Antiaromatic spacer-bridged bisfluorenyl dications generated by superacid induced ionization

Article abstract of DOI:10.1021/ja711501k

Derivatives of the dication of tetrabenzo[5.5]fulvalene were prepared with phenyl and ethynyl spacers through ionization of the appropriate bis-methylethers. The antiaromaticity shown by the parent dication was demonstrated for these dications with spacer

Full text of DOI:10.1021/ja711501k

Published on Web 07/12/2008
Antiaromatic Spacer-Bridged Bisfluorenyl Dications
Generated by Superacid Induced Ionization
Bart J. Dahl and Nancy S. Mills*
Department of Chemistry, Trinity UniVersity, San Antonio, Texas 78212
Received December 31, 2007; E-mail: nmills@trinity.edu
Abstract: Derivatives of the dication of tetrabenzo[5.5]fulvalene were prepared with phenyl and ethynyl
spacers through ionization of the appropriate bis-methylethers. The antiaromaticity shown by the parent
dication was demonstrated for these dications with spacers, although it was attenuated by the presence of
the spacer. It was substantially greater than that of fluorenyl monocations with similar substituents.
Antiaromaticity was evaluated through comparison of 1H NMR shifts with those of acyclic analogues, through
nucleus independent chemical shifts, and through magnetic susceptibility exaltation. Although the fluorenyl
systems are separated by spacers, the antiaromaticity of one system is affected by the other remote fluorenyl
system. An explanation for this interaction may lie in the ability of a remote cationic substituent to attenuate
delocalization in the spacer. The use of spacers is designed to prevent side reactions in less stable
antiaromatic dications, allowing exploration of a number of species that have previously been inaccessible.
Introduction
(4), have not been successfully observed using these methods
because of their tendencies to cyclize under the reaction
While often subjects of great theoretical interest, the inherent
instability of compounds of an antiaromatic nature such as 1,3-
cyclobutadiene has severely limited the collection of experi-
mental data on these often fleeting molecules.1–4 However,
numerous antiaromatic bisfluorenyl (1),5–7 3-phenylindenyl-
fluorenyl 1,2-carbodicationic systems (2),8 and other similar
compounds,9–12 in which the two cations are located on directly
adjacent carbon atoms, have been prepared and studied experi-
mentally due to their surprising kinetic stability. The method
for generating carbodications in each of these systems is low-
temperature oxidation of the corresponding neutral olefin
precursors with the highly reactive SbF5/SO2ClF. While this
route has been very successful in producing dications that further
our understanding of antiaromaticity, its overall scope has been
currently limited to the aforementioned systems. Potentially
interesting less stabilized antiaromatic systems, including bis-
indenyl (3) and unsubstituted indenyl-fluorenyl carbodications
conditions. Cyclization of unstable dications has been commonly
observed in similar systems.13–15 Additionally, synthetic methods
to achieving the appropriate neutral precursors are dependent
on the ability to form sometimes highly strained16,17 or
electronically unstable18–22 double bond species.
Superacids such as fluorosulfonic acid and Magic acid have
been staple reagents for Olah et al.23,24 in their groundbreaking
and ongoing experimental studies of carbocationic molecules
(1) Breslow, R. Acc. Chem. Res. 1973, 6, 393–398.
(2) Tidwell, T. Chem. ReV. 2001, 101, 1333–1348.
(3) Bally, T. Angew. Chem., Int. Ed. 2006, 45, 6616–6619.
(4) Minkin, V. I.; Glukhovtsev, M. N.; Simkin, B. Y. Aromaticity and
Antiaromaticity; John Wiley and Sons: New York, 1994.
(5) Mills, N. S.; Malandra, J. L.; Burns, E. E.; Green, A.; Unruh, K. E.;
Kadlecek, D. E.; Lowery, J. A. J. Org. Chem. 1997, 62, 9318–9322.
(6) Mills, N. S.; Malinky, T.; Malandra, J. L.; Burns, E. E.; Crossno, P.
J. Org. Chem. 1999, 64, 511–517.
(13) Klumpp, D. A.; Baek, D. N.; Prakash, G. K. S.; Olah, G. A. J. Org.
Chem. 1997, 62, 6666–6671.
(14) Mills, N. S.; Malandra, J. L.; Hensen, A.; Lowrey, J. E. Polycyclic
Aromat. Compd. 1998, 12, 239–247.
(15) Olah, G. A.; Klumpp, D. A.; Neyer, G.; Wang, Q. Synthesis 1996,
321–3.
(16) Ter Wiel, M. K. J.; Kwit, M. G.; Meetsma, A.; Feringa, B. L. Org.
Biomolec. Chem. 2007, 5, 87–96.
(17) ter Wiel, M. K. J.; Vicario, J.; Davey, S. G.; Meetsma, A.; Feringa,
B. L. Org. Biomolec. Chem. 2005, 3, 28–30.
(7) Levy, A.; Rakowitz, A.; Mills, N. S. J. Org. Chem. 2003, 68, 3990–
98.
(18) Montgomery, L. K.; Huffman, J. C.; Jurczak, E. A.; Grendze, M. P.
J. Am. Chem. Soc. 1986, 108, 6004–11.
(8) Mills, N. S.; Llagostero, K. B.; Tirla, C.; Gordon, S.; Carpenetti, D.
J. Org. Chem. 2006, 71, 7940–7946.
(19) Al-Dulayymi, A.; Li, X.; Neuenschwander, M. HelV. Chim. Acta 2000,
83, 1633–1644.
(9) Mills, N. S. J. Am. Chem. Soc. 1999, 121, 11690–11696.
(10) Mills, N. S.; Benish, M. M.; Ybarra, C. J. Org. Chem. 2002, 67, 2003–
2012.
(20) Becker, H. D.; Gustafsson, K. J. Org. Chem. 1976, 41, 214–21.
(21) Ipaktschi, J.; Hosseinzadeh, R.; Schlaf, P.; Dreiseidler, E.; Goddard,
R. HelV. Chim. Acta 1998, 81, 1821–1834.
(11) Mills, N. S.; Levy, A.; Plummer, B. F. J. Org. Chem. 2004, 69, 6623–
33.
(22) Ipaktschi, J.; Hosseinzadeh, R.; Schlaf, P.; Dreiseidler, E. HelV. Chim.
Acta 2000, 83, 1224–1238.
(12) Mills, N. S.; Tirla, C.; Benish, M. A.; Rakowitz, A. J.; Bebell, L. M.;
Hurd, C. M. M.; Bria, A. L. M. J. Org. Chem. 2005, 70, 10709–
10716.
(23) Olah, G. A. Angew. Chem., Int. Ed. Engl. 1995, 34, 1393–405.
(24) Olah, G. A.; Schlosberg, R. H. J. Am. Chem. Soc. 1968, 90, 2726–7.
9
10.1021/ja711501k CCC: $40.75
2008 American Chemical Society
J. AM. CHEM. SOC. 2008, 130, 10179–10186 10179

A R T I C L E S
Dahl and Mills
Scheme 1
several bisfluorenyl spacer-separated diether systems 6a-c were
synthesized (Scheme 1). The corresponding nonadjacent dica-
tions 6a2+ and 6b2+ were generated by Magic acid ionization
and characterized at low temperatures with 1H NMR spectros-
copy. Dication 6c2+ could not be reliably generated. The
1
1
experimental H NMR shifts along with calculated H NMR
shifts of 6a2+ and 6b2+, nucleus independent chemical shift
(NICS) values, and magnetic susceptibility exaltation (Λ)
calculations for 6a2+-c2+ were analyzed to evaluate the relative
antiaromaticity of the dications and compared to that of the well
studied tetrabenzo[5,5]fulvalene dication system.5,9,32–34 The
quality of the calculated NICS values was evaluated by
comparison of H NMR shifts, which could be compared to
experimental values to determine their accuracy, calculated by
the same method.
1
by way of ionization reactions. These superacids could be useful
in generating antiaromatic dications of neutral precursors
containing readily ionizable lone pair containing functionalities
such as alcohols and ethers, thus expanding the library of
compounds capable of being studied. In theory, this approach
to 1,2-dications is thermodynamically more favorable than
alkene oxidation because the bond breaking is partially com-
pensated by the creation of the new protonated bonds. However,
although the presence of such intermediates is often indirectly
supported, experimentally obserVed 1,2-carbodicationic species
from 1,2-diols or diethers is unknown in the literature, presum-
ably due to the propensity for fast cyclization, rearrangement,
or elimination reactions of those intermediates in the superacidic
media.25 In unpublished studies by our group, attempted
preparation of dications such as 3, R ) 3-CH3, resulted in rapid
cyclization when the dications were formed by oxidation with
SbF5/SO2ClF, and would also be predicted to be inaccessible
by superacid ionization of its 1,2-diol analogue. Therefore, in
order to employ superacidic ionization routes for the production
of experimentally obserVable antiaromatic dications, neutral
precursors with ionizable moieties on nonadjacent carbon atoms
must be utilized.
Possible precursors that would be suitable to afford antiaro-
matic carbodications via superacid ionization are neutral di-
methyl ethers with conjugating phenyl, ethynyl, or ethenyl
“spacer” units connecting the ionizable moieties, such as 5a-c.
Conjugated spacer separated dications and dianions of an
aromatic nature have been reported in the literature,26–31but
similar antiaromatic species are thus far unknown. The diether
functionality has been chosen due to the relative ease of
purification for the neutral precursors and the potential for fewer
side reactions upon superacid ionization when compared to the
corresponding diol species. The spacer unit would serve three
purposes: (1) it hinders cyclization of less stable antiaromatic
dications by rendering the products more sterically demanding
in comparison to the cyclized products of dications with no
spacer unit, (2) it avoids the aforementioned rapid decomposition
of 1,2-dications generated with superacids, and (3) it provides
extra stabilization to the cationic species through delocalization.
Results and Discussion
Synthesis of Precursors to Dications. Phenyl bridged dimethyl
ether 6a was synthesized by double lithium-halogen exchange
on p-dibromobenzene and subsequent addition to 9-fluorenone
to achieve diol 7 (Scheme 2). The diol was then methylated
with NaH and MeI resulting in 6a. Ethynyl and ethenyl bridged
dimethyl ethers 6b and 6c were synthesized by the addition of
ethynylmagnesium bromide to 9-fluorenone. Alkyne deproto-
nation of 8 and a second addition of 9-fluorenone gave diol 9.
The diol 9 was methylated with NaH and MeI, resulting in 6b.
Lithium aluminum hydride reduction of 9 resulted in olefin diol
10 which was methylated to produce 6c.
In order to make a direct comparison between olefin oxidation
and methyl ether ionization as routes toward antiaromatic
dications, the syntheses of the corresponding bridged olefins
were examined. Cyclohexadiene-bridged bisfluorene 14, the
neutral olefin precursor that should produce 6a2+ upon oxidation,
proved to be elusive following two independent synthetic routes.
The first route was a stepwise double Peterson olefination of
monoketal protected 1,4-cyclohexadione with fluorene to ulti-
mately achieve 13. Unfortunately, oxidation of 13 with DDQ
under several different conditions produced no desired product
(Scheme 3). SnCl2 reduction of phenyl bridged diol 7 was also
unsuccessful, although the product was observed transiently as
evidenced by the bright purple color that was observed
immediately after the Sn(II) addition. To our disappointment,
the product decomposed immediately upon reaction workup.
In other reports synthetic attempts to achieve 1435 and similar
species21,22 were also unsuccessful via Zn or Hg reduction of
the dibromo analogue of 7. These reports suggest that 14 is
very difficult to isolate due to the formation of a cyclic tetramer
(25) Nenajdenko, V. G.; Shevchenko, N. E.; Balenkova, E. S.; Alabugin,
I. V. Chem. ReV. 2003, 103, 229–282.
(26) Kagayama, A.; Komatsu, K.; Nishinaga, T.; Takeuchi, K.; Kabuto,
C. J. Org. Chem. 1994, 59, 4999–5004.
(27) Eicher, T.; Berneth, H. Tetrahedron Lett. 1973, 2039–42.
(28) Gilbertson, R. D.; Weakley, T. J. R.; Haley, M. M. J. Org. Chem.
2000, 65, 1422–1430.
(29) Eicher, T.; Berneth, H. Tetrahedron Lett. 1973, 14, 2039–2042.
(30) Komatsu, K. A., M.; Arai, M.; Okamoto, K. Tetrahedron Lett. 1982,
23, 91–94.
(31) Komatsu, K.; Arai, M.; Hattori, Y.; Fukuyama, K.; Katsube, Y.;
Okamoto, K. J. Org. Chem. 1987, 52, 2183–2192.
(32) Malandra, J. L.; Mills, N. S.; Kadlecek, D. E.; Lowery, J. A. J. Am.
Chem. Soc. 1994, 116, 11622–11624.
(33) Mills, N. S.; Burns, E. E.; Hodges, J.; Gibbs, J.; Esparza, E.; Malandra,
J. L.; Koch, J. J. Org. Chem. 1998, 63, 3017–3022.
(34) Mills, N. S.; Benish, M. A. J. Org. Chem. 2006, 71, 2207–2213.
(35) Wittig, G.; Dreher, E.; Reuther, W.; Weidinger, H.; Steinmetz, R.
Leibigs Ann. Chem. 1969, 726, 188–200.
In order to establish proof of concept for this novel and
potentially scope-expanding route to antiaromatic carbodications,
9
10180 J. AM. CHEM. SOC. VOL. 130, NO. 31, 2008

Antiaromatic Spacer-Bridged Bisfluorenyl Dications
A R T I C L E S
Scheme 2
Scheme 3
Scheme 4
1
from a diradical intermediate in the reduction process. SnCl2
reduction was successful, however, in producing butatriene 15
and diene 16 from their corresponding diols (9 and 10), which
should produce 6b2+ and 6c2+, respectively, upon oxidation.
Additionally, ethynyl bridged bisdibenzocycloheptatriene
dimethyl ether 17 and its corresponding butatriene 18 (Scheme
4) were synthesized in order to determine whether formation
of aromatic ethynyl spacer-separated dication 172+ using either
superacid ionization or oxidation methods was also feasible.
samples were immediately analyzed by H NMR at -50 °C.
The experimental H NMR shifts for 6a2+ and b2+ as well as
1
the calculated chemical shifts are given in Table 1. Dication
6c2+ could not be reliably observed in a reproducible manner
1
due to its probable instability. The average experimental H
NMR shift for the fluorenyl protons of 6a2+, 5.865 ppm, shows
a substantial paratropic shift compared to the average shift for
the phenyl protons of 19a2+, 8.22 ppm. Dication 6b2+ with an
average shift for the fluorenyl system of 5.520 shows an even
larger paratropic shift in comparison with the phenyl protons
of 19b2+, 8.28 ppm, as has been seen for related species
previously.32
1H NMR Shifts for Dications Formed by Ionization. The
dimethoxy (6a-c) precursors to 6a2+-c2+ were reacted with
Magic acid (1:1 FSO3H/SbF5) at -78 °C in an NMR tube. The
9
J. AM. CHEM. SOC. VOL. 130, NO. 31, 2008 10181

A R T I C L E S
Dahl and Mills
Table 1. Experimentala and Calculatedb 1H NMR Shifts (ppm) for
6a2+and b2+ and 1, as well as the Estimated Shift of 1 in Magic
Acid, Vide infra
6a2+
6b2+
1
expt
calcd
expt
calcd
exptc estimatedd
calcd
1,8
2,7
3,6
4,5
5.980 6.830 5.638 6.714 5.33
5.670 6.780 5.346 6.491 5.16
6.161 7.478 5.903 7.106 5.77
5.650 6.734 5.196 6.371 4.97
5.23
5.06
5.67
4.87
6.523
6.328
7.044
6.228
spacer
fluorenyl
proton ave.
6.885 8.035
5.865
-
-
5.520
5.21
The relationship between the experimental shifts and those
calculated for 6a2+-b2+ and 1 are shown in Figure 1. It is
apparent that the agreement between the experimental and
calculated shifts is reasonable, based on the quality of the
correlation coefficient.
a Spectra obtained in a mixture of SO2ClF, FSO3H, and SbF5, with an
external capillary of d6-acetone, -50 °C. b NMR shifts calculated using
the GIAO method in Gaussian 03 with basis set B3LYP/6-311+g(d,p)
on geometries optimized with basis set B3LYP/6-31 g(d). c Spectra
obtained in SbF5/SO2ClF; see refs 5 and 24. d Calculated from the
experimental shifts for 1, using the upfield shift seen for 6b2+ when
formed in Magic acid compared to SbF5/SO2ClF, Vide infra.
The comparison of the calculated 1H NMR shifts to the
experimental ones is important in the validation of the calculated
NICS values, Vide infra, but they also serve to support the
geometry calculated for each species. NMR shifts have been
shown to be very sensitive to the geometry of the species under
examination.36,37 For 6a2+, geometry optimization resulted in
two geometries, one in which the phenyl spacer had a dihedral
angle of 90° and one with a dihedral angle of 135°, the latter
having the best agreement with experimental chemical shifts
(Figure 1a). This geometry, which frequency calculations
showed to be a minimum on the potential energy surface, was
also about 8 kcal/mol lower in energy. The plot of the calculated
and experimental shifts for the other geometry are given in the
Supporting Information. Similarly, optimization of 6b2+gave
two geometries, one with the fluorenyl systems perpendicular
to each other and another in which the fluorenyl systems had a
dihedral angle of 113°. The energy difference between the two
geometries was <1 kcal/mol, with the geometry in which the
fluorenyl systems was perpendicular of slightly lower energy.
The agreement between the calculated and experimental shifts
was approximately the same for each species, and the calculated
shifts for the slightly lower energy species are shown in Figure
1b. Finally, comparison of experimental shifts of 1 with the 1H
NMR shifts also calculated with solvent are shown in Figure
1c, along with the composite plot for all species in Figure 1d.
Plots of experimental vs calculated shifts in the Supporting
Information also demonstrate that inclusion of solvent in the
calculated chemical shifts is important for good agreement with
experimental shifts.
unsaturated precursor to 6c2+, 16, resulted in a very complex
spectrum; see Supporting Information, from which no useful
information could be obtained but which is consistent with our
failure to reliably prepare 6c2+ by ionization. Oxidation of 15
also gave a complex mixture, but the predominant species
formed immediately upon mixing 15 with the oxidizing solution,
SbF5/SO2ClF, appeared to be 6b2+. With time, these resonances
disappeared, to be replaced by apparent decomposition products.
Comparison of the data for 6b2+ from ionization with that from
oxidation showed that the peaks in the Magic acid/SO2ClF
solution were shifted upfield by ∼0.10 ppm relative to the peaks
in the SbF5/SO2ClF solution. To the authors’ knowledge, these
appear to be the first reported data on a dication formed both
by ionization in Magic acid/SO2ClF and by oxidation with SbF5/
SO2ClF. Based on the chemical shift difference for 6b2+ in the
two solvents, we estimated the chemical shift for 1 in Magic
acid, which is given in Table 1 and which was used in Figure
1d.
The average chemical shift of the fluorenyl system in 6a2+and
6b2+ and in 1 shows the following order of antiaromaticity,
6a2+ < 6b2+ < 1. The shift for protons 1,8 and, to a lesser
extent, protons 2,7 of 1 are further downfield than would be
expected based solely on the antiaromaticity of the fluorenyl
ring system; the two fluorenyl rings of 1 are almost perpen-
dicular to each other, allowing the ortho protons on one ring
system to feel the effect of the center of the opposing
antiaromatic fluorenyl system.5,32 This results in an additional
diatropic shift for those protons.
1H NMR Shifts for Dications Formed by Oxidation. One goal
in this study was to evaluate the effect of spacers on the
magnitude of the antiaromaticity of dications related to species
such as 1. It is difficult to compare the chemical shift of 1 with
the chemical shifts of 6a2+and 6b2+ because the media, Magic
acid/SO2ClF for ionization and SbF5/SO2ClF for oxidation, are
different. We attempted the ionization of 9,9-dihyroxy-
tetrabenzo[5.5]fulvalene with Magic acid/SO2ClF but were
unable to obtain 1 by this method. Olah et al. were also unable
to form dications from 1,2-diols,38 because of other avenues
for reaction in these species. Thus, in order to make a proper
comparison, formation of 6a2+-c2+ by oxidation was attempted.
As described previously, attempts to make the unsaturated
precursor to 6a2+, 14, were unsuccessful. Oxidation of the
Because the quality of the spectra for the antiaromatic dication
6b2+ formed by oxidation was relatively poor, we examined
the feasibility of forming an analogous bis-aromatic dibenzot-
ropylium dication with a spacer by the oxidation of cumulene
18. While dication formation was successful for this system,
1
the H NMR spectrum (see Supporting Information) for the
dication formed by ionization of dimethoxy 17 had much better
resolution. Comparison of calculated with experimental 1H NMR
shifts, given in the Supporting Information, shows the same
excellent relationship as that for the antiaromatic dications with
spacers. Thus, the ionization method appears to be advantageous
over oxidative methods in the generation of spacer-separated
dications.
(36) Schleyer, P. l. v. R.; Laidig, K.; Wiberg, K. B.; Saunders, M.;
Schindler, M. J. Am. Chem. Soc. 1988, 110, 300–1.
(37) Siehl, H.-U.; Fuss, M.; Gauss, J. J. Am. Chem. Soc. 1995, 117, 5983–
91.
Nucleus Independent Chemical Shifts. A second way of
evaluating the antiaromaticity of the 6a2+and 6b+ is through
(38) Olah, G. A.; Grant, J. L.; Spear, R. J.; Bollinger, J. M.; Serianz, A.;
Sipos, G. J. Am. Chem. Soc. 1976, 98, 2501–2507.
9
10182 J. AM. CHEM. SOC. VOL. 130, NO. 31, 2008

Antiaromatic Spacer-Bridged Bisfluorenyl Dications
A R T I C L E S
Figure 1. Calculated vs experimental 1H NMR shifts for 6a2+-b2+: (a) 6a2+; (b) 6b2+; (c) 1. (d) Composite for 6a2+, 6b2+, and 1. Experimental shifts
for 1 are those estimated for Magic acid solution, Vide infra.
Scheme 5. NICS(1)zz for 6a2+, 6b2+, and 1, including ΣNICS(1)zz
the use of nucleus independent chemical shifts39(NICS) which
for the Fluorenyl Systemsa
evaluate the chemical shift of a dummy atom placed at the center
of the aromatic or antiaromatic ring system. Because the
electrons of the σ-system may affect the chemical shift
calculated for the dummy atom, it is normally placed 1 Å above
the plane of the ring, NICS(1). In addition, because the ring
current is due to the electrons in the π-system, the magnetic
tensor of the chemical shift perpendicular to the plane of the
ring, NICS(1)zz, is used to evaluate aromaticity or antiaroma-
ticity.40 In the traditional model of a ring current, protons on
the periphery of an aromatic system are shifted downfield and
have more positive values for the chemical shift, while a probe
a Calculated with the GIAO method, basis set B3LYP/6-311+g(d,p) for
a dummy atom 1 Å above the plane of the ring, magnetic tensor
perpendicular to the ring system.
in the center of the ring system would be shifted upfield and
would have a more negative value for the chemical shift. Thus
aromatic compounds have negative NICS while aromatic
systems have positive NICS.
The NICS(1)zz values calculated for the optimized geometries
with the lowest energies are shown in Schemes 1 and 5.
According to this method of evaluation, the order of antiaro-
maticity is 6a2+ < 6b2+ < 1.
Magnetic Susceptibility Exaltation, Λ. A third method for
the evaluation of aromaticity or antiaromaticity is through the
enhancement of the magnetic susceptibility of the species due
to the existence of the ring current, magnetic susceptibility
exaltation, or Λ.41–43 Briefly, the magnetic susceptibility of the
species in question is calculated with the CSGT method. The
magnetic susceptibility, ꢀ, for the species with localized bonds
is obtained through a summation of bond increments (see
(39) Schleyer, P. v. R.; Maerker, C.; Dransfeld, A.; Jiao, H.; Hommes,
N. J. v. E. J. Am. Chem. Soc. 1996, 118, 6317–6318.
(41) Snyder, J. P. Nonbenzenoid Aromatics; Academic Press: New York,
1971; Vol. 2, pp 167-180.
(40) Fallah-Bagher-Shaidaei, H.; Wannere, C. S.; Corminboeuf, C.; Puchta,
R.; Schleyer, P. v. R. Org. Lett. 2006, 8, 863–866.
(42) Gayoso, J.; Ouamerali, O. ReV. Roum. Chim. 1981, 26, 1035–40.
(43) Haberditzl, W. Angew. Chem., Int. Ed. Engl. 1966, 5, 288–298.
9
J. AM. CHEM. SOC. VOL. 130, NO. 31, 2008 10183

A R T I C L E S
Dahl and Mills
Table 2. Magnetic Susceptibility, ꢀ, Magnetic Susceptibility
Exaltation, Λ, and ΣNICS(1)zz for the Fluorenyl Systems of
6a2+-c2+ and 1a
spacer to delocalize charge results in an attenuation of the
antiaromaticity of the fluorenyl system. Of greater importance
is the observation that with the inclusion of a spacer, even
one that is capable of a great deal of stabilization such as phenyl,
the antiaromaticity of the fluorenyl system is maintained. This
suggests that derivatives of dications previously unknown, such
as 3 and 4, should be accessible through the use of spacers.
Interaction of the fluorenyl ring systems. We have demon-
strated that the antiaromaticity of the fluorenyl system in dication
1 is markedly greater than that in the fluorenyl monocation 20
and attributed this to the effect of a positively charged substituent
in position 9 of the fluorenyl cation, as shown in general terms
as 21.9,32 We have successfully demonstrated that the degree
of antiaromaticity in the fluorenyl cation of 21 is affected by
the nature of Z. In many of the systems we have explored, Z
was a ring system, such as an indenyl cation,8 or dibenzopy-
rylium, dibenzotropylium, or dibenzocyclohexadienyl cation,5,32
and geometry optimization showed that the ring system of Z
was often perpendicular to the fluorenyl ring system. Since these
systems appear to be much more antiaromatic than 20, the
cationic substituent Z in 21 must be affecting the electronic
nature of the fluorenyl cation. However, since resonance bond
delocalization is not possible for these effectively perpendicular
and thus orthogonal systems, other possibilities for communica-
tion between the ring systems were explored. We had proposed
two models for interaction of the systems, σ to p donation,32
also known as cross-hyperconjugation as shown below for 1,
or electronically enforced delocalization as a consequence of
the positive charge of Z.12 In both cases, this requires that the
fluorenyl cation and positively charged substituent be in close
proximity. This is obviously not the case for 6a2+ and 6b2+, in
which each cation might be expected to behave independent of
each other and more like fluorenyl monocations, 20,45 due to
separation of charge. The fluorenyl monocation with a phenyl
substituent is known (20a),46 but we also attempted the
preparation of 20b with an ethynyl substituent and 20c with an
ethenyl group. Ionization of either the alcohol precursor to 20b
and 20c or the methoxy precursor resulted only in a polymeric
product, presumably through reaction of small amounts of the
initially formed monocation with an un-ionized precursor.
ꢀ, cgs ppm
Λ, cgs ppm
ΣNICS(1)zz
6a2+
6b2+
1
-136.40
-83.61
-62.05
65.57
82.76
92.19
191.32
233.71
264.90
a Magnetic susceptibility calculated with the CSGT method with basis
set B3LYP/6-311+g(d,p) on geometries optimized with basis set
B3LYP/6-31g(d). NICS calculated in a similar manner with the GIAO
method.
Figure 2. Comparison of Λ/ring area2 with ΣNICS(1)zz: (a) 6a2+, 6b2+
,
and 1; (b) 6a2+, 6bc2+, and 1 as well as neutral PAH, aromatic and
antiaromatic cations, dications, anions, and dianions; see ref 35.
The average proton shifts for 20a, the sum of the NICS(1)zz
values, and the magnetic susceptibility exaltation, Λ, are shown
Supporting Information), and the difference between the two
gives Λ. Table 2 gives ꢀ and Λ for 6a2+, 6b2+, and 1.
The values of Λ show the order of antiaromaticity to be 6a2+
< 6b2+ < 1, the same order shown by the NICS(1)zz calcula-
tions. We have recently shown44 an excellent correlation
between the sum of NICSS(1)zz and Λ. Magnetic susceptibility
exaltation is known to be dependent upon the square of the ring
area,42 and we demonstrated that the ΣNICS(1)zz is also
in Table 3 along with the corresponding data for 6a2+
.
Calculated data (NICS(1)zz and Λ) for 6b2+ and 20b are
included in the table for comparison purposes. To allow proper
comparison, the values for the fluorenyl systems of the dications
with spacers must be divided by the number of fluorenyl
systems. Comparison of the data in Table 3 shows that neither
6a2+ nor 6b2+ is behaving like a fluorenyl monocation. The
fact that the dications with spacers are substantially more
antiaromatic than the analogous monocations suggests that there
must be some mechanism for the fluorenyl systems to com-
dependent on ring area squared. That relationship for 6a2+, 6b2+
,
and 1 is shown in Figure 2a. Inclusion of neutral PAH, aromatic
and antiaromatic cations, dications, anions, and dianions gives
the plot shown in Figure 2b, confirming that there is nothing
unusual about the behavior of 6a2+, 6b2+, and 1.
(45) We appreciate the comments of a reviewer who suggested that we
consider the behavior of fluorenyl monocations.
(46) Olah, G. A.; Prakash, G. K. S.; Liang, G.; Westerman, P. W.; Kunde,
K.; Chandrasekhar, J.; Schleyer, P. v. R. J. Am. Chem. Soc. 1980,
102, 4485–4492.
The order of antiaromaticity from 1H NMR shifts, NICS, or
Λ is 6a2+ < 6b2+ < 1. It is apparent that the ability of the
(44) Mills, N. S.; Llagostera, K. B. J. Org. Chem. 2007, 72, 9163–9169.
9
10184 J. AM. CHEM. SOC. VOL. 130, NO. 31, 2008

Antiaromatic Spacer-Bridged Bisfluorenyl Dications
A R T I C L E S
Table 3. Comparison of the Average 1H NMR Shift, Where
Available, Summation of NICS(1)zz and Λ for the Fluorenyl
1.4596 Å compared to 1.4383 Å for 20a. Since the bond in the
dicationic system is significantly longer, the presence of a
cationic substituent in 6a2+ appears to effectively reduce the
delocalization of the charge into the phenyl system compared
to 20a. A similar trend was shown for 6b2+ and 20b, 1.4030 Å
compared to 1.3957 Å. The shorter bond length in the species
with ethynyl substituents compared to those with aryl substit-
uents is expected due to the sp-hybridization of the ethynyl
carbon compared to the sp2-hybridization of the aryl carbon.
Systems of 6a2+ a 20a, 6b2+ a
, , and 20b
6a2+b
20ab
6b2+
20b
ΣNICS(1)zz
96.55
32.78
5.865
56.33
18.62
7.42
116.86
41.38
5.520
83.30
27.13
NA
Λ
average proton shift
a Values of ΣNICS(1)zz and Λ for 6a2+ and 6b2+ were divided by
the number of fluorenyl rings. b The optimized geometry for 6a2+ and
20a showed that the dihedral angle of the phenyl ring with the fluorenyl
system was ∼135° for both systems.
Summary
municate with each other that is of necessity different from the
mechanisms proposed previously for 1,2-dications such as 1.
Derivatives of the dication of tetrabenzo[5.5]fulvalene with
phenyl, 6a2+, and alkynyl spacers, 6b2+, have been prepared
by ionization of the dimethoxy precursors with Magic acid in
SO2ClF. The dication with an ethynyl spacer, 6b2+, could also
be prepared by oxidation of the appropriate unsaturated precur-
sor in SbF5/SO2ClF, allowing evaluation of the effect of the
1
different media on the H NMR shifts of the dication. This
allowed the shift of the parent dication 1, prepared by oxidation
to be directly compared to the shifts of 6a2+ and 6b2+, prepared
by ionization of the corresponding diether. The antiaromaticity
of the fluorenyl systems of 6a2+ and 6b2+ was assessed through
experimental measurement of their 1H NMR shifts and through
calculation of their nucleus independent chemical shifts and
magnetic susceptibility exaltation, Λ. A comparison of NICS
and Λ to that of the parent dication of tetrabenzo[5.5]fulvalene
reveals that the antiaromaticity of the fluorenyl system is
maintained although it is attenuated by the spacers, with an order
of increasing antiaromaticity of 6a2+ < 6b2+ < 1. In addition,
the ability of remote fluorenyl cations to interact through spacers
appears to be related to the effect of the remote substituent on
the delocalization of the positive charge of the fluorenyl
substituent. The ability to prepare dications with spacers which
maintain their antiaromaticity should allow the preparation of
dications which contain antiaromatic systems such as unsub-
stituted indenyl cations. We are currently examining this type
of species in our laboratories.
Interaction of Fluorenyl Systems Separated by Spacers. A
possible explanation for the effect that a distal cationic fluorenyl
system has on the antiaromaticity of another could lie in the
effect of that substituent on the delocalization of the positive
charge of the dication. That is, the presence of a second charge
might inhibit the delocalization of charge in the spacer.47 Even
though the phenyl is not coplanar with the fluorenyl system for
either 6a2+ or 6b2+, others have demonstrated that compounds
in which the interacting units have dihedral angles similar to
those in 6a2+ and 20a show interactions consistent with
delocalization between the units.48 The extent of delocalization
might be examined through a comparison of the 13C NMR shifts
of the phenyl substituent of 6a2+ and 20a. However, the analysis
of this (See Supporting Information) was complicated by the
comparison of the shift of the para carbon of the phenyl ring
which is adjacent to a hydrogen in 20a with the corresponding
carbon on 6a2+ which is adjacent to a positively charged carbon
on the fluorenyl system. A similar comparison of the 13C NMR
shifts of the carbons of 2238 with those of the dimethylphenyl-
methyl cation49 showed a similar variation in shift, suggesting
that the nature of the positively charged substituent, antiaromatic
vs nonaromatic, is not the key effect here; see Supporting
Information.
Experimental Section
Calculation of 1H NMR Shifts, NICS, and Λ. Calculation
1
of NICS(1)zz and Λ. H NMR shifts for 6a2+, 6b2+, and 1 were
calculated using the GIAO method in Gaussian 03 with basis set
B3LYP/6-311+g(d,p) on structures optimized with basis set
B3LYP/6-31g(d). Geometries for hydrocarbons calculated with
density functional theory at this level have been shown to agree
well with experimental data.50–52 The 1H NMR shifts were
calculated with the inclusion of solvent using the PCM method in
Gaussian 03, with DMSO chosen as the solvent. NICS(1)zz was
calculated with the GIAO method with basis set B3LYP/6-31 using
the component of the magnetic shift tensor in the z direction,
perpendicular to the plane of the ring, for a dummy atom 1 Å above
the plane of the ring.40
Magnetic susceptibility exaltation, Λ, is the difference between
the calculated magnetic susceptibility, ꢀ, for the molecule of interest
and that of a reference system. The reference system is formed by
the summation of ꢀ for increments representing the bonds of the
localized species. Details of the calculations can be found in the
Supporting Information.
An alternative measure of relative delocalization would be
the comparison of the bond lengths between the positively
charged carbon on the fluorenyl system and the adjacent ipso
carbon of the phenyl substituent in the monocationic and
dicationic systems. The calculated bond length for 6a2+ was
(50) Evdokimov, A. G.; Kalb, A. J.; Koetzle, T. F.; Klooster, W. T.; Martin,
J. M. L. J. Phys. Chem. A 1999, 103, 744–753.
(47) Klumpp, D. A. Chem.sEur. J. 2008, 14, 2004–2015.
(48) Allen, B. D.; Benneston, A. C.; Harriman, A.; Llarena, I.; Sams, C. A.
J. Phys. Chem. A 2007, 111, 2641–2649.
(51) Rasul, G.; Prakash, G. K. S.; Olah, G. A. J. Org. Chem. 2000, 65,
8786–8789.
(49) Olah, G. A.; Westerman, P. W. J. Am. Chem. Soc. 1973, 95, 7530–
7531.
(52) Nendel, M.; Houk, K. N.; Tolbert, L. M.; Vogel, E.; Jiao, H.; Schleyer,
P.v. R. J. Phys. Chem. A 1998, 102, 7191–7198.
9
J. AM. CHEM. SOC. VOL. 130, NO. 31, 2008 10185

A R T I C L E S
Dahl and Mills
for alternative geometries of 6a2+and 6b2+; plots of experi-
mental 1H NMR shifts and shifts calculated without solvent for
6a2+ and 6b2+; increments for calculation of magnetic suscep-
tibility for localized dications; calculated total energies, and
[x,y,z] coordinates for 6a2+-c2+. NMR spectra of 6a-c, 7-13,
and 16-18. This material is available free of charge via the
Internet at http://pubs.acs.org.
Acknowledgment. We thank the Welch Foundation (Grant
W-794) and the National Science Foundation (Grants CHE-
0242227, CHE-0553589, CHE-0552292, and CHE-0138640) for
their support of this research.
Supporting Information Available: Experimental procedures
for the preparation of all compounds. Spectra of 6a2+-c2+ from
ionization of dimethoxy compounds 6a-c and 17 and from
ionization of 7, 9, and 10; spectra from oxidation of 15, 16,
and 18; plots of the calculated and experimental 1H NMR shifts
JA711501K
9
10186 J. AM. CHEM. SOC. VOL. 130, NO. 31, 2008