Cross-hyperconjugation: An unexplored orbital interaction between π-conjugated and saturated molecular segments

Article abstract of DOI:10.1002/anie.201206030

Crossing a barrier: Molecules with saturated ER₂ units (E=C or Si, R=electron-releasing group) inserted between two π-conjugated segments have electronic and optical properties that resemble those of cross-conjugated molecules (see figure). This cross-hyperconjugation provides a deeper understanding of the conjugation phenomenon, and is an alternative to cross-conjugation in the design of molecules for nano and materials applications. Copyright

Full text of DOI:10.1002/anie.201206030

Angewandte
Chemie
DOI: 10.1002/anie.201206030
Conjugation
Cross-Hyperconjugation: An Unexplored Orbital Interaction between
p-Conjugated and Saturated Molecular Segments**
Rikard Emanuelsson, Andreas Wallner, Eugene A. M. Ng, Joshua R. Smith, Djawed Nauroozi,
Sascha Ott, and Henrik Ottosson*
p-Conjugation, which is the stabilizing interaction between
local orbitals of p-symmetry at multiple bonds, is a core
concept in chemistry. In the classical sense, a p-conjugated
compound has a structure with alternating single and multiple
(double and/or triple) bonds.[1] However, this model is
deceiving in its simplicity because conjugation also depends
on the connectivity in the bond alternating path.[2] Thus, p-
conjugation is divided into linear conjugation (vicinal con-
nectivity), cross-conjugation (geminal connectivity), and
1) positive, that is, donation of electron density from filled s-
orbitals to empty p*-orbitals; 2) negative, that is, interaction
between p- or pp-orbitals with s*-orbitals; and 3) neutral
hyperconjugation, with no strong directional interaction.[7]
The interaction between pp- or p- and s-bond orbitals in
a linear hyperconjugative connectivity is well-known and
often found in charged species.[8] However, the fusion of two
neutral hyperconjugated paths to a cross-hyperconjugated
molecule with geminal connectivity between the two paths
should also be considered.[9] This interaction would occur in
species with two linearly p-conjugated fragments linked by an
sp3 hybridized ER2 unit (E = Group 14 element) that con-
tributes with two localized orbitals of p-symmetry; a bonding
pseudo-p and an antibonding pseudo-p* (Figure 1). These
orbitals must be symmetry-adapted combinations of either
ꢀ
two E R s- or s*-bond orbitals, and in the C2 point group
they belong to the b irreducible representation.
We now report on a study of compounds that are suitable
for probing the strength of such potential cross-hyperconju-
gation. Regular cross-conjugation has recently received
increased attention in the fields of nano and materials
sciences.[10,11] Thus, if strongly cross-hyperconjugated com-
pounds are realizable, they could serve as design alternatives
omniconjugation. The first provides a path for delocalization
=
over, for example, a C C double bond from one end to the
other, the second provides strong delocalization paths only
=
from the groups attached to the C C bond but not between
the groups, and the third is a branched system with all parts
linearly conjugated.[3]
Conjugative interaction is not limited to p-conjugated
molecular segments but can also be found in s-bonded species
(s-conjugation)[4] and between s- and p-systems (hypercon-
jugation).[5,6] Hyperconjugation can further be divided into
[*] R. Emanuelsson, Dr. A. Wallner, E. A. M. Ng, Dr. H. Ottosson
Department of Chemistry–BMC, Uppsala University
Box 576, 75123 Uppsala (Sweden)
E-mail: henrik.ottosson@kemi.uu.se
Prof. J. R. Smith
Department of Chemistry, Humboldt State University
1 Harpst St., Arcata, CA 95521 (USA)
Dr. D. Nauroozi, Dr. S. Ott
Department of Chemistry–ꢀngstrçm, Uppsala University
Box 523, 75120 Uppsala (Sweden)
[**] We are grateful for support from the Swedish Research Council
(Grant Nos. 2008-3710, 2009-362, and 2010-4858), Uppsala
University through the U3MEC KoF07 initiative, and the Wenner-
Gren Foundations, Gçran Gustafsson Foundation, and American
Scandinavia Foundation for fellowships to A.W. and J.R.S. The
National Supercomputer Center (NSC) in Linkçping, Sweden, and
the Uppsala Multidisciplinary Center for Advanced Computational
Science (UPPMAX), Uppsala, Sweden are acknowledged for
generous allotment of computer time.
Figure 1. Qualitative molecular orbital diagram displaying both the
pseudo-p and pseudo-p* orbitals (b-symmetric) of H2ER2, and the in-
phase (b-symmetric) and out-of-phase (a-symmetric) fragment orbital
combinations of the phenylacetylene frontier orbitals (HOMO and
LUMO) of the phenylethynyl–phenylethynyl fragment.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201206030.
Angew. Chem. Int. Ed. 2013, 52, 983 –987
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
983

.
Angewandte
Communications
to rigid and planar p-conjugated species. Owing to their non-
rigid structure they should display higher solubilities than the
latter compounds, a fact that can be beneficial from applica-
tions point of view. Cross-hyperconjugation would also
expand the fundamental understanding of the conjugation
phenomenon.
We first investigated bis(phenylethynyl)compounds 1–4,
where 2–4 are substituted at the central Si or C atom. By
varying the substituents R from electron-releasing to elec-
tron-withdrawing, the energy levels of the pseudo-p(ER2) and
pseudo-p*(ER2) orbitals of the saturated segment are altered
and brought closer to the frontier p-orbitals of the phenyl-
ethynyl arms (Figure 1). This should afford a varying degree
of cross-hyperconjugation as a local filled pseudo-p(ER2)
orbital is oriented similarly as a local filled p-bond orbital of
=
a central C C bond in a cross-conjugated compound. The out-
Figure 2. UV absorption spectra of solutions of 1 and 2 in cyclohexane.
of-phase combination of the pseudo-p(ER2) and the b-
symmetric orbital on the phenylethynyl–phenylethynyl frag-
ment leads to the highest occupied molecular orbital
(HOMO) of the assembled molecule. The question then
arises as to how extensive this interaction can become.
excitation energies of 4.17 and 4.42 eV. The second and third
absorptions are found at 274 (1) and 265 nm (2), and at 261 (1)
and 253 nm (2), respectively.
Evaluations of conjugation based on features in UV
spectra should however be carried with caution.[19] Thus, to
further elucidate the similarities the compounds were studied
by cyclic voltammetry as well as by DFT calculations, with
geometries optimized at B3LYP/6-311G(d) and M062X/6-
311G(d) levels and vertical excitations obtained at these
geometries by time-dependent DFT (TD-DFT) at TD-
M062X/6-311 + G(2d,p) level.[20]
The calculations reveal that the analogies persist to the
frontier orbitals of 1 and 2. For example, in the HOMO there
is significant orbital density on the Si(SiMe3)2 moiety of 2
The heavier Group 14 elements are often found in studies
on hyperconjugation, especially in charged species.[12] Herein
we show that they also provide opportunities for synthetically
viable molecules by which the neutral cross-hyperconjugation
strength can be varied. The ethynyl segments inserted
between the phenyl and ER2 groups were incorporated to
minimize the steric interaction between the substituents on
the E atom and the p-system, and thus, to safe-guard against
poor orbital overlap as a result of conformational restrictions
that originate from steric congestion.
Initially we sat out to match the HOMO energy (eHOMO) of
phenylacetylene with the pseudo-p(ER2) orbitals of various
H2ER2 molecules (in all cases HOMO). Calculations using
M062X/6-311G + (2d,p)//B3LYP/6-311G(d) hybrid meta den-
sity functional theory (DFT)[13–16] reveal that eHOMO of
=
aligned in a cross-conjugated manner, similar as on the C
CMe2 segment of 1 (Figure 3). With regard to orbital energies,
eHOMO of 2 is within 0.3 eV from that of 1 (1: ꢀ6.87, 2:
ꢀ7.13 eV). The a-symmetric HOMOꢀ1 orbitals are nearly
isoenergetic (1: ꢀ7.78, 2: ꢀ7.81 eV) and display strong
resemblance which stem from the a orbital symmetry impos-
=
ing a nodal plane bisecting the ER2 and C CMe2 moieties.
With regard to the lowest and second lowest unoccupied
molecular orbitals (LUMO and LUMO + 1), they are clearly
analogous in the two compounds, yet in 2 they are isoener-
getic but split by 0.70 eV in 1. The splitting reflects the overlap
of the in-phase combination of the LUMO of the phenyl-
=
ethynyl–phenylethynyl fragment with the p*(C CMe2) of
phenylacetylene
(ꢀ7.96 eV)
and
2-methylpropene
1 versus the pseudo-p*(Si(SiMe3)2) of 2. It also reflects the
distance between the closest two sp-hybridized C atoms of the
phenylethynyl arms (2.412 vs. 2.986 ꢀ in 1 and 2) and thus the
local bonding overlap. This clearly leads to a lower LUMO in
1 than in 2. With TD-M062X/6-311 + G(2d,p)//B3LYP/6-
311G(d) the first transitions in 1 and 2 are 298.9 nm
(4.28 eV) and 269.0 nm (4.60 eV), respectively. However,
the transitions are to states of opposite symmetries (21A vs.
11B, Table 1), and this repeats for the second transitions,
which are to 11B in 1 and 21A in 2. Thus, the similarity in the
UV absorption spectra of the two compounds is deceiving, as
the seemingly analogous transitions involve different excita-
tions.
(ꢀ8.25 eV), corresponding to fragments of 1, differ by less
than 0.3 eV (for computational details, see Supporting
Information). The pseudo-p(ER2) of H2Si(SiMe3)2 is even
closer (ꢀ7.98 eV), while the analogous orbitals of H2SiMe2
(ꢀ9.99 eV) and H2CMe2 (ꢀ10.49 eV) are deeper down. With
two trifluoromethyl groups at silicon, leading to H2Si(CF3)2,
eHOMO is ꢀ11.58 eV. Thus, bis(trimethylsilyl)bis(phenylethy-
nyl)silane (2) should show strong cross-hyperconjugation, and
we therefore synthesized 1 and 2, both reported earlier[17,18]
(for details see Supporting Information).
The UV absorption spectra of 1 and 2 reveal similar
spectral features (Figure 2), with the first absorption should-
ers at 298 (1) and 281 nm (2), respectively, corresponding to
984
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 983 –987

Angewandte
Chemie
Figure 4. UV absorption spectra of solutions of 1–4 in cyclohexane.
the molecules to planar C2v symmetric structures so as to
optimize the orbital overlap gives more orbital density at the
ER2 segment in 3’ and 4’ (Figure 5), but brings only little
change in the calculated first excitation energies.
Figure 3. Frontier orbitals of 1 and 2 at the M062X/6-311+G(2d,p)//
B3LYP/6-311G(d) level.
Table 1: The first two excitations of 1–4 from UV absorption spectros-
copy and the corresponding vertical excitations as calculated with TD-
M062X/6-311+G(2d,p)//B3LYP/6-311G(d).[a]
Transition 1
Transition 2
nm (eV), state sym.
nm (eV), state sym.
1
2
3
4
exp.
comp.
exp.
comp.
exp.
comp.
exp.
298 (4.16)
274 (4.53)
289.9 (4.28), 21A
281 (4.41)
268.2 (4.62), 11B
265 (4.68)
Figure 5. The HOMOs of the optimal C2-symmetric 3 and 4 and the
planar C2v-symmetric 3’ and 4’ calculated at the M062X/6-311+G-
(2d,p)//B3LYP/6-311G(d) level.
269.0 (4.61), 11B
261 (4.75)
265.1 (4.68), 21A
249 (4.98)
250.6 (4.95), 11B
253 (4.91)
250.1 (4.96), 21A
241 (5.15)
Further than UV spectroscopy, electrochemically deter-
mined redox potentials can provide substantial information
on the nature of the frontier orbitals of 1–4. Thus, the
compounds were investigated by means of cyclic voltammetry
(CV), and the anodic scans of the CVs of 1, 2, 3, and 4
(Figure 6) feature electrochemically irreversible oxidative
processes at 1.17, 1.22, 1.92, and 1.83 V vs. Fc+/Fc0, respec-
tively. It is thus apparent that compounds 1 and 2 are oxidized
at similar potentials, while the oxidations of 3 and 4 are
anodically shifted by more than 600 mV. These findings
strongly support the results obtained in the UV analysis as
well as the DFT calculations in that the HOMOs of 1 and 2
are considerably higher in energy compared to those of 3 and
4.
comp.
241.6 (5.13), 11B
239.1 (5.18), 21B
[a] All transitions are visible, with calculated oscillator strengths above
0.20. For 4, there are four forbidden transitions between the first and
second visible excitation.
To determine whether other bis(phenylethynyl)silanes
and methanes display similar features as 2, we synthesized 3
and 4 with ER2 as SiMe2 and CMe2.[21] Compared to 2, their
first UV absorptions are hypsochromically shifted to 261 (3)
and 253 nm (4), respectively, corresponding to raises in
energy by 0.34 and 0.49 eV (Figure 4). The TD-M062X
calculations reproduce the blue-shift as the first excitation
of 4, a HOMO!LUMO transition leading to the 11B state, is
found at 241.6 nm (5.13 eV). However, in 3 the lowest
transition is calculated at 250.6 nm (4.95 eV) and it is much
more complex. Also, the contribution of the pseudo-p(ER2)
orbital to the HOMOs of 3 and 4 is negligible, indicating that
3 and 4 are much less cross-hyperconjugated than 2. Locking
The calculated geometries also reveal that 2 is more
related to 1 than is 3 and 4, as the C2-C1-C1’-C2’ dihedral
angle V involving the inner ortho- and the ipso-C atoms of the
phenyl groups (Table 2), which measures the out-of-plane
rotation of the phenyl groups and thus reflects the linear p-
Angew. Chem. Int. Ed. 2013, 52, 983 –987
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
985

.
Angewandte
Communications
this level is 270.4 nm (4.58 eV). With E as C and R as either
SiMe3 (7), SnMe3 (8), or CF3 (9) excitations at 264.3 nm
(4.69 eV), 286.5 nm (4.33 eV), and 235.0 nm (5.28 eV) are
found. Thus, the computations reveal possibilities for wider
tuning of the first transition when E = C than when E = Si,
and compound 8 is the best cross-hyperconjugated analogue
of 1. Yet, an important difference between 1 and the cross-
hyperconjugated species, excluding 6,[23] is that the first
transition in 1 is to the 21A state, whereas in 2–5 and 7–9 it
is to the 11B state.
Figure 6. Cyclic voltammograms (anodic scans) of 1–4 (1 mm solu-
tions in CH3CN) containing 0.1m NBu4PF6 versus Fc+/Fc0,
n=100 mVsꢀ1. ? Fc+/Fc0 couple as internal standard.
A striking geometry feature of 6–8 is that these species are
either planar C2v-symmetric (6 and 8, Table 2), or nearly so
(7), supporting strong cross-hyperconjugation. It is also
Table 2: Dihedral angle V and essential CC bond lengths calculated at
B3LYP/6-311G(d) (normal) and M062X/6-311G(d) (italics) levels.[a]
Molecule, sym.
V [8]
r C [ꢀ]
ꢁ
C
rCꢀC(Ph) [ꢀ]
Drcc(Ph) [ꢀ]
ꢁ
notable that the C C triple bonds in 6 and 8 are notably
elongated when compared to 1. From the calculated lowest
excitation energies and the geometries, it is thus apparent that
cross-hyperconjugation strength can be tuned through choice
of element E and substituents R, in a similar manner as found
1
2
C2
C2
C1
C2
C2
C2
C2
C1
C2
C1
C2
C1
C2
10.7
49.6
44.6
89.3
84.5
49.7
0.0
0.0
8.2
8.2
0.0
1.210
1.218
1.213/1.213
1.215
1.208
1.214
1.230
1.226/1.226
1.210
1.210
1.423
1.425
1.431/1.431
1.426
1.428
1.423
1.433
1.438/1.438
1.426
1.427
1.394–1.407
1.390–1.406
1.388–1.400
1.390–1.406
1.390–1.406
1.389–1.406
1.399–1.414
1.395–1.406
1.390–1.407
1.390–1.407
1.399–1.415
1.395–1.408
1.390–1.405
3
4
5
6
earlier for the hyperconjugative donor and acceptor abilities
[24]
ꢀ
of C X s-bonds. In our species, electropositive R groups
raise the energy of the pseudo-p(ER2) orbital as well as
ꢀ
7
8
9
polarize the E R s-bond electron density towards the E
atom.
1.226
1.223/1.223
1.203
1.433
1.437/1.437
1.426
In conclusion, through a series of easily accessible and
persistent compounds we have shown that a properly sub-
stituted ER2 segment has similar effect on the electronic and
optical properties of a molecule when inserted between two
2.3
54.2
[a] V denotes the C2-C1-C1’-C2’ dihedral angle involving the inner ortho-
ꢀ
and the ipso-C atoms of the phenyl groups, Drcc(Ph) denotes the C C
=
linearly p-conjugated segments as a C C double bond with
bond length range of the Ph groups. Here, only values of species with
bulky R=SiMe3 or SnMe3 are given with both B3LYP and M062X; for the
full table see the Supporting Information.
geminal connectivity. This cross-hyperconjugation provides
new design motifs for compounds with applications in nano
and materials sciences. A next question is now whether
omnihyperconjugated molecules can also be designed.
conjugation/hyperconjugation strengths between the phenyl-
=
Received: July 27, 2012
Revised: September 26, 2012
Published online: November 26, 2012
ethynyl arms and either the C CMe2 or ER2 segments, has
values with B3LYP of 10.78 (1), 49.68 (2), 89.38 (3), and 84.58
(4), respectively. Interestingly, the dihedral angle of 2 is even
smaller with M062X, a dispersion corrected DFT method
suitable to handle steric congestion such as between the two
SiMe3 groups of 2. On the other hand, no significant
differences are seen in analogous CC bond distances of
these compounds, even not when comparing the regularly
cross-conjugated 1 with 3 and 4 (Table 2). The same holds for
experimentally determined NMR chemical shifts, as there are
no significant differences in the aromatic region of 1 when
compared to those of 2–4.
Keywords: conjugation · cross-conjugation ·
.
Group 14 elements · hyperconjugation · optical tuning
[1] R. Hoffmann, C. Janiak, C. Kollmar, Macromolecules 1991, 24,
3725 – 3746.
[2] M. Bruschi, M. G. Giuffreda, H. P. Lꢁthi, Chem. Eur. J. 2002, 8,
4216 – 4227.
[3] Linear conjugation: H. Wang, R. Helgeson, B. Ma, F. Wudl, J.
Org. Chem. 2000, 65, 5862 – 5867; Cross-conjugation: N. F.
Phelan, M. Orchin, J. Chem. Educ. 1968, 45, 633 – 637; M.
Trætteberg, H. Hopf, Acta Chem. Scand. 1994, 48, 989 – 993;
Omniconjugation: M. H. van der Veen, M. T. Rispens, H. T.
Jonkman, J. C. Hummelen, Adv. Funct. Mater. 2004, 14, 215 –
223.
We also computed how cross-hyperconjugation varies
with atom E and a wider set of substituents R. For E = Si,
a span in the excitation energies is provided by R = CF3 (5)
and SnMe3 (6) giving a wavelength (energy) range for the first
visible excitation of 249.6–276.3 nm (4.49–4.97 eV). Here,
compound 6 was calculated at TD-M062X/LANL2DZdp//
B3LYP/LANL2DZdp level,[22] and the first excitation of 2 at
[4] R. D. Miller, J. Michl, Chem. Rev. 1989, 89, 1359 – 1410.
986
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 983 –987

Angewandte
Chemie
[5] a) R. S. Mulliken, J. Chem. Phys. 1939, 7, 339 – 352; b) R. S.
Mulliken, C. A. Rieke, W. G. Brown, J. Am. Chem. Soc. 1941, 63,
41 – 56.
[6] For thorough reviews on p- and s-delocalization, see Chem. Rev.
2005, 105, all of issue 10.
[7] For a recent review of hyperconjugation, see: I. V. Alabugin,
K. M. Gimore, P. W. Peterson, WIREs Comput. Mol. Sci. 2011, 1,
109 – 141.
[8] a) N. Muller, R. S. Mulliken, J. Am. Chem. Soc. 1958, 80, 3489 –
3497; b) P. v. R. Schleyer, A. J. Kos, Tetrahedron 1983, 39, 1141 –
1150; c) R. Hoffmann, J. Chem. Phys. 1964, 40, 2480 – 2488.
[9] We use the term cross-hyperconjugation instead of hypercross-
conjugation to stress that the geminal connectivity of the
conjugated paths should have higher priority than the specific
conjugation type.
10246 – 10248; g) M. A. Cook, C. Eaborn, D. R. M. Walton, J.
Organomet. Chem. 1970, 24, 293 – 299; h) U. Weidner, A.
Schweig, Angew. Chem. 1972, 84, 550; Angew. Chem. Int. Ed.
1972, 11, 536 – 537.
[13] Gaussian09 (RevisionA.1). Frisch, M. J. et al. Gaussian, Inc.,
Wallingford CT, 2009. For the complete reference, see the
Supporting Information.
[14] a) A. D. Becke, J. Chem. Phys. 1993, 98, 5648 – 5652; b) P. J.
Stevens, F. J. Devlin, C. F. Chablowski, M. J. Frisch, J. Phys.
Chem. 1994, 80, 11623 – 11627.
[15] a) R. Krishnan, J. S. Binkley, R. Seeger, J. A. Pople, J. Chem.
Phys. 1980, 72, 650 – 654; b) A. D. McLean, G. S. Chandler, J.
Chem. Phys. 1980, 72, 5639 – 5648.
[16] a) Y. Zhao, D. G. Truhlar, Theor. Chem. Acc. 2008, 120, 215 –
241; b) Y. Zhao, D. G. Truhlar, J. Chem. Theory Comput. 2008, 4,
1849 – 1968.
[10] a) M. Gholami, R. R. Tykwinski, Chem. Rev. 2006, 106, 4997 –
5027; b) P. A. Limacher, H. P. Lꢁthi, WIREs Comput. Mol. Sci.
2011, 1, 477 – 486; c) H. Hopf, M. S. Sherburn, Angew. Chem.
2012, 124, 2346 – 2389; Angew. Chem. Int. Ed. 2012, 51, 2298 –
2338.
[17] W. Yan, X. Ye, N. G. Akhmedov, J. L. Petersen, X. Shi, Org. Lett.
2012, 14, 2358 – 2361.
[18] C. Mechtler, M. Zirngast, J. Baumgartner, C. Marschner, Eur. J.
Inorg. Chem. 2004, 3254 – 3261.
[11] a) G. C. Solomon, D. Q. Andrews, R. P. Van Duyne, M. A.
Ratner, J. Am. Chem. Soc. 2008, 130, 7788 – 7789; b) G. C.
Solomon, D. Q. Andrews, R. H. Goldsmith, T. Hansen, M. R.
Wasielewski, R. P. Van Duyne, M. A. Ratner, J. Am. Chem. Soc.
2008, 130, 17301 – 17308; c) G. C. Solomon, D. Q. Andrews, R. P.
Van Duyne, M. A. Ratner, ChemPhysChem 2009, 10, 257 – 264;
d) A. B. Ricks, G. C. Solomon, M. T. Colvin, A. M. Scott, K.
Chen, M. A. Ratner, M. R. Wasielewski, J. Am. Chem. Soc. 2010,
132, 15427 – 15434.
[12] a) J. B. Lambert, C. E. Shawl, E. Basso, Can. J. Chem. 2000, 78,
1441 – 1444; b) J. B. Lambert, Y. Zhao, R. W. Emblidge, L. A.
Salvador, X. Liu, J.-H. So, E. C. Chelius, Acc. Chem. Res. 1999,
32, 183 – 190; c) J. B. Lambert, G. T. Wang, R. B. Finzel, D. H.
Teramura, J. Am. Chem. Soc. 1987, 109, 7838 – 7845; d) J. B.
Lambert, Tetrahedron 1990, 46, 2677 – 2689; e) J. B. Lambert,
G. T. Wang, D. H. Teramura, J. Org. Chem. 1988, 53, 5422 – 5428;
f) J. B. Lambert, R. A. Singer, J. Am. Chem. Soc. 1992, 114,
[19] a) M. Klokkenburg, M. Lutz, A. L. Spek, J. H. van der Maas,
C. A. van Walree, Chem. Eur. J. 2003, 9, 3544 – 3554; b) N. N. P.
Moonen, F. Diederich, Org. Biomol. Chem. 2004, 2, 2263 – 2266.
[20] a) E. Runge, E. K. U. Gross, Phys. Rev. Lett. 1984, 52, 997 – 1000;
b) D. Jacquemin, E. A. Perpꢂte, I. Ciofini, C. Adamo, J. Chem.
Theory Comput. 2010, 6, 1532 – 1537.
[21] H. E. Zimmerman, J. A. Pincock, J. Am. Chem. Soc. 1973, 95,
3246 – 3250.
[22] a) W. R. Wadt, P. J. J. Hay, Chem. Phys. 1985, 82, 284 – 298;
b) C. E. Check, T. O. Faust, J. M. Bailey, B. J. Wright, T. M.
Gilbert, L. S. Sunderlin, J. Phys. Chem. A 2001, 105, 8111 – 8116.
[23] Compound 6 is analogous to 1, with the first transition to the 21A
state and the second to the 11B state.
[24] a) I. V. Alabugin, T. A. Zeidan, J. Am. Chem. Soc. 2002, 124,
3175 – 3185; b) I. V. Alabugin, M. Manoharan, J. Org. Chem.
2004, 69, 9011 – 9024.
Angew. Chem. Int. Ed. 2013, 52, 983 –987
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
987