Preparation and structure determination of a stable cis-Bis-σ- homobenzene derivative

Article abstract of DOI:10.1002/ejoc.201000114

By cyclopropanation of tetramethyl[2,2]paracyclophane 9 the tetramethylene adduct 10 has been prepared. This hydro carbon is the first stable cis-bis-σ-homobenzene for which a structure has been determined by X-ray diffraction.

Full text of DOI:10.1002/ejoc.201000114

FULL PAPER
DOI: 10.1002/ejoc.201000114
Preparation and Structure Determination of a Stable cis-Bis-σ-homobenzene
Derivative[‡]
Dieter Wullbrandt,[a] Henning Hopf,*[a] and Peter G. Jones[b]
Keywords: Cyclophanes / Cyclopropanation / Homobenzenes
By cyclopropanation of tetramethyl[2.2]paracyclophane 9 the
tetramethylene adduct 10 has been prepared. This hydro-
carbon is the first stable cis-bis-σ-homobenzene for which a
structure has been determined by X-ray diffraction.
Introduction
withdrawing substituents such as the nitrile function favor
The formal addition of methylene carbene to benzene (1) the “closed” norcaradiene form 2.[3] If another equivalent
results in the generation of homobenzene (2, norcaradiene). of the carbene is added to 2, bis-σ-homo-benzenes result;
It has long been known that this hydrocarbon prefers its since methylenation can take place either at the same or the
valence-isomeric form, 1,3,5-cycloheptatriene (5, R = H: opposite side of the three-membered ring already present,
tropylidene)[2] into which it is converted by an electrocyclic cis- (3) and trans-bishomobenzene (6) can be produced. Fi-
ring opening process. In fact, the equilibrium between these nally, if a third methylenation is carried out either all-cis-
tautomers can be influenced by various factors, among (4) or trans-tris-σ-homobenzene derivatives result, respec-
them the electronic nature of the substituents R at the tively. In fact the information available on 3 and 4 is rather
methylene group of the three-membered ring in 2. Electron- scarce, especially as far as X-ray structures are concerned. A
Scheme 1. The cyclopropanation of benzene (1) and the [2.2]paracyclophane 9.
[‡] Cyclophanes, LXV. Part LXIV: Ref.[1]
CCDC[4] search carried out for the tricyclic core framework
[a] Institut für Organische Chemie, Technische Universität
tricyclo[5.1.02,4]octane yields 35 hits, none including the X-
Braunschweig,
Hagenring 30, 38106 Braunschweig, Germany
Fax: +49-531-391-5388
ray structure of a cis-bis-σ-homo-benzene. Most derivatives
of this particular polycyclic hydrocarbon are trans-config-
ured (6),[5] and those prepared in a cis-configuration are
evidently only generated as reactive and/or thermally labile
intermediates. As far as the all-cis-tris adducts are con-
cerned, the parent hydrocarbon has escaped all attempts at
E-mail: H.Hopf@tu-bs.de
[b] Institut für Anorganische und Analytische Chemie, Technische
Universität Braunschweig,
Postfach 3329, 38023 Braunschweig, Germany
Fax: +49-531-391-5387
E-mail: P.Jones@tu-bs.de
4046
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 4046–4048

Preparation and Structure of a Stable cis-Bis-σ-homobenzene
preparation so far. The only stable adducts characterized so C7 and C8, 119.2–120.0° at C4 and C5 (junctions to the
far (including X-ray analyses) are the tris-nitrile 4 (R = CN)[6] cyclopropane rings) and 116.7–117.6° at the bridgehead
and the triscyclopropanation product of tris(benzocyclo- atoms C3 and C6 (see below). Within the cyclopropane
butadieno)cyclohexatriene.[7] Stability has also been intro- rings, the “outer” bonds C3–C9 and C6–C10 are the longest
duced into these systems by removing the “internal” meth- at 1.518–1.529 Å; the other bond lengths are 1.501–1.511 Å.
ylene hydrogen substituents by placing a bridging carbon The cyclopropane rings subtend interplanar angles of 109–
atom into the center of 4 (R = H) as shown by diademane 110° to the best planes defined by the four atoms including
or mitrane (7).[8] The price paid here, of course, is that any the hinge bond, e.g. C5,4,3,8 for the ring C3,4,9.
interaction between these crowded atoms (or substituents)
can no longer be studied (Scheme 1).
Despite the partial saturation of the six-membered rings,
both molecules show some, but not all, of the distortions
normally associated with the [2.2]paracyclophane frame-
work. The flattened boat form of the six-membered rings is
present, but less pronounced than is usual; the deviations
of the bridgehead atoms C3 and C6 from the best plane of
the other four atoms amount to 0.097–0.115 Å, whereas the
normal amount is ca. 0.15 Å (mean of 326 hits for [2.2]-
paracyclophanes from the Cambridge Database,[4] not fur-
ther sorted). Similarly, the distance between bridgehead
atoms of opposite rings is 3.11, 3.12 Å (cf. 2.77 Å from the
above search) and the bridge single bonds (C1–C2) are not
significantly lengthened (1.536, 1.539 Å in 10, cf. 1.578 Å
from the above search). The angles within the six-membered
rings at the bridgeheads are 116.7–117.6°, which is normal
for paracyclophanes (search result 116.9°). The main strain
of the molecules in 10 appears to be taken up in the for-
mally sp3 C–C–C bridge angles, for which the average
search value of 112.8° is greatly increased – to as much as
119.4–120.6° in 10.
Results and Discussion
To prepare stable homobenzenes we have developed an-
other concept: to force all methylene groups into one orien-
tation, viz. into the cis-configuration, we bridged one of the
hemispheres of the benzene ring, i.e. employed a cyclophane
as the substrate to be methylenated. The concept is illus-
trated in Scheme 1 for the tetramethyl derivative 9, a hydro-
carbon readily obtained by reduction of the four ester sub-
stituents of 8.[9] When 9 was cyclopropanated with diazo-
methane/CuCl in dichloromethane, two products were
formed in 3 and 7% yield, which according to spectro-
scopic and analytical data were bis and tetra adducts of
methylene to 9.[10] Clearly a fully cyclopropanated deriva-
tive of 9 is not produced under these conditions. Both of
our products are stable cis-bis-σ-homobenzenes that can be
kept unchanged at room temperature for years, and for the
tetra adduct 10, crystals suitable for X-ray diffraction could
be grown from trichloromethane/methanol.
The structure reveals two closely similar, inversion-sym-
metric, independent molecules, one of which is shown in
Figure 1. The only remaining double bonds in the six-mem-
bered rings, C7–C8, are clearly the shortest at 1.349,
1.339 Å (esd’s of bond lengths and angles are 0.003 Å and
0.2° respectively; where ranges of values are given, the indi-
vidual values can be found in the supplementary crystallo-
graphic data). All other bond lengths of these rings corre-
spond to shortened single bonds (1.488 Å for the bond join-
ing the cyclopropyl rings, 1.497–1.508 Å for the other
bonds). The bond angles within the six-membered rings are
all close to sp2 values, with 122.3–123.2° at the sp2 atoms
Conclusions
[2.2]Paracyclophanes have been shown to be suitable pre-
cursors for cis-bis-σ-homobenzenes. To the best of our
knowledge the structure reported here for the tetra methyl-
enation product 10 is the first for a representative of this
interesting class of hydrocarbons.
Experimental Section
General: N-Nitroso-N-methylurea (17 g, 0.16 mol) was added in
0.5 g portions over 10 h to an aqueous solution of potassium hy-
droxide (40%; 250 mL), which had been covered by a layer of dec-
alin (80 mL). The diazomethane generated at the phase boundary
was purged by a stream of nitrogen (250 to 300 mL/min) through
a drying tower filled with KOH pellets into a solution of 4,5,12,13-
tetramethyl[2.2]paracyclophane (9; 2.0 g; 7.6 mmol) in anhydrous
dichloromethane (60 mL) to which 0.4 g of CuCl had been added.
After completion of the cyclopropanation the reaction mixture was
passed through a short column filled with silica gel, the solvent was
removed, and the oily residue separated by preparative thick layer
chromatography (silica gel; n-hexane/dichloromethane = 97:3) to
yield two fractions. Whereas the first fraction (66 mg, 3%) was the
bis adduct, the second fraction (185 mg, 7%) was the tetra adduct
10; colorless needles (CHCl3/MeOH), m.p. 114–116 °C (decomp.).
IR (KBr): ν = 3050 (m), 3020 (s), 2920 (s), 2850 (m), 1445 (m),
˜
1
1375 (w), 1340 (m), 1067 (s), 1010 (s), 928 (w), 795 (m) cm–1. H
NMR (400.1 MHz): δ = –0.45 (m, 4 H, endo-cyclopropane-CH2),
0.27 (m, 4 H, exo-cyclopropane-CH2), 0.93 (m, 4 H, CH2CH2),
1.11 (m, 4 H, cyclopropane-CH), 1.73 (s, 12 H, CH3), 2.22 (m, 4
Figure 1. One of the two independent molecules of compound 10
in the crystal. Ellipsoids represent 50% probability levels.
Eur. J. Org. Chem. 2010, 4046–4048
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4047

D. Wullbrandt, H. Hopf, P. G. Jones
FULL PAPER
H, CH2CH2) ppm. 13C NMR (100.6 MHz): δ = 16.21 (q, CH3),
18.44 (d, cyclopropane-CH), 20.09 (s, cyclopropane-C), 21.27 (t,
cyclopropane-CH2), 34.15 (t, CH2CH2), 127.56 (s, C=C) ppm. UV
(cyclohexane): λmax (lgε) = 212 nm (3.14). MS (EI, 70 eV): m/z (%)
= 320 (1.6) [M+], 305 (1.6), 201 (11), 159 (11), 145 (24), 128 (32),
119 (32), 105 (19), 91 (41), 44 (100), 41 (50). C24H32 (320.53): calcd.
C 90.00, H 10.00; found C 89.79, H 10.29.
[1] C. Werner, H. Hopf, J. Grunenberg, L. Ernst, P. G. Jones, F.
Köhler, R. Herges, Eur. J. Org. Chem. 2009, 2621–2626.
[2] For a still valuable summary of the literature, see: G. Maier,
Valenzisomerisierungen, Verlag Chemie, 1972, chapter 6.2.2, pp
105–117; cf. J. J. Gajewski, Hydrocarbon Thermal Isomeriza-
tions, 2nd ed., Elsevier, p. 2004, chapter 2.1, pp. 176–180.
[3] E. Ciganek, J. Am. Chem. Soc. 1965, 87, 652–653; cf. E. Ci-
ganek, J. Am. Chem. Soc. 1967, 89, 1454–1458.
[4] F. H. Allen, Acta Crystallogr., Sect. B 2002, 58, 380–388.
[5] M. Yang, T. R. Webb, P. Livant, J. Org. Chem. 2001, 66, 4945–
4949; H. D. Fühlhuber, C. Gousetis, J. Sauer, H. J. Lindner,
Tetrahedron Lett. 1979, 20, 1299–1302.
[6] C. Rücker, H. Müller-Bötticher, W.-D. Braschwitz, H.
Prinzbach, U. Reifenstahl, H. Irngartinger, Liebigs Ann./Re-
cueil 1997, 967–999; cf. W.-D. Braschwitz, Th. Otten, C.
Rücker, H. Fritz, H. Prinzbach, Angew. Chem. 1989, 101, 1383–
1386; Angew. Chem. Int. Ed. Engl. 1989, 28, 1348–1351; For
the structure of a cis,cis,trans-hexamethyl derivative see C.
Krüger, P. J. Roberts, Cryst. Struct. Commun. 1974, 3, 459–462.
[7] D. L. Mohler, K. P. C. Vollhardt, S. Wolff, Angew. Chem. 1995,
107, 601–603; Angew. Chem. Int. Ed. Engl. 1995, 34, 563–565.
[8] S. P. Verevkin, M. Kummerlin, E. Hickl, H.-D. Beckhaus, C.
Rüchardt, S. I. Kozhushkov, R. Boese, R. Haag, J. Benet-Buch-
olz, K. Nordhoff, A. de Meijere, Eur. J. Org. Chem. 2002, 2280–
2287.
Crystal Structure Determination: Crystal data for 10: Triclinic,
¯
space group P1, a = 7.2772(8), b = 8.7524(10), c = 15.3361(17) Å,
α = 95.978(13), β = 90.467(14), γ = 114.338(15)°, Z = 2, crystal
0.5ϫ 0.25ϫ 0.1 mm, 29976 intensities to 2θmax 52.7°; refinement
to wR2 = 0.113, R1 = 0.043 for 222 parameters and 3589 unique
reflections; max. ∆ρ = 0.22 eÅ–3, S = 1.00.
The data were recorded on an Oxford Diffraction Xcalibur dif-
fractometer at –173 °C using Mo-Kα radiation. The structure was
refined anisotropically on F2.[11] Hydrogen atoms were included
using a riding model.
The determination of the correct unit cell and space group was not
trivial. The cell as found by the automatic diffractometer routines
was C-centred monoclinic, but we discarded this cell in favour of
the triclinic cell for two reasons. First, the monoclinic α and γ
angles deviated from 90° by some tenths of a degree; secondly, the
R(int) value for monoclinic symmetry was rather high at 0.16.
[9] M. Nagel, R. Allmann, S. El-Tamany, H. Hopf, Chem. Ber.
1982, 115, 3203–3207, and references cited therein.
¯
Structure solution was successful in P1, but the refinement none-
[10] For earlier studies on the cyclopropanation of cyclophanes, see:
K. Menke, H. Hopf, Angew. Chem. 1976, 88, 152–153; Angew.
Chem. Int. Ed. Engl. 1976, 15, 165–166; R. Näder, A. de Mei-
jere, Angew. Chem. 1976, 88, 153–154; Angew. Chem. Int. Ed.
Engl. 1976, 15, 166–167; K.-L. Noble, H. Hopf, L. Ernst,
Chem. Ber. 1984, 117, 474–488; A. de Meijere, C.-H. Lee, B.
Bengtson, E. Pohl, S. I. Kozhushkov, P. R. Schreiner, R. Boese,
Th. Haumann, Chem. Eur. J. 2003, 9, 5481–5488.
[11] G. M. Sheldrick, Acta Crystallogr., Sect. A 2008, 64, 112–122.
[12] These first experiments were carried out in 1982 together with
Prof. Dr. W. S. Sheldrick, Gesellschaft für Biotechnologische
Forschung in Braunschweig. We thank Prof. Sheldrick for the
help at that time although the results were never published.
Received: January 28, 2010
theless led to unsatisfactorily high R values. We therefore assumed
pseudo-merohedral twinning via 180° rotation about the “mono-
clinic” b axis; use of the corresponding twin matrix (1 0 0/–1 –1 0/
0 0 –1) allowed the refinement to be successfully completed. The
BASF parameter (fraction of minor twinning component) refined
to 0.318(2). This structure had been investigated in 1980,[12] but
the techniques available at the time were probably inadequate for
discovering the true cell and twinning law.
CCDC-763450 contains the supplementary crystallographic data
for this paper. These data can be obtained free of charge from The
Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/
data_request/cif.
Published Online: June 1, 2010
4048
www.eurjoc.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 4046–4048