Twin triptycyl spinning tops: A simple case of molecular gearing with dynamic C2 symmetry

Article abstract of DOI:10.1002/ejoc.200800202

Lithiation and subsequent oxidation of the recently described prototype molecular paddlewheel 9-(2-indenyl)triptycene (1) leads to the racemic dimer 2,2′-bis(9-triptycyl)-1,1′-biindenyl (4). In the solid state this very congested molecule adopts a conformation in which the two triptycenyl paddlewheel fragments are attached to the C₂-symmetric biindenyl core. However, the orientation of the triptycene blades, dictated by their size and shape, is gear-like and breaks the nascent C₂-symmetry of the whole molecule. Nevertheless, effective C₂-symmetry is restored in solution when the triptycene paddlewheels, although apparently tightly interlocked, undergo rapid contra-rotation on the NMR time-scale. The analogous reaction of (2-phenylindenyl)lithium (6), likewise yields racemic (RR/SS) 2,2′-diphenyl-1,1′-biindenyl(7). Interestingly, although the treatment of 6 with a Cu^II salt leads to both rac-7 and its diastereomer meso-7, the reaction of 6 and 2-phenylindenyl bromide furnishes selectively rac-7. Overall, these data suggest that the oxidative dimerisation to yield racemic products occurs through a nucleophilic substitution mechanism.

Full text of DOI:10.1002/ejoc.200800202

FULL PAPER
DOI: 10.1002/ejoc.200800202
Twin Triptycyl Spinning Tops: A Simple Case of Molecular Gearing with
Dynamic C2 Symmetry
Kirill Nikitin,*[a] Helge Müller-Bunz,[a] Yannick Ortin,[a] Wilhelm Risse,[a] and
Michael J. McGlinchey*[a]
Keywords: C–C coupling / Molecular modelling / Oxidation / Radical reactions
Lithiation and subsequent oxidation of the recently described
prototype molecular paddlewheel 9-(2-indenyl)triptycene (1)
leads to the racemic dimer 2,2Ј-bis(9-triptycyl)-1,1Ј-biindenyl
(4). In the solid state this very congested molecule adopts a
conformation in which the two triptycenyl paddlewheel frag-
ments are attached to the C2-symmetric biindenyl core. How-
ever, the orientation of the triptycene blades, dictated by
their size and shape, is gear-like and breaks the nascent C2-
symmetry of the whole molecule. Nevertheless, effective C2-
symmetry is restored in solution when the triptycene pad-
dlewheels, although apparently tightly interlocked, undergo
rapid contra-rotation on the NMR time-scale. The analogous
reaction of (2-phenylindenyl)lithium (6), likewise yields race-
mic (RR/SS) 2,2Ј-diphenyl-1,1Ј-biindenyl(7). Interestingly, al-
though the treatment of 6 with a CuII salt leads to both rac-
7 and its diastereomer meso-7, the reaction of 6 and 2-phen-
ylindenyl bromide furnishes selectively rac-7. Overall, these
data suggest that the oxidative dimerisation to yield racemic
products occurs through a nucleophilic substitution mecha-
nism.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2008)
Introduction
Results and Discussion
The appeal of the triptycene system to molecular me-
chanical engineering is based on several factors: (a) the
threefold symmetry of its aromatic blades, (b) the rigidity
of its bicyclo[2.2.2]octatriene core, and (c) the proximity of
any substituent at C(9) to the H(1), H(8) and H(13) sites
on the blades. When connected by a short rigid moiety, e.g.
a methylene bridge or an ether linkage, two triptycyl moie-
ties readily form so-called “molecular gears”.[3–7] In such
systems, two or more triptycyl units are positioned so
closely to each other that internal disrotatory motion about
the linking bonds is rapid and mutually correlated. How-
ever, we are unaware of any examples in which correlated
rotation was observed between two triptycyl units con-
nected by a flexible linker.
In continuation of our quest for an organometallic deriv-
ative that might lead to restricted rotation about the trip-
tycyl-indenyl C(9)–C(17) single bond,[2] we were investiga-
ting the chemistry of the cyclopentadiene ring in 1 which,
as expected, can be deprotonated by a strong base. Despite
its sparing solubility in diethyl ether (less than 0.01  at
ambient temperature), when treated with n-butyllithium for
1 h, 1 forms the lithium derivative 2, which upon the ad-
dition of a deuteriated solvent regenerates the 1-deuterated
hydrocarbon 1-d.
Two-fold symmetric molecules are often readily access-
ible, in racemic form, by the dimerisation of achiral starting
materials. For instance, we recently reported the [2+2]
cyclodimerisation of a series of fluorenylidene-allenes lead-
ing to an interesting family of trans-3,4-diaryl-1,2-bis(fluor-
enylidene)cyclobutanes that are rendered C2-symmetric as a
result of the twisting and overlapping of the fluorenylidene
moieties with their very large wingspans.[1] Alternatively,
the dimerisation of free radicals derived from a prochiral
centre can lead to a mixture of the C2-symmetric (R,R and
S,S) enantiomers as well as the meso (R,S) product. The
product distribution may be dictated by such factors as ac-
cessibility of the reaction centre and total steric strain.
In this short paper we focus on the dimerisation of the
recently prepared molecular pro-gearing system 9-(2-in-
denyl)-triptycene (1).[2] Because the barrier to rotation
about the single bond connecting the indenyl and triptycyl
moieties in 1 is low (Ͻ 9 kcalmol–1), it was tempting to
investigate how dimerisation, if it were to occur, could affect
or even arrest internal rotation in such a picturesque struc-
ture.
[a] School of Chemistry and Chemical Biology, University College
Dublin,
However, when left at ambient temperature for an ex-
tended period of time, the lithium salt 2 was gradually oxi-
dised by the slow diffusion of air into the reaction vessel,
and the oxidation products were separated by chromatog-
Belfield, Dublin 4, Ireland
Fax: +353-1-716-1178
E-mail: michael.mcglinchey@ucd.ie
kirill.nikitin@ucd.ie
Eur. J. Org. Chem. 2008, 3079–3084
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3079

K. Nikitin, H. Müller-Bunz, Y. Ortin, W. Risse, M. J. McGlinchey
FULL PAPER
raphy and characterised. While the minor product was iden-
tified as 2-(9-triptycyl)-1-indenone (3), the major product,
1
4, displayed a H NMR spectrum very similar to that of
the starting material, 1. The most obvious difference was
the loss of the singlet resonance at δ = 4.36 ppm attributable
to the indene CH2, and appearance of a new singlet at δ =
6.01 ppm with intensity corresponding to only one H per
indenyl moiety. These data, together with a comparison of
the 13C NMR spectra of 1 and 4, indicated the formation of
the dimer 2,2Ј-bis(9-triptycyl)-1,1Ј-biindenyl (4) (Scheme 1).
Figure 1. (a) Mercury[10] representation of the X-ray crystal struc-
ture of 2,2Ј-bis(9-triptycyl)-1,1Ј-biindenyl (rac-4); aromatic hydro-
gen atoms have been removed for clarity; (b) space-filling view of
rac-4 showing the C2 character of the indenyl orientations. (c)
space-filling view of rac-4 showing the gear-meshing of the triptycyl
groups.
Scheme 1. Oxidation of the lithium derivative 2 leading to the di-
mer 4.
135.1°, 122.2°, 102.7°). This exceeds the perturbation of in-
The structure of 4 was unequivocally established by X- terplanar angles previously seen in 3-(9-triptycyl)indene
ray crystallography which revealed that the unit cell con- when one of the blades bears a Cr(CO)3 substituent (126°,
tains enantiomeric pairs of (1R,1ЈR) and (1S,1ЈS) mole- 126°, 108°),[11] or in (η6-triptycene)Co4(CO)9 (124.1°,
cules. Careful scrutiny of the molecular structure of racemic 118.9°, 117.0°).[12] Moreover, the degree of intramolecular
4 (Figure 1, a) shows that the molecule is not strictly C2- overcrowding is even more evident in the angle distortions
symmetric in the solid state: while the core 1,1Ј-biindenyl about the bridgehead carbon atoms, C(9) and C(32) of the
moiety has a local C2 axis, the Cs-type orientation of the triptycyl fragments, away from the normal tetrahedral
intermeshed triptycene blades breaks this symmetry, as il- value: C(17)–C(9)–C(9a) 121.5°, C(17)–C(9)–C(8a) 112.8°,
lustrated by the space-filling representations in Figure 1, b C(17)–C(9)–C(12) 108.9°; C(40)–C(32)–C(31a) 120.0°,
and Figure 1, c, respectively. The overall geometry of the C(40)–C(32)–C(24a) 117.2°, C(40)–C(32)–C(35) 106.9°.
biindenyl moiety is very reminiscent of bifluorenyl which Furthermore, the indenyl–triptycyl bonds are now some-
adopts a gauche conformation such that the central H–C– what longer in 4 [1.528(4) and 1.529(4)] than in 1
C–H dihedral angle is 60°;[8,9] in rac-4, the corresponding [1.514(4) Å]; even more striking is the length of the central
torsion angle, H–C(23)–C(46)–H, is 73°. As suggested pre- C(23)–C(46) linkage at 1.580(4) Å. Finally, it is important
viously,[8] the rationale for the favoured gauche conforma- to reiterate that the configurations of the directly connected
tion in bifluorenyl (and presumably also in the biindenyl stereogenic C(23) and C(46) centres are identical, i.e. the
dimer) is based on the “back-clamping” which disallows dimer 4 is chiral.
aryl ring stacking. In the anti conformation, the hydrogen
Remarkably, despite the severe steric congestion in rac-4,
atoms in the 1,8 positions of one fluorenyl are forced to the rotation about the C(9)–C(17) and C(40)–C(32) in-
point directly at their counterparts in the other half of the denyl–triptycenyl single bonds in solution is apparently al-
1
molecule. These severe H···H nonbonded interactions are most unhindered, as evidenced by the simplicity of its H
greatly relieved in the gauche conformation, and contrast and 13C NMR spectra in which all six triptycyl blades are
with the situation for unclamped tetraarylethanes, which equivalent at room temperature. Moreover, even at 193 K,
1
uniformly favour the anti conformation.
the 500 MHz H and 125 MHz 13C NMR spectra exhibit
Both triptycyl units are very severely distorted from their only minimal peak broadening, and fully decoalesced spec-
normal threefold symmetry, as indicated by the interplanar tra cannot be obtained. However, to provide an approxi-
angles between the blades (124.9°, 120.0°, 114.9° and mate value for the rotational barrier, one could assume a
3080
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 3079–3084

Molecular Gearing with Dynamic C2 Symmetry
similar chemical shift separation (ca. 1.3 ppm) between the
H1/H8 and H13 sites as was observed in 3-indenyltriptycene
at low temperature[11] and, estimating the maximum co-
alescence temperature as 183 K, the barrier would appear
to be no more than 9 kcalmol–1. Evidently, in solution the
molecular gearing system rac-4 exhibits dynamic C2-sym-
metry because of the low triptycyl rotation barrier, and the
intrinsic C2-symmetry of the interconnecting biindenyl moi-
ety. Examination of the space-filling representation in Fig-
ure 1, c, or of scale molecular models, suggests that the trip-
tycyl units must execute correlated disrotatory motion with
concomitant interplanar bending to compensate for the un-
usually large range of angles (from 103° to 135°) between
the blades of the paddlewheels.
Interestingly, when kept in solution for a prolonged
period of time, the gradually changing NMR spectra indi-
cated the conversion of rac-4 into a closely related product
in which all the triptycyl blades were again equivalent. The
1H and 13C NMR and mass spectra indicated that this
product was the alcohol 2-(9-triptycyl)inden-1-ol (5), pre-
sumably arising from homolysis of the long C(23)–C(46)
central linkage. This is supported by the presence of a
strong OH absorption band in the IR spectrum of 5.
To probe the generality of these interesting observations,
2-phenylindene was converted into its lithium derivative 6
Scheme 2. Reactions of 6 leading to rac-7, meso-7, and the alcohol
10.
(Scheme 2) under an atmosphere of nitrogen, and air was anions in the presence of oxygen raises interesting mechan-
allowed to diffuse in slowly. The only dimeric product was istic questions. Moreover, in marked contrast to the preced-
rac-2,2Ј-diphenyl-1,1Ј-biindenyl, rac-7, as confirmed by X- ing result, the treatment of (2-phenylindenyl)lithium 6 with
ray crystallography (Figure 2, a). As with the di-triptycyl a CuII salt – a typical radical process[14] – led to both rac-7
system rac-4, the molecule adopted the gauche conforma- and meso-7 isomers in a 47:53 ratio. These observations
tion; however, in the case of rac-7, the H–C(1)–C(14)–H suggest that the exclusive formation of rac-4 and rac-7 does
dihedral angle was 64° and the central C(1)–C(14) bond not arise from a simple radical dimerisation process.
length was a more normal 1.549(2) Å. The torsion angle
X-ray crystallographic data were acquired on several
between the local C2-axes of the phenyl groups in rac-7 is samples of meso-7; however, because of the insufficient
86°, implying much less steric strain than in rac-4 for which quality of the crystals, the molecular parameters cannot be
the corresponding angle between the local C3 axes of the quoted with a high degree of accuracy, except that the mole-
triptycenes is only 30°. We note that an energy-minimised cule clearly adopts a gauche conformation (Figure 3, a) with
geometry calculation at the B3LYP/6-31G** level[13] also a dihedral angle of ca. 70°. Gratifyingly, the DFT calculated
predicts that rac-7 should adopt a gauche conformation, energy-minimised molecular geometry at the B3LYP/6-
with a H–C(1)–C(14)–H dihedral angle of 70°. Further- 31G** level also yields a gauche structure with a dihedral
more, the calculated structure has a C(1)–C(14) bond length angle of 73° (Figure 3, b). However, unlike dimeric rac-7
of 1.568 Å, and a torsion angle of 77° between the local C2- that retains its C2-symmetry whatever the rotamer confor-
axes of the phenyl groups.
mation, meso-7 loses all symmetry elements unless the cen-
The initial exclusive formation of the racemic dimers de- tral H–C–C–H dihedral angle is 0° or 180°, giving rise to
rived from the 2-(9-triptycyl)indenide and 2-phenylindenide either a single mirror plane, (Cs) or an inversion centre (Ci),
Figure 2. (a) X-ray crystal structure of rac-7; (b) Mercury[10] representation of S,S-7; (c) DFT energy-minimised structure of S,S-7.
Eur. J. Org. Chem. 2008, 3079–3084
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
3081

K. Nikitin, H. Müller-Bunz, Y. Ortin, W. Risse, M. J. McGlinchey
FULL PAPER
respectively. Consequently, the indenyl moieties and the nucleophilic displacement process must be heavily influ-
phenyl substituents in meso-7 become non-equivalent, and enced by the steric hindrance arising from the bulky substit-
this is reflected in its variable-temperature NMR spectra uent at the indenyl-C(2) position. Moreover, because both
which reveal very significant line broadening in the carbon dimers, rac- and meso-7, adopt stable gauche-conforma-
spectra at room temperature, and gradual sharpening of tions, it is not unreasonable to assume that the phenylin-
those signals as temperature is raised to 80 °C. Evidently, denyl moieties also favour a gauche-orientation in the tran-
the steric crowding in meso-7 causes restricted rotation sition state. Consideration of the least hindered possible ap-
about the C1–C14 central single bond on the NMR time- proach trajectories for both SN2 and SN2Ј mechanisms led
scale, thus indicating atropisomerism.
to the conclusion that the formation of rac-7 is the more
favourable.
Scheme 3. A possible mechanistic representation of an SN2Ј route
to rac-7.
Figure 3. (a) X-ray crystal structure of meso-7, showing the molecu-
lar connectivity; (b) DFT energy-minimised structure of meso-7.
In seeking precedents for such proposals, we note not
only the reaction of indenyllithium with oxygen to form the
corresponding hydroperoxide,[16] but also the report by
Abeles[17] on the mechanism of action of dioxygenases in
the methionine salvage pathway whereby a crucial step in-
volves attack by a carbanion on dioxygen to form the corre-
sponding alkyl peroxide anion. Although peroxide has fre-
quently been used as an attacking nucleophile,[18] there are
also examples where peroxide (or hydroperoxide) can be a
leaving group, but we are well aware that cleavage of the
oxygen–oxygen bond is the more normal observation.[19]
Following the method described by Hock,[16b] we at-
tempted the oxidation of (2-phenylindenyl)lithium 6 with
dioxygen at –70 °C. As anticipated, acidic workup yielded
the hydroperoxide 8, for which iodometric titration indi-
cated a 65% yield. This correlates with the isolation of 2-
phenyl-1-indenol (10), in 65% yield, thereby establishing
that 6 readily forms the corresponding peroxide when
treated with dioxygen.
To further test the plausibility of a non-radical nucleo-
philic pathway leading to dimers 7, the cross-coupling of 1-
bromo-2-phenylindene (9) with the lithium derivative 6 was
studied (Scheme 2). It transpired that, when this reaction
was carried out at 0 °C in ether, it resulted in the selective
formation of rac-7 in 72% yield while meso-7 was formed
in just 4% yield, i.e. a 95:5 ratio. Accordingly, it can be
concluded that a nucleophilic mechanism is responsible for
the selective formation of rac-dimers in the oxidative dimer-
isation of lithiated indenes and, possibly, other substrates.
We envisage further applications of this significant finding
in the area of selective syntheses of chiral, sterically hin-
dered C2-symmetrical molecules, and trust that these obser-
vations prompt others to undertake more detailed studies
to verify, or modify, these mechanistic proposals.
Although the barrier to interconversion of the enantio-
mers of meso-7 was not accessible experimentally, it was
probed computationally by calculating the relative energies
of the energy-minimised ground state, and both the mirror-
symmetric (Cs) and centrosymmetric (Ci) potential transi-
tion states, all at the B3LYP/6-31G** level. In the former,
the barrier was evaluated as ca. 20 kcalmol–1, and the cen-
tral carbon–carbon bond, C1–C14, appears to have length-
ened somewhat (from 1.57 Å to 1.64 Å) as it passes through
the Cs structure. An imaginary frequency was found
(25i cm–1), and its vibrational twist motion is consistent
with enantiomer conversion via this mirror-symmetric tran-
sition state. In the latter case, the Ci structure lies only ca.
11 kcalmol–1 above the ground state; however, during the
rotation about the central C–C bond required to attain this
geometry, both phenyl substituents would have to rotate
simultaneously so as to avoid a steric confrontation with
the indenyl moiety in the other half of the molecule, thus
raising the barrier substantially and making this a rather
unlikely eventuality.
In the light of these observations, one might envisage at
least two viable scenarios to rationalise predominant forma-
tion of the racemic products in the presence of oxygen.
Firstly, one could invoke initial reaction of the 2-substituted
indenyllithium with dioxygen to form the corresponding
peroxide; subsequent SN2 displacement of the peroxide ion
by a second anion of the opposite configuration, would thus
generate an R,R (or S,S) product. A second, and in some
ways more attractive proposal, that might alleviate the se-
vere steric problems that would be encountered in a normal
SN2 reaction could proceed instead via an SN2Ј mechanism.
In the latter case, the incoming nucleophile (which must
again adopt the opposite configuration to that of the perox-
ide-bearing carbon centre) approaches the indenyl-C(3) po-
sition in a syn fashion, possibly involving an interaction of
its lithium cation with a peroxide oxygen (Scheme 3).
Furthermore, in cases where it has been possible to deter-
mine the stereochemistry of the SN2Ј process, the syn ap-
Conclusions
The oxidative dimerisation of 2-substituted lithiated in-
proach predominates.[15] Clearly, the stereochemistry of the denes leads to racemic hydrocarbons with a C2-symmetric
3082
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 3079–3084

Molecular Gearing with Dynamic C2 Symmetry
7.1 Hz, 2 H, H21), 7.34 (d, J = 7.7 Hz, 3 H, H1,8,13), 7.29 (d, J =
7.3 Hz, 1 H, H19), 7.00, (t, J = 7.3 Hz, 3 H, H3,6,15), 6.95 (t, J =
7.7 Hz, 3 H, H2,7,14), 5.40 (s, 2 H, H10) ppm. 13C NMR (100 MHz,
CDCl3, 25 °C): δ = 197.4 (C23), 148.6 (C18), 145.9 (C8a,9a,12), 145.6
(C4a,10a,11), 143.9 (C22a), 137.1 (C17), 134.5 (C21), 130.0 (C18a), 129.6
(C20), 124.8 (C2,7,14), 125.5 (C3,6,15), 123.9 (C4,5,16), 122.5 (C19), 55.9
(C9), 55.0 (C10) ppm. MS (ES): m/z = 383.1 (100) [M + H]. IR
bis-1,1Ј-indenyl core. The sterically constrained triptycene
derivative, 4, can be prepared by an oxidative dimerisation
of 2-(9-triptycyl)indene. In solution, the molecule 4 un-
dergoes correlated gear-like contra-rotation with a low acti-
vation energy leading to NMR equivalence of its “cogs”
thus resulting in dynamic C2-symmetry. However, in the so-
lid state, the intermeshing of the pairs of triple pad-
dlewheels breaks the C2-symmetry of the system and gives
rise to severe geometric perturbations. The analogous oxi-
dative dimerisation of 2-phenylindenyl-lithium, 6, likewise
furnished the dimer rac-7, whereas the reaction with cupric
(KBr): ν = 1714 cm–1.
˜
2-(9-Triptycyl)inden-1-ol (5): When a solution of 4 (23 mg) in
dichloromethane was maintained under ambient conditions in air
for 10 d, the products were purified to yield 5 (20 mg, 80%) as a
yellow solid; m.p. 188–189 °C. 1H NMR (500 MHz, CDCl3, 25 °C,
chloride led to the formation of both rac- and meso-7 di- atom numbering as for 4): δ = 7.20 (d, J = 7.3 Hz, 1 H, H19), 7.09
(s, 1 H, H18), 7.58 (d, J = 7.6 Hz, 3 H, H1,8,13), 6.30 (d, J = 7.4 Hz,
mers. Evidence for the possible involvement of a nucleo-
philic displacement, rather than a simple radical dimeris-
ation pathway, leading to the formation of rac-4 and rac-7
is presented.
1 H, H22), 7.42 (d, J = 7.2 Hz, 3 H, H4,5,16), 7.06 (t, J = 7.4 Hz, 1
H, H20), 6.34 (t, J = 7.3 Hz, 1 H, H21), 6.97, (t, J = 7.3 Hz, 3 H,
H3,6,15), 6.85 (t, J = 6.5 Hz, 3 H, H2,7,14), 6.50 (s, 1 H, H23), 5.43
(s, 1 H, H10) ppm. 13C NMR (125 MHz, CDCl3): δ = 146.1
(C4a,10a,11), 146.5 (C8a,9a,12), 141.0 (C18a), 139.8 (C17), 142.6 (C22a),
139.9 (C18), 128.2 (C20), 124.6 (C2,7,14), 126.2 (C21), 125.2 (C3,6,15),
Experimental Section
123.3 (C4,5,16), 120.6 (C19), 88.2 (C23), 58.7 (C9), 55.2 (C10) ppm.
MS (ES): m/z = 367 (65) [M – OH]. IR (KBr): ν = 3400 cm–1.
˜
General: All reactions were carried out under a nitrogen atmo-
sphere unless otherwise stated. Column chromatography separa-
tions were carried out with a Buchi Sepacor machine with UV ab-
sorbance detector using silica gel particle size 40–63 mm. NMR
spectra were acquired with Varian Inova 400 or 500 MHz spec-
trometers. Assignments were based on standard 1H-1H and 1H-13C
two-dimensional techniques, and NOE measurements. 2-(9-Trip-
tycenyl)indenyl (1), was prepared by benzyne addition to 9-(2-in-
denyl)anthracene[20] according to the previously described pro-
cedure.[2] Infrared spectra were recorded with a Perkin–Elmer Para-
gon 1000 FT-IR spectrometer and were calibrated with polystyrene.
Melting points were determined with an Electrothermal ENG in-
strument and are uncorrected. Elemental analyses were carried out
by the Microanalytical Laboratory at the University College Dub-
lin.
rac-2,2Ј-Diphenyl-1,1Ј-biindenyl (rac-7): 2-Phenylindene (38 mg,
0.2 mmol) and ether (6 mL) were transferred under a nitrogen at-
mosphere to a round-bottom flask (25 mL) equipped with a rubber
septum. A solution of nBuLi (0.2 mmol) in hexanes was then added
to the flask. After 1 h, a clear solution of 2-phenylindenyllithium
(6) was formed. The reaction mixture was stirred for 1 d while the
nitrogen inlet was removed. The mixture was quenched with meth-
anol (1 mL), extracted with dichloromethane and separated by
chromatography by eluting with 5% dichloromethane in cyclohex-
1
ane to give rac-7 (9.6 mg, 25%) as a white solid, m.p. 191 °C. H
NMR (400 MHz, CDCl3, 25 °C, numbering in accord with Fig-
ure 2, a): δ = 7.84 (d, J = 6.8 Hz, 4 H, H9,13,22,26), 7.58 (t, J =
7.4 Hz, 4 H, H10,12,23,25), 7.40 (t, J = 7.4 Hz, 2 H, H11,24), 7.19 (s,
2 H, H3,16), 7.14 (d, J = 7.3 Hz, 2 H, H4,17), 7.00 (t, J = 7.3 Hz, 2
H, H5,18), 6.99 (d, J = 7.3 Hz, 2 H, H7,20), 6.81 (t, J = 7.3 Hz, 2 H,
H6,19), 4.49 (s, 2 H, H1,14) ppm. 13C NMR (100 MHz, CDCl3): δ
= 149.6 (C2,15), 143.8 (C3a,16a), 143.4 (C7a,20a), 136.0 (C8,21), 129.2
(C3,10,12,16,23,25), 127.7 (C11,24), 127.2 (C9,13,22,26), 126.7 (C5,18),
rac-2,2Ј-Bis(9-triptycyl)-1,1Ј-biindenyl (rac-4): A suspension of 9-(2-
indenyl)-triptycene (1) (37 mg, 0.1 mmol) in diethyl ether (6 mL)
was transferred under a nitrogen atmosphere to a round-bottom
flask (25 mL) equipped with a rubber septum. Then, a solution of
nBuLi (0.2 mmol) in hexanes was added, and the reaction mixture
was stirred for 1 d while the nitrogen inlet was removed. The mix-
ture was quenched with methanol (1 mL), extracted with dichloro-
methane and separated by chromatography by eluting with 5%
dichloromethane in cyclohexane to give ketone 3 (5 mg, 8%) as a
yellow glassy solid, and 4 (23 mg, 70%) as a white solid, m.p. 222–
126.2 (C21), 124.1 (C6,19), 122.9 (C7,20), 120.7 (C4,17), 49.3 (C1,14
)
ppm. C30H22 (382.51): calcd. C 94.20, H 5.80; found C 94.14, H
5.84. No isomeric dimer, meso-7, was found indicating that its yield
was below 1%.
meso-2,2Ј-Diphenyl-1,1Ј-biindenyl (meso-7):
A solution of 6
(0.4 mmol) in diethyl ether (6 mL) was added slowly to a cold
(–20 °C) suspension of cupric chloride (54 mg, 0.4 mmol) in THF
(3 mL). The mixture was warmed to 0 °C and stirred for 15 h after
which it was quenched with methanol (0.5 mL), extracted with
dichloromethane and separated by chromatography by eluting with
5% dichloromethane in cyclohexane to give meso-7 (18 mg, 23%)
as a white solid, m.p. 181–182 °C. C30H22·0.2C4H10O (397.33):
calcd. C 93.11, H 6.09; found C 93.53, H 5.94. 1H NMR (500 MHz,
(CDCl2)2, 80 °C, numbering as in rac-7): δ = 7.30 (d, J = 7.3 Hz,
4 H, H9,13,22,26), 7.28 (m, 4 H, H4,17,6,19), 7.24 (d, J = 7.5 Hz, 2 H,
H7,20), 7.15 (t, J = 7.5 Hz, 2 H, H5,18), 7.05 (br. s, 4 H, H10,12,23,25),
1
224 °C. H NMR (500 MHz, CDCl3, 25 °C, numbering in accord
with Scheme 1, only positions C1–C23 listed): δ = 7.84 (d, J =
7.3 Hz, 2 H, H19), 7.70 (s, 2 H, H18), 7.56 (d, J = 7.7 Hz, 6 H,
H1,8,13), 7.36 (d, J = 7.4 Hz, 2 H, H22), 7.31 (d, J = 7.3 Hz, 6 H,
H4,5,16), 7.25 (t, J = 7.4 Hz, 2 H, H20), 7.05 (t, J = 7.1 Hz, 2 H,
H21), 6.84, (t, J = 7.3 Hz, 6 H, H3,6,15), 6.54 (t, J = 7.7 Hz, 6 H,
H2,7,14), 6.01 (s, 2 H, H23), 5.29 (s, 2 H, H10) ppm. 13C NMR
(125 MHz, CDCl3, 25 °C): δ = 147.0 (C4a,10a,11), 146.3 (C8a,9a,12),
142.4 (C18a), 144.6 (C17), 145.3 (C22a), 141.1 (C18), 123.0 (C20),
124.6 (C2,7,14), 125.0 (C21), 125.1 (C3,6,15), 123.5 (C4,5,16), 122.6
(C19), 61.7 (C9), 55.5 (C10), 54.5 (C23) ppm. MS (ES): m/z = 735
[MH+]. C58H38·CH2Cl2 (819.87): calcd. C 86.43, H 4.92; found C
85.90, H 5.01.
7.00 (br. s, 2 H, H11,24), 6.58 (s, 2 H, H3,16), 4.69 (s, 2 H, H1,14
)
ppm. 13C NMR (125 MHz, CDCl3): δ = 149.3 (C2,15), 145.5
(C7a,20a), 144.3 (C3a,16a), 136.0 (C8,21), 128.2 (C3,16), 127.7 (C6,19),
127.2 (C4,17,10,12,23,25), 127.0 (C9,13,22,26), 124.5 (C5,18), 122.6
(C11,24), 120.8 (C7,20), 50.2 (C1,14); and rac-7 (16 mg, 20%) ppm.
Ketone 3: 1H NMR (400 MHz, CDCl3, 25 °C, numbering as for 4):
δ = 8.53 (s, 1 H, H18), 7.60 (d, J = 7.4 Hz, 1 H, H22), 7.49 (t, J = Coupling of 6 with 1-Bromo-2-phenylindene (9):[21] A solution of 9
7.4 Hz, 1 H, H20), 7.40 (d, H4,5,16, J = 7.3 Hz, 3 H, d), 7.37 (t, J = (108 mg, 0.4 mmol) in diethyl ether (1 mL) was added to a stirred
Eur. J. Org. Chem. 2008, 3079–3084
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
3083

K. Nikitin, H. Müller-Bunz, Y. Ortin, W. Risse, M. J. McGlinchey
FULL PAPER
solution of 6 (0.4 mmol) at –20 °C. The mixture was treated as
above to give rac-7 (110 mg, 72%) and meso-7 (6 mg, 4%).
Computational Details: Computational geometry optimisations
were carried out using B3LYP (a density functional theory method)
for determining equilibrium geometries of rac-7 and meso-7, and
for calculating the barrier to interconversion of the enantiomers of
meso-7. This interconversion barrier was calculated from the rela-
tive energies of the energy-minimised ground state, and the mirror-
symmetric and centrosymmetric transition states. The basis set 6-
31G** was employed for all calculations, and the software package
used was Titan (Jaguar, version 3.5).[13]
2-Phenyl-1-indenol (10): A solution of 6 (0.4 mmol) was cooled to
–70 °C and the nitrogen was replaced with dioxygen for 4 h after
which diluted hydrochloric acid (0.8 mL, 1 ) was added. The mix-
ture was extracted at 5 °C with ether, the organic layer was treated
with an excess potassium iodide and titrated with a sodium thiosul-
fate solution indicating the presence of 0.26 mmol (65%) of the
peroxide 8. The residue was separated by chromatography to give
the alcohol 10 (54 mg, 65%) as a white solid; m.p. 150–151 °C, lit.
152 °C.[22]
Acknowledgments
X-ray Measurements for rac-4, rac-7: Crystal data were collected
with a Bruker SMART APEX CCD area detector diffractometer,
and are listed in Table 1. A full sphere of the reciprocal space was
scanned by phi-omega scans. Pseudo-empirical absorption correc-
tion based on redundant reflections was performed by the program
SADABS.[23] The structures were solved by direct methods using
SHELXS-97[24] and refined by full-matrix least-squares on F2 for
all data using SHELXL-97.[25] All hydrogen atoms, except for the
ones of solvent molecules, were located in the difference Fourier
map and allowed to refine freely with isotropic thermal displace-
ment factors. Hydrogen atoms of the solvents were added at calcu-
lated positions and refined using a riding model. Their isotropic
displacement parameters were fixed to 1.2 times the equivalent iso-
tropic displacement parameters of the parent carbon atom. Aniso-
tropic temperature factors were used for all non-hydrogen atoms.
We thank Science Foundation Ireland and University College Dub-
lin for generous financial support.
[1] E. V. Banide, B. C. Molloy, Y. Ortin, H. Müller-Bunz, M. J.
McGlinchey, Eur. J. Org. Chem. 2007, 2611–2622.
[2] K. Nikitin, H. Müller-Bunz, Y. Ortin, M. J. McGlinchey, Org.
Biomol. Chem. 2007, 5, 1952–1960.
[3] H. Iwamura, K. Mislow, Acc. Chem. Res. 1988, 21, 175–182.
[4] S. Brydges, L. E. Harrington, M. J. McGlinchey, Coord. Chem.
Rev. 2002, 233, 75–105, and references therein.
[5] F. Chiaravalloti, L. Gross, K.-H. Rieder, S. M. Stojkovic, A.
Gourdon, C. Joachim, F. Moresco, Nat. Mater. 2007, 6, 30–33.
[6] R. E. Bulo, F. Allaart, A. W. Ehlers, F. J. J. de Kanters, M.
Schakel, M. Lutz, A. L. Spek, K. Lammertsma, J. Am. Chem.
Soc. 2006, 128, 12169–12173.
[7] E. R. Kay, D. A. Leigh, F. Zerbetto, Angew. Chem. Int. Ed.
2007, 46, 72–191.
[8] D. A. Dougherty, F. M. Llort, K. Mislow, J. F. Blount, Tetrahe-
dron 1978, 34, 1285–1300.
[9] L. E. Harrington, J. F. Britten, M. J. McGlinchey, Tetrahedron
Lett. 2003, 44, 8057–8060.
[10] Mercury 1.4.2, available free from http://www.ccdc.cam.ac.uk/
mercury
Table 1. Crystallographic data for rac-4 and rac-7.
rac-4
rac-7
Formula
M
Crystal system
Space group
a [Å]
b [Å]
c [Å]
α [°]
C58H38·2CH2Cl2
907.74
C30H22
382.48
monoclinic
P21/c (#14)
8.7327(7)
10.8598(9)
21.6999(18)
90
monoclinic
P21/c (#14)
19.2854(13)
14.3423(10)
15.9639(11)
90
[11] L. E. Harrington, L. S. Cahill, M. J. McGlinchey, Organome-
tallics 2004, 23, 2884–2891.
[12] R. A. Gancarz, J. F. Blount, K. Mislow, Organometallics 1985,
4, 2028–2032.
[13] Titan. Wavefunction Inc. Irvine, CA and Schrödinger Inc.
Portland, OR, USA.
β [°]
90.129(2)
90
4415.6(5)
4
1.361
100(2)
94.226(2)
90
2052.3(3)
4
1.238
100(2)
0.070
[14] a) P. Nicolet, J. Y. Sanchez, A. Benaboura, M. J. M. Abadie,
Synthesis 1987, 202–203; b) U. Simmross, K. Müllen, Chem.
Ber. 1993, 126, 969–973.
Γ [°]
V [Å3]
Z
[15] J. March, Advanced Organic Chemistry, 4th ed., John Wiley &
Sons, New York, 1992, pp. 328–329.
ρcalcd. [gcm–3]
T [K]
µ [mm–1]
2θmax [°]
0.311
49.58
[16] a) E. C. Friedrich, D. B. Taggart, J. Org. Chem. 1975, 40, 720–
723; b) H. Hock, F. Ernst, Chem. Ber. 1959, 92, 2723.
[17] Y. Dai, T. C. Pochapsky, R. H. Abeles, Biochemistry 2001, 40,
6379–6387.
52.00
Reflns. measured
Reflns. used
Rint
Parameters
Final R values [IϾ2σ
(I)]:
33803
7571
0.0416
729
17291
4025
0.0239
359
[18] B. Plesnicar in The Chemistry of the Functional Groups, Perox-
ˇ
ides (Ed.: S. Patai), John Wiley & Sons, New York, 1983, ch.
17, pp. 521–584.
[19] G. Merényi, J. Lind, J. Am. Chem. Soc. 1991, 113, 3146–3153.
[20] D.-W. Lee, J. Yun, Bull. Korean Chem. Soc. 2004, 25, 29–30.
[21] P. C. Jocelyn, J. Chem. Soc. 1954, 1641–1642.
[22] a) T. Matsuda, M. Makino, M. Murakami, Chem. Lett. 2005,
34, 1416–1417; b) D. Kuck, T. Lindenthal, A. Schuster, Chem.
Ber. 1992, 125, 1449–1460.
R1
wR2
0.0542
0.1322
0.0366
0.0896
R values (all data):
R1
0.0673
0.1401
1.042
0.0415
0.0930
1.029
wR2
GOF on F2
[23] G. M. Sheldrick, SADABS, Bruker AXS Inc., Madison, WI
53711, 2000.
[24] G. M. Sheldrick, SHELXS-97, University of Göttingen, 1997.
[25] G. M. Sheldrick, SHELXL-97-2, University of Göttingen,
1997.
CCDC-677070 (for rac-4) and CCDC-677071 (for rac-7) contain
the supplementary crystallographic data for this paper. These data
can be obtained free of charge from the Cambridge Crystallo-
graphic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
Received: February 21, 2008
Published Online: May 8, 2008
3084
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 3079–3084