Doi:10.1016/j.elspec.2003.11.007

Journal of Electron Spectroscopy and Related Phenomena 135 (2004) 1–5 The molecular and electronic structure of N,N -ethylenebis(acetylacetonylideiminato)oxovanadium(IV) and the electronic structure of its thio analogue Barry L. Westcott , Guy Crundwell , Thomas R. Burkholder , Laura A. Michelsen , Caroline B. Gardner , Nadine E. Gruhn , Allen D. Hunter , Penny Miner , Mathias Zeller a Department of Chemistry, Central Connecticut State University, 1615 Stanley Street, New Britain, CT 06050, USA b The Center for Gas Phase Electron Spectroscopy, Department of Chemistry, The University of Arizona, Tucson, AZ 85721, USA c Department of Chemistry, Youngstown State University, Youngstown, OH 44512, USA Received 30 August 2003; received in revised form 21 November 2003; accepted 24 November 2003 Abstract
The electronic structures of the title complexes—VO(acen) and VS(acen)—and the free H2(acen) ligand are probed using gas-phase UV-photoelectron spectroscopy [acen = N,N -ethylenebis(acetylacetonylideiminato)]. The effect of the different axial donors on the metalcenter is examined, as is the effect that the oxo and thio ligands have on the acen orbitals. We find that the oxo and thio donors primarily affectthe metal center and that the ligand periphery remains mostly unchanged.
2003 Elsevier B.V. All rights reserved.
Keywords: Photoelectron spectroscopy; Vanadium; Vanadyl; Acen; Petroleum 1. Introduction
technique gives a picture of the molecular orbital energiesand, consequently, insight into the bonding and reactiv- Complexes of vanadium can account for over 75% of ity of complexes. Previously we reported the electronic the transition metal impurities found in crude petroleum structure of VO(oep) and VO(pc) [oep: octaethyl por- phyrin, and pc: pthalocyanin] and complexes of the been identified as derivatives of vanadyl porphyrins type VO(ac) [ac = 2,4-pentanedione and its fluorinated and and complexes of this type have been the primary focus alkylated derivatives] In the porphyrin complexes, the of previous studies. The structural nature—both molecular singly-occupied vanadium d orbital did not lie highest in and electronic—of the non-porphyrin complexes, however, energy as would be expected from the Aufbau principle de- remains a mystery; since these impurities have deleterious scribed by Bohr the HOMO in these complexes effects on catalysts in the petroleum refining process, a bet- was a porphyrin-based ␲-orbital, which contributes to the ter understanding of these impurities is necessary. We have unusual stability of these complexes. The singly-occupied been studying the electronic structure of complexes of vana- orbital in the acac complexes (O4 coordination sphere) dium in order to correlate the effect that ligand connectivity was the HOMO, following the expectations of the Aufbau has on the electronic structure and reactivity of the metal principle. Here we present the photoelectron spectra of two complexes—VO(acen) and VS(acen)—that are accepted Our primary method of studying electronic structure models for non-porphyrin complexes found in crude oil is gas-phase ultraviolet photoelectron spectroscopy. This deposits, according to Crans et al. Both complexes pos-sess an N2O2 coordination sphere about the metal center allowing comparison between the electronic structures of Corresponding author. Tel.: +1-860-832-2677; these two complexes with the acac complexes that have E-mail address: [email protected] (B.L. Westcott).
0368-2048/$ – see front matter 2003 Elsevier B.V. All rights reserved.
doi:10.1016/j.elspec.2003.11.007 B.L. Westcott et al. / Journal of Electron Spectroscopy and Related Phenomena 135 (2004) 1–5 band profile and the number of peaks necessary for a sta-tistically good fit. For the HeII fits the peak positions andhalf-widths were fixed with respect to those of the HeI fit.
Only the peak amplitudes were allowed to vary to account forchanges in photoionization cross-section. Confidence limitsfor the relative integrated peak areas are about 5% with theprimary source of uncertainty being the determination of thebaseline. The baseline arises from electron scattering and istaken to be linear over the small energy range of these spec-tra. We have described the fitting procedures in more detailelsewhere Fig. 1. ORTEP representation of VO(acen). Ellipsoids at 50% probablility.
Two types of ab initio quantum mechanical calculations were performed using the GAMESS-US suite of programs 2. Experimental
described by Shmidt et al. In the first set, effectivecore potentials (ECP’s) using valence basis sets (VBS) at the -31G level were used for heavy atoms using the SKBJCbasis set With this method, the 10 core electrons All starting materials were used as received from of V are combined to form an effective core charge and the Fisher Chemicals. All three compounds—acen, VO(acen), contraction scheme (8s,8p,6d) → [4s,4p,3d] is employed for VS(acen)—were synthesized according to the procedures the remaining 13 electrons. For the main group elements, of McCarthy et al. Boucher et al. and Callahan two core electrons of C, O and N are combined to form an and Durand Green–black plates of VO(acen) effective core charge and the contraction scheme (4s,4p) → suitable for X-ray diffraction were obtained via sublima- [2s,2p] is employed for the remaining valence electrons. The tion at 80 ◦C, and X-ray datawas collected according to H basis set is an unscaled -31 basis.
In the second set, the Ahlrichs VTZ basis set was used This method uses a (14s,9p,5d) → [8s,5p,3d] con- traction scheme for V and (10s,6p) → [6s,3p] for the sec-ond row elements. The hydrogen basis uses a (5s) → [3s] The HeI (21.2 eV) and HeII (40.8 eV) photoelectron spectra were recorded using instrumentation and proce- The VO(acen) and VS(acen) energies were calculated dures described previously Samples of H2(acen) and based on the experimental crystal structures (this work, VO(acen) showed no signs of impurity during sublimation optimized geometries using SBKJC ECP and Ahlrichs VTZ for VO(acen). The spectrum of VS(acen) showed signs of VO(acen) impurity below temperatures of 150 ◦C; however,no further impurities were observed in the range from 155to 180 ◦C. A small amount of non-volatile solid remained 3. Results and discussion
in the sample cell at the end of each experiment. Duringdata collection the instrument resolution (measured using FWHM of the argon 2P3/2 peak) was 0.020–0.025 eV HeIand 0.022–0.028 eV for HeII.
The crystal structure of VO(acen) was previously re- The spectra were fit analytically with asymmetric Gaus- ported by Bruins and Weaver however, in that case sian peaks with a confidence limit of peak positions and the VO(acen) samples were obtained by recrystallization width deviations generally considered as ±0.02 eV (≈3σlevel) by Lichtenberger and Copenhaver or the differ-ent complexes, the HeI spectrum was fit first. The number 2 Ahlrichs’ and Roos’ basis sets were obtained from the Extensi- of peaks used in each fit was based on the features of the ble Computational Chemistry Environment Basis Set Database, Version12/18/02, as developed and distributed by the Molecular Science Comput-ing Facility, Environmental and Molecular Sciences Laboratory which is 1 Crystallographic data: VO(C12H18N2O2), M = 289.22 g/mol, mon- part of the Pacific Northwest Laboratory, P.O. Box 999, Richland, Wash- oclinic P21/n, a = 15.4274(12), b = 11.77868(8), c = 8.0647(6) Å, ington 99352, USA, and funded by the US Department of Energy. The b = 118.0610(10)◦, V = 1293.19(16) Å3, Z = 4, ρcalc = 1.486 g/cm3, Pacific Northwest Laboratory is a multi-program laboratory operated by λ(MoKα) = 0.71073 Å, T = 90(2) K, 12,580 reflections measured, 3,177 Battelle Memorial Institue for the US Department of Energy under con- unique (Rint = 0.0239), R1 on F (wR2 on F 2) = 0.0338 (0.0909) on tract DE-AC06-76RLO 1830. Contact David Feller or Karen Schuchardt B.L. Westcott et al. / Journal of Electron Spectroscopy and Related Phenomena 135 (2004) 1–5 Table 1GAMESS calculated energies and orbital compositions for the valenceorbitals of H2(acen) Atom numbering scheme is consistent with that of the crystal structureof VO(acen) from gen lone pairs, in agreement with the computational data(A more specific assignment is difficult becauseof the covalency of this system. Attempts to collect datawith different ionizing sources—i.e., HeII or NeI—wouldnot lead to a more specific assignment, as the photoioniza-tion cross-sections of N and O—from the data of Yeh andLindau not significantly different. Since our fo-cus is on the effect that the ligand has on the metal center,a more unambiguous assignment of the ligand spectrum isnot necessary, as it would not provide further insight intothe electronic structure at the metal center.
The spectra of the metallated complexes are quite similar.
The first spectral feature—6.80 eV for VO(acen), 6.52 eV Fig. 2. HeI photoelectron spectra of acen, VO(acen), and VS(acen).
for VS(acen)—corresponds to ionization from the primarilymetal-based dx2−y2-orbital, in agreement with the results of the EPR studies of Callahan and Durand 0.28 eV from benzene, and the room temperature crystal structure destabilization of the metal-based SOMO is attributed to the was based on 954 reflections that were uncorrected for ab- lower electronegativity of the S atom versus O, and is in sorption effects. Although the refinement model contained agreement with electrochemical studies of Seangprasertkij all 36 atoms, only the vanadium and vanadyl oxygen were and Riechel HeII spectra of VO(acen) and VS(acen) refined anisotropically. Furthermore, all hydrogen atoms exhibit growth in the band labeled 1 for both complexes, were located by difference maps and were not constrained consistent with the photoionization cross-section changes to ideal distances or geometries e.g. C7 formed two bonds with hydrogens at distances of 1.122 and 0.874 Å with a the assignment of band 1 with the results from the SBKJC ∠H–C–H bond angle of 73◦. Therefore, the crystal structure basis set showing 98% V dx2−y2 character for the HOMO of VO(acen) has been redetermined, and this new model was used as the basis geometry for our DFT calculations.
At higher ionization potentials, the features from the lig- There are several differences between our structure and and are clearly observed and can be correlated to the free that reported previously. First, the bond lengths between ligand spectrum, with an expected destabilization of these atoms in VO(acen) have been determined to a greater degree orbitals due to the interaction with the metal center. Due to of accuracy; many differ significantly from the values pub- the low symmetry of the complex (Cs idealized, C1 in the lished previously. For example, the V=O bond was initially solid state), an unambiguous assignment of the ligand-based reported to be 1.585 ± 0.007; however, this study finds theV=O bond to be 1.6094 (±0.0011). As for the V–acen inter- actions, in our determination, the V–O bonds to the ligand Relative areas of peaks upon changing photon source are longer on average, and the V–N bonds to the ligand are The HeI photoelectron spectra of the acen ligand and the two vanadium complexes are shown in The free lig- and shows three distinct features in the region from 7.5 to 11.5 eV. These features are assigned as ionizations arising from orbitals that are an admixture of oxygen and nitro- B.L. Westcott et al. / Journal of Electron Spectroscopy and Related Phenomena 135 (2004) 1–5 Fig. 4. MOLEKEL representation of the SOMO of VO(acen) (7.36 eV) the acen complex is the most easily ion-ized (6.80 eV) Stability of porphyrin complexes hasbeen established by Gruhn et al. and can be attributedto the exceptional stability of the ligand itself. The relative Fig. 3. HeI/HeII spectra of VS(acen).
ionizations is difficult. In the spectrum of VS(acen), a spec-tral feature is seen at 8.44 eV (band 5a) that is not observedin the spectrum of VO(acen). This feature is assigned toionization from an orbital predominately localized on thethio moiety because upon changing ionizing source fromHeI to HeII, band 5a shrinks significantly, suggesting a largeamount of sulfur character in this band data for orbitals other than the HOMO are inconclusive,showing substantial mixed character with contributions fromthe axial ligand—either O or S—and the atoms involved inthe delocalized ␲-system of the acen ligand ( relative stability of metal complexes can be inferred by theenergy required to remove the d electrons. When comparingthe series VO(oep), VO(acac)2, and VO(acen) we find thed1 electron in the porphyrin complex is the most difficultto remove (7.50 eV) the acac complex is intermediate Table 3GAMESS calculated energies and orbital compositions for the valenceorbitals of VO(acen) 26% C3, 23% C10, 12% N1, 11% N2, 7%O2, 7% O3 28% C10, 25% C3, 10% N2, 9% N1, 6%O3, 5% O2, 5% C11 24% O2, 11% O1, 9% O3, 8% N2, 7%N1,5% C1, 5% V1, 5% C3 Atom numbering scheme is consistent with that of the crystal structure Fig. 5. MOLEKEL representations of the HOMO-1 (top) and HOMO-2 B.L. Westcott et al. / Journal of Electron Spectroscopy and Related Phenomena 135 (2004) 1–5 References
GAMESS calculated energies and orbital compositions for the valenceorbitals of VS(acen) [1] A.J.G. Barwise, Energy Fuels 4 (1990) 647–652.
[2] C.B. Pearson, J.B. Green, Energy Fuels 7 (1993) 338–346.
[3] M.E. Pena, A. Manjarrez, A. Campero, Fuel Process. Tech. 46 (1996) [4] R.H. Fish, J.J. Komlenic, Anal. Chem. 56 (1984) 510–517.
[5] B.L. Westcott, N.G. Gruhn, L.J. Michelsen, D.L. Lichtenberger, J.
Am. Chem. Soc. 122 (2000) 8083–8084.
[6] N.E. Gruhn, L.J. Michelsen, B.L. Westcott, Inorg. Chem. 41 (2002) [7] N. Bohr, Philos. Mag. 26 (1913) 476.
Atom numbering scheme is consistent with that of the crystal structure [8] D.C. Crans, S.S. Amin, A.D. Keramidas, in: J.O. Nriagu (Ed.), Vanadium in the Environment. Part I: Chemistry and Biochemistry,Wiley, New York, 1998, pp. 73–75.
[9] P.J. McCarthy, R.J. Hovey, K. Ueno, A.E. Martell, J. Am. Chem.
stability of the acac complex with respect to the acen [10] L.J. Boucher, T.F. Yen, E.C. Tynan, Inorg. Chem. 7 (1968) 731–736.
complex can be explained through simple electronegativ- [11] K.P. Callahan, P.J. Durand, Inorg. Chem. 19 (1980) 3211–3217.
ity arguments. Since the O atoms are more electronegative [12] (a) G. Crundwell, B.L. Westcott, R. Coffey, M. Zeller, A.D. Hunter, than the N atoms, the N atoms will necessarily be bet- (b) J. Woltz, B.L. Westcott, G. Crundwell, M. Zeller, A.D. Hunter, ter donors to the metal center, leading to a destabilization S.O. Sommerer, Acta Crystallogr. E 58 (2002) 609–610.
in the orbital energy of the metal-based orbital. Similar [13] B.L. Westcott, N.E. Gruhn, J.H. Enemark, J. Am. Chem. Soc. 120 behavior has been observed previously by Lichtenberger (1998) 3382–3386, and references therein.
et al. the photoelectron spectra of dimetal com- [14] D.L. Lichtenberger, A.S. Copenhaver, J. Electron. Spectrosc. Rel.
plexes with O and N donors, as well as in Re complexes [15] M.W. Schmidt, K.K. Baldridge, J.A. Boatz, S.T. Elbert, M.S. Gordon, where a halogen ligand is varied amongst Cl, Br, and I J.H. Jensen, S. Koseki, N. Matsunaga, K.A. Nguyen, S.J. Su, T.L.
Windus, M. Dupuis, J.A. Montgomery, J. Comput. Chem. 14 (1993)1347.
[16] (a) W.J. Stevens, H. Basch, M. Krauss, J. Chem. Phys. 81 (1984) 4. Supplementary material
(b) W.J. Stevens, M. Krauss, H. Basch, P.G. Jasien, Can. J. Chem.
70 (1992) 612; Listings of results and input parameters for the com- (c) T.R. Cundari, W.J. Stevens, J. Chem. Phys. 98 (1993) 5555.
putation studies are available from the authors. The full [17] T.R. Cundari, L. Saunders, L.L. Sisterhen, J. Phys. Chem. A. 102 details of the crystal structure of VO(acen) have been de- [18] T.R. Cundari, M.S. Gordon, Coord. Chem. Rev. 87 (1996) 87.
posited with the Cambridge Structural Data Centre, CCDC [19] A. Schafer, H. Horn, R. Ahlrichs, J. Chem. Phys. 97 (1992) 2571.
No. 218164. Copies of this information may be obtained [20] M. Sato, K.M. Miller, J.H. Enemark, C.E. Strouse, K.P. Callahan, free of charge from: The Director, CCDC, 12 Union Road, Cambridge, CB2 1EZ, UK (fax: +44-1233-336033; [21] D. Bruins, D.L. Weaver, Inorg. Chem. 9 (1970) 130–135.
[22] J.J. Yeh, I. Lindau, At. Data Nucl. Data Tables 32 (1985) 1–14.
[23] R. Seangprasertkij, T.L. Riechel, Inorg. Chem. 23 (1984) 991–994.
[24] G.M. Bancroft, Y.F. Hu, in: E.I. Solomon, A.B.P. Lever, (Eds.), In- organic Electronic Structure and Spectroscopy, Volume I: Method-ology, Wiley, New York, 1999, pp. 443–512.
Acknowledgements
[25] J.C. Green, Acc. Chem. Res. 27 (1994) 131–137.
[26] N.E. Gruhn, D.L. Lichtenberger, H. Ogura, F.A. Walker, Inorg. Chem.
38 (1999) 423–427, and references therein.
We wish to thank Dr. T. Cundari (University of North [27] D.L. Lichtenberger, M.A. Lynn, M.H. Chisholm, J. Am. Chem. Soc.
Texas) for helpful discussions regarding the application of GAMESS to vanadyl complexes. G.C., C.G., L.M., and B.W.
[28] D.L. Lichtenberger, A. Rai-Chadhuri, M.J. Seidel, J.A. Gladysz, S.K.
were supported by CSU-AAUP and CCSU Student-Faculty Agbossou, A. Igau, C.H. Winter, Organometallics 10 (1991) 1355– Research grants. M.Z. was supported by NSF grant [29] E.E. Russel, Honors Thesis, The University of Arizona, 1997.
0111511, and the diffractometer was funded by NSF grant [30] MOLEKEL 4.0, P. Flukiger, H.P. Luthi, S. Portmann, J. Weber, 0087210, by Ohio Board of Regents grant CAP-491, and by Swiss Center for Scientific Computing, Manno (Switzerland), 2000.

Source: http://www.as.ysu.edu/~scic/publications/J%20Electr%20Spec%20Rel%20Phen%202004,%20135,%201-5.pdf