A reassignment of the EPR spectra previously attributed to Cu@C 60

Article abstract of DOI:10.1039/b700320j

EPR spectra attributed to the endohedral metallofullerene Cu@C₆₀ are better explained by the previously characterized Cu(ii) dithiocarbamate family of compounds. The Royal Society of Chemistry.

Full text of DOI:10.1039/b700320j

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
A reassignment of the EPR spectra previously attributed to Cu@C60
a
Bevan Elliott, Keqin Yang, Apparao M. Rao, Hadi D. Arman, William T. Pennington and
b
b
a
a
a
Luis Echegoyen*
Received (in Cambridge, UK) 9th January 2007, Accepted 26th January 2007
First published as an Advance Article on the web 22nd February 2007
DOI: 10.1039/b700320j
EPR spectra attributed to the endohedral metallofullerene
Cu@C60 are better explained by the previously characterized
Cu(II) dithiocarbamate family of compounds.
In an effort to duplicate the published results, a modified K–H
4
apparatus was utilized. In a 300 Torr helium atmosphere, an 80 A
dc arc was struck between a graphite anode and a core-drilled
cathode into which a piece of copper metal had been inserted. The
soluble portion of the soot produced from this synthesis was
Though many endohedral mono-metallofullerenes having rela-
tively large carbon cages have been isolated, characterized, and
chemically functionalized, M@C60 and M@C70 species have so far
remained largely elusive due to their low solubility in common
organic solvents and their general instability in air. Recently, a
report of the synthesis, isolation, and characterization of a new
endohedral fullerene, Cu@C60, was published by Huang et al in
extracted in CS , and EPR spectra showed the presence of the
2
copper radical previously reported, but in very small amounts. To
increase the yield (monitored by the intensity of the EPR signal),
variations in the elemental composition and configuration of the
rods, helium pressure, arc current, and extraction solvents were
explored. Cathode composition included pure graphite, core-
drilled graphite with an inserted copper slug, and pure copper.
Anodes were either pure graphite or composite carbon/copper
powder annealed molded rods. Extraction solvents consisted of
CS , pyridine, or mixtures of the two, since M@C species have
1
this journal. In their synthesis, copper ions sputtered into a
nitrogen rf-plasma collided with sublimated C60, and the resulting
2
deposit was solubilized in CS and subsequently examined by EPR
spectroscopy. The two observed overlapping patterns of four
completely resolved hyperfine lines having slightly different
hyperfine splittings and the correct isotopic intensities gave
unequivocal proof of paramagnetism arising from unpaired
2
60
been shown to be preferentially extractable in pyridine and
5
aniline. Yields of the copper-centered radical remained low until
both yttrium and nickel were added in catalytic amounts to the
anode. Milligram quantities of the radical compound were then
6
3,65
electron density on
spectrum with a main peak at m/z 783, the g-value of the EPR
hyperfine pattern, and its m -dependent line widths, the assignment
Cu nuclei. On the basis of a TOF-MS
2
collected when a 1 : 1 mixture of CS –pyridine was used to extract
I
soot produced by arcing a carbon/Y/Ni composite anode against a
pure copper cathode.
2+
to a carbon-60 cage containing an off-center Cu ion was made.
In 2005, Dinse and co-workers published both room-temperature
The compound was purified by column chromatography in a
1 : 1 eluent ratio of toluene : CS after first evaporating the CS –
and liquid nitrogen-frozen CS solution EPR spectra of a radical
2
2
2
which was produced in a standard Kr a¨ tschmer–Huffman (K–H)
arcing apparatus by unintentional addition of copper from the
2
graphite electrode mounts. Since their room-temperature solution
pyridine mixture and redissolving and filtering in CS . Two yellow
2
f
bands having similar R values by TLC were eluted separately, and
the band containing the paramagnetic species was collected, dried,
spectrum was identical to that in the previous study, Dinse and co-
workers also assigned it to Cu@C60, though they did not report a
and redissolved in CS for EPR studies. Fig. 1 shows the X-band
2
first-derivative EPR absorption spectrum of the radical in solution
I
mass spectrum of the compound. The m -dependent spectral line
at room temperature. With g-values of 2.0483 and 2.0477 and
63
widths and the temperature behavior, as well as DFT structure
optimizations of the radical, were believed to be indications that
65
hyperfine coupling constants of 77.4 and 82.7 G for Cu and Cu
nuclei, respectively, in an isotopic ratio of 2.25, this spectrum
2
+
the off-center Cu endohedral ion was ‘internally docking’ to the
inner carbon cage on the EPR timescale. Also in 2005, a
theoretical paper motivated by the first report of Cu@C was
1
is identical to those reported by Huang et al, and Dinse and
2
co-workers.
60
MALDI-TOF-MS of this compound did not yield the expected
peak corresponding to Cu@C60 despite repeated attempts. Instead,
shown in Fig. 2, the most intense feature of the spectrum was a
double peak at m/z 383 and 385 with an isotopic ratio indicative of
a copper-containing species. Cyclic voltammetry in o-dichloroben-
3
published. In it, energy optimization calculations using the
Hartree–Fock method yielded a model of Cu@C60 with an off-
center copper ion inside a deformed carbon cage and a small but
positive binding energy. In this paper, we report attempts to
synthesize Cu@C60 by K–H arc in large enough quantities for
complete characterization, and the surprising finding that the
isolated radical corresponds to a dithiocarbamate copper complex,
not Cu@C . However, the radical’s EPR spectrum is identical to
zene (Fig. 3) showed only two reversible waves at 0.11 and
+
1.04 V vs. Fc/Fc within the entire solvent window, a behaviour
2
that is not typical of fullerenes. At this point, it was discovered that
the compound was highly soluble in many organic solvents
ranging in polarity from CS2 to methanol. Again, no known
pristine fullerenes behave this way. The compound was then
dissolved in THF, and UV-Vis spectra collected. The spectrum in
Fig. 4 shows an intense absorption at 272 nm with a shoulder at
287 nm which dominates the spectrum, and a broad band with
60
that assigned to Cu@C60
.
a
Department of Chemistry, Clemson University, Clemson, SC, USA.
E-mail: luis@clemson.edu; Fax: +1 (864) 656 6613;
Tel: +1 (864) 656 0778
Department of Physics, Clemson University, Clemson, SC, USA
b
This journal is ß The Royal Society of Chemistry 2007
Chem. Commun., 2007, 2083–2085 | 2083

View Article Online
Fig. 4 UV-Vis spectrum in THF of the purified copper compound.
2
+
Fig. 1 X-Band EPR spectra in CS
2
of the purified Cu radical
lmax at 435 nm followed by a very broad, weaker, tailing band that
extends past 700 nm.
compound.
As no clear evidence indicating the presence of Cu@C60 was yet
present except for the EPR spectrum in our characterization of this
compound, crystallization and X-ray diffraction were attempted.
2
Slow evaporation of a concentrated solution in CS yielded flat
reddish-black parallelpipeds which were examined by single-crystal
X-ray diffraction at room temperature. The resulting structure, a
copper ion coordinated to two thiocarbamate ligands with
terminal piperidine groups, shown in Fig. 5, was quite
unexpected.{ A search of single-crystal structures yielded a result
6
identical to ours, namely, bis(piperidine-1-dithiocarbamato)
copper(II), the divalency of the copper causing the paramagnetism
2
+
2
of the compound. The complexation of CS ligands by Cu is
easily explained, but the origin of our piperidine N-terminal
ligands is not as clear. One possible source is the pyridine used in
the extraction.
2+
Fig. 2 MALDI-TOF-MS spectrum of the purified Cu
radical
More information about the Cu(II) dithiocarbamate family of
compounds was obtained from an extensive literature search.
Those reports mentioned herein are merely representative, not
exhaustive, since the literature abounds with such examples. The
first report of electron spin resonance of a divalent copper
compound.
7
dithiocarbamate compound was published in 1959 in Nature. The
reported X-band benzene solution spectrum of the Cu(II)
diisopropyl dithiocarbamate complex is in all respects identical
Fig. 5 Single-crystal X-ray structure of bis(piperidine-1-dithiocar-
bamato)copper(II). Thermal ellipsoids are shown at 50% probability and
solvent molecules are omitted for clarity. The central metal atom lies on an
inversion center and the atoms generated by the inversion are labelled with
suffix (A).
Fig. 3 Cyclic voltammogram in oxygen-free THF with TBAPF
6
as the
+
supporting electrolyte. Pt working and auxiliary electrodes with a Ag/Ag
pseudo-reference electrode were utilized in a one-compartment cell.
2
084 | Chem. Commun., 2007, 2083–2085
This journal is ß The Royal Society of Chemistry 2007

View Article Online
to that published by Huang et al., and Dinse and co-workers,
and as shown in this report. According to this early article,
exchanging the terminal groups on the dithiocarbamate
ligand (ethyl, isopropyl and methyl phenyl) had little or no effect
on the solution EPR spectrum. Also interesting in this article
We thank the U.S. National Science Foundation (Grants
CHE0509989 and DMI0304019) for support of this work. Thanks
to Amit Palkar for assistance in the early stages of this
investigation. This material was based on work supported by the
National Science Foundation, while working at the Foundation.
Any opinions, findings, conclusions, or recommendations
expressed in this material are those of the authors, and do not
necessarily reflect the views of the National Science Foundation.
was the authors’ explanation of the
I
m -dependent line
widths, which were attributed to insufficient averaging of
anisotropy in the resonance structure. Another solution EPR
study of a series of Cu(II) dithiocarbamate complexes offered a
different interpretation: that the variation in linewidths is due to
the orientation of the copper nuclear spin with respect to the
Notes and references
8
6 2 2
{ The unit cell parameters for [Cu(C H10NS ) ] synthesized in this study
applied magnetic field.
are as follows: space group = P2 /c, a = 6.1726(16), b = 5767(2), c =
1
3
EPR spectra of solid-phase cupric dithiocarbamate complexes
have also been reported in the literature. Room-temperature
EPR spectra of thin films of PVC doped with 0.5% Cu(II)
˚
˚
1
5.266(4) A, b = 95.442(10)u, V = 804.5589 A .
1 H. Huang, M. Ata and Y. Yoshimoto, Chem. Commun., 2004, 1206.
2 C. Knapp, N. Weiden and K. Dinse, Magn. Reson. Chem., 2005, 43,
S199.
9
diethyldithiocarbamate are identical to the low-temperature
spectra reported by Dinse and co-workers, as are liquid-
nitrogen solid-solution spectra of the same molecule in another
3
V. Gurin, Int. J. Quantum Chem., 2005, 104, 249.
4 K–H arcing apparatus in the lab of Prof. A. M. Rao, Department of
Physics, Clemson University.
1
0
article. Previously reported voltammetry and UV-Vis spectra of
cupric dithiocarbamate complexes with a series of different
N-terminal ligands showed qualitative similarity with our results,
with the N,N-dipiperidine species quantitatively identical to our
5
L. Wang, J. Alford, Y. Chai, M. Diener, J. Zhang, S. McClure, T. Guo,
G. Scuseria and R. Smalley, Chem. Phys. Lett., 1993, 207, 354;
Y. Kubozono, T. Ohta, T. Hayashibara, H. Maeda, H. Ishida,
S. Kashino, K. Oshima, H. Yamazaki, S. Ukita and T. Sogabe, Chem.
Lett., 1995, 457; Y. Kubozono, T. Noto, H. Maeda, S. Kashino,
S. Emura, S. Ukita and T. Sogabe, Chem. Lett., 1996, 453;
Y. Kubozono, H. Maeda, Y. Takabayashi, K. Hiraoka, T. Nakai,
S. Kashino, S. Emura, S. Ukita and T. Sogabe, J. Am. Chem. Soc.,
1996, 118, 6998.
11
data.
Though theoretically the endohedral mono-metallofullerene
3
Cu@C60 is a stable molecule, the only evidence to date of the
existence of Cu@C60 is the mass spectrum reported by Huang
1
et al. However, the EPR spectra assigned to Cu@C60 can be
6 A. Radha, M. Seshasayee and G. Aravamudan, Acta Crystallogr., Sect.
C: Cryst. Struct. Commun., 1988, 44, 1378.
accounted for by the compounds formed by the complexation of
˚
T. V a¨ nngard and S. Akerstr o¨ m, Nature, 1959, 184, 183.
7
2
+
CS
2
in solution by Cu ions. Despite repeated attempts under a
8 J. Gibson, Trans. Faraday Soc., 1964, 60, 2105.
9 B. Jeliazkowa, I. Petkov, G. Sarova, T. Deligeorgiev and M. Evstatiev,
Polym. Degrad. Stab., 1996, 51, 301.
wide variety of conditions, we failed to detect the presence of
Cu@C . More research, both experimental and theoretical, is
60
1
1
0 H. Gersmann and J. Swalen, J. Chem. Phys., 1962, 36, 3221.
1 A. Hendrickson, R. Martin and N. Rohde, Inorg. Chem., 1976, 15,
2115.
needed to ascertain the stability and existence of divalent d-block
metals encapsulated by small fullerenes.
This journal is ß The Royal Society of Chemistry 2007
Chem. Commun., 2007, 2083–2085 | 2085