Chemistry 380.37 Fall 2015 Dr. Jean M. Standard August 17, 2015

Introduction to Computational Chemistry What is computational chemistry? Computational chemistry involves the use of computers to model the structure, energy, physical and chemical properties, and reactivity of atomic and molecular systems. Computational chemistry may involve the study of individual molecules at the microscopic level, or it may involve the simulation of the bulk properties of molecular systems at the macroscopic level. It may involve modeling of gas, solution, or solid phase systems, or even processes at the interface between phases (for example, adsorption of a gas on a solid substrate). In which areas of chemistry is computational chemistry used? Computational chemistry methods are employed in all areas of chemistry: analytical, inorganic, organic, and physical chemistry, as well as biochemistry. In addition, computational chemistry is very prevalent in other subfields of chemistry, including materials and polymer chemistry. Journal articles in all areas publish research that involves computational chemistry. Attached on the last several pages are some samples of title pages of some articles from the ISU Chemistry Department published in scientific journals during the last several years. These samples are just a few out of the numerous articles containing computational chemistry research to appear in current journals. Computational chemistry is truly widespread in its utility as a predictive and analytical tool in all areas of chemistry. Where are computational chemists employed? Computational chemists are hired by academic institutions, government agencies, and all types of industries. The pharmaceutical industry in particular has embraced computational chemistry as an effective tool in the design of new drugs. Computational chemists are employed to study all types of chemical processes from catalytic reactions occurring on metal surfaces to gas phase reactions involved in destroying the earth’s stratospheric ozone layer. Listed on the last page of this handout are a few job advertisements for computational chemists found in previous issues of Chemical and Engineering News. These listings are just a random sampling of the types of job opportunities available for computational chemists. What sort of educational preparation do computational chemists need? Computational chemistry often requires a Ph.D. However, more and more companies today also are hiring computational chemists at the Master’s level. Students who want to pursue a degree in computational chemistry should generally start with some extra courses in math (differential equations is particularly helpful) as well as at least one course in a computer programming language (C or Fortran are the most useful). Are there advanced degrees awarded in computational chemistry? Most universities do not award a specific Ph.D. degree in computational chemistry. A Ph.D. degree is usually awarded in physical chemistry for a student carrying out research under the direction of a physical chemistry faculty member who specializes in computational chemistry; however, faculty members with specializations in computational chemistry also may be found in the other subdisciplines of chemistry. Some universities have special programs in computational chemistry and computer visualization. Two examples are Wayne State University and Northeastern University. These programs may lead either to certificates or Master’s degrees in computational chemistry. What is molecular modeling? Molecular modeling is sometimes used as a synonym for computational chemistry. However, computational chemistry is often meant as a broader term that includes computer simulation of molecular systems at both the microscopic and macroscopic levels. Molecular modeling may often refer solely to computer simulation of systems at the microscopic level.

2

Brief Description of Methods Covered in CHE 380.37 I. Definitions and Terms A. What is molecular modeling?

• the generation, representation, analysis, and prediction of molecular structures, properties, interactions, and reactions via computer simulation.

B. What is the purpose of molecular modeling?

• predictive tool • analytic tool

C. Who uses molecular modeling?

• molecular modeling is employed in every area of chemistry • molecular modeling is used in academia, government, and industry

II. Force Field Methods - Molecular Mechanics A. Description of the Method

• Molecular mechanics (MM) is based on an empirical force field representation of the interactions between atoms.

• A force field is made up of stretching, bending, torsional, electrostatic, and nonbonded interactions which yield the energy of the system as a function of atomic coordinates.

• The force field is parametrized using experimental properties such as heats of formation, vibrational frequencies, and dipole moments.

B. Utility and Applications

• MM is particularly useful for large molecular systems (biomolecules, polymers) where quantum calculations are not possible.

• MM may be used for prediction of relative energies of conformers. III. Conformation searching A. Description of the Method

• Conformation searching involves calculations to find the lowest energy conformer of a molecule (or the lowest conformers within a specified energy range).

• An empirical force field is usually employed to represent the interactions between atoms. However, quantum mechanical methods may also be used.

• Several methods, including systematic, Monte Carlo, distance geometry, and molecular dynamics, are employed to carry out conformer searches.

B. Utility and Applications

• Conformation searching is useful for large, flexible molecules with many low-energy conformers.

3 IV. Quantum Mechanical Methods A. Description of the Methods

• All the methods which employ quantum mechanics (QM) are based on solving the Schrödinger equation (to some level of approximation) for the molecular system of interest.

• Ab initio ("from the beginning") methods involve no empirical parameters and therefore are the most accurate techniques (and the most expensive computationally).

• Semiempirical methods rely on parametrization of some of the integrals that occur in the solution of the Schrödinger equation using experimental data.

• Density functional methods are based on the specification of a certain functional form for the electron density in the molecule.

B. Utility and Applications

• QM is most useful for smaller molecular systems, although semi-empirical methods have been used for larger molecules. Density functional theory has proven especially useful not only for organic molecules but also for molecules containing transition metals.

• Molecular geometries, energies, electron density, orbitals, and other properties (such as vibrational frequencies) may be determined.

• Transition states, reaction paths, and mechanisms may be studied. V. Molecular Dynamics A. Description of the Method

• In molecular dynamics (MD), Newton's classical equations of motion are used to solve for the motion of the atoms as a function of time.

• An empirical force field can be employed to represent the interactions between atoms. Some recent studies have employed quantum mechanical force fields, or a mixture of classical and quantum mechanics.

B. Utility and Applications

• Like molecular mechanics, MD is useful for large systems such as biomolecules and polymers where quantum mechanics is prohibitive.

• MD is useful in the study condensed phase systems, such as biomolecules in the presence of explicit solvent molecules.

• MD is also good for looking at condensed phases of smaller systems but with a large number of particles (e.g. liquid Ar or H2O).

• MD yields not only information about how the molecular configuration varies in time, but it may be used to calculate thermodynamic information, such as free energies, heat capacity, etc.

• MD can be used as a searching tool to find lowest energy conformation in large molecular systems.

Published: January 31, 2011

r 2011 American Chemical Society 1243 dx.doi.org/10.1021/jp107688v | J. Phys. Chem. A 2011, 115, 1243–1249

ARTICLE

pubs.acs.org/JPCA

Multireference Configuration Interaction Study of BromocarbenesJean M. Standard,* Rebecca J. Steidl, Matthew C. Beecher, and Robert W. Quandt

Department of Chemistry, Illinois State University, Normal, Illinois 61790-4160, United States

ABSTRACT: Multireference configuration interaction (MRCI) calculations of the lowestsinglet ~X(1A0) and triplet ~a(3A00) states as well as the first excited singlet ~A(1A00) state havebeen performed for a series of bromocarbenes: CHBr, CFBr, CClBr, CBr2, and CIBr. TheMRCI calculations were performed with correlation consistent basis sets of valence triple-ζplus polarization quality, employing a full-valence active space of 18 electrons in 12 orbitals(12 and 9, respectively, for CHBr). Results obtained include equilibrium geometries andharmonic vibrational frequencies for each of the electronic states, along with ~a(3A00) r~X(1A0) singlet-triplet gaps and ~A(1A00)r ~X(1A0) transition energies. Comparisons have been made with previous computational andexperimental results where available. The MRCI calculations presented in this work provide a comprehensive series of results at aconsistent high level of theory for all of the bromocarbenes.

I. INTRODUCTION

It has long been known that chlorine-containing hydrocarbonshave large ozone depletion potentials (ODPs) due to consider-able release rates and long atmospheric lifetimes.1 It would beexpected that bromine-containing halocarbons, with release ratesand atmospheric lifetimes that are generally much smaller thantheir chlorine-containing counterparts, would have negligibleODPs. However, unstable reservoir molecules and synergisticeffects with chlorine lead to much greater than expected ODPsfor these species.2 Indeed, it has been estimated that, on a peratom basis, bromine is almost 60 timesmore destructive to ozonethan chlorine.3 These larger than expected ODPs have led to arenewed interest in bromine-containing hydrocarbons in recentyears. It has been found that, in addition to better known sourcessuch as Halons and methyl bromide, bromoform is a significantsource of reactive bromine in the stratosphere.4-6 Dependingupon the excitation wavelength, photodissociation of bromocar-bons such as bromoform can follow two different routes:dissociation into atomic bromine and a substitutedmethyl radicalor dissociation into molecular bromine and a singlet carbene. Forexample, McGivern et al. observed that at 193 nm the primaryphotoproduct was atomic bromine.7 They also observed second-ary dissociation of the excited photoproduct, CHBr2*, to formCBr and HBr. Xu et al. observed the formation of atomic halogenin both the 2P1/2 and

2P3/2 spin states upon excitation at 234 and267 nm.8 They also found significant formation of molecularbromine with branching ratios into that channel of 0.16 and 0.26at 267 and 234 nm, respectively. Quandt and co-workers recentlystudied the 2 ! 193 nm photodissociation of CBr4 and CHBr3via photoproduct emission.9 Observed emission was attributedto the Swan system (d3Πgf a3Πu) of C2, which was formed viareaction of electronically excited radicals, CH(A2Δ) and CBr-(A2Δ). In addition, formation of CBr2 (or CHBr) and Br2 wasimplied by secondary evidence. A computational study of thisdark channel was undertaken, and the results showed the pre-sence of three transition states and an ion-pair isomer inter-mediate for both CBr4 and CHBr3 dissociation.

A number of previous computational studies have been per-formed on bromocarbenes. A comprehensive study of the lowestsinglet ~X(1A0) and triplet ~a(3A00) states of all halocarbenes wascarried out by Schwartz and Marshall in 1999 in which equilib-rium geometries, harmonic vibrational frequencies, and ~a(3A00)r ~X(1A0) singlet-triplet gaps obtained at the QCISD/6-311G-(d) level were reported.10 A few years later, Drake et al. employedCISD, CASSCF, and CASPT2 methods along with basis sets ofDZP quality to determine geometries and adiabatic ~A(1A00) r~X(1A0) transition energies for the ground ~X(1A0) and first excited~A(1A00) singlet states of a series of bromo- and iodocarbenes.11 Inparticular, theCASPT2(18,12)/DZPmethodwas shown to providea good balance of computational cost and predictive accuracy for theadiabatic ~A(1A00) r ~X(1A0) transition energies.11 CASSCF andCASPT2 calculations also have been employed in other studies ofthe bromocarbenes CFBr 12 and CBr2

13 in order to obtain equili-brium geometries, harmonic vibrational frequencies, and other spec-troscopic parameters for the ground and first excited singlet states aswell as the lowest triplet state.

Higher level multireference configuration interaction (MRCI)calculations have been previously completed only for CHBr,14,15

CFBr,12 andCClBr.16 Inwork byYu et al.,14MRCI calculationswerecarried out on CHBr with a cc-pVTZ basis set using state-averagedfull-valence active space CASSCF reference functions. Equilibriumgeometries and a detailed analysis of the potential surfaces of theground ~X(1A0) and excited ~A(1A00) singlet states as well the lowest~a(3A00) triplet state of CHBr were presented. In recent work byBurrill and Grein,15 a TZP basis set with polarization and diffusefunctions was employed in order to carry out MRCI calculations onthe lowest six singlet and triplet electronic states of CHBr. For CFBr,the MRCI method with an active space of two electrons in twoorbitals, MRCI(2,2), and a TZP quality basis set was employed tostudy only the first excited ~A(1A00) singlet state.12 For CClBr, asimilar MRCI(2,2) study was carried out to determine the geometry

Received: August 13, 2010Revised: December 31, 2010

8164 J. Org. Chem. 2009, 74, 8164–8173 Published on Web 10/07/2009 DOI: 10.1021/jo9016474r 2009 American Chemical Society

pubs.acs.org/joc

β-Amino Alcohol Derived β-Hydroxy- and β-(o-Diphenylphosphino)benzoyloxy(o-diphenylphosphino)benzamides: AnEster-Amide Ligand Structural Model for the Palladium-Catalyzed

Allylic Alkylation Reaction

Geetanjali S. Mahadik, Stanley A. Knott, Lisa F. Szczepura, Steven J. Peters,Jean M. Standard, and Shawn R. Hitchcock*

Department of Chemistry, Illinois State University, Normal, Illinois 61790-4160

hitchcock@ilstu.edu

Received July 29, 2009

A commercially available collection of β-amino alcohols have been converted to their correspondingβ-hydroxy- and β-(o-diphenylphosphino)benzoyloxy(o-diphenylphosphino)benzamides 11a-f and12a-f and have been employed in the Tsuji-Trost asymmetric alkylation reaction with 1,3-diphenyl-propenyl acetate. With the exception of ligands 11b and 11f, the β-hydroxybenzoyloxy(o-diphenylphosphino)benzamide ligands 11a-fprimarily afforded the (R)-enantiomerof the product.In contrast, the bis(phosphine) ligands 12a-f consistently afforded the (S)-enantiomer. The best ligand(12c) was derived from cis-(1R,2S)-2-amino-1,2-diphenyl-1-ethanol, and when applied in the asym-metric allylic alkylation reaction, it yielded the product in an enantiomeric ratio of 97.8.22 favoring the(S)-enantiomer. A computational study was conducted on the conformation that this ligand mightadopt in the palladium-catalyzed alkylation reaction as compared to that of the Trost ligand 1a.

1. Introduction

The palladium-catalyzed asymmetric allylic alkylation reac-tion1 knownas theTsuji-Trost reactionhas been the subject ofintense studies2 directed toward the design, synthesis, andapplication of a myriad of chiral, nonracemic ligand scaffolds.Of the ligands that have been prepared and applied in this

reaction, the Trost modular phosphine ligands (1a,b)3 haveproven tobe the benchmark for evaluating the efficacy of newlydeveloped phosphine ligands in the asymmetric allylic alkyla-tion reaction. In fact, on the basis of their successful use in avariety of applications, many of these modular ligands arecommercially available.4 The success of these ligands hasencouraged the synthesis and application of structurally novelphosphines based on the binaphthyl type ligands,5 tartratederived systems,6 carbohydrates,7 paracyclophanes,8 and(1) Tsuji, J. InPalladiumReagents and Catalysts: NewPerspectives for the

21st Century; John Wiley & Sons: Chichester, UK, 2004; pp 431-518.(2) (a) Lu, Z.; Ma, S. Angew. Chem., Int. Ed. 2008, 47, 258. (b) Trost, B.

M. J. Org. Chem. 2004, 69, 5813. (c) Trost, B.M.; Crawley,M. L.Chem. Rev.2003, 103, 2921. (d) Trost, B.M.;Machacek,M. R.; Aponick, A.Acc. Chem.Res. 2006, 39, 747. (e) Trost, B.M.; VanVranken, D. L.Chem. Rev. 1996, 96,395. (f) Trost, B.M.Acc. Chem. Res. 1980, 13, 385. (g) Trost, B.M.; Breit, B.;Organ, M. G. Tetrahedron Lett. 1994, 35, 5817.

(3) (a) Trost, B. M.; Van Vranken, D. L.; Bingel, C. J. Am. Chem. Soc.1992, 114, 9327. (b) Trost, B.M.; VanVranken, D. L.Angew. Chem., Int. Ed.1992, 31, 228.

(4) (a) Trost, B. M.; Fandrick, D. R. Aldrichim. Acta 2007, 40, 59.(5) Feng, J.; Bohle, D. S.; Li, C. Tetrahedron: Asymmetry 2007, 18, 1043–

1047.(6) Marques, C. S.; Burke, A. J. Tetrahedron: Asymmetry 2007, 18, 1804–

1808.(7) Khiar, N.;Navas, R.; !Alvarez, E.; Fern!andez, I. ARCHIVOC 2008, 8,

211-224.(8) Jiang, B.; Lei, Y.; Zhao, X. J. Org. Chem. 2008, 73, 7833–7836.

Metal-Olefin Bond Energies in M(CO)5(C2H4-nCln) M ) Cr, Mo, W; n ) 0-4:Electron-Withdrawing Olefins Do Not Increase the Bond Strength

Darin N. Schlappi and David L. Cedeno*Department of Chemistry, Illinois State UniVersity, Box 4160, Normal, Illinois 61790-4160

ReceiVed: March 26, 2009; ReVised Manuscript ReceiVed: July 17, 2009

Metal-olefin bond dissociation enthalpies have been calculated for the series of complexes M(CO)5(C2H4-nCln),M ) Cr, Mo, W; n ) 0-4 using density functional theory. Experimental values of the bond enthalpies havebeen measured for M(CO)5(C2H4-nCln) M ) Cr, Mo, W; n ) 2 (vinyl chloride), 3, and 4 using laserphotoacoustic calorimetry in n-hexane solution. Experimental and calculated values indicate that the trend inmetal-olefin bond energies is opposite to the electron-withdrawing ability of the olefin, which is counter toexpectations based on the Dewar-Chatt-Duncanson model for metal-olefin bonding. An in-depth analysisof the metal-olefin interaction using a bond energy decomposition scheme implies that the observed andcalculated decreasing trend is influenced by the increase in steric interactions and olefin reorganizationalenergy which is concomitant to the increase of the number of electron-withdrawing halogen atoms.

Introduction

Many significant chemical processes such as olefin hydro-genation, isomerization, hydrocarbonlyation, hydroformylation,polymerization, and metathesis among others are driven by thepresence of a metal catalyst and involve the formation of anintermediate that contains a metal-olefin bond.1-6 It is beneficialto be able to synthesize catalysts for these reactions that arefine-tuned to the needs of a particular reaction or process becausethe use of olefins and olefin-related products in industry hasbecome prevalent. The ability to control the properties of thesecatalysts relies heavily on a complete understanding of thethermodynamic factors that influence the strength of the bondbetween a given metal complex and an olefin.

Contributing to a level of understanding that would allowfor an accurate prediction of the bond strength between a metalcomplex and an olefin is the primary goal of our research. Thecurrent picture of metal-olefin bonding is based on frontiermolecular orbital theory introduced by Dewar in 19517 andexpanded by Chatt and Duncanson in 1953.8 The approach isknown as the Dewar-Chatt-Duncanson (DCD) model ofmetal-olefin bonding. The DCD model details the metal-olefinbond as being a two way synergistic electron exchange betweena metal complex and an olefin. The bond consists of a σinteraction in which the highest occupied molecular orbital(HOMO) of the olefin donates electron density to an empty dσorbital on the metal complex. Additionally, there is a π bondinginteraction in which the metal donates electron density back tothe olefin from an occupied dπ orbital to the unoccupiedantibonding π* orbital of the olefin.

The electron population changes in the π and π* orbitals ofthe olefin have the physical consequence of decreasing the bondorder of the carbon-carbon double bond. This is equivalent toa partial sp2 to sp3 rehybridization of the olefinic carbons thatcauses the lengthening of the CdC bond and the back-bendingof the substituents around the CdC bond away from the metalcomplex and outside of the plane of the CdC bond. The DCDmodel has been commonly used to rationalize the bondingstrength between a metal complex and an olefin.2,5,9 A wide-

spread expectation of this rationalization is that for some metalsthe π (or back-) bonding interaction is the dominating contribu-tion to the metal-olefin bond. Therefore, if the hydrogen atomsin ethylene were to be replaced with a more electron-withdraw-ing substituent such as a halogen (X ) F, Cl), then the back-bonding would increase because a halogenated ethylene is abetter π acceptor than ethylene. Based on this rationalization,the bonding energies between a metal and the olefin series C2H4,C2F4, and C2Cl4 would be a function of the electron-withdrawingability of the substituents around the CdC bond of the olefinand decrease in the order C2F4 > C2Cl4 > C2H4. Experimentaldata on Cr(CO)5(C2X4) (where X ) H, F, Cl) indicate, however,that the metal-olefin bond strength follows the trend Cr-C2H4

> Cr-C2F4 > Cr-C2Cl4.10 Cedeno and Weitz10 carried out aDensity Functional Theory (DFT) study of the seriesFe(CO)4(C2X4) and Cr(CO)5(C2X4) where X ) H, F, Cl, thatprovides an explanation for the discrepancy between theexperimental data and the expectations based on the DCDmodel.

Along the same line of reasoning, the DCD model may beused to predict the trend of the metal bond strength for the olefinseries C2H4-nXn (X ) F or Cl). Given that the back-bondingability of an olefin is enhanced as halogenation increases, thenthe metal-olefin bond strength should increase with an increasein the number of halogens. Back in 1974, Tolman11 determinedthe equilibrium bonding constants between bis(tri-o-tolyl phos-phite)nickel(0) and 38 different olefins including the C2H4-nFn

(n ) 0-4) series. In his paper, Tolman found that none of thefluoro olefins examined (with the exception of CH2dCHCF3)were as good as C2H4 in coordinating to nickel(0), a surprisingresult that was out of line with his expectation. Tolman hintedthat the reason for the inadequacy of the DCD picture ofmetal-olefin bonding was due to the reorganization that occursin the olefin as the carbons of the double bond are forced torehybridize from sp2 to sp3. A computational DFT study bySchlappi and Cedeno12 examined the bonding of the olefinsC2XnH4-n (X ) F or Cl, and n ) 0-4) to Ni(PH3)2(CO). It wasfound that the olefins bound to the nickel with dissociationenergies that follow a trend very similar to the one shown inTolman’s study and confirmed his presumption. We concluded* Corresponding author. Email: dcedeno@ilstu.edu.

J. Phys. Chem. A 2009, 113, 9692–96999692

10.1021/jp9027468 CCC: $40.75 © 2009 American Chemical SocietyPublished on Web 08/07/2009

Catalytic Dioxygen Activation by(Nitro)(meso-tetrakis(2-N-methylpyridyl)porphyrinato)cobalt(III) CationDerivatives Electrostatically Immobilized in Nafion Films: AnExperimental and DFT Investigation

John A. Goodwin,*,† Jennifer L. Coor,† Donald F. Kavanagh,† Mathieu Sabbagh,† James W. Howard,†John R. Adamec,† Deidre J. Parmley,† Emily M. Tarsis,† Tigran S. Kurtikyan,‡Astghik A. Hovhannisyan,‡ Patrick J. Desrochers,§ and Jean M. Standard|

Department of Chemistry and Physics, Coastal Carolina UniVersity, P.O. Box 261954,Conway, South Carolina 29526-6054, Molecular Structure Research Centre, National Academy ofSciences, YereVan, Armenia, Department of Chemistry, UniVersity of Central Arkansas,Conway, Arkansas 72035, and Department of Chemistry, Illinois State UniVersity,Normal, Illinois 61790

Received January 15, 2008

Complexes of the (nitro)(meso-tetrakis(2-N-methylpyridyl)porphyinato)cobalt(III) cation, [LCoTMpyP(2)(NO2)]4+, in which L )water or ethanol have been immobilized through ionic attraction within Nafion films (Naf). These immobilized six-coordinatespecies, [LCoTMPyP(2)(NO2)/Naf], have been found to catalyze the oxidation of triphenylphosphine in ethanol solution bydioxygen, therefore retaining the capacity to activate dioxygen catalytically without an additional reducing agent as was previouslyobserved in nonaqueous solution for the non-ionic (nitro)cobalt porphyrin analogs. Heating these immobilized six-coordinatespecies under vacuum conditions results in the formation of the five-coordinate nitro derivatives, [CoTMPyP(2)(NO2)/Naf] at 85°C and [CoTMPyP(2)/Naf] at 110 °C. The catalytic oxidation of gas-phase cyclohexene with O2 is supported only by theresulting immobilized five-coordinate nitro complex as was previously seen with the corresponding solution-phase catalyst indichloromethane solution. The simultaneous catalytic oxidation of triphenylphosphine and cyclohexene with O2 in the presenceof the Nafion-bound six-coordinate ethanol nitro complex is also observed; however, this process is not seen for the CoTPPderivative in dichloromethane solution. The oxidation reactions do not occur with unmodified Nafion film or with Nafion-supported[BrCo(III)TmpyP]/Naf or [Co(II)TmpyP]/Naf, indicating the necessity for the nitro/nitrosyl ligand in the oxidation mechanism. Theexistence of a second reactive intermediate is indicated because the two simultaneous oxidation reactions depend on twodistinct oxygen atom-transfer steps having different reactivity. The absence of homogeneous cyclohexene oxidation by thesix-coordinate (H2O)CoTPP(NO2) derivatives in the presence of Ph3P and O2 in dichloromethane solution indicates that thesecond reactive intermediate is lost by an unidentified route only in solution, implying that the immobilization of it in Nafionallows it to react with cyclohexene. Although direct observation of this species has not been achieved, a comparitive DFT studyof likely intermediates in several catalytic oxidation mechanisms at the BP 6-31G* level supports the possibility that this intermediateis a peroxynitro species on the basis of relative thermodynamic accessibility. The alternate intermediates evaluated include thereduced cobalt(II) porphyrin, the dioxygen adduct cobalt(III)-O2

-, the oxidized cobalt(II) π-cation radical, and the nitrito complex,cobalt(III)-ONO.

Introduction

Development of robust heterogeneous catalysts for activa-tion of molecular oxygen is important for a wide range ofapplications including environmentally benign synthesis,water purification, and the oxygen reduction reaction in fuel-cell technology.1 Many heterogeneous catalysts based onmetalloporphyrins that are effective in electrocatalytic andphotocatalytic oxidations with dioxygen have been studied,

and porphyrin-catalyzed oxidation reactions using peroxides,N-oxides, and other oxidants or co-reductants are also wellknown.2 A relatively small subset of these metalloporphyrinsystems carries out catalytic oxidation reactions with mo-lecular oxygen as the oxidant in the absence of a co-reductant; although some of these have been shown to occurby auto-oxidation radical-chain mechanisms.3 The use ofmetalloporphyrin catalysts in supported heterogeneous sys-tems4,5 and in microporous porphyrin assemblies6 has alsobeen widely investigated.1

The reactivity of five-coordinate (nitro)cobaltporphyrinsin the catalytic oxidation of alkenes, apparently throughsecondary oxo-transfer from the coordinated nitro ligand, was

* To whom correspondence should be addressed. E-mail:jgoodwin@coastal.edu.

† Coastal Carolina University.‡ National Academy of Sciences.§ University of Central Arkansas.| Illinois State University.

(1) Centi, G.; Misono, M. Catal. Today 1998, 41, 287–296.

Inorg. Chem. 2008, 47, 7852-7862

7852 Inorganic Chemistry, Vol. 47, No. 17, 2008 10.1021/ic8000762 CCC: $40.75 © 2008 American Chemical SocietyPublished on Web 07/30/2008