Photo of Karl

Karl Freed

Born Brooklyn, New York, 1942. Columbia University, B.S., 1963. Harvard University, A.M., 1965; Ph.D., 1967. University of Manchester, England, NATO Postdoctoral Fellow, 1967-68. University of Chicago, 1968-.

E-mail address:


Figure describing recent work on protein folding

Our research interests cover several areas of theoretical chemistry, including the electronic structure of molecules, the statistical mechanics of polymers in the liquid phase, and the long time dynamics of proteins and polymers in solution.

We have developed a highly correlated ab initio electronic structure method that is designed to tackle the difficult problem of describing electronic excited states of molecules. The method is a multiconfigurational generalization of the widely used MPn single reference configuration methods that are available in many commercial electronic structure packages. These new ab initio methods have been applied to describe the excited states of a number of conjugated pi-electron systems, where our computed energies and oscillator strengths rival in accuracy the most advanced ab initio methods. Additional applications have been made to computing two-dimensional methyl mercaptan and three-dimensional hydrogen sulfide potential energy surfaces for the electronically excited states that are accessed in non-adiabatic photodissociation experiments carried out by Professor Butler's group. (In fact, the computations have been performed by a theory-experiment student working jointly with Prof. Butler and me.)

Our electronic structure methods are unique in enabling us to derive from first principles the true valence shell effective Hamiltonian that is mimicked by the model Hamiltonians of purely semiempirical molecular orbital theories of molecular electronic structure. We have computed the first fully correlated "ab initio" pi-electron Hamiltonian that demonstrates why some assumptions of semiempirical pi-electron theories are correct, but our computations for small conjugated pi-electron systems indicate deficiencies of these older methods along with theoretically justified methods for their improvement.

We have been developing a theory of the statistical thermodynamics of polymers in the liquid state. Our theory is the first and only one to describe the influence of monomer molecular structure on the thermodynamic properties of polymer mixtures. Several applications explain small angle neutron scattering and thermodynamic experiments for mixtures of polymers in the liquid state. Our theoretical predictions of a strong pressure dependence to the small angle neutron scattering intensities has been verified. Likewise, we have predicted the possibility that certain block copolymers will form mesoscopically ordered self-assembled structures in the liquid phase upon heating, a bold prediction subsequently verified experimentally. Recent extensions of the theory consider random copolymers, the influence of short chain branching and chain semiflexibility on miscibilities of polymers in the liquid phase, as well as the phase behavior of liquid crystalline systems. Other work in this area is devoted to developing a theory of interfaces in polymer systems. Particular examples include the interfaces between phase separated polymers, surface segregation profiles of polymers near an impenetrable surface, and the interfaces in self-assembling block copolymer systems.

A key motivation is to understand the molecular features governing the rich array of observed phenomena.

Flexible aqueous proteins and solution polymers have important dynamical processes occurring on time scales far exceeding current capabilities for computer simulations of these systems. Thus, we are developing a theory of the long time protein and polymer dynamics which can use input information that is accessible to current computer capabilities. The theory is able succesfully uses this information to provide a realistic description of the longer time dynamics. While the initial applications of the theory compare well with experiments made by Professor Fleming's group, current research has considered simpler alkanes, polypeptides, and small neurotransmitting peptides for testing and refining various components of the theory. This work focuses on the internal chain dynamics and on the nonequilibrium dynamics modeling protein unfolding, but future extensions are planned to consider the fundamentals of molecule-solvent interactions.

Further Reading

Mode Coupling Theory for Calculating the Memory Functions of Flexible Chain Molecules. Influence on the Long Time Dynamics of Oligioglycines. , J. Chem. Phys., 106, 771 (1997).

Ab Initio Computation of Semiempirical Pi-electron Methods. V Geometry Dependence of Pi-electron Integrals. J. Chem. Phys., 105, 1437 (1996).

A Lattice Model Molecular Theory for the Properties of Polymer Blends. Trends Polym. Sci., 3, 248 (1995).

Influence of Monomer Structure and Interaction Asymmetries on the Miscibility and Interfacial Properties of Polyolefin Blends. Macromolecules, 29, 8960 (1996).

Building a Bridge between Ab Initio and Semiempirical Theories of Molecular Electronic Structure. in Structure and Dynamics of Atoms and Molecules: Conceptual Trends, Eds. J. L. Calais and E. Kryachko (Kluwer, Netherlands, 1995).

Analytic theory of Surface Segregation in Compressible Polymer Blends. J. Chem. Phys., 105, 10572 (1996).

Global Three-dimensional Potential Energy Surfaces of H2S from the Ab Initio Effective Valence Shell Hamiltonian Method. J. Chem. Phys., 105, 8754 (1996).

Application of the Effective Valence Shell Hamiltonian Method to Accurate Estimation of Valence and Rydberg States Oscillator Strengths and Exictation Energies for Pi Electron Systems. J. Chem. Phys., (in press).

University home page Go
to Chemistry Department home page Go up one level