (e.g., transfer-RNA). The measurement of internuclear nuclear Overhauser effects provides a set of distance constraints on various proton dyads, and this information is used in conjunction with a force-field minimization to generate a plausible nucleic acid structure. A more complete study of a structurally characterized alanine t-RNA using a tetraloop analog included the full environment and took advantage of recent advantages in simulation technology. Moreover, experimental data is being used for a mutant version of the tetraloop known to have different properties with respect to charging the RNA with alanine, and dynamic behavior of the two in simulations are being compared in order to understand their different biological activity.

This group has also pioneered methods for modeling solvent as a surrounding dielectric continuum. Advantages of this approach, opposed to explicit-solvent models, is its considerably greater speed and its quantum mechanical treatment of the solute, which allows all aspects of chemical reactions to be addressed. Early work has been successful in developing and calibrating models for any solvent at various theoretical levels. This research involves the implementation of two-response-time theory for the excitation of molecules in solutions using INDO/S-CIS method and generalized Born model. As an important first step, a new charge model has been developed, which is used to calculate atomic partial charges for both ground states and excited states. The primary results are quite encouraging. This provides an excellent starting point for the calculation of spectroscopic shifts in solution. A SM5 type solvation model has been implemented into Zerner's semiempirical program. Further research on this project is focusing on the calculations of spectral shifts in solution. Finally, a generalized Born equation is being developed for continuing the work on studying conformational issues in sugars, aqueous acceleration of electrocyclic reactions, bonding/stacking interactions in DNA base-pairing, and longer range, enzyme inhibitor interactions. Finally, on the technical side, new algorithms are being developed to calculate analytical gradients of the energy with respect to nuclear motion to more efficiently optimize molecular geometries.

Nitrenium ions are reactive intermediates implicated as carcinogenic products from aromatic amine catabolism. The divalent nitrenium ions, which may exist in either singlet or triplet spin states, covalently modify DNA in the carcinogenic event. Early theoretical work (and there has been very little) suggested arylnitrenium ions to be uniformly ground state singlets. However, recent higher level calculations and experimental work done by this group has indicated that the triplet spin state may be preferred in certain circumstances-in particular when the aromatic ring is substituted with electron withdrawing groups or when the nitrogen atom is substituted with a very bulky ligand. This is quite critical since the mechanisms for reaction of the singlet compared to the triplet are distinct. These researchers intend to probe the electronic effects of substitution on preferred spin state and confirmation at the ab initio level. With these results, faster semi-empirical methods can be validated and interaction of the reactive species with DNA and other biological macromolecules can be modeled. Finally, there is an interest in exploring the ability of Density Functional techniques for the calculation of structures and the prediction of singlet-triplet energy gaps. The group is also in the process of comparing very accurate levels of theory to more efficient levels in order to ascertain how useful the lower levels of theory are.

This group is also interested in the reaction of nerve agent VX with various nucleophiles. It is known that VX reacts with hydroperoxide ion in aqueous solution to give a non-lethal phosphonate salt. Furthermore, it is known that significant amounts of a toxic byproduct are formed when the nucleophile is hydroxide ion. The precise mechanism of solvolysis at phosphorus in VX is unknown, nor is it known whether this change in nucleophile leads to a change in mechanism, or simply a change in the relative energies of the intermediates and transition states involved on the potential energy hypersurface. Studies of these two reactions are being done using state-of-the-art ab initio molecular orbital methods. Solvation effects are included. Through elucidation of these mechanisms, a better understanding of the neutralization process of nerve agent VX can be achieved.

This group is also in the process of exploring the illustrated dicoper structures using Hartree-Fock and Density Functional Theories. The goals are to establish the spectroscopic properties of these molecules to make comparison with experiment, determine whether the molecules are well described as ground-state singlets with complete spin-pairing or whether symmetry breaking to localize one electron on each copper atom is advantageous, more fully understand the effects of nature and location of counterions, more fully understand the effects of solvation, and to gain a more firm understanding of the molecular orbital issues that control the relative stabilities of the two forms. Future plans are to focus on the reactivity of these systems, in particular the computational characterization of arene oxidation and hydrogen-atom abstraction reactions. The group will characterize the reaction coordinate for this process to determine whether both species are in fact active, or whether a common species intermediate to the two best represents the transition state structure. A similar analysis will be performed for H-atom abstractions. In this case, kinetic data are available against which to benchmark the theory, to include kinetic isotope effects against which the results of dynamics calculations, including tunneling, can be compared. For both of these reactions, effects of varying the ligands on the copper atoms will be examined.

This project also consists of a class taught at the University of Minnesota with state-of-the-art hardware and software employed in modern molecular modeling. Students performed several experiments using the IBM SP and software packages including gaussian 98 and amsol.

This information is available in alternative formats upon request by individuals with disabilities. Please send email to alt-format@msi.umn.edu or call 612-624-0528.

HOME | QUESTIONS | FEEDBACK
Employment | Events | Links | People | Programs | Publications | Support | Welcome