Unprecedented thermal rearrangement of push-pull-chromophore-[60]fullerene conjugates: Formation of chiral 1,2,9,12-tetrakis-adducts

Article abstract of DOI:10.1039/c0cc00881h

Push-pull-chromophore-[60]fullerene conjugates featuring N,N-dimethylanilino-substituted 1,1,4,4-buta-1,3-dienes directly attached to the carbon sphere are transformed into chiral 1,2,9,12-tetrakis-adducts by a novel thermal rearrangement pathway.

Full text of DOI:10.1039/c0cc00881h

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Unprecedented thermal rearrangement of push–pull-chromophore–
[60]fullerene conjugates: formation of chiral 1,2,9,12-tetrakis-adductsw
Michio Yamada, W. Bernd Schweizer, Franziska Schoenebeck* and Franc¸ ois Diederich*
Received 12th April 2010, Accepted 24th May 2010
First published as an Advance Article on the web 17th June 2010
DOI: 10.1039/c0cc00881h
i
Push–pull-chromophore–[60]fullerene conjugates featuring
N,N-dimethylanilino-substituted 1,1,4,4-buta-1,3-dienes directly
attached to the carbon sphere are transformed into chiral
1,2,9,12-tetrakis-adducts by a novel thermal rearrangement
pathway.
larger steric hindrance of the bulkier vicinal Pr group on the
fullerene surface. The molecular structures of (ꢀ)-2a,b
were characterised by 1H and 13C NMR spectroscopy.
Atropisomerism about the CC single bond connecting the
buta-1,3-diene moiety to the fullerene core is observed by
1H NMR analysis of (ꢀ)-2a. The two atropisomers of
C1-symmetric (ꢀ)-2a appear in a 1.0 : 0.19 ratio at 296.6 K.
Increased broadening of the peaks is observed upon heating
the sample to 366.0 K (see Fig. 4 (ESIw)). In contrast, the
1H NMR spectrum of (ꢀ)-2b only shows one C1-symmetric
atropisomer: the equilibrium is apparently shifted because of
Chemical functionalisation of fullerenes remains of great
interest not only for deciphering fundamental aspects
concerning the reactivity of curved p-electron systems but also
for the production of covalent derivatives for use as organic
materials in photovoltaic devices and solar cells.1 We recently
reported the synthesis of push–pull-chromophore–fullerene
conjugates by [2+2] cycloaddition of electron-deficient olefins,
such as TCNE (tetracyanoethene), to the electron-rich CC
triple bond of 4-(N,N-dimethylamino)phenylethynyl[60]fullerene,
followed by cycloreversion.2 In the presence of a vicinal
methyl group on the surface of the carbon sphere, the barrier
for rotation around the central CC single bond of the
fullerene-appended, N,N-dimethylanilino (DMA)-substituted
1,1,4,4-tetracyanobuta-1,3-diene moiety becomes slow and the
axially chiral conjugate can be optically resolved at ambient
temperature and the chiroptical properties of the enantiomers
characterised. This finding motivated us to introduce bulkier
vicinal addends to further raise the barrier for rotation and
increase the optical stability of the conjugates. During this
work, we encountered an unprecedented thermal rearrange-
ment yielding the formation of C1-symmetric 1,2,9,12-tetrakis-
adducts of [60]fullerene,3 which is reported below.
i
the steric repulsion between the Pr group on C60 and the
push–pull-chromophore moiety.
In sharp contrast, heating a solution of 1a,b in chloro-
benzene to reflux (132 1C) in the presence of TCNE led to a
drastic color change from brown to green and afforded the
formation of the 1,2,9,12-tetrakis-adducts (ꢀ)-3a,b in 58%
and 38% yield, respectively. Mass spectrometry clearly
indicated that the products are isomers of (ꢀ)-2a,b because
the spectra showed the same molecular masses for the parent
ions. It is noteworthy that the green products (ꢀ)-3a,b can also
be obtained by heating a chlorobenzene solution of (ꢀ)-2a,b,
suggesting that (ꢀ)-2a,b are kinetic products whereas (ꢀ)-3a,b
are thermodynamic products. The thermal conversion from
(ꢀ)-2a,b to (ꢀ)-3a,b was almost quantitative and no degradation
of the starting materials was observed (TLC). The thermo-
1
dynamic products (ꢀ)-3a,b were also characterised by H and
13C NMR spectroscopy, however, any further structural
information was not gained because they are C1-symmetric.
The new fullerene derivatives (ꢀ)-3a,b exhibited charge-
transfer (CT) bands in their UV/Vis spectra at 601 and
604 nm, respectively. Clear hypochromic shifts were observed
after addition of CF3COOH, and the original CT bands were
fully recovered upon addition of Et3N (see Fig. 14 (ESIw) and
The synthesis of the new conjugates is depicted in Scheme 1.
The precursors 1a,b with vicinal ethyl (Et) or isopropyl (iPr)
addends next to the ethynyl residue, were prepared in 22% and
24% yield, respectively, by addition of lithiated DMA-
acetylene to C60, followed by ethylation (EtI) or isopropylation
(iPrI).4 Stirring a chlorobenzene solution of 1a,b in the
presence of TCNE at 20 1C led to a color change of the
solutions from brown to red and afforded the formation of
the charge-transfer (CT) chromophore–fullerene conjugates
(ꢀ)-2a,b featuring nonplanar push–pull-substituted buta-1,3-
dienes directly attached to the carbon sphere. The decreased
yield of (ꢀ)-2b (21%) compared to (ꢀ)-2a (55%), is due to the
Laboratorium fur Organische Chemie, ETH Zurich, Honggerberg,
¨
¨
HCI, CH-8093 Zurich, Switzerland.
¨
E-mail: diederich@org.chem.ethz.ch; Fax: +41 44 632 1109;
Tel: +41 33 632 2992
w Electronic supplementary information (ESI) available: Synthesis
and characterisation of new compounds, details of the calculations
and full details of reference 5. CCDC 770342, 770343 and 770541. For
ESI and crystallographic data in CIF or other electronic format see
DOI: 10.1039/c0cc00881h
i
Scheme 1 Synthetic overview. a: R = Et, b: R = Pr.
ꢁc
This journal is The Royal Society of Chemistry 2010
5334 | Chem. Commun., 2010, 46, 5334–5336

View Article Online
Fig. 15 (ESIw)). Similarity between (ꢀ)-3a and (ꢀ)-3b was also
found in their IR spectra (see Fig. 20 (ESI) and Fig. 21 (ESI)).
Finally, the molecular structures of (ꢀ)-2a, (ꢀ)-3a and
(ꢀ)-3b were elucidated by X-ray crystallography as depicted
in Fig. 1.z As for (ꢀ)-2a (R = Et) the two addends are
attached to the [6,6] junction (addition to the double bond
shared by two six-membered rings). The two dicyanovinyl
groups are highly distorted with the torsion angle
y(C71–C63–C64–C76) = 75.561. The DMA ring and the
neighbouring dicyanovinyl group are less distorted with a
torsion angle y(C70–C65–C64–C76) = 44.631.
Scheme 2 Proposed mechanism for the thermal rearrangement of
push–pull-chromophore–fullerene conjugate (ꢀ)-2 to 1,2,9,12-tetrakis-
i
adduct (ꢀ)-3. R = Me, Et, Pr.
The X-ray crystallographic analysis disclosed that (ꢀ)-3a,b
are 1,2,9,12-tetrakis-adducts of C60, with a cyclopentene ring
fused to
a
[5,6] junction. In the structure of (ꢀ)-3a
the fullerene surface strongly protrude from the spherical
surface because of their sp3-character. Indeed, the bonds
C19–C1 (1.612 A), C1–C2 (1.644 A) and C2–C22 (1.602 A)
are single bonds, whereas the C20–C21 bond (1.356 A)
possesses double bond character.
(the structure of (ꢀ)-3b is very similar), one dicyanovinyl
and a third cyano group are found almost in the plane with
the cyclopentene ring, whereas the fourth cyano group is
attached to the neighbouring carbon atom on the fullerene
surface. The DMA ring is almost perpendicular to the cyclo-
pentene ring with a torsion angle y(C68–C67–C71–C72) = 84.561.
The functionalised carbon atoms (C1, C2, C19 and C22) on
The proposed mechanism for the rearrangement of (ꢀ)-2a,b
to (ꢀ)-3a,b is shown in Scheme 2. The dicyanovinyl group
adjacent to the DMA ring in (ꢀ)-2 is located very closely to the
fullerene surface (distance between C76 and C13 = 3.308 A
(see Fig. 1)) and attacks the fullerene. This ring-closing step
generates a fullerenyl anion, which subsequently attacks a
cyano group to afford, by an addition–elimination-type
mechanism, the tetrakis-adduct (ꢀ)-3. This thermal rearrange-
ment was also observed for the methylated analogue
(R = Me) of (ꢀ)-2a,b while it does not proceed when R = H.
Computational studies5 were applied to gain deeper insights
into the mechanism of the rearrangement. We applied
the ONIOM6 approach, using DFT B3LYP/6–31G(d) for
the high-level layer and SVWN/STO–3G or AM1 for the
low-level layer (see Fig. 2 and ESIw for the precise partitioning
of the ONIOM layers). These ONIOM approaches were
previously applied to reactivity studies of fullerenes and
nanotubes.7 Both approaches resulted in essentially identical
activation and reaction energies. Fig. 2 summarises the results.
Our findings support the mechanism proposed in Scheme 2.
The attack of the N,N-dimethylanilino-substituted buta-1,3-
diene onto the fullerene sphere in (ꢀ)-2a has an activation
barrier of DGz = 117.6 kJ molꢂ1 and is accompanied with a
charge transfer8 of ꢂ0.17e from the dimethylanilino-
substituted butadiene to the fullerene moiety. A partial negative
charge results on the carbon atom C22 (ꢂ0.22e) adjacent to
the cyclopentyl addend in Int-2a.9 An orbital analysis of the
intermediate (Int-2a) shows a significant lobe of the HOMO
also on the latter carbon atom (see ESIw for further information).
Thus, cyanide transfer subsequently takes place with a relatively
low activation barrier (DGz = 38.5 kJ molꢂ1). Overall, the
rearrangement is exergonic by DGrxn = ꢂ31.4 kJ molꢂ1
.
The rearrangement of the analogue of 2 with R = H is
similarly exergonic (DGrxn = ꢂ24.3 kJ molꢂ1), but proceeds
with a roughly 18 kJ molꢂ1 higher ‘rate-controlling’ activa-
tion barrier (DGz
= 159.1 kJ molꢂ1 for R = H and
2 - TS2
141.5 kJ molꢂ1 for R = Et). Thus, the rearrangement is
disfavoured under the experimental conditions applied. The
increase in barrier (if R = H) is primarily due to the greater
stabilisation of the reactant as a result of decreased steric
interaction.
Fig. 1 ORTEP plots of (ꢀ)-2a (a), (ꢀ)-3a (b), and (ꢀ)-3b (c) with
thermal ellipsoids at 123 K shown at the 50% probability level. Solvate
molecules are omitted for clarity.
ꢁc
This journal is The Royal Society of Chemistry 2010
Chem. Commun., 2010, 46, 5334–5336 | 5335

View Article Online
at 123 K (CCDC-770541): C79H17N5 ꢃ3 C6H5Cl, Mr = 1373.717,
= ,
1.552 Mg cmꢂ3
ꢀ
Triclinic, space group P1 (no. 2), rc
Z = 2, a = 9.9620(7), b = 17.8492(9), c = 18.4538(12) A, a =
64.282(3), b = 84.813(2), g = 85.112(2)1, V = 2940.3(3) A3, m =
0.222 cmꢂ1. Numbers of measured and unique reflections were 9701
and 5697, respectively (Rint = 0.145). Final R(F) = 0.1200, wR(F2) =
0.2542 for 882 parameters and 3469 reflections with I > 2s(I) and
4.92oyo20.681 (corresponding R-values based on all 5697 reflections
are 0.1841 and 0.2873, respectively).
1 (a) A. Hirsch and M. Brettreich, Fullerenes—Chemistry and
Reactions, Wiley-VCH, Weinheim, 2004; (b) D. M. Guldi and
N. Martın, Fullerenes: From Synthesis to Optoelectronic Properties,
´
ed. Kluwer Academic Publishers, Dordrecht, 2002; (c) N Martin
and J.-F. Nierengarten, Guest ed.; Supramolecular Chemistry of
Fullerenes, Tetrahedron Symposium in Print, 2006, 62, 1923–2131.
2 M. Yamada, P. Rivera-Fuentes, W. B. Schweizer and F. Diederich,
Angew. Chem., Int. Ed., 2010, 49, 3532.
Fig. 2 Reaction path for the rearrangement of (ꢀ)-2a to (ꢀ)-3a. Free
energies in kJ molꢂ1 are given, calculated at the ONIOM(B3LYP/
6–31G(d):AM1) level. The ONIOM partitioning is illustrated with
(ꢀ)-2a —the atoms highlighted in green correspond to the high-level
layer.9
3 (a) A. L. Balch, D. A. Costa, J. W. Lee, B. C. Noll and
M. M. Olmstead, Inorg. Chem., 1994, 33, 2071; (b) G. Schick,
A. Hirsch, H. Mauser and T. Clark, Chem.–Eur. J., 1996, 2, 935;
(c) F. Djojo, A. Herzog, I. Lamparth, F. Hampel and A. Hirsch,
Chem.–Eur. J., 1996, 2, 1537; (d) A. L. Balch, D. A. Costa, C. Noll
and M. M. Olmstead, Inorg. Chem., 1996, 35, 458; (e) B. Gross,
V. Schurig, I. Lamparth, A. Herzog, F. Djojo and A. Hirsch, Chem.
Commun., 1997, 1117; (f) T.-Y. Hsiao, S. K. Chidambareswaran and
C.-H. Cheng, J. Org. Chem., 1998, 63, 8617; (g) Y. Rubin,
P. S. Ganapathi, A. Franz, Y.-Z. An, W. Qian and R. Neier,
Chem.–Eur. J., 1999, 5, 3162; (h) L.-W. Guo, X. Gao,
D.-W. Zhang, S.-H. Wu, H.-M. Wu and Y.-J. Li, Chem. Lett.,
1999, 411; (i) L.-W. Guo, X. Gao, D.-W. Zhang, S.-H. Wu,
H.-M. Wu, Y.-J. Li, S. R. Wilson, C. F. Richardson and
D. I. Schuster, J. Org. Chem., 2000, 65, 3804; (j) G. P. Miller,
M. C. Tetreau, M. M. Olmstead, P. A. Lord and A. L. Balch, Chem.
Commun., 2001, 1758; (k) L. Ulmer, C. Siedschlag and J. Mattay,
In summary, we have described a novel thermal rearrange-
ment of push–pull-chromophore–fullerene conjugates to yield
new CT chromophore–fullerene conjugates possessing
a
1,2,9,12-tetrakis-addition pattern, which was unambiguously
elucidated by X-ray crystallography. This reaction will now be
further exploited and its broad scope explored by introducing
different R-residues on the surface, substituting different
donors (such as ferrocene, tetrathiafulvalene) for the DMA
moiety and by investigating the rearrangements of fullerene
conjugates with 7,7,8,8-tetracyano-p-quinodimethane (TCNQ)
and tetrafluoro-TCNQ-based push–pull chromophores.
Eur. J. Org. Chem., 2003, 3811; (l) N. Martı
´
S. Filippone and A. Martı
´
n, M. Altable,
´
n, M. Altable, S. Filippone, A. Martı
n-Domenech, Chem. Commun., 2004,
´
1338; (m) N. Martı
´
´
n-
Domenech, A. Poater and M. Sola, Chem.–Eur. J., 2005, 11, 2716;
(n) N. Martın, M. Altable, S. Filippone, A. Martın-Domenech,
M. Guell and M. Sola, Angew. Chem., Int. Ed., 2006, 45, 1439;
This work was supported by the Swiss National Science
Foundation and the ERC Advanced Grant NO. 246637
(‘‘OPTELOMAC’’). We thank the C4 Competence Center
´
´
¨
(o) M. Altable, S. Filippone, A. Martı
M. Sola and N. Martın, Org. Lett., 2006, 8, 5959; (p) M. Guell,
N. Martın, M. Altable, S. Filippone, A. Martı
M. Sola, J. Phys. Chem. A, 2007, 111, 5253; (q) M. Izquierdo, S. Osuna,
S. Filippone, A. Martın-Domenech, M. Sola and N. Martın, Eur. J.
Org. Chem., 2009, 6231; (r) M. Izquierdo, S. Osuna, S. Filippone,
A. Martın-Domenech, M. Sola and N. Martın, J. Org. Chem., 2009, 74,
1480; (s) M. Izquierdo, S. Osuna, S. Filippone, A. Martın-Domenech,
M. Sola and N. Martın, J. Org. Chem., 2009, 74, 6253.
´
n-Domenech, M. Guell,
¨
´
¨
for Computational Chemistry at ETH Zurich and the ETH
¨
´
´
´
n-Domenech and
High Performance Cluster Brutus for the allocation of
computational resources. M. Y. acknowledges the receipt of
a JSPS Research Fellowship for Young Scientists.
´
´
´
´
´
´
Notes and references
4 (a) K. Komatsu, Y. Murata, N. Takimoto, S. Mori, N. Sugita and
T. S. M. Wan, J. Org. Chem., 1994, 59, 6101; (b) H. L. Anderson,
R. Faust, Y. Rubin and F. Diederich, Angew. Chem., Int. Ed. Engl.,
1994, 33, 1366.
5 M. J. Frisch et al., Gaussian03 and Gaussian09, Revision A.01, see
ESIw for full reference.
´
6 S. Dapprich, I. Komaromi, K. S. Byun, K. Morokuma and
z Crystal data of (ꢀ)-2a at 123 K (CCDC-770343): C78H15N5 ꢃ 3 CꢂS23,
ꢀ
Mr = 1250.430, triclinic, space group P1 (no. 2), rc = 1.596 Mg cm
,
Z = 2, a = 13.0651(3), b = 13.5710(3), c = 15.9971(5) A, a =
74.0161(11), b = 77.3303(12), g = 75.1841(12)1, V = 2602.02(12) A3,
m = 0.325 cmꢂ1. Numbers of measured and unique reflections were
19258 and 11674, respectively (Rint = 0.061). Final R(F) = 0.0748,
wR(F2) = 0.2109 for 893 parameters and 8566 reflections with I >
2s(I) and 3.27oyo27.461 (corresponding R-values based on all 11674
reflections are 0.1008 and 0.2294, respectively). Crystal data of (ꢀ)-3a
at 123 K (CCDC-770342): C78H15N5 ꢃ 1.5 C6H5Cl, Mr = 1190.851,
monoclinic, space group P21/c (no. 14), rc = 1.578 Mg cmꢂ3, Z = 4,
a = 15.9413(3), b = 18.6916(3), c = 17.0131(3) A, b = 98.6182(6)1,
V = 5012.1(2) A3, m = 0.170 cmꢂ1. Numbers of measured and unique
reflections were 19793 and 10912, respectively (Rint = 0.057). The
structure contains 1.5 chlorobenzene molecules. Final R(F) = 0.0664,
wR(F2) = 0.1874 for 902 parameters and 8574 reflections with I >
2s(I) and 5.91oyo27.481 (corresponding R-values based on all 10912
reflections are 0.0864 and 0.2032, respectively). Crystal data of (ꢀ)-3b
M. J. Frisch, THEOCHEM, 1999, 462, 1.
7 (a) S. Osuna, J. Morera, M. Cases, K. Morokuma and M. Sola,
J. Phys. Chem. A, 2009, 113, 9721; (b) S. Osuna and K. N. Houk,
Chem.–Eur. J., 2009, 15, 13219; (c) J. L. Delgado, S. Osuna,
P.-A. Bouit, R. Martinez-Alvarez, E. Espildora, M. Sola and
N. Martin, J. Org. Chem., 2009, 74, 8174; (d) X. Lu, F. Tian,
X. Xu, N. Wang and Q. Zhang, J. Am. Chem. Soc., 2003, 125,
10459.
8 Mulliken charge analysis.
9 See ESIw for the reaction path of the rearrangement of 2 with
R = H and the energies at the level of theory: ONIOM(B3LYP/
6–31G(d):SVWN/STO–3G).
ꢁc
This journal is The Royal Society of Chemistry 2010
5336 | Chem. Commun., 2010, 46, 5334–5336