Cundari Th.R. -- Computational Organometallic Chemistry-0824704789

Freida S. Dale, B.S. Department of Chemistry, The University of Alabama at Birmingham, Birmingham, Alabama

Michael Diedenhofen, Dipl.Chem. Department of Chemistry, Philipps-Uni- versita¨t Marburg, Marburg, Germany

Warthen Douglass, M.S. Department of Chemistry, University of North Carolina at Wilmington, Wilmington, North Carolina

Dmitri G. Fedorov, Ph.D. Department of Applied Chemistry, School of Engineering, University of Tokyo, Tokyo, Japan

Gernot Frenking, Ph.D. Department of Chemistry, Philipps-Universita¨t Marburg, Marburg, Germany

Mark S. Gordon, Ph.D. Department of Chemistry, Iowa State University, Ames, Iowa

Tracy P. Hamilton, Ph.D. Department of Chemistry, The University of Alabama at Birmingham, Birmingham, Alabama

Jeremy Noel Harvey, D.Sc. School of Chemistry, University of Bristol, Bristol, England

Amy Hirsh, B.S. Department of Chemistry, The University of Memphis, Memphis, Tennessee

Angela Jolly, B.S. Department of Chemistry, The University of Memphis, Memphis, Tennessee

Karsten Krogh-Jespersen, Ph.D. Department of Chemistry, Rutgers, The State University of New Jersey, New Brunswick, New Jersey

Takako Kudo, Ph.D. Department of Fundamental Studies, Faculty of Engineering, Gunma University, Kiryu, Japan

Ohyun Kwon, M.S. Department of Chemistry, Auburn University, Auburn,

Alabama

Pascual Lahuerta, Ph.D. Department of Inorganic Chemistry, University of Valencia, Valencia, Spain

Douglass F. Taber, Ph.D. Department of Chemistry and Biochemistry, University of Delaware, Newark, Delaware

Thomas Wagener, Ph.D. Department of Chemistry, Philipps-Universita¨t Marburg, Marburg, Germany

Yanong Wang, Ph.D. Division of Chemical Sciences, Wyeth-Ayerst Research, American Home Products, Pearl River, New York

Simon P. Webb, Ph.D. Department of Chemistry, Pennsylvania State University, University Park, Pennsylvania

David P. White, Ph.D. Department of Chemistry, University of North Carolina at Wilmington, Wilmington, North Carolina

Takeyce K. Whittingham, B.A. Department of Chemistry, Rutgers, The State University of New Jersey, New Brunswick, New Jersey

Soon S. Yoon, B.S. Department of Chemistry, The University of Alabama at Birmingham, Birmingham, Alabama

Kimberly K. You, Ph.D. Plastics Application Center, BASF Corporation, Wyandotte, Michigan

Wei Zhang, M.S. Process Research and Development, Bristol-Myers Squibb Pharmaceutical Research Institute, New Brunswick, New Jersey

Another major motivation for this volume is to organize in a single place much of the hard-won experience that speaks to the ‘‘how to’’ of computational organometallic chemistry. This monograph brings together experts in the field and is designed to combine instructional aspects with cutting-edge applications. The former are intended to introduce this exciting research field to those, experimentalists and theorists alike, who might wish to try their hand at computational organometallic chemistry, while the latter should provide motivation for embarking on the journey.

As we start the new millennium, we see that the face of scientific research has changed dramatically in just the past decade. Two of the most important trends are the growing importance of computers in all aspects of scientific research and the increasing interdisciplinary nature of the science being undertaken. These tendencies are well represented in the present volume. Computational chemistry and organometallic chemistry are, almost by definition, interdisciplinary endeavors. The latter exists at the interface between inorganic and organic chemistry, providing erstwhile inorganic chemists a chance to try their hand at making new organometallic compounds by manipulation of the metal and its environs. Closet organic chemists also play a major role in organometallic chemistry through their attention to the organic functionalities. Computational chemistry has also metamorphosed from its origins as a branch of physical chemistry to embracing all traditional and nontraditional chemical disciplines. Computational chemists now routinely tackle problems in organic, inorganic, analytical, materials, and biological chemistry, and the list goes on.

In many respects, progress in computational organometallic chemistry has traditionally lagged behind other areas, because it combines the inherent challenges of both organic and inorganic modeling. An organometallic compound, as the name implies, is made up of two chemical regions—a metallic ‘‘core’’ and an organic ‘‘coating’’—if I might be allowed a little poetic license. The organic coating is often characterized by its large size, large in terms of the number of atoms, orbitals, and/or conformational possibilities. It takes very few t- butyl substituents before a calculation on an organometallic compound becomes onerous! For the metallic core, i.e., the metal (or metals) and its surrounding inner coordination sphere, the inherent challenges for the computational organometallic chemist are different. Metals, particularly those of the d- and f-block, typically give rise to three main challenges in their chemical modeling: the large number of orbitals (many of them core), the so-called electron correlation problem (which is exacerbated by the presence of low-energy excited states), and relativistic effects for the heaviest metals.

Two techniques for dealing with these challenges, effective core potentials (or pseudopotentials) and density functional theory, have quickly transformed themselves from marginal techniques, once primarily the domain of solid-state chemists and physicists, to almost de rigueur standards for the computational

organometallic chemist. This is due in part to computational improvements but perhaps, more importantly, to the inclusion of these techniques into powerful, yet user-friendly, computational chemistry packages.

Another trend, and a very welcome one at that, in modern computational organometallic chemistry is in some respects a return to the roots of computational chemistry. In the Stone Age (at least according to some of the students who have worked in my research group), hardware and software limitations forced the utilization of less qualitative methodologies. This is best typified by the unparalleled work of Hoffmann and his colleagues employing extended Hu¨ckel methods. Much of this work spoke to the ‘‘how’’ and ‘‘why’’ of organometallic chemistry, with less concern for ‘‘how much.’’ For a while, it seemed that the only trend in computational organometallic chemistry was to be more quantitative, particularly for nongeometric quantities, such as reaction energies. This increase toward what some have termed chemical accuracy was certainly needed for the field to realize its full promise, but in many cases quantitative concerns overshadowed qualitative insight. Chapter 5, by Pierloot, shows that fundamental chemical insight, and not just accurate energies and bond lengths, can be extracted from even the most high-level calculations. Likewise, Chapters 3 and 6, by White and Maseras, respectively, tackle an age-old problem in chemistry, quantification of steric effects, in the former using molecular mechanics techniques and in the latter with hybrid quantum mechanics/molecular mechanics approaches.

In putting together this volume, the overriding theme was diversity—diver- sity of methods, diversity of applications, and diversity of chemistry. The ‘‘something for everyone’’ approach is not only an attempt to attract the largest possible audience for this book, but is also meant to highlight the amazing breadth and depth of computational organometallic chemistry. Chapters 2, 3, and 10, respectively by Norrby, White, and White and Douglass, focus primarily on classical (molecular mechanics) descriptions of chemical bonding. Of course, quantum mechanical approaches receive attention. Diedenhofen et al. (Chap. 4) and Gordon et al. (Chap. 11) address the accuracy of different quantum chemical techniques.

At one extreme of quantum chemical methodology lie approximate methods. Such techniques (for example, semiempirical quantum mechanics) typically involve great latitude in the number and type of approximations made to the full Schro¨dinger treatment. Approximations generally involve either the replacement of difficult-to-calculate quantities with experimental or theoretical estimates or the neglect of interactions (typically between electrons) thought to be of less chemical importance. Hence, the tradeoff for approximate methods is one of computational efficiency versus accuracy. The balance between accuracy and speed can be quite problematic for semiempirical quantum calculations on organometallic compounds because of the challenges discussed earlier for modeling metal species. The development or extension of any approximate method (molecular

mechanics included) has a prerequisite parameterization phase. In this process, one seeks to determine those parameters that maintain computational efficiency (not to mention realistic chemistry and physics) while maximizing the descriptive and predictive power of the model. Ideally, the parameterization process should take into account the full range of motifs that characterize a chemical family. One major issue in the parameterization of approximate methods for metal-containing species is therefore the development of a robust parameterization that can handle what our group has termed ‘‘chemical diversity.’’ Progress has been made in this field, not only for the molecular mechanics approaches alluded to earlier (see, for example, Chapter 2, by Norrby), but also for semiempirical quantum mechanics, as typified by the chapters of Taber (Chaps. 8, 9).

Chemical diversity can be defined as the ability of metals to stabilize distinct bonding environments involving different bond (e.g., dative, single, and multiple bonds) and ligating-atom (e.g., hard and soft donors) types, spin and formal oxidation states, coordination numbers, and geometries. Chapter 12, by Harvey, is an excellent example of the challenges inherent in modeling organometallic species and processes in which ‘‘spin flips’’ occur. As has become apparent, as computational organometallic chemists have explored all regions of the periodic table, this chemical diversity is also part and parcel of elements other than those of the transition series. This is plainly evident in the contributions by Kwon and McKee (Chap. 16) and McGee et al. (Chap. 15) on main group chemistry and by Li and Bursten (Chap. 14) on organoactinides.

It can be argued that the tremendous growth in the popularity of research into organometallic chemistry, experimental and computational, is due in large part to their utility in industrial and academic applications. As the field of computational organometallic chemistry has matured it has become evident that it is the chemical diversity that characterizes these entities that gives rise to many of the challenges in their reliable and rapid modeling. One need only consider some of the myriad catalytic transformations involving organometallic species to appreciate the chemical gymnastics that alter oxidation states, coordination numbers, ligand types, etc. Thus, it is this very property of chemical diversity that makes organometallics so very interesting (and at times quite frustrating) as computational targets.

In putting together this volume, the traditional description of organometallics as entities with a metal–carbon bond has been expanded to include any entities with an organic and metallic functionality, whether they be joined by a direct metal–carbon bond or not. I have also tried to go beyond applications other than just those related to industrial catalysis, as admirably demonstrated by Czerw et al. in Chapter 13. Chapter 7, by Parrill and coworkers, with its biomedical bent, is a good demonstration of this philosophy, as are Chapters 8 and 9, by Taber et al., on the computer-aided design of organometallic catalysts for carrying out useful organic synthetic transformations.

I would like to conclude this introductory chapter by thanking the chapter authors, individually and as a group, for their good humor and spirit, particularly in dealing with the inadequacies of a first-time editor. I would also like to thank Anita Lekhwani (Acquisitions Editor), Moraima Suarez (Production Editor), and Jennifer Paizzi (Administrative Assistant) of Marcel Dekker for their encouragement and for answering my numerous questions. Much of the planning for Computational Organometallic Chemistry occurred while I was on a Professional Development Assignment (PDA), for which opportunity I am grateful to The University of Memphis College of Arts and Sciences and Chemistry Department. I’d also like to thank the Chemistry Department at Bristol University (UK), for providing a relaxing yet stimulating environment during this PDA, and the United States National Science Foundation Office of International Programs, for their support of travel between Memphis and Bristol. It would not have possible to become an ‘‘expert’’ (real or imagined) in computational organometallic chemistry without the hard work and dedication of a fabulous bunch of graduate and undergraduate research students at The University of Memphis. I thank the various agencies (American Chemical Society—Petroleum Research Foundation, Los Alamos National Laboratory, National Science Foundation, and U.S. Department of Energy) for their generous support of these students during their careers at The University of Memphis.

Saving the best for last, I would like to thank my lovely wife, Mary Anderson, for her support, suggestions, and spirited Texan ways. She has done more than help improve this monograph; she has improved my life in immeasurable ways. For these reasons, I dedicate this volume to her.

Finally, I take full responsibility for any errors of commission or omission that may exist in this volume.