Angewandte
Chemie
DOI: 10.1002/anie.201107490
Cage Compounds
Chemistry at the Nanoscale: Synthesis of an N@C60–N@C60 Endohedral
Fullerene Dimer**
B. J. Farrington, M. Jevric, G. A. Rance, A. Ardavan, A. N. Khlobystov, G. A. D. Briggs, and
K. Porfyrakis*
Endohedral fullerenes—carbon cages with one or more
heteroatoms incarcerated within them—are one of the most
exotic classes of molecules. They possess a wealth of
fascinating functional properties, including magnetism and
photoactivity. While the properties of individual endohedral
fullerene molecules are remarkable,[1] the full functional
potential of many of these species will likely be realized in
systems where two or more of these molecules are connected.
One prominent example is the endohedral fullerene radical
species N@C60. The highly symmetric central location of the
nitrogen atom within the C60 cage,[2] with minimal mixing
between the nitrogen atom and fullerene electron wave-
functions, imparts a degree of isolation to the nitrogen radical
that is usually only attainable using an ion trap or in an atomic
including the creation of controlled entanglement between
electron spins, an array of at least two N@C60 centers with
fixed separation is required to understand the interaction of
neighboring spin centers with one another and investigate
methods of selective manipulation of one of the spin centers.
Herein we describe a one-pot method, using a double 1,3-
dipolar cycloaddition, for the synthesis of a two-center N@C60
molecule with fixed spatial separation (Scheme 1), thus
providing a platform for future experiments to probe the
nature of the electron interaction between two N@C60
molecules to assess the potential capabilities of a 2 qubit
N@C60 quantum computer. To the best of our knowledge, this
is the first endohedral fullerene dimer comprising two
chemically linked N@C60 spin centers.
4
gas. This isolation allows the nitrogen center to retain a S3/2
A Prato 1,3-dipolar cycloaddition reaction was chosen to
link the two N@C60 molecules owing to its proven compat-
ibility with the stability[10] of N@C60 and the wide range of
amino acid and aldehyde derivatives that it has been reported
to be compatible with. A dibenzaldehyde-terminated oligo(p-
phenylene polyethylene) (OPE) molecule (Ald-3Bz) was
synthesized and used as the spacing unit because of its rigidity
and modular construction. The incorporation of long alkyl
chains on the central phenyl group enhances the inherent
poor solubility of the rigid, linear molecule, thus allowing the
reaction to be performed at high concentration and making
the product fullerene dimer (dimer-3Bz) soluble in common
solvents such as chloroform. Previous work has shown that
the structure of the amino acid derivative used makes
a significant difference to the rate and yield of the 1,3-dipolar
cycloaddition reaction.[11] Work within our group has found
that the N-(4-(hexyloxy)benzyl)glycine amino acid derivative
greatly enhances the dimer formation reaction rate compared
to more commonly used amino acid derivatives such as N-
(ethyl)glycine (see the Supporting Information). In addition,
the amino group produces a fullerene dimer that can be
purified using standard single-pass HPLC methods.
electron spin ground state[3] and to have an extraordinarily
long electron spin coherence time (T2e of 250 ms).[4] This
remarkable spin coherence time has lead many research
groups to study the potential and feasibility of an N@C60-
based quantum computer over the past decade.[5] Owing to
a number of synthetic challenges associated with N@C60,
experimental electron spin resonance studies have so far been
restricted to probing the interaction of ensembles of mole-
cules each containing a single N@C60 spin center. These
challenges center on the low-yielding production methods
available to synthesize N@C60, thus producing at best
a 500 ppm N@C60/C60 mixture,[2,6] and the time consuming
processing techniques currently required to enrich the ratio of
N@C60 to C60.[7] In addition, compared to C60, chemically
functionalized N@C60 derivatives have significantly lower
thermal and photo stability.[8,9] To further explore the
potential of N@C60 as a quantum computing element,
[*] B. J. Farrington, Prof. G. A. D. Briggs, Dr. K. Porfyrakis
Department of Materials, University of Oxford
Oxford OX1 3PH (UK)
E-mail: kyriakos.porfyrakis@materials.ox.ac.uk
We synthesized three grams of a N@C60/C60 mixture using
an ion implantation method,[12] thus yielding an average
purity of 50 ppm of N@C60 in C60. Extensive purification of
this sample followed. After 19 recycling HPLC runs,[13] a high
purity N@C60/C60 peak was isolated. The integrated area of
the N@C60/C60 peak indicated the total mass of N@C60/C60 to
be 10 mg. This sample then underwent quantitative ESR
analysis of the total number of N@C60 spin centers present,
thereby giving a lower bound for the mass of N@C60 to be
4.6 mg.
Dr. M. Jevric, Dr. G. A. Rance, Prof. A. N. Khlobystov
School of Chemistry, University of Nottingham
Nottingham NG7 2RD (UK)
Dr. A. Ardavan
CAESR, The Clarendon Laboratory, Department of Physics
Oxford University, Oxford OX1 3PU (UK)
[**] We acknowledge the EPSRC and NSF for funding (EP/F028806/01).
B.J.F. is supported by the EPSRC through a DTA studentship. A.A.
and A.N.K. are supported by the Royal Society. We thank Dr. Simon
Plant for valuable suggestions and discussions during the course of
this work.
Optimized reaction conditions and an efficient purifica-
tion method for the one-pot dimer synthesis were identified
and scaled down to work using only 20 mg of the C60 starting
Supporting information for this article is available on the WWW
Angew. Chem. Int. Ed. 2012, 51, 3587 –3590
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3587