Occupancies of "Frontier Orbitals" of Compound I; Nature of Nonbonding Electrons in the Fe=O Oxygen

These diagrams show the orbital occupations of a commonly-encountered electronic state of perferryl heme (the quartet A2u state of compound I), and many of the "orbital occupancies" are, in other iron(IV) or iron(III) hemes, the same as these or similar to these. The nature of the "lone pair" of oxygen has been a source of confusion in articles, and it's almost certainly a sigma-type bonding interaction of the 2p(z) atomic orbital of oxygen and the 3d(z2) atomic orbital of iron, with minimal 3d(z2) character and somewhat minimal bonding character (i.e., a "nonbonding MO"). [Also, I forgot to put it on the diagram, but the formal charge on oxygen, in this species, would be 6 - 2 (lone pair electrons) - 1 [nonbonding radical in pi*(xz or yz) MO] - 1/2(4 electrons in pi(xz) and pi(yz) bonding MO's) = +1. I think the formal charge can be +2 or something else, in some of the transition states, but that's not the actual charge. For example, in one of the major iron(II)-heme species, porphyrin-Fe(II)-O-O(-), the actual charge is about -0.2 and not the formal charge of -1, apparently [found experimentally, as reported in the abstract of Jensen et al., 2005: (http://www.ncbi.nlm.nih.gov/pubmed/15598490)(http://www.cicum.cup.uni-muenchen.de/ac/kluefers/homepage/L/BAC/heme1.pdf)]. With regard to the oxygen's lone-pair electrons, though, there may be some 2s character, too, and part of the confusion stems from the fact that, in dioxygen, the "nonbonding MO's" have predominantly 2s character and mix to only a minimal extent with the 2p(z) atomic orbitals of oxygen (they're not "sp3" orbitals) [this article discusses the absence of the supposed "rabbit ears," produced by sp3-hybridized, "lone-pair" MO's in the water molecule: Laing, 1987: (http://pubs.acs.org/doi/abs/10.1021/ed064p124), and this article includes a discussion of nonbonding MO's in general, I think: Hurst et al., 1999: (http://pubs.acs.org/doi/abs/10.1021/jp984565h)]. But other articles refer to dioxygen's nonbonding MO's as being the pi*(yz) and pi*(xz) antibonding MO's, which are singly-occupied in ground-state triplet dioxygen (there's also at least one excited triplet state of dioxygen). There's no single way to determine, without looking at experimental data [data from the use of different types of photoelectron spectroscopy (PES): (http://scholar.google.com/scholar?hl=en&q=%22bonding+character%22+%22lone+pair%22+nonbonding+molecular+orbital+%22photoemission+spectroscopy%22+OR+%22photoelectron+spectroscopy%22&btnG=Search&as_sdt=100000001&as_ylo=&as_vis=0)], the extent to which each of the various MO's exhibits "bonding character", within a given molecule. In general, the highest-occupied molecular orbitals (HOMO's, meaning the HOMO, the HOMO-1, the HOMO-2, etc.), which may be either bonding or antibonding MO's, are the nonbonding ("lone-pair") MO's [the highest-energy MO's, in which there is more separation of electron density throughout different parts of a molecule (more "charge separation," for example, in resonance structures within a given MO--not as much delocalization of electrons)]. But in a lot of nitrogen-containing compounds, there may be essentially no MO's that exhibit nonbonding character, as revealed by PES. Anyway, these supplementary data show some of the strange pi- and sigma-type MO's that describe the bonding interactions in the Fe(II)-O-O moiety of iron(II)-heme bound to dioxygen [the two MO's on the right (the 2nd and 3rd) of p.7 (the 2nd one is probably most similar to the "lone-pair" sigma-type interaction in the FeO moiety of iron(IV)-heme, as sketched in the second-to-last diagram below, but those MO's on p.7 of that pdf are pi-type interactions, not sigma-type interactions, as I've shown): (http://www.rsc.org/suppdata/CC/b7/b704871h/b704871h.pdf)]. Anyway, transition metals will tend ("prefer") to "obtain," via overlap with and donation from ligands, 18 electrons in organometallic complexes [the "18-electron rule" (http://scholar.google.com/scholar?hl=en&q=%2218+electron+rule%22+transition+metal&btnG=Search&as_sdt=100000001&as_ylo=&as_vis=0)], and the rest of the electrons are in other MO's of the O-Fe-S moiety and MO's of the porphyrin ring that are shared with the heme iron(IV) (or iron in another oxidation state). In one article that describes orbital occupancies of an iron(II)-heme, the iron(II) shares a total of 20 electrons with the porphyrin ring and the distal (i.e., O2) and proximal (i.e., histidine or cysteine) ligands. I've only shown the electrons that are in the HOMO's, or "frontier orbitals," whose occupancies are most likely to shift during the courses of Cytochrome P450-enzyme-catalyzed reactions, etc. Note that the diagrams should say "3pi(xz)" or "3pi(yz)" constructive overlap of the sulfur's 3px or 3py atomic orbitals (with the porphyrin's a2u molecular orbital).



These diagrams show the higher-energy steric (i.e., repulsion by electron clouds), as the porphyrin molecular orbitals are interacting, via the nitrogens, with the 3d(xy) atomic orbital of iron, in Coordinate System 1 [shown below and, in terms of the relative energy levels of the d(xy) and d(x2-y2) MO's of transitional electronic states of hydrogen abstraction reactions catalyzed by compound I, depicted in Ogliaro et al., 2000, p. 8984, Scheme 4: (http://pubs.acs.org/doi/abs/10.1021/ja991878x)(http://theochem.chem.rug.nl/~filatov/Pubs/JACS_122_8977.pdf)], or, in Coordinate System 2 [not shown below but shown above, in the diagrams of the "frontier-orbital" occupations of commonly-encountered iron(IV)-heme species], with the 3d(x2-y2) atomic orbital of iron. In coordinate System 2, the relative energy levels of the d(x2-y2) and d(xy) atomic orbitals [and the corresponding delta(d(x2-y2)) and delta(d(xy)) MO's of heme] are reversed, in comparison to the situation in Coordinate System 1. Another potential source of confusion is that the delta(3d(x2-y2)) MO of heme is sometimes referred to, in diagrams, as d(x2-y2) (as if it's an atomic orbital of iron), and the pi*(xz) and pi*(yz) MO's of heme are often depicted as being "d(xz)" and "d(yz)" (parentheses mean subscripts in the blogger software that I'm using to type this) orbitals (as if they're atomic orbitals of iron). The pi(xz) and pi(yz) bonding MO's are not-infrequently referred to as being "pi orbitals" or the like [see Lehnert et al., 2001, p. 8289, Scheme 4: (http://www.ncbi.nlm.nih.gov/pubmed/11516278)]. It's always clear that these are MO's, but it's potentially a source of confusion, in my opinion. I have an article in which the authors specifically address this issue of the two different coordinate systems, but I can't find it on my computer. It's potentially a serious source of confusion, partly because there is sometimes no mention of the assignment, of either one or the other coordinate system, that the authors of various articles might be making, in their computer modeling of electron distributions in MO's ("probability density functions," as in "density functional theory," or DFT) or the like.





These are approximations for the "lone-pair" MO (sigma(d(z2)), etc.) of oxygen, a MO of heme that is likely to exhibit minimal 3d(z2) character, and the frequently-unoccupied sigma*(d(z2)) (antibonding) MO of heme that becomes occupied during charge-transfer states (ferryl-to-ring and thiolate ligand-to-ring and substrate-to-oxo transitions during high-energy, intermediate electronic states, such as in the electronic states of transition states of overall Cytochrome P450-dependent reactions, or during the multitude of "configuration interaction orbitals" that form during the absorption of visible or UV wavelengths by heme, etc.):