Special Issue Preface pubs.acs.org/JPCB
Autobiography of Marshall D. Newton
■
EARLY HISTORY I spent most of the first 25 years of my life in the Boston area (and somehow managed to avoid a pronounced regional accent). I suppose the first pivotal event in my life as a chemist (as for so many chemists) was being given a chemistry set at age 13. Harmless as far as that goes, but due to the fact that my father was mentoring an East German immigrant graduate student at MIT, I got access to many more chemicals than was wise, or safe. However, as a happy survivor, I’m here to tell this tale. Even though I did not think about it much at the time, it is clear in retrospect that my trajectory in chemistry had been set. With my informal training, and the opportunity to spend a very inspiring senior year at Phillips Andover Academy, I was able to emerge after 4 accelerated years at Dartmouth with not only a BA in Chemistry but also a master’s degree in synthetic organic chemistry. This latter was helped greatly by NSF undergraduate stipends for summer research and also due to the masterful guidance of Thomas Spencer (who during the early 60s managed to launch a number of productive careers in chemistry). Working with Spencer gave me not only immense intellectual stimulation but also a much needed dose of discipline (the closest I ever got to a “basic training” boot camp). My path forward was a bit indirect. While my experimental activities gave me an invaluable gut sense of what chemistry is about, I decided that my emotional attraction to chemistry probably lay more in theoretical directions, as I began to look at graduate school. Since my fiancée (and future wife) Natika Waterman had a final year at Wellesley, I focused on Harvard and MIT and finally chose Harvard since there seemed to be more opportunities for my particular growing interest in electronic structure theory, which I pursued in the research group of Bill (“Colonel”) Lipscomb. My timing at this point in time was very fortunate for two reasons: Sputnik had already caused the funding spigots to open, great for U.S. science, in general, and also, in particular, helping to support needy graduate students like me; also, thanks to the IBM 7090 mainframe, it became possible to start doing “realistic” ab initio Hartree−Fock (HF) calculations for some small polyatomic molecules and also to employ the ab initio results for parametrizing approximate HF models. In addition to training in traditional quantum chemistry, Harvard gave me a broad background which turned out to be extremely valuable as my research in physical chemistry evolved. This included physical organic chemistry, statistical mechanics, and electricity and magnetism. (I was even “forced” to do the laboratory work in an E&M coursegood for the soul, but reinforcing my conviction that experimental science was not the career for me.) After Harvard (1966) I spent a stimulating year in Oxford as an NSF postdoctoral fellow (in the group of the Charles Coulson, housed in the Mathematical Institute since the authorities at Oxford did not think the time was yet right for a self-standing department of theoretical chemistry, notwithstanding the worldwide prominence of Coulson as a leader in © 2015 American Chemical Society
quantum chemistry), and then 3 years at Carnegie Mellon University as a “Pople Person”, associated with John Pople’s research group. Once again, the timing was favorable. Just before I arrived, Warren Hehre had joined the group, bringing with him a copy of the historic “Polyatom” code, which he had been playing with as an undergraduate at Cornell working in Roald Hoffmann’s group. The upshot is history. Together with Warren, Pople quickly switched from semiempirical models to new, efficient Gaussian orbital-based Hartree−Fock methods, leading to the family of “Gaussian” codes. Pople set a formidable example for the conduct of research, including everything from punching cards to devising and coding new atomic orbital integral methods as well as having spectacular knowledge of many fields of physical science and mathematics. I consider that this was when I really “learned the trade” of computational quantum chemistry. The computational efforts placed strong emphasis on evaluation and analysis of many important chemical properties based on the results of the newly available ab initio capabilities. In 1969, at the late age of 29, I finally got a “real” job, selecting the Chemistry Department at Brookhaven National Laboratory (BNL), after realizing what a wonderful opportunity this offered for very basic physical research, under a broad programmatic framework sponsored then by the AEC, subsequently ERDA, and since then, DOE.
■
THE PATH TO A RESEARCH PROGRAM When I arrived at BNL, I had many interests and a fair amount of technical training but still no clear goal for a specific research program. Although I had switched from experimental organic chemistry to theory, I still had an active involvement in organic bonding theory, which had started in the late 60s and continued at BNL well into the 70s, focusing mostly on the use of localized molecular orbitals (LMOs) to analyze bonding in highly strained organic polycyclic molecules. This work had started with Lipscomb and Gene Switkes, a Lipscomb student, and led to a very productive collaboration with Jerry Schulman (Queens College). While my organic phase was playing out, the main thrust of my career was rapidly falling into place thanks to the influence of my colleagues at BNL (including Stan Ehrenson, Lew Friedman, the radiation chemistry group, and then a bit later, a major mentor in the person of Norman Sutin) and nearby Stony Brook University (especially my mentor Harold Friedman and later his close associate Fernando Raineri). In fairly rapid succession, my research, with a heavy reliance on the evolving Gaussian code as well as a version of Polyatom enhanced with effective core potentials from Bill Goddard’s group, progressed from molecular level studies of aqueous excess protons and hydroxide ions (1970−1976), to excess electrons in water and liquid ammonia (1972−1975), exploiting Special Issue: John R. Miller and Marshall D. Newton Festschrift Published: June 18, 2015 7129
DOI: 10.1021/jp5113087 J. Phys. Chem. B 2015, 119, 7129−7131
The Journal of Physical Chemistry B
Special Issue Preface
a very early ab initio HF reaction field approach which we developed by combining a discrete cluster with a surrounding equilibrium or nonequilibrium dielectric continuum (DC), and finally (1976 to the present) to mechanistic analysis of electron and hole transfer (ET) kinetics in molecular systems, later extended to excitation energy transfer (EET). The innocuous label “electron transfer” actually covers a large number of fundamental processes and mechanisms and has never led to a dull moment over the roughly 40 years I have been grappling with this topic (since 2008, as an active BNL Emeritus Senior Chemist). There is, of course, a large class of ET processes central to chemical science: thermal/optical/ photoinitiated; adiabatic/nonadiabatic; homogeneous/interfacial; intermolecular/intramolecular; and occurring in inorganic/ organic/organometallic assemblies. The mechanistic analysis of ET leads one to invoke numerous theoretical concepts and techniques. These include (to name some of the most important): formulation of initial (“prepared”) and final states (the “diabatic” states, in contrast to the “adiabatic” electronic eigenstates), electronic structure models for determining molecular structure and energetics, and the electronic Hamiltonian coupling (HDA) of distant donor (D) and acceptor (A) sites (my work has employed primarily wave function approaches but also, to some extent, DFT methods); statistical mechanics and molecular dynamics (both classical and quantal) for characterizing the nonequilibrium states of the media in which ET processes occur; and electrochemical aspects of ET kinetics taking place, for example, at self-assembled monolayer (SAM) film-modified electrodes. I mention here a few milestones in our efforts that come to mind: 1980−1995: Exploiting symmetry broken HF wave functions, based on both ab initio and semiempirical many-electron methods, to formulate diabatic states and evaluate coupling elements for a wide variety of intermolecular and molecular systems. A first big step was an ab initio study carried out with postdoctoral student Jean Logan (1983) for the prototype selfexchange of hexa-aquo ferrous and ferric ions (a process first studied experimentally by Silverman and Dodson in the BNL Chemistry Department (1952)). Starting in 1991, we demonstrated quantitatively that proper many-electron HDA values can typically be approximated to within ∼10% by effective orbital coupling elements, thus greatly simplifying the analysis of coupling and facilitating a number of detailed assessments of superexchange-mediated electron versus hole transfer pathways in organic and organometallic systems (e.g., with postdoc Congxin Liang (1992−1993); similar studies were reported by Curtiss and Miller et al. and Paddon-Row and Jordan et al. (1991−1993)). 1995−present: Introducing with my longtime friend and collaborator Bob Cave the generalized Mulliken−Hush model (GMH) for defining diabatic states and their properties, especially the Hamiltonian coupling elements and effective D/A separation distances. This approach, now widely used by both theorists and experimentalists, is applicable to an arbitrary molecular configuration (not limited to the Franck−Condon crossing point and thus permitting tests of the Condon approximation) and electronic state space (i.e., not limited to the traditional 2-state model) and can be applied to any electronic structure model which provides adiabatic energies and dipole matrix elements and as well to any experimental probe yielding the same information.
Aside from 1-particle (electron or hole) coupling, we have also taken a look at 2-particle coupling as pertains to electronic excitation transfer (with BNL Goldhaber postdoc Seogjoo Jang (currently at Queens College) and Bob Silbey (2004 and 2007) and with Joe Subotnik and coworkers (2010)) and also one venture (1992) calculating magnetic exchange coupling in copper oxides with my BNL colleague Jim Davenport. 1986−2006: Formulation and application of general methods for nonequilibrium behavior of polar media (using both discrete molecular level and DC models) in controlling the energetics, dynamics, and kinetics of ET in polar media. This multifaceted effort involved several collaborators, including: •With Harold Friedman, starting in 1982, and then also with Fernando Raineri (1996), applying molecular-level integral equation and molecular dynamics (MD) techniques to characterize equilibrium and nonequilibrium solvent energetics for ET kinetics in both polar and also nondipolar solvents (e.g., benzene), with a detailed analysis of the role of activation entropy for ET occurring in solutions with appreciable ionic strength. In 1988, Harold and I presented a rigorous expression for nonequilibrium medium response at the linear (but not limited to local) level in terms of a dielectric Green’s function, correcting and extending some prior models for solvent reorganization free energy. •With postdoctoral student Yi-Ping Liu, we derived a 3-zone DC model (1994) ideally matched to the modeling of ET at SAM film-modified electrodes (see below) and a unified ab initio HF self-consistent reaction field model (1995) which yielded reorganization energy and electronic coupling for a homologous series of ET through alkyne spacers in water. •For over a decade starting in 1995, with Misha Basilevsky and his colleagues in Moscow, we formulated and implemented general DC models for ET kinetics (applicable to arbitrary molecular cavity shapes), including a novel frequency-resolved continuum model and a spinoff (with Igor Leontyev), providing a way to screen parameters in nonpolarizable molecular force fields so as to account implicitly for the full electronic response contribution to the energetics (2004). •With Greg Voth and his postdoctoral student Lowell Ungar (1999), we applied a quantum path approach, which when combined with quantum chemical estimates of coupling elements allowed a full mechanistic analysis of the reduction of a Co3+ complex by a Ru(bpy)32+ complex, mediated through a tetraproline spacer, studied experimentally by Stephan Isied et al. The calculations yielded separate mechanistically crucial factors in the rate constant prefactor not available from experiment. This analysis was subsequently extended working with Dmitry Matyushov et al. (2006), using very general solvation models. Related work with Dmitry (2001) helped to elucidate the line shapes for optical ET in coumarin-153. With Mike Elliott and coworkers, a pair of papers (1996, 1998) addressed the variation in HDA magnitudes by comparing calculated and experimentally inferred values in a binuclear mixed valence Fe2+/Fe3+ complexes held together with three alkane spacers of variable length. The second paper was a great privilege for me to participate in since it reported the first direct experimental comparison of D/A coupling obtained by optical and thermal ET in the same system, with the very satisfying result that the two coupling magnitudes agreed to within ∼20%. Since 1993, thanks to the encouragement and on-the-job training due to BNL colleagues Steve Feldberg and John Smalley and Chris Chidsey (then at Bell Laboratories, now at Stanford), I got heavily involved in mechanistic analysis (and 7130
DOI: 10.1021/jp5113087 J. Phys. Chem. B 2015, 119, 7129−7131
The Journal of Physical Chemistry B
Special Issue Preface
relevant coupling and energetics) of ET kinetics through oligomeric organic thiolate films (SAMs) on Au (111) electrodes, doped with terminal ferrocene and other redox groups. Exploring several saturated and unsaturated SAM spacers as a function of length and in different electrolyte media, offered a wealth of fascinating challenges for basic chemical dynamics and kinetics, which resulted in a steady stream of papers for nearly two decades. A related study with postdoc Vasili Perebeinos (2005) used DFT band structure calculations to characterize the structure and work functions of phenylthiolate SAMs on Au(111) and Cu(111) surfaces.
■
FINAL THOUGHTS The work described above, while nominally all theory and related computation, has maintained a strong focus in varying degrees on the analysis and elucidation of chemical behavior directly accessible to contemporary experimental probes. Nearly 20% of our work (including that with Mike Elliott highlighted above) has involved collaboration with experimentalists in joint efforts to interpret new experimental data. Having a long-term experimental colleague, John Miller, a member of our Chemistry Department since 1998 and my Festschrift partner, has been of great help in keeping me in close touch with the experiments which have driven so much of our theoretical investigations. Another experimentalist colleague whose insights have been of great benefit is Piotr Piotrowiak, one of the coeditors of this Festschrift together with Bob Cave. It is worth emphasizing the distinct dual roles of theory within the realm of chemical theory and computation, whereby the computations are generally bookended by theory on both ends: i.e., the theory both underlies the computational models and, of equal importance, the analysis of the often large amount of data emerging from the calculations, which are sometimes referred to as in silico numerical experiments. In dealing with these interrelated aspects of theoretical research, I have greatly valued for nearly half a century my interactions with colleague and fellow department member, Jim Muckerman, whose office has always been just a few yards from mine. Chemistry employs to a large extent a common “language” in spite of occasional parochial tendencies and has for some time been a truly international enterprise (long before “globalization” was a catchword), and the value of face-to-face contact between scientists within the U.S. and around the world cannot be overestimated. Aside from greatly enhancing the opportunities for productive research collaboration, the more general cultural benefits and the many wonderful personal relationships established with foreign scientists have made my scientific activities especially rewarding. I am also deeply grateful to the BNL Chemistry Department and the Department of Energy Office of Basic Sciences (BES) and, in particular, the Solar Photochemistry Program for their continuing support over the years. Finally, I am very indebted to my family for enriching my life in so many ways: my daughter Erika (a physician) and son Joel (a guitarist) and my wife Natika, who has tolerated whatever obsessions my quests for knowledge have entailed. She has her own, as a philosopher and professor, and provides me with much intellectual stimulation, but from the perspective of a field sufficiently distant from my own that we do not get in each other’s hair. Marshall D. Newton 7131
DOI: 10.1021/jp5113087 J. Phys. Chem. B 2015, 119, 7129−7131