Fullerides: Heterometallic superconductors with composition M2M′C60 (M = K, Rb; M′ = Yb, Lu, Sc)

Article abstract of DOI:10.1007/s11172-005-0018-9

One- or two-step reactions of potassium and rubidium fullerides with composition M_k C₆₀ (M = K, Rb; k = 3-6) and K ₆C₆₀ + m K mixtures (m = 1, 3) with anhydrous salts M′Cl₃ (M′ = La, Pr, Nd, Sm, Gd, Tb, Yb, Lu, Y, Sc) and YbI2 in a toluene-THF medium afforded heterometallic fullerides M3- _n M′_n C₆₀ (n = 1-3). Among these compounds, substituted fullerides with composition M₂M′C₆₀ (M′ = Yb, Lu, Sc) display superconducting properties with critical temperatures of 14-20 K.

Full text of DOI:10.1007/s11172-005-0018-9

1686
Russian Chemical Bulletin, International Edition, Vol. 53, No. 8, pp. 1686—1692, August, 2004
Fullerides: heterometallic superconductors with composition
M2M´C60 (M = K, Rb; M´ = Yb, Lu, Sc)
b
b
a
c
B. M. Bulychev,a R. A. Lunin, V. A. Kul´bachinskii, R. V. Shpanchenko, and V. I. Privalov
ꢀ
aDepartment of Chemistry, Moscow State University,
Leninskie Gory, 119992 Moscow, Russian Federation.
Fax: +7 (095) 932 8846. E_mail: b.bulychev@highp.chem.msu.ru
bDepartment of Physics, Moscow State University,
Leninskie Gory, 119992 Moscow, Russian Federation.
Fax: +7 (095) 932 8876. Eꢀmail: kulb@mig.phys.msu.ru
cN. S. Kurnakov Institute of General and Inorganic Chemistry, Russian Academy of Sciences,
31 Leninsky prosp., 119991 Moscow, Russian Federation.
Eꢀmail: privalov@igic.ras.ru
Oneꢀ or twoꢀstep reactions of potassium and rubidium fullerides with composition MkC60
(M = K, Rb; k = 3—6) and K6C60 + m K mixtures (m = 1, 3) with anhydrous salts M´Cl3
(M´ = La, Pr, Nd, Sm, Gd, Tb, Yb, Lu, Y, Sc) and YbI2 in a toluene—THF medium afforded
heterometallic fullerides M3–nM´nC60 (n = 1—3). Among these compounds, substituted
fullerides with composition M2M´C60 (M´ = Yb, Lu, Sc) display superconducting properties
with critical temperatures of 14—20 K.
Key words: fullerenes, fullerides, lanthanides, potassium, rubidium, ytterbium, lutetium,
scandium, superconductivity.
The insertion of metal atoms into voids of the lattice
serves as the superconducting phase in the Ba—C60 sysꢀ
tem. The superconducting transition temperatures of these
compounds having different compositions and strucꢀ
tures are lower than the lowest temperature found for
K3C60 (19 K).19
In the present study, we dealt with homoꢀ and heteroꢀ
metallic fullerides of scandium, yttrium, lanthanides, and
alkali metals, which were prepared by homogeneousꢀ
heterogeneous redox reactions of potassium and rubidium
fullerides with lanthanide(III) chlorides and ytterbium(II)
iodide in a toluene—THF solution at 300—340 K.
of C60 fullerite leads to successive filling of the t1u elecꢀ
tron sublevel of the fullerene molecule and gives rise to
fullerides possessing the superconducting properties. These
properties have been discovered for the first time in homoꢀ
and heteronuclear trimetallic fullerides of heavy alkali
metals1—5 with general composition MM´2C60 (M = K,
Rb; M´ = K, Rb, Cs) having three electrons in the t1u
sublevel of the fullerene molecule. These compounds crysꢀ
tallize in the faceꢀcentered cubic (fcc) lattice.
It was hypothesized6,7 that fullerides could regain their
superconducting properties upon transfer of more than
six electrons to the C60 molecule, i.e., upon filling of the
next higherꢀlying t1g sublevel of the lowest unoccupied
MO of fullerene, and the rareꢀearth fullerides could take
on even unique superconducting properties.8 These exꢀ
pectations were only partially realized. Among alkalineꢀ
earth fullerides, the maximum superconducting transiꢀ
tion temperature (8.4 K) was found for faceꢀcentered cuꢀ
Experimental
Toluene and THF were purified and dehydrated according
to standard procedures. Anhydrous chlorides M´Cl3 (M´ = La,
Pr, Nd, Sm, Gd, Tb, Yb, Lu, Y, Sc) were prepared by heating
the corresponding crystal hydrates in the presence of ammoꢀ
nium chloride according to a known procedure;20 YbI2 was synꢀ
thesized by the reaction of ytterbium metal with iodine in THF21
and desolvated by heating in vacuo. Potassium and rubidium
fullerides were prepared by the reaction of a suspension of C60 in
toluene with the calculated amount of the corresponding metal
at 120—130 °C according to a procedure described earlier.22
The synthesis of the starting alkali fullerides, mixing of suspenꢀ
sions of the fullerides with suspensions of M´Cl3 or YbI2 in
THF, the exchange reactions, and all intermediate operations
were carried out in vacuo in allꢀsealed evacuated devices equipped
9
bic Ca5C60 (4, 7, and 5.3 K for bodyꢀcentered cubic
Sr6C60,10 Ba6C60,11 and K3Ba3C60,12 respectively). In a
series of rareꢀearth fullerides, the maximum superconꢀ
13
ducting transition temperature is 12.5 K for La3+xC60
(6 and 8 K for orthorhombic Yb2.75C60 and Sm3C60,15
14
respectively). However, the superconducting transition
was not confirmed for Yb2.75C60 16 and Ba6C60.17,18 In the
studies,16,17 it was suggested that orthorhombic Ba4C60
Published in Russian in Izvestiya Akademii Nauk. Seriya Khimicheskaya, No. 8, pp. 1623—1628, August, 2004.
1066ꢀ5285/04/5308ꢀ1686 © 2004 Springer Science+Business Media, Inc.

Fullerides: heterometallic superconductors
Russ.Chem.Bull., Int.Ed., Vol. 53, No. 8, August, 2004
1687
with a system of breakable glass valves and metallic anvils. Below,
one of procedures is exemplified by the synthesis of K2YbC60.
Synthesis of K2YbC60 from K5C60 and YbCl3. A reactor А
containing a suspension of K5C60 (0.25 g) in toluene (15 mL)
was sealed to a reactor B containing a suspension of YbCl3
(0.076 g) in THF (10 mL). The system was evacuated and sealed
off from the vacuum line. Then all necessary operations associꢀ
ated with the replacement of the solvent, mixing of the reagents,
and removal of wastes were performed. Finally, the intermediꢀ
ate reactor B was sealed off. The mixture of the compounds in
the reactor А was kept in an oven at 75—85 °C for 15—20 days.
The suspension over the crimsonꢀred solution was stirred at
regular intervals by shaking. After completion of the synthesis, a
special system was sealed to the reactor А. This system made it
possible to remove the solvent, dry the precipitate in vacuo, pack
the dry substance in tubes, and fill the system with helium (at
350—400 Torr) necessary for thermal exchange between a cryoꢀ
stat and a sample in determining the superconducting propꢀ
erties.
The Xꢀray diffraction spectra of the compounds were reꢀ
corded on a STADI/P (STOE) diffractometer (CuꢀKα radiaꢀ
tion) in vacuumꢀsealed glass capillaries. The structures of the
fullerides were refined using the GSAS program.23 The atomic
coordinates published in the study24 were used as the starting
models. The background was approximated graphically and
refined using polynoms of different degrees. The occupancies
of the K atom positions were refined with fixed thermal paꢀ
rameters.
The 13C NMR spectra were recorded on a Bruker ACꢀ200
spectrometer according to a singleꢀpulse program (excitation
pulse duration was 4 µs, the pulse interval was 2 s). The number
of accumulations was varied from 400 to 4000 depending on the
amount and composition of the compounds. The chemical shifts
in the 13C NMR spectra were determined relative to the signal
of Me4Si.
The superconducting transition temperatures in the fullerides
were determined by a standard lowꢀfrequency inductive method
from the temperature dependence of the magnetic susceptibility
in a temperature range of 4.2—300 K (accuracy of the determiꢀ
nation of the transition temperature was 1 K) using the same
tubes as those used for NMR spectroscopy.
Twoꢀstep procedure for the synthesis of K2YbC60. In the first
step, YbC60 was prepared by the reaction of K3C60 with a susꢀ
pension of YbCl3 in THF (10 mL). Then the calculated amount
of potassium metal was added.
The compounds synthesized according to the aboveꢀmenꢀ
tioned two procedures were found to have the same Xꢀray difꢀ
fraction patterns and electrophysical properties.
Other heterofullerides of rareꢀearth metals and alkali metals
were synthesized analogously.
Results and Discussion
Heterofullerides of lanthanides and alkali metals were
synthesized on the assumption that the reactions (1)—(5)
Table 1. Tentative compositions and selected properties of heterofullerides of alkali metals and lanthanides
Speꢀ
Composiꢀ
Tentative
composition
of heteroꢀ
fulleride
Tc/K
Color of
compound configuration
of the
Electronic
13C NMR, δ
cimen tion of the
starting
mixturea
metal atom
1
2
3
4
5
6
7
8
5 K + ScCl3
5 K + YCl3
K2ScC60
K2YC60
K3LaC60
K2LaC60
14
—
—
—
Black
Black
Black
Darkꢀbrown
Black
Black
Black
Black
Black
Black
Black
Black
Black
Black
Black
Black
Black
Black
Black
Black
3d14s2
190
189
190 (br)
216 (v.br)
191
4d15s2
6 K + LaCl3
5 K + LaCl3
5 K + PrCl3
5 K + NdCl3
5 K + SmCl3
9 K + 3 GdCl3
5 K + GdCl3
5 K + TbCl3
4 K + YbI2
5d16s2
5d16s2
b
K2PrC60
—
4f35d06s2
4f45d06s2
4f65d06s2
4f75d16s2
4f75d16s2
4f95d06s2
4f145d06s2
4f145d06s2
4f145d06s2
4f145d06s2
4f145d06s2
4f145d06s2
4f145d06s2
4f145d06s2
4f145d06s2
4f145d16s2
K2NdC60
K2Sm3+C60
Gd3C60
K2GdC60
K2TbC60
K2Yb2+C60
K2Yb2+C60
Rb2Yb2+C60
K2Yb3+C60
K2Yb3+C60
Rb2Yb3+C60
Rb2Yb2+C60
K2Yb2+2C60
Na2Yb3+C60
K2LuC60
—
—
—
187
187
194 (v.br)
196 (v.br)
b
9
—
c
10
11
12
13
14
15
16
17
18
19
20
—
15
16.5
14
13
16
22
19
—
—
198
4 K + YbI2
193 (br.asymm)
c
4 Rb + YbI2
5 K + YbCl3
5 K + YbCl3
5 Rb + YbCl3
5 Rb + YbI2
5 K + 2 YbI2
5 Na + YbCl3
5 K + LuCl3
—
196 (br.asymm)
—
c
—
—
c
206
182
191
—
20
a The number of K atoms per C60 molecule.
b The compound contains a superconducting phase, but its amount is too small to allow us to determine
the superconducting transition temperature within the framework of the method used for measurements.
c Polymer.

1688
Russ.Chem.Bull., Int.Ed., Vol. 53, No. 8, August, 2004
Bulychev et al.
T/K
5
10
15
20
25
proceed. The reactions (1) and (3) involve two successive
steps.
K3C60 + M´Cl3
({[M´C60] + 3 KCl} + 2 K)
K2M´C60 + 3 KCl
–0.004
–0.008
–0.012
–0.016
a
(1)
(2)
Tc = 16 K
K5C60 + M´Cl3
K2M´C60 + 3 KCl
2
1
K6C60 + 2 M´Cl3
({[M´2C60] + 6 KCl} + K)
KM´2C60 + 6 KCl
Tc = 16.5 K
(3)
(4)
(5)
{K6C60 + K} + 2 M´Cl3
KM´2C60 + 6 KCl
χ/cm3 g–1
K4C60 + YbI2 К2YbC60 + 2 KI
5
10
15
20
25
T/K
Note. The values in braces and brackets indicate tentative comꢀ
positions of the mixtures of fulleride with potassium chloride,
which formed in the intermediate step of the reaction and were
not isolated from the solution.
–0.004
–0.008
–0.012
–0.016
–1.5
b
As in the earlier study,25 alkali fulleride serves as a
reducing agent in these reactions, which indicates that
the bonds in this compound are of a radically different
kind from the purely ionic bond. Taking into account this
fact, the charges on the ytterbium atoms and the ionic
radii of lanthanides should be considered only as a distinꢀ
guishing characteristic of the starting state of these atoms
(their charge in the halides). Estimation of the degree of
charge transfer in fullerides is beyond the scope of the
present study.
Tc = 14 K
–3.0
Tc = 22 K
1
–4.5
2
χ/cm3 g–1
χ•10–3/cm3 g–1
30 T/K
10
20
The bulk chemical compositions of selected comꢀ
pounds synthesized in the present study and the results of
their investigation are given in Table 1. It can be seen that
only compounds synthesized by the reactions (1), (2),
and (5) possess superconducting properties. The tentative
composition of these compounds is M2M´C60 (M = K,
Rb; M´ = Yb, Lu, Sc). This composition assumes that, on
the one hand, the fcc lattice is retained and, on the other
hand, the t1u sublevel of the fullerene molecule is only
partially filled, no more than five electrons being formally
transferred from the metal atoms.
The temperature dependences of the specific magꢀ
netic susceptibility (χ) for potassiumꢀytterbium, rubidiumꢀ
ytterbium, and potassiumꢀlutetium fullerides are shown
in Fig. 1, a—c. The critical temperatures (Tc) of these
specimens were determined from the onset of superconꢀ
ductivity. Their specific magnetic susceptibilities are comꢀ
parable to the value determined under analogous condiꢀ
tions for individual K3C60, i.e., the percentage of a superꢀ
conducting phase in these specimens was at most 50%.
For potassiumꢀcontaining heterofullerides, the maximum
Tc (20 K) was observed for K2LuC60 (see Fig. 1, c).
Unfortunately, it is presently impossible to unambiguꢀ
ously answer the question of whether substances with comꢀ
position M2M´C60 are individual compounds or the reacꢀ
tions (1), (2), and (5) follow another path to form mixꢀ
tures of phases with partially substituted superconducting
–0.004
–0.008
–0.012
–0.016
c
Tc = 20 K
χ/cm3 g–1
Fig. 1. Temperature dependences of the specific magnetic susꢀ
ceptibility (χ) for K2Yb2+C60 (1) and K2Yb3+C60 (2) (a);
Rb2Yb2+C60 (1) and Rb2Yb3+C60 (2) (b); and K2LuC60 (c).
fullerides K3C60 (Tc = 18—19 K) or Rb3C60 (Tc = 28 K)
containing an impurity of lanthanide fullerides with unꢀ
known compositions.
The broadline 13C NMR spectra of the compounds
are indicative of the presence of fulleride in the solid
phase (single line at δ 188—190 typical of most comꢀ
pounds of this class) and the absence of free fullerite in
the mixture of the products (δ 144—145). Hence, it can
be reasoned that the fullerite introduced into the reaction
occurs completely in the bound state. However, the
fullerite can be both involved in fullerides and consumed
for the formation of polymer fullerite phases or Xꢀray
amorphous fulleride phases, whose presence is observed

Fullerides: heterometallic superconductors
Intensity (rel. units)
Russ.Chem.Bull., Int.Ed., Vol. 53, No. 8, August, 2004
1689
4000
2000
1
2
0
20
40
2θ/deg
Fig. 2. Experimental (1, dotted line), calculated (1, solid line), and difference (2) Xꢀray diffraction patterns of specimen 11.
Table 2. Unit cell parameters (a) and occupancies of the octaꢀ
hedral (8c) and tetrahedral (4b) sites in the structures of homoꢀ
and heterofullerides of potassium and rareꢀearth metals
in Xꢀray diffraction patterns as a halo. A phase similar to
tripotassium fulleride and, depending on the composition
of the starting halide, potassium chloride or iodide were
observed as the major crystalline phases in the Xꢀray difꢀ
fraction patterns (Figs 2 and 3).
b
Specimena
a/Å
Occupancy
of position
AnC60 KHalb
wt.%
The parameter of the fcc lattice of fulleride K3C60,
which was prepared in the presence of toluene according
to a known procedure,22 (Table 2) appeared to be subꢀ
stantially larger than that published in the literature for the
fulleride synthesized by the gasꢀphase method (14.24 Å).24
The refinement of occupancies of the cation positions in
this specimen by the Rietveld method gave the composiꢀ
tion K2.47C60 similar to that calculated in the study24
(K2.41C60), the superconducting transition temperature
(18 K) remaining unchanged. By contrast, fullerides,
which were prepared according to a procedure described
in the study22 but with the use of the reactant ratios
K : C60 = (2—2.5) : 1, do not possess superconducting
properties. In the study,26 fullerite was synthesized by
pyrolysis of C60D24. Alkali heterofullerides characterized
by "expanded" fcc lattices were prepared by intercalation
from solutions in ammonia.27 In the latter case, no inꢀ
crease in the superconducting transition temperature was
8c
4b
K3C60 [K2.41C60]c 14.24
0.8
0.81
—
100
88
—
85
—
—
12
—
15
13
K3C60 [K2.47C60]d 14.3110(5) 0.87 0.73
K2Yb2+C60 (11)d
K2Yb2+C60 (12)d
K2Yb3+C60 (15)d
K2Lu3+C60 (20)d
14.310(1) 1.26 0.75
14.226
14.336(2) 0.75 1.0
14.249(2) 0.75 0.7
—
—
87
a The specimen number used in Table 1 is given in parentheses.
b Estimates of the weight contents of the phases are tentative
because of the presence of an amorphous impurity in the speciꢀ
mens (A is the cation in fulleride).
c According to the published data.24
d Based on the results of the present study.
also observed in spite of an increase in the a parameter,
which is in contrast with the assumption made earlier28,29
based on the linear dependence of Tc on a.

1690
Russ.Chem.Bull., Int.Ed., Vol. 53, No. 8, August, 2004
Bulychev et al.
The most substantial changes in the occupancies of
the positions were observed in the specimen 11. The ocꢀ
cupancy of the 8c position increased to 1.26 (calculated
for the K atoms) and the intensity of the 200 reflection
increased by a factor of more than two, the unit cell
parameter being the same and the occupancy of the
4b position being virtually unchanged. This result can be
attributable to the insertion of Yb atoms, which are subꢀ
stantially smaller than the K atoms, into the posiꢀ
tion 1/4; 1/4; 1/4. In spite of the fact that the speciꢀ
mens 11 and 12 were synthesized under similar condiꢀ
tions, the Xꢀray diffraction pattern of the latter was subꢀ
stantially different. The parameter of the cubic unit cell
given in Table 2 is approximate and was calculated from
the interplanar distances of two reflections. Most likely,
the lattice of the fulleride 12 undergoes orthorhombic
distortion (one parameter is 14.23 Å). This assumption is
supported by the difference in the halfꢀwidths of the reꢀ
flections and the appearance of a large number of addiꢀ
tional reflections compared to those observed for the cuꢀ
bic phase.
a
b
c
d
The unit cell parameter of the specimen 20 is again
close to that of K2.41C60. The distribution of the K atoms
over the octahedral and tetrahedral sites is virtually idenꢀ
tical to that published in the literature, although the superꢀ
conducting transition temperature for this specimen is
maximum in the series of the compounds under considerꢀ
ation and is higher than that determined for K3C60.
Therefore, analysis of the occupancies of the K atom
positions in the fullerides synthesized in the present study
demonstrates that rareꢀearth metals can be involved in
these compounds (most probably, in the octahedral sites).
This process is accompanied by redistribution of the catꢀ
ions in the voids of the lattice and, as a consequence, by
changes in the unit cell parameters and, in some cases, by
lowering of the symmetry of the starting cubic unit cell.
However, the available data did not allow us to estimate
the precise amount of the heteroatoms of rareꢀearth metꢀ
als in the fulleride lattice. These data also leave unexꢀ
plained the differences between, on the one hand, the
mechanisms of the reactions (1), (2), and (5) involving
the Yb, Lu, and Sc atoms and, on the other hand, the
mechanisms of the analogous reactions performed under
similar conditions but with the participation of other rareꢀ
earth chlorides. The role of the inserted atoms in the
changes in the electrophysical properties of heteroꢀ
fullerides remains unclear.
12
20
28
36
2θ/deg
Fig. 3. Xꢀray diffraction patterns of pure K3C60 (a) and speciꢀ
mens 15 (b), 11 (c), and 12 (d).
The partial replacement of the K atoms, which occupy
the tetrahedral and octahedral voids of the lattice of the
fulleride structure, with rareꢀearth metals should be acꢀ
companied by redistribution of the intensities of reflecꢀ
tions (in particular, for pairs of the reflections 111, 200
and 311, 222). However, it should be noted that the reꢀ
sults of refinement of the occupancies of atomic positions
based on powder Xꢀray diffraction data cannot provide
unambiguous evidence of whether rareꢀearth metals are
involved or not in the heterofulleride structure.
The results of calculations of the occupancies of the 8c
and 4b positions by metal atoms and the unit cell paramꢀ
eters of fullerides with composition K2M´C60 are given
in Table 2. The occupancies of these positions for most
of the heterofullerides under study and for the aboveꢀ
considered K3C60 compound are lower than those exꢀ
pected for the stoichiometric compositions. At the same
time, the parameter of the cubic unit cell of the heteroꢀ
metallic specimen 15 is even more increased compared
to the value published in the study24 for tripotassium
fulleride. This fact can be attributable to an increase in
the occupancy of the tetrahedral site and, as a conseꢀ
quence, to an increase in the size of the tetrahedral void.
However, the type of atoms inserted into this void is
not clear.
The main and evident difference between these metal
atoms and the atoms of other rareꢀearth metals used in
the present study is their electronic structure. The highest
superconducting transition temperatures were observed
for K2LuC60, in which the Lu atom has the 4f145d16s2
configuration (Tc = 20 K, r(Lu3+) = 0.84 Å), and
for isoelectronic K2Yb2+C60 (average Tc ≈ 16 K,
r(Yb2+) = 1.05 Å), i.e., for fullerides in which the comꢀ

Fullerides: heterometallic superconductors
Russ.Chem.Bull., Int.Ed., Vol. 53, No. 8, August, 2004
1691
pletely filled 4f sublevel of the lanthanide atoms remains
unchanged.
All the compounds formed in this case are dielectrics.
Hence it follows that the presence of a stronger base (comꢀ
pared to THF) in a solution can change either the mechaꢀ
nism of the exchange reaction or the structure of fulleride
through solvation of the metal ion, as exemplified by
(THF)K3C60.30 Therefore, the smaller radii of the multiꢀ
charged ions compared to those of the K+ and Rb+ ions
are, apparently, partially compensated by the transfer of
additional electrons to the fullerene molecule. However,
both the appearance of the superconducting properties of
heterofullerides and the value of Tc depend on the nature
of these heteroatoms.
Although unexpected, the behavior of scandium in the
series of the reagents under consideration is explainable.
Many structural and chemical properties of scandium
compounds are close to those of lutetium compounds.
Some of these compounds are isomorphous. The apparꢀ
ent difference between the electronic structure of the Sc3+
ion and the structures of the Y3+ and La3+ ions is associꢀ
ated with the fact that the Sc3+ ion has neither valence d
electrons nor the lowꢀlying completely filled d10 shell. In
the Yb2+ and Lu3+ atoms, the latter is shielded by the
compact f shell and, apparently, influences only slightly
the electrophysical properties of heterofullerides. If this is
the case, one would not expect all heterofullerides involvꢀ
ing metals with the completely filled d shell, i.e., all postꢀ
transition, late transition, and mainꢀgroup metals, to exꢀ
hibit noticeable superconducting properties. Predictions
as to metals with the unfilled 3d shell and early mainꢀ
group metals having no d electrons are not so unamꢀ
biguous.
In the case of the reaction of K5C60 with Yb3+Cl3
(4f145d06s2, r(Yb3+) = 0.85 Å), the possible involvement
of only one f electron in the bond formation, most likely,
changes the reaction mechanism and leads to the formaꢀ
tion of partially (or weakly) substituted K3C60 and Xꢀray
amorphous ytterbium fulleride. A further decrease in the
number of f electron causes even more substantial changes
in the composition of the reaction products. The reacꢀ
tions of all other lanthanide salts listed in Table 1 with
K5C60 afforded several compounds, but none of these
compounds crystallized in the fcc lattice with the paramꢀ
eter close to that of K3C60. Therefore, these reactions
were accompanied by even more substantial changes in
the pathways of the reactions (1) and (2) compared to
those observed in the reaction of K5C60 with Yb3+Cl3.
Heterofullerides involving the yttrium and lanthanum
atoms, which were prepared according to the aboveꢀdeꢀ
scribed procedure and have valence orbitals analogous to
those of the Lu atom (nd1ms2) but the unfilled f levels, do
not possess superconductivity. It is unlikely that this difꢀ
ference is associated exclusively with the difference in the
sizes of these metal ions, because the radius of the Y3+ ion
(0.93 Å) differs only slightly from those of the Yb and Lu
ions, and the radius of the La3+ ion (1.15 Å) is maximum
close (in the series of lanthanides) to that of the K+ ion
(1.33 Å). Potassium chloride was detected in the precipiꢀ
tates remained after the synthesis. Reduction of halides
and the reaction of these metals with a carbon matrix are
beyond doubt. However, as mentioned above, the mechaꢀ
nism of this reaction can differ from that of intercalation,
and the differences are, presumably, determined by higher
Lewis acidity of yttrium and lanthanum halides compared
to that of ytterbium and lutetium halides. Judging from
the shape and position of the resonance band in the
13C NMR spectra, virtually all compounds involving lanꢀ
thanum are polymeric (see Table 1), i.e., the formation of
compounds with La—C covalent bonds is, apparently,
more typical of this element. At the same time, the posiꢀ
tion and shape of the band in the 13C NMR spectrum of
the product of the reaction between K5C60 and YCl3 are
identical to those in the spectrum of the starting pentaꢀ
potassium fulleride, i.e., this compound, unlike the La
derivative, is not polymeric, but it also does not possess
superconducting properties.
In conclusion, it should be emphasized that the comꢀ
positions of fullerides found in this study are tentative
because they are based exclusively on the stoichiometry of
the reactions, which could afford these compounds. Hence
it also follows that all the aboveꢀconsidered results and
conclusions refer only to compounds prepared by exꢀ
change reactions. One would expect that other properties,
including superconducting properties, will be imparted to
these compounds by varying the procedure for the synꢀ
thesis and taking into account the possibility of the existꢀ
ence of a wide homogeneity region for heterofullerides.
This study was financially supported by the Russian
Foundation for Basic Research (Project No. 02ꢀ03ꢀ
32575).
By contrast, the reaction of K5C60 with anhydrous and
unsolvated ScCl3, which is characterized by the smallest
size of the metal ion (r(Sc3+) = 0.81 Å) in the series of
Group IIIB metals, afforded a compound with tentative
composition K2ScC60 and Tc = 14 K. The analogous reꢀ
action of K5C60 with ScCl3•3MeCN was accompanied by
a partial change in the composition of the fulleride, as
evidenced by the fact that the 13C NMR spectrum has a
weak band at δ 145, which can be assigned to free fullerite.
References
1. A. F. Hebard, M. J. Rosseinsky, R. C. Haddon, D. W.
Murphy, S. H. Glarum, T. T. M. Palistra, A. P. Ramirez,
and A. R. Kartran, Nature, 1991, 350, 600.
2. M. J. Rosseinsky, A. P. Ramirez, S. H. Glarum, D. W.
Murphy, R. C. Haddon, A. F. Hebard, T. M. Palistra, A. R.
Kartran, S. M. Zahurak, and A. W. Makhija, Phys. Rev.
Lett., 1991, 66, 2830.

1692
Russ.Chem.Bull., Int.Ed., Vol. 53, No. 8, August, 2004
Bulychev et al.
3. K. Holczer, O. Klein, S.ꢀM. Huang, R. B. Kaner, K.ꢀJ. Fu,
R. L. Whetten, and F. Diederich, Science, 1991, 252, 1154.
4. K. Tanigaki, I. Mirosawa, T. W. Ebbesen, J. Mizuki,
Y. Shimakawa, Y. Kubo, J. S. Tsai, and S. Kuroshima,
Nature, 1992, 356, 419.
19. V. Buntar, Physica C, 1998, 309, 98.
20. Handbuch der preparativen anorganischen Chemie, Herausg.
G. Brauer, Studgardt, 1975.
21. A. L. Waida, S. Cheng, and I. Mukerji, J. Organomet. Chem.,
1987, 330, C17.
5. T. T. M. Palistra, O. Zhou, Y. Iwasa, P. E. Sulewski, R. M.
Fleming, and B. R. Zegarski, Solid State Commun., 1995,
93, 327.
22. B. M. Bulychev, V. I. Privalov, and A. A. Dityat´ev,
Zh. Neorg. Khim., 2000, 45, 103 [Russ. J. Inorg. Chem., 2000,
45, 931 (Engl. Transl.)].
6. S. C. Erwin and M. R. Pederson, Phys. Rev. B, 1993,
47, 14657.
7. S. Saito and A. Oshiyama, Phys. Chem. Lett., 1993, 71, 121.
8. R. S. Ruoff, Y. Wang, and D. Tománek, Chem. Phys. Lett.,
1993, 203, 438.
9. A. R. Kortan, N. Kopylov, S. Glarum, E. M. Gyorgy, A. P.
Ramirez, R. M. Fleming, F. A. Thiel, and R. C. Haddon,
Nature, 1992, 355, 529.
10. A. R. Kortan, N. Kopylov, E. Özdas, A. P. Ramirez, R. M.
Fleming, and R. C. Haddon, Chem. Phys. Lett., 1994,
223, 501.
11. A. R. Kortan, N. Kopylov, S. Glarum, A. P. Ramirez, R. M.
Fleming, O. Zhou, F. A. Thiel, P. L. Trevor, and R. C.
Haddon, Nature, 1992, 360, 566.
12. Y. Iwasa, H. Hayashi, T. Furudate, and T. Mitani, Phys.
Rev. B, 1996, 54, 14960.
13. A. R. Kortan, N. Kopylov, and E. Özdas, Recent Adv. Chem.
Phys. Fuller. Relat. Mater., 1997, 3, 423.
14. E. Özdas, A. R. Kortan, N. Kopylov, A. P. Ramirez,
T. Siegrist, K. M. Rabe, H. E. Bair, S. Schupples, and P. H.
Citrin, Nature, 1995, 375, 126.
23. A. C. Larson and R. B. Von Dreele, General Structure Analyꢀ
sis System (GSAS), Los Alamos National Laboratory, Report
LAUR, 2000, 86ꢀ748.
24. P. W. Stephens, L. Mihaly, P. L. Lee, R. L. Whetten, S.ꢀM.
Huang, R. Kaner, F. Deiderich, and K. Holczer, Nature,
1991, 351, 632.
25. T. Pradeep, G. U. Kulkarni, K. P. Kannan, N. Y.
Vasanthacharya, T. N. Guru Row, and C. N. R. Rao, Ind.
J. Chem., Sect. A, 1992, 31, P17.
26. B. P. Tarasov, Yu. M. Shul´ga, V. N. Vokin, V. N. Vasilets,
N. Yu. Shul´ga, D. V. Schur, and V. A. Yartys, J. Alloys
Comp., 2001, 314, 296.
27. P. Dahlke, M. S. Denning, P. F. Henry, and M. J.
Rosseinsky, J. Am. Chem. Soc., 2000, 122, 12352.
28. Fullerene. Chemistry, Physics, and Technology, Eds K. M.
Kadish and R. S. Ruoff, Wiley—Interscience, New
York, 2000.
29. K. Tanigaki, I. Hirosawa, Y. Shimakawa, T. W. Ebbesen,
J. Mizuki, Y. Kubo, J. S. Tsai, and S. Kuroshima, Synth.
Metals, 1993, 55—57, 3240.
30. C. Janiak, S. Mühle, R. Hemling, and K. Köhler, Polyꢀ
hedron, 1996, 15, 1559.
15. X. H. Chen and G. Roth, Phys. Rev. B, 1995, 52, 15534.
16. Y. Ksary and G. Chouteau, Full. Sci. Technol., 1996, 4, 1227.
17. M. Baenitz, M. Heinze, K. Lüders, H. Werner, and
R. Schlögl, Solid State Commun., 1995, 96, 539.
18. B. Gopdia, K. Kordatos, H. Suematsu, K. Tanigaki, and
K. Prassides, Phys. Rev. B, 1998, 58, 1077.
Received December 15, 2003;
in revised form March 12, 2004