William A. Goddard III

William A. Goddard III was born in El Centro California (March 29, 1937) and lived part of most years until college in El Centro, Delano, Indio, and occasionally McFarland, Lodi, and the bad side of Bakersfield (Oildale), all in California. He worked as a draftsman for Imperial County the last two years of high school and for a year afterward. He obtained his BS Engineering (highest honors) from UCLA in June 1960 during which time he worked full time for 6 months as a tool designer for Douglas Aircraft Company. He finished his PhD in Engineering Science (minor Physics) at Caltech in Oct. 1964. During his PhD he consulted for the Autonetics Division of North American Aviation (NAA) Company the summers of 1961,1962, and 1963 and worked as an engineer for the Atomics International division of NAA summer 1960. Goddard joined the chemistry faculty in November 1964, where he is now Charles and Mary Ferkel Professor of Chemistry, Materials Science, and Applied Physics and Director of the Materials and Process Simulation Center (MSC). Goddard has been and continues to be a pioneer in developing methods for quantum mechanics (QM), force fields (FF), reactive dynamics (RD), molecular dynamics (MD), and Complete Sampling predictions of protein-ligand systems. In particular he extended the applications of QM accuracy from the level of 100’s of atoms for 10’s of picoseconds to reactive force fields (ReaxFF, cited 4291 times) capable of predicting reaction kinetics on systems with 100,000 to 3 million atoms and to nonreactive force fields (Dreiding for main group atoms cited 6105 times, UFF for inorganic compounds up to Z=103 cited 8404 times), including methods for predicting charge transfer dynamics (QEq cited 3229 times) and polarization dynamics (PQEq), and Poisson Boltzmann solvation (PBF method cited 1159 times) He has used these methods for numerous applications to nanotechnology (e.g., starburst dendrimer paper cited 4520 times), catalysis (heterogeneous, homogeneous, electrocatalysis), ceramics, and composites, mechanical bonding (42 papers with Fraser Stoddart, including linear artificial muscles, cited 734 times), Covalent Organics Frameworks (cited 730 times), thermoelectrics (Si NW cited 3052 times), including applications to 3D structures of membrane proteins, ligand-binding sites on proteins, and DNA self-assembly of carbon nanotubes (cited 618 times). He has collaborated with 100’s of experimental groups around the world in all of these areas, where he develops and applies whatever computational tools are needed to explain and understand the experimental results. The goal of the Goddard research has been to make the methods sufficiently accurate that the need for experimental validation can be severely restricted to the predicted best systems. and sufficiently efficient that they can be applied to realistic systems with millions of atoms (now referred to as Materials Genomics). This required improving the QM (X3LYP cited 934 times), particularly for nonbond or van der Waals interactions (XYGJ-OS) and for band gap (B3PW) while also improving the methods for matching the FF to QM to describe large scale reactive systems. He uses hierarchical approaches (multiscale, multiparadigm) to couple between the electronic states of QM with dynamics of macroscale systems, enabling first-principles based accuracy of realistic systems (millions of atoms, nanosecond time scales). A particular focus has been to determine the detailed reaction mechanisms underlying heterogeneous and homogeneous catalysts, including electrocatalysis. This led to the recent development of the Grand Canonical QM (GC-QM) or Grand Canonical Potential Kinetics (GCP-K) method of calculating the free energy of activation and of reaction at constant applied potential rather than at constant charge as in normal QM. This leads to turn-over-frequency and current as a function of applied potential in excellent agreement with experiment. Another particular focus has been predicting the 3D structures for G-Protein Coupled Receptors (GPCRs) and the activation of the G-Protein through coupling to agonists. For protein-protein and protein-ligand structures the energies are reliable for selecting the best protein structures and the best ligand-protein complexes. To predict these optimal protein and protein-ligand structures, his methods use hierarchical approaches to enable exhaustive sampling of protein structures or ligand-protein complexes coupled with methods to eliminate rapidly the worst cases using coarse level methods, successively pruning with more accurate methods, finally ending with an ensemble of thermally accessible structures. Validation studies show that these methods now achieve the accuracy needed to design reliably new stronger binding ligands.