For a brief review of the research that is examined here, one can feel free to look a brief power point presentation of it. Computational Host/Guest Chemistry Review

Below: A screen shot of the macrocycle propsed by Dr.Duffy-Matzner, sought by her reseearch group.

Much of the computational work in this lab was set up around the research that took place at Augustana College over the summer of 2001. It deals primarily with the production of a novel macrocycle that is theorized to have antibiotic properties. This novel macrocycle is similar in structure to a known protease inhibitor Nonactin (an antibiotic that limits transfer across cell membranes through capturing a group I metal ion). Because the synthesis of this molecule is still in the early stages, the opportunity to better "design" the molecule has arisen. Ideally, the novel macrocycle should act much like the Nonactin in the way it bonds to metal ions (in this case potassium). Initial work was done on small molecules to check the accuracy of the various calculations including geometry and thermodynamic data. The host-guest chemistry was initially looked at using crown ethers as examples. They are simple cyclic structures that, not unlike the antibiotics aforementioned, capture group I metal ions by donating its lone pairs to the metal's empty orbitals in a form of coordination bond. An actual analysis was done on Nonactin models to try to ascertain its unique properties calculated by the various computational methods. Criteria like pocket size, electron availability, and thermodynamic energies and effects become important parameters to search for. One specific topic of interest concerning the antibiotic Nonactin is its affinity for potassium over sodium. Once these criteria are examined for the Nonactin, the novel macrocycle will be examined likewise and altered depending on the computational results.

I. The Examination of Crown Ethers and their Host-Guest Abilities

II. The Structure and Properties of Nonactin

III. A brief look at the proposed macrocycle

Clockwise from the upper left: A PDB file of 12Crown4 ether with a Li+ ion, 15Crown5 ether with a Na+ ion, 18Crown6 ether with a Cs+ ion, and 18Crown6 ether with a K+ ion.

A brief background look at crown ethers:

The first number referred to the number of non-hydrogen atoms are in the molecule. The number following the "crown" is the number of those atoms which are oxygens. Crown ethers are always arranged in a O--C--C--O--C--C-O pattern. This, and their natural zig-zag formation, is what gives this unique type of ether their "Crown" appearance Crown ethers were invented by Charles J. Pedersen, Donald J. Cram, and Jean-Marie Lehn in 1967. They won the Nobel prize in chemistry in 1987 for there work on these and other selective host guest chemistry.

The examination of crown ethers was done on multiple aspects of the molecules. One pertinent aspect to this project involves the geometry of the analyzed molecules. Because of the nature of this host-guest chemistry, the distance from the metal ion is of certain interest. As is seen in crown ethers, the size of pocket that can form around the ion plays a major role in the selectivity. Also, the energy of the ether with and without the bonding to the metal ion where of interest for trying to decipher or determine a thermodynamic preference or driving force in the binding of the metal ion. Besides the usual heat of formation and other thermodynamic factors that go into these types of reactions, the entropy that could be obtained from IR simulations may be of interest because of the change in mobility that would ensue from multi atom bonding to a single metal ion. A brief examination of the molecular orbitals of the ethers will also be examined. This is based on the assumption that the lone pair which donate themselves to the empty orbitals of the metal ion come from the highest occupied orbitals and will become of greater interest in the antibiotics where there are multiple combinations of oxygens that could, in theory, bind to the ion. The three most common types of crown ethers were examined: 14crwon4, 15crwon5, and 18crown6. Each molecule also had calculations done with it weak bonded to at least on one metal ion that it is known to bind. Lithium ions were used for the coordination with the smallest crown ether while both potassium and rubidium were coordinated to the 18crown5. Subsequently, sodium was used for the 15crown5. For the ethers, which are known to prefer different earth metal ions, calculations where done using both the new PM5 computational method and the Becke88 LYP density functional theory. These computations where carried out for three main crown ethers and different earth metal ions.

The initial geometry calculations turned out to be quite accurate for the crown ethers when done by either the DFT aforementioned or the new PM5 semi-empirical method. The DFT was slightly more accurate when the metal ions were introduced to the systems but the PM5 was faster by a significant factor. The thermodynamic information that was generated through the infrared simulations via the PM5 was also less accurate but far quicker than the DFT method. The examination of the molecular orbitals of the crown ethers revealed what was already thought reasonable. The highest occupied molecular orbitals were found to be on the binding oxygens, with the LUMO located on the metal ion. This seems to be a reasonable conclusion given the nature of the oxygen/metal bonding. Overall the results from the crown ether computations seemed reasonable.

Early computational work on the known antibiotic, Nonactin, mirrored many of the same similarities that where inherent in the crown ethers. One interesting point is the way the molecule was predicted to bend in on itself. Although further calculations still must be completed, the area of most interest is the antibiotics vast preference to bind to potassium rather than sodium, both of which are important in cell membrane transfer. Pocket size becomes less of an issue when the molecule bend and twist around the atom rather than holding it directly in the center as in the mostly planer crown ethers. However, it is well documented which of the possible oxygens in the Nonactin actively bind to the metal ion. From this fact, more assumptions about the role of the molecular orbital energies, conformational analysis, and change in energy from one ion to another, can be made.

The calculations of the novel macrocycle take an extremely long time, making the many aspects of the molecule the are being examined quite a dubious task to analyze. Using the relatively new technique referred to as CONFLEX, the large macrocycle was bend around and each geometry's energy was taken by a simple molecular mechanics program (MM3). This calculation took nearly a week, but revealed a reasonable starting point for the more accurate and consuming calculations of PM5 and DFT methods. Another aspect of the novel proposed macrocycle that will take considerable calculation time is the fact that there are multiple combinations of oxygens that can bond to the metal ions. The initial binding oxygens, which are pictured in the PDF file, are bonded by the most reasonably close oxygens as predicted by the initial geometry calculations.

Because of the large size of the molecules, this work is still in progress. The initial findings based on the crown ether calculations and work on Nonactin provide the idea that the prediction of simple host guest chemistry and their capabilities/unique qualities, possible.

Back to Quantum Chemistry Page