Purity effects on the charge-transport properties in one-dimensional TPP[CoIII(Pc)(CN)2]2 (TPP = tetraphenylphosphonium and Pc = phthalocyaninato) conductors

Article abstract of DOI:10.1246/bcsj.82.692

Purity effects on the temperature dependence of electrical resistivity in TPP[Co(Pc)(CN)2]2 with a typical one-dimensional metallic electronic system have been examined. In this system, no dependence on the impurity content has been observed, suggesting that the apparent thermally activated behavior is dominated by charge disproportionation.

Full text of DOI:10.1246/bcsj.82.692

692
Bull. Chem. Soc. Jpn. Vol. 82, No. 6, 692–694 (2009)
Short Articles
and Br) is utilized as a component.8 In this case, only the
periphery of the Pc ring overlaps with the neighboring ring.
A characteristic of this system is that the ³-³ stacking
network varies by the cationic species. When TPP is utilized
as the cationic component, a 1-D partially oxidized salt of
TPP[M(Pc)L2]2 is obtained.11 The starting mono-valent anion
Purity Effects on the Charge-
Transport Properties in One-
Dimensional TPP[CoIII(Pc)(CN)2]2
(TPP = Tetraphenylphosphonium
and Pc = Phthalocyaninato)
Conductors
¹
of [M(Pc)L2] is oxidized to [M(Pc)L2]1/2¹ in this salt. In the
crystal, this unit forms a 1-D ³-³ stacking network. The Pc
planes in the neighboring 1-D networks are nearly perpendic-
ular to each other so that there is negligible electronic
interaction between them; the anisotropy of the conductivity
is about 103 for M = Co and L = CN.11 Since the ³-³ stacking
is uniform, this crystal has a metallic 3/4-filled conduction
band. However, the conductivity shows thermally activated
behavior below room temperature. On the other hand, the
thermoelectric power is metallic above 100 K, and the band
width estimated from the temperature dependence is about
0.5 eV, which is almost one-half of that in [Ni(Pc)]I.
Satoshi Yamashita, Toshio Naito, and
Tamotsu Inabe*
Department of Chemistry, Graduate School of Science,
Hokkaido University, Sapporo 060-0810
In order to investigate whether this thermally activated
behavior is due to lattice distortion, precise X-ray diffraction
experiments on TPP[Co(Pc)(CN)2]2 were performed at 21 K.
Neither super-lattice diffraction nor diffuse streaks were
observed. On the other hand, 59Co NQR measurements detected
an asymmetric signal at low temperature, suggesting charge
disproportionation in the crystal.12
This interpretation of the ground state is reasonable since the
electronic system is highly 1-D. However, since there still
remains a possibility that impurities affect the charge transport
as found in [Ni(Pc)]I, it is vital to examine the impurity effects
on the transport properties in TPP[Co(Pc)(CN)2]2. In this
study, we carried out preparation of TPP[Co(Pc)(CN)2]2 single
crystals from starting material with different impurity levels
and measurement of the transport properties.
The purification methods are shown in Scheme 1, and the
amount of paramagnetic impurity detected by ESR measure-
ments in the 1:1 salt of TPP[Co(Pc)(CN)2] (starting material for
the electrolysis) for each batch is shown in Table 1. Since CoIII
in the Pc unit is in the low-spin state, the material should be
diamagnetic. Therefore, the paramagnetic species detected in
the ESR measurements correspond wholly to impurities; the
g-factor was always nearly 2.
Purification efficiency in each process is summarized as
follows. SCo (sublimation of [Co(Pc)]) reduces the amount of
impurities effectively, but repetition (represented by superscript
n) is not effective. RK (recrystallization of the potassium salt)
reduces the amount of impurities, while RTPP increases the
impurity level. Since it is not likely that CoIII is further
oxidized during recrystallization, the ³-ligand is thought to be
chemically decomposed to form a free radical.
Received December 3, 2008; E-mail: inabe@sci.hokudai.ac.jp
Purity effects on the temperature dependence of
electrical resistivity in TPP[Co(Pc)(CN)2]2 with a typical
one-dimensional metallic electronic system have been
examined. In this system, no dependence on the impurity
content has been observed, suggesting that the apparent
thermally activated behavior is dominated by charge
disproportionation.
There are several ground states of one-dimensional (1-D)
electronic systems. The Peierls-type insulating state with
lattice distortion was once thought to be the most typical
ground state. However, charge disproportionation (charge
ordering) due to Coulombic interaction without distinct lattice
distortion has recently been recognized as a ground state
of quarter-filled electronic systems.1-7 The phthalocyaninato
(Pc) system in this study is also known as a highly 1-D
conductor.
There are two types of Pc conductors. One is [M(Pc)]Xy,
¹
¹
¹
M = Ni, Cu, H2, etc., X = I3 , BF4 , PF6 , etc., and 0.33 <
y < 0.5, in which planar [M(Pc)] units form a 1-D column with
metal-over-metal stacking.8 The crystal system is tetragonal,
and the 1-D column is surrounded by anions. The electronic
interactions between the columns are rather weak, and the
electronic system is highly 1-D; the anisotropy of the
conduction has been reported to be 500. The conductivity
was reported to diminish at low temperature in an early report.9
¹
However, in [Ni(Pc)]I (the charge is [Ni(Pc)]1/3+(I3
)
and
1/3
the Pc ring is partially oxidized), metallic conductivity was
observed down to 5 K when the samples were prepared from
high purity [Ni(Pc)].10 The conductivity did not decrease so
Commercially available [M(Pc)]s are known to contain
impurity metals. In order to exclude the effect of the impurity
metals, high purity [Co(Pc)] was synthesized from purified
[H2(Pc)] and Co2+ with 99.999% purity. TPP[Co(Pc)(CN)2]
was then synthesized through route (b) in Scheme 1, which was
considered to be the route to yield the purest sample. The purity
of TPP[Co(Pc)(CN)2] thus obtained was however, not so much
improved. From this result, it is strongly suggested that the
paramagnetic species detected by ESR are mainly chemically
decomposed ligand. Although some foreign metal ions may be
¹1
much at lower temperature; the high conductivity of 104 S cm
was maintained even at 1.85 K. The ground state in [Ni(Pc)]I is
still puzzling, but this study indicates that the temperature
dependence of the conductivity of the Pc system is sensitively
affected by impurities.
Another ³-³ stacking structure is obtained when the axially
ligated Pc unit ([M(Pc)L2], M = Co and Fe, and L = CN, Cl,
Published on the web June 12, 2009; doi:10.1246/bcsj.82.692

S. Yamashita et al.
Bull. Chem. Soc. Jpn. Vol. 82, No. 6 (2009)
693
10000
(a)
commercial
Co(Pc)
A
B
K[Co(Pc)(CN)2]
TPP[Co(Pc)(CN)2]
(II) (400 ppm)
0
[RK]0[RTPP
]
(I)
1000
100
10
1
(100 ppm)
(I)
(V)
RK
RTPP
(30 ppm)
1
[RK]0[RTPP
]
(II)
(60 ppm)
(VIII)
S
B
[RK]1[RTPP]0 (III)
[RK]1
RTPP
1
[RK]1[RTPP
]
(IV)
A
B
n
[SCo]n[RK]0[RTPP]0 (V)~(VII)
[SCo
]
n (number of repetition) = 1~3
Purification method (represented by
S: vacuum sublimation at ~400 °C
)
1
RK: recrystallization from acetone
RTPP: recrystallization from acetonitrile
1
(b)
commercial
H2(Pc)
purified
H2(Pc)
S
C
B
1
Co(Pc)
0
50
100 150 200 250
300
S
(VIII)
T / K
RK
A
[SH ]1[SCo]1[RK]1[RTPP
]
0
2
Figure 1. Temperature dependence of the normalized
resistivity of TPP[Co(Pc)(CN)2]2 prepared from starting
material with varying purity. The vertical axis for each
sample is shifted for clarity.
Chemical reaction (represented by
)
A: reflux in EtOH with KCN, and recrystallization from acetone
.
B: metathesis with TPP I in acetonitrile
C: reflux with CoCl2 (99.999% purity)
in quinoline (10%)/1-chloronaphthalene
¹2
within «50% of 10 ³ cm. The values have no correlation
Scheme 1.
to the impurity level, and disperse even in the same batch.
The results are in contrast to the case of [Ni(Pc)]I. For
[Ni(Pc)]I, the paramagnetic impurity in starting [Ni(Pc)]
seriously affected the temperature dependence of the conduc-
tivity, and high conductivity at low temperature was not
achieved when the purity level was not high enough.10 In other
words, the apparent thermally activated behavior of the
conductivity at low temperature in this 1-D system was
dominated by the scattering of carriers by impurities. On the
other hand, the 1-D conductor based on the [Co(Pc)(CN)2] unit
always shows thermally activated behavior of the conductivity
at low temperature irrespective of the impurity level. This fact
strongly suggests that there is another mechanism that induces
charge localization in this system. Even if charge scattering by
impurities occurs, the effect could not be seen because another
effect that localizes the carriers dominates the charge transport.
Such charge localization may occur by band gap formation
at the Fermi level due to 2kF or 4kF lattice distortion. However,
this scenario is contradicted by the temperature dependence of
the ESR signal, magnetic susceptibility, and low-temperature
X-ray diffraction.12 Consequently, charge disproportionation
driven by the electron-electron correlation effect is thought to
be most probable.
The temperature dependence of the paramagnetic suscepti-
bility of TPP[Co(Pc)(CN)2]2 prepared from two different
starting materials ((V) (30 ppm) and (I) (100 ppm)) is shown
in Figure 2. Both samples show similar susceptibility; the main
component is temperature independent Pauli-like paramagnet-
ism. It should be noted that the Curie tailing at low temperature
is almost 1% (assuming g = 2 and S = 1/2) in both samples. If
this component corresponded to paramagnetic impurities,
impurity concentration would be increased by more than 100
times. Since the electrolysis was performed under mild
Table 1. Impurity Level in TPP[Co(Pc)(CN)2] Purified
Following Scheme 1
Impurity level
/ppm
Impurity level
/ppm
Sample
Sample
I
II
III
IV
100
400
50
V
VI
VII
VIII
30
60
40
60
250
ESR silent, their presence must be much lower in the sample
prepared from Co2+ with 99.999% purity.
TPP[Co(Pc)(CN)2]2 single crystals were prepared by electro-
chemical oxidation of the above-mentioned TPP[Co(Pc)(CN)2]
samples. The impurity levels may be reduced during electro-
crystallization, since the process may eliminate foreign species
that occurred in recrystallization. However, the free radical
species contained in the present system is rather readily
incorporated in the host lattice, since recrystallization of the
TPP salt increased the impurity levels. Therefore, the purity of
the product must reflect the purity of the starting material as
demonstrated in the [Ni(Pc)]I system.10
The temperature dependence of resistivity in some crystals
(normalized to the value at room temperature) is shown in
Figure 1. The difference between the highest (400 ppm) and
lowest (30 ppm) impurity levels is practically negligible. The
resistivity of the crystal obtained from the salt prepared from
Co2+ with 99.999% purity is nearly constant in the high-
temperature region, whereas it shows the same increase at low
temperature. The room-temperature values are distributed

694
Bull. Chem. Soc. Jpn. Vol. 82, No. 6 (2009)
© 2009 The Chemical Society of Japan
TPP[Co(Pc)(CN)2] in the route shown in Scheme 1b was prepared
from [Co(Pc)] obtained by the reaction of sublimed commercial
[H2(Pc)] and CoCl2 with 99.999% purity. Paramagnetic impurity
levels were estimated from the ESR signal intensity assuming that
the paramagnetic spins obey Curie-like behavior with S = 1/2.
Measurements. ESR spectra were obtained with a Bruker
EMX spectrometer at room temperature. TPP[Co(Pc)(CN)2]2 was
used as a standard for spin concentration. Electrical resistivity was
measured along the needle axis of the crystal by a standard four-
probe method in the temperature range of 5-300 K. Magnetic
susceptibility was measured with a Quantum Design MPMS
SQUID susceptometer for randomly oriented polycrystalline
samples.
2.0
(I) (100 ppm)
(30 ppm)
(V)
1.5
1.0
0.5
0.0
The authors are grateful to Prof. N. Hanasaki (Okayama
University) for valuable discussion and to Profs. R. Wakeshima
and Y. Hinatsu for their help in susceptibility measurements.
This work was supported in part by Grants-in-Aid for Scientific
Research from the Ministry of Education, Culture, Sports,
Science and Technology, Government of Japan and Grants-in-
Aid for Scientific Research from the Japan Society for the
Promotion of Science.
0
20
40
60
80
100
T / K
Figure 2. Temperature dependence of the paramagnetic
susceptibility of TPP[Co(Pc)(CN)2]2. The solid lines
correspond to the fitting based on the [Pauli paramag-
netism + Curie spin] model. The concentration of the
paramagnetic Curie spin (S = 1/2) is 0.9% for (V) and is
1.1% for (I).
References
1
2
3
H. Seo, H. Fukuyama, J. Phys. Soc. Jpn. 1997, 66, 1249.
K. Hiraki, K. Kanoda, Phys. Rev. Lett. 1998, 80, 4737.
T. Yamamoto, H. Tajima, R. Kato, M. Uruichi, K. Yakushi,
J. Phys. Soc. Jpn. 2002, 71, 1956.
K. Yamamoto, T. Yamamoto, K. Yakushi, C. Pecile, M.
Meneghetti, Phys. Rev. B 2005, 71, 045118.
D. S. Chow, E. Zamborszky, B. Alavi, D. J. Tantillo, A.
Baur, C. A. Merlic, S. E. Brown, Phys. Rev. Lett. 2000, 85, 1698.
F. Zamborszky, W. Yu, W. Raas, S. E. Brown, B. Alavi,
C. A. Merlic, A. Baur, Phys. Rev. B 2002, 66, 081103.
4
oxidation conditions, it is not likely that the paramagnetic
impurity increases during crystal growth. Therefore, the
relatively large Curie component is considered to be due to
localized spins originating from charge disproportionation in
microscopic domains.
In conclusion, we have studied how purity affects the
transport behavior in 1-D metal TPP[Co(Pc)(CN)2]2 which
shows thermally activated conductivity, and found that para-
magnetic impurity has little effect on the transport behavior.
These results strongly suggest that the temperature dependence
of the conductivity is dominated by charge disproportionation,
agreeing with the conclusion drawn from the asymmetric 59Co
NQR signal in TPP[Co(Pc)(CN)2]2.12
5
6
7
8
9
S. Fujiyama, T. Nakamura, Phys. Rev. B 2004, 70, 045102.
T. Inabe, H. Tajima, Chem. Rev. 2004, 104, 5503.
C. J. Schramm, R. P. Scaringe, D. R. Stojakovic, B. M.
Hoffman, J. A. Ibers, T. J. Marks, J. Am. Chem. Soc. 1980, 102,
6702.
10 J. A. Thompson, K. Murata, D. C. Miller, J. L. Stanton,
W. E. Broderick, B. M. Hoffman, J. A. Ibers, Inorg. Chem. 1993,
32, 3546.
11 H. Hasegawa, T. Naito, T. Inabe, T. Akutagawa, T.
Nakamura, J. Mater. Chem. 1998, 8, 1567.
Experimental
Purification Methods. Commercial [Co(Pc)] was used in the
preparation and purification routes in Scheme 1a. Seven kinds of
starting materials for electrolysis, TPP[Co(Pc)(CN)2], were ob-
tained by combining vacuum sublimation and recrystallization.
12 N. Hanasaki, K. Masuda, K. Kodama, M. Matsuda, H.
Tajima, J. Yamazaki, M. Takigawa, J. Yamaura, E. Ohmichi, T.
Osada, T. Naito, T. Inabe, J. Phys. Soc. Jpn. 2006, 75, 104713.