Electrons 1D

12
Interacting electrons in one

dimension: the two chain
perylene-metal dithiolate series
Luís Alcácer1 , Rui T. Henriques1 and Manuel Almeida2

1
Instituto de Telecomunicações, Instituto Superior Técnico,
Universidade de Lisboa, Portugal
2
Centro de Ciências e Tecnologias Nucleares,
Campus Tecnológico e Nuclear, Pólo de Loures do IST,
Universidade de Lisboa, Portugal
Corresponding author contact: [email protected]
Abstract
In this review, we attempt to present the most notable aspects
of the story of the two chain, highly one-dimensional, organic con-
ductors of the perylene metal-dithiolate series, which have been
under investigation since the end of the 1960’s until the present
time, with the recent discovery of a superconducting phase. We
will start with an overview of the developments of the research on
these materials and then, after a brief introduction and a survey
of the theoretical background, we will describe and discuss the
structure, transport and magnetic properties, as well as the low
temperature instabilities and phase transitions. Finally we will
focus on the study of their behaviour under extreme conditions of
low temperatures, high pressures and high magnetic fields up to
45 tesla.
304 Luís Alcácer, Rui T. Henriques and Manuel Almeida
Overview
”One-dimensional thought is systematically promoted by the
makers of politics and their purveyors of mass information.
Their universe of discourse is populated by self-validating
hypotheses which, incessantly and monopolistically repeated,
become hypnotic definitions of dictations.”
Herbert Marcuse, One-Dimensional Man, Beacon Press, 1964.
In 1964, William Little, from Stanford University, proposed a model

to synthesize a room temperature organic superconductor [1], which, a
few years later, led to a new field of research entitled one-dimensional
conductors. In those materials, the electrical conductivity is high along
one direction and much lower in the other directions. This is possible in
two cases: i) in solids in which flat organic molecules are stacked on top of
each other, and through intermolecular interactions, electrons can travel
along the staking axis, and ii) in organic polymers in which single carbon-
carbon bonds alternate with double bonds allowing electrons to travel
along the polymer chain. Nowadays, one can confine electrons in one
or two dimensions by means of top-down approaches to nanotechnology,
but that is not the subject of the present paper. The present paper deals
with organic crystalline solids, with extreme anisotropy, in which the one-
dimensional character of the interactions of the electrons with each other
and with the lattice vibrations (the phonons) are the source of various
types of competing instabilities and phase transitions. Take, for example,
a linear chain of N sites, separated from each other by a distance a, the
unit cell size, and with one electron per site, or unit cell. Since each site
can accommodate two electrons, with opposite spins, the system would
have 2N states available, but only N would be occupied. One would
have, therefore, a half-filled band metal. However, the total energy of the
electrons would decrease if the sites would arrange themselves in pairs,
doubling the unit cell, which corresponds to opening a gap at the Fermi
level, the last occupied energy level. In one-dimension, below a certain
critical temperature, the energy needed to distort the lattice, doubling
the unit cell, becomes lower than the gain in energy of the electrons, and
a phase transition may occur. This is the Peierls transition, one of many
that may occur.
A good example of a one-dimensional solid is the one, best known
as (TMTSF)2 PF6 , which was the first organic superconductor to be dis-
covered, in 1979, and has been a centre of attention since then. In the
words of Paul Chaikin, a well known physicist, now at the New York
University, this is ”the most remarkable electronic material ever discov-
Interacting electrons in one dimension 305
ered. In one single crystal, one can observe all the usual competitions,
all electron transport mechanisms known to man, plus somethings new”.
In an instance of lucky serendipity, one of the present authors (LA),
synhtesized the first members of a family of organic conductors based
on the perylene molecule and metal-dithiolate complexes [2] (see Fig-
ure 12.1). At the time, it was not possible to prove that any of these
materials was a metal and much less a superconductor, in spite of the
many evenings spent cooling them down to liquid helium temperature.
They always turned insulators! The first papers with a full description of
their electrical and magnetic properties appeared only in 1974 and 1976
[3, 4] as the outcome of more detailed studies. This family of materials
of general formula (Per)2 [M(mnt)2 ], with M = Pt, Au, etc., became of
considerable interest later on for exhibiting metallic behaviour and very
high one-dimensionality. The crystal structures are formed by chains
of [M(mnt)2 ] units surrounded by perylene chains, which are 3/4 filled
band metals and therefore highly conducting at room temperature. On
the other hand, the [M(mnt)2 ] chains are insulating, with localized spins
in the case of M=Ni and Pt, (S=1/2), for example, spins that exchange
fast, with the itenerant electrons on the perylene chains. For more than
40 years, now, these materials have also been amazing us, with their
remarkable properties, some of them typical of the extreme anysotropy,
but mostly not expected as they came. The nature of the metal-insulator
transition in the Pt compound, for example, is particularly compelling.
At the temperature of 8 kelvin, the perylene chain tetramerizes, develop-
ing a charge density wave (CDW), with the opening of a gap. This lattice
distortion drives a distortion of the [M(mnt)2 ] chains, which dimerize,
leading to a spin-Peierls ground state (SP). The spin-Peierls state in the
[M(mnt)2 ] chains is coupled to the charge density wave state in the pery-
lene chain, up to magnetic fields of the order of 20 tesla, a situation not
foreseen in the present theory, which predicts a decoupling at much lower
fields. Under extreme conditions of low temperature, high pressure and
high magnetic fields, these materials exhibit a rich variety of phenom-
ena, including field induced charge density waves (FICDW), interference
effects of electrons from “trajectories” on different Fermi surface sheets,
and for our delight, superconductivy was finally observed, under pres-
sure, in 2009, evolving in a rather unusual way, from the charge density
wave state. This paper is about the wonderful materials of this family.
12.1 Introduction
On page 111, of his 1955 ”Quantum Theory of Solids”, [5], Rudolf Peierls
states that ”it is therefore likely that a one-dimensional model could
never have metallic properties”. This he concludes, after arguing that
in a one-dimensional metal with a partly filled band the regular chain
structure could not be stable, since one would always find a distortion
corresponding to a suitable number of atoms in the unit cell for which a
break would occur at or near the edge of the Fermi level [6].
In the early 1960’s, a few years after the success of the BCS theory of
superconductivity [7], there was the hope to learn how to produce ma-
terials which could be superconductors at room temperature. In 1964,
William Little’s seminal paper entitled ”Possibility of Synthesizing an Or-
ganic Superconductor” [1] stimulated the search for new materials which
could comply with his model system. Based on London’s idea [8] that
superconductivity might occur in organic macromolecules, he proposed
a mechanism similar to the BCS theory of superconductivity and even
suggested model molecules that could lead to room temperature super-
conductivity. His model structure was composed of a conjugated polymer
chain dressed with highly polarizable molecules as side groups. The poly-
mer chain would be a one-dimensional metal with a single mobile electron
per C-H unit; electrons on separate units would be Cooper-paired by in-
teracting with the exciton field on the polarizable side groups. However,
in spite of much effort, nobody succeeded in making Little’s molecule,
nor room temperature organic superconductors!
The first organic conductor, a perylene-bromine complex, had been
reported in 1954 by H. Akamatu, H. Inokuchi and Y. Matsunaga [9].
As many important discoveries, this was a case for serendipity. At the
time, nobody would believe that an organic substance could ever be a
conductor. Linus Pauling, who was visiting Japan at the time, was truly
surprised. In the late 1960’s and early 1970’s, several organic semicon-
ductors were reported and the subject became of much interest, espe-
cially due to William Little’s paper.
Before going into more detail on the (Per)2 [M(mnt)2 ] family, (see Fig-
ure 12.1), it should be mentioned that, in the meanwhile, several organic
conductors were synthesized, in particular TTF-TCNQ, a charge-transfer
salt of tetrathiafulvalene (TTF) and tetracyanoquinodimethane (TCNQ)
reported to exhibit superconducting fluctuations and Peierls instability
[10]. Superconductivity was not confirmed in this compound, but the
discussion it triggered led to considerable experimental and theoretical
developments, and the search for organic superconductivy was on its way.
At last, in 1979, the first organic superconductor was synthesized
NC S S CN
M
NC S S CN
M(mnt) 2
(mnt=maleonitriledithiolate)
M=Ni, Pt, Au, Co, Fe, Cu, Pd Perylene
Figure 12.1: The constituent molecules of the (Per)2 [M(mnt)2 ] charge

transfer salts (M = Ni, Pt, Au, Co, Fe, Cu, Pd).
by Klaus Bechgaard and found to have a transition temperature of

Tc = 1.1 K, under an external pressure of 6.5 kbar [11]. Such find-
ing gave rise to the dawn of a new era in the field of organic conductors
and superconductors. Several series of organic superconductors were
synthesized and the needed theory was developed.
Little’s idea was based on conjugated polymers, for which single
carbon-carbon bonds alternate with double bonds, and, in principle,
they should be metals or semiconductors, depending on the unit cell.
Polyacetylene was the paradigm compound and had been under investi-
gation for some time [12], but it was shown to become highly conducting
only through heavy doping [13]. That accidental discovery led to the
thorough study of conducting polymers, which were the subject of the
2000 Nobel Prise in Chemistry awarded to Alan Heeger, Alan MacDi-
armid and Hideki Shirakawa. A new technology emerged and is now the
important field of Organic Electronics.
12.2 The Interesting Physics: 1 D Instabilities

In a one-dimensional crystal, fluctuations destroy all long-range order,
and since there are always fluctuations at any finite temperature, such
a system cannot be ordered except at zero temperature. This is what
makes the 1D real systems interesting—they are not strictly 1D and they
exhibit many interesting kinds of fluctuations, instabilities and phase
transitions. This is what happens in solids with extreme anisotropy, as
in most organic conductors and superconductors, where the electron-
electron interactions mediated by phonons originate structural instabil-
ities and phase transitions, namely of the metal-insulator and supercon-
ducting type. The needed theoretical framework is identical to that of

superconductivity and all these critical phenomena benefit of a common
formalism. The anisotropy in the properties of these solids, such as the
electrical conductivity along the chain, which can attain values as high as
105 times the value in the perpendicular directions, is due to the highly
directional ⇡ ⇡ interactions along the stacks of flat molecules or along
the chains of carbon atoms in conjugated polymers.
One can consider several types of instabilities in a 1 D crystal, namely:
1. In the presence of electron-phonon interactions, the ground state
is unstable relative to the development of charge density waves
(CDW) with wave vector q twice the Fermi wave vector (q = 2kF ).
This is the Peierls instability and competes, in general, with the
BCS superconducting instability.
2. In the presence of magnetic interactions, the ground state is unsta-

ble relative to the development of spin density waves, (SDW), with-
out lattice modulation, and spin-Peierls instabilities, when there is
a lattice distortion.
3. In a 1 D system with short range interactions, the thermal fluctu-

ations destroy the long range order at any temperature T > 0.
4. In a 1 D system, an arbitrarily small disorder induces electron

localization and the consequent transition from a metal into an
insulator.
Figure 12.2 illustrates how, due to the electron-phonon interaction, a
regular linear chain of molecules with one electron per molecule, in a
metallic state, (half filled band system) undergoes a Peierls distortion to
a charge density wave (CDW) state, by doubling the size of the unit cell,
and becomes an insulator.
In Figure 12.2 a), a chain of molecules along the x-axis, equally spaced
by a lattice parameter a, with one electron per unit cell, is a metal at
high temperature with constant electron density, n. If there is one elec-
tron per unit cell in the high temperature state, the energy band is
half-filled and kF = ⇡/2a, as shown in Figure 12.2 b). At low tempe-
rature, the lattice parameter is modulated and tends to double due to
the electron-phonon interaction, and a charge density wave (CDW) is
generated (Figure 12.2 c)). Whenever the number of electrons per unit
cell is less than two, the band is partially filled and an identical situation
can occur. In general, if the degree of band filling is 1/n0 , kF = n10 ⇡a
and the stable charge density wave will have a wave length n0 a. How-
ever, we should take into account that, in a band more than half filled,
a 2a
a) x c) x
n n
(High T) Metallic state (Low T) CDW state
b) d)
E E
EF q=2kF
2Δ
k k
−π/a -kF 0 kF π/a −π/a -kF 0 kF π/a
Figure 12.2: a) 1 D lattice, electron density n, corresponding to one

electron per site, and b) electron dispersion on the regular chain with
a half filled band. c) and d) Peierls modulated in the electron-phonon
system. States with |k| . kF gain energy by the q = 2kF distortion
(lattice doubling), leading to the opening of an energy gap at kF .
holes should be considered instead of electrons, and the degree of fill-

ing would be 1 n0 . A 3/4 filled band would correspond to a 1/4 hole
filled band, and the charge density wave (of positive charge) would have
a wave length 4a, giving rise to a tetramerization (see below). Since in
one dimension the elastic energy needed to modulate the crystal lattice
is less than the energy gain in the conduction electrons, when the tempe-
rature is lowered, the opening of the gap, 2 , at kF , drives a structural
instability and eventually a structural phase transition.
12.2.1 The Peierls Instability

As mentioned before, Peierls [5] and Fröhlich [6] predicted that a 1 D
electron system in a deformable lattice is unstable relative to a modula-
tion of the lattice with wave vector q = 2kF , in which kF is the Fermi
wave vector. For an overview of the theoretical background we will follow
the excellent theoretical introduction of D. Jérôme and H. J. Schulz [14]
to the instabilities of quasi-one-dimensional electron systems originally
published in 1982 in Advances in Physics and which should be referred
to for more detail. A one-dimensional electron system can be described
by the following Hamiltonian, in second quantization,
X X g X +
H= "k a +
ks a ks + ! q b +
q b q + p ak+q,s aks bq + b+q (12.1)
ks q L kqs
where a+ks and aks are the creation and annihilation operators for an
electron in state k, with spin s, and energy "k ; b+
q and bq are the creation
and annihilation operators for a phonon with wave vector q and energy
!q ; L is the length of the system. For simplicity we make "kF = 0 and
~ = 1. The first two terms describe the non-interacting electron and
phonon systems, respectively, and the third term is the electron-phonon
interaction, with coupling constant g, usually a function of k and q. In
the present context, electrons with |k| ⇡ kF and phonons with |q| ⇡ 2kF
are the most important and, therefore, the wave vector dependence on g
can be ignored. The sums are limited to the Brillouin zone.
The lattice oscillations u can be written in terms of the phonon op-
erators, as
X 1
u(x) = p bq + b+q eiqx (12.2)
q
2L! q
A modulation of theplattice with wave vector q = 2kF is described by

hb2kF i = hb+2kF i / L, i.e., the phonon modes with wave vector ±kF
are occupied.
In the mean field approximation, one can approximately describe the
system by the Hamiltonian
X 2
H= "k a +
ks aks + 2!2kF |hb2kF i| +
ks
2g X ⇣ + ⌘ (12.3)
+p ak+2kF ,s aks hb2kF i + a+
k
+
2kF ,s aks hb2kF i
L ks
this meaning that the lattice modulation mixes the k states with the k ±
2kF states. More important is the mixing of the quasi-degenerate states
with |k| ⇡ kF that can be approximately treated in the free electron
model and by linearizing "k in the vicinity of the Fermi points, i.e.,
"k = vF (|k| kF ), vF being the Fermi velocity, giving
X !2kF
H= Ek c +
ks cks + L 2g 2 | |
2
(12.4)
ks
with h i1/2
Ek = sign (|k| kF ) vF2 (|k| kF ) 2 + | | 2 (12.5)
c+
ks and cks are now the creation and annihilation operators for electrons
in states which are linear combinations of the old states k e k ± 2kF , and
in which we define the order parameter
2 g hb2kF i
= p (12.6)
L
As it can be seen in Figure 12.2 d), the energy of the electrons is lowered
and the lattice modulation entails the opening of a gap of width 2| | at
the Fermi level, as shown from equation (12.5). At the same time, the
elastic energy increases due to the modulation of the lattice described
by the second term of equation (12.4).
The overall effect on the total energy of the electronic system can be
calculated to give
kF
X
Eel ( ) = 2 Ek
k= kF
 ✓ ◆
L n EF | |2 2 EF
= 1 + 2 ln + higher order terms
2 EF | |
(12.7)
In this equation, EF is the Fermi energy, measured relative to the bottom
of the band, and n is the electron density. From equation (12.7), one can
see that the energy gained by the electrons is, for small | |, proportional
to | |2 ln | |, which is always bigger than the loss in elastic energy
which is only proportional to | |2 . From this, one can conclude that,
in the mean field approximation, the system is unstable relative to a
modulation of the lattice, for an arbitrarily small coupling. To get the
value for the gap, one can minimize the energy and obtain
1/ 2 g2
| | = 2 EF e , = (12.8)
⇡ vF !2kF
where is the wave length of the lattice modulation. It should be noted
that the logarithmic term in (12.7) only appears if the gap opens exactly
at ±kF . In other words, the wave length of the lattice modulation is
determined by the electronic band filling and is equal to ⇡/kF . The
expectation value of the modulation in the ground state is
s✓ ◆
2 | |
hu(x)i = cos(2kF x + ) (12.9)
!2kF g
where is the phase of (i.e., | |ei ): there is a modulation of the elec-

tron density as a consequence of the lattice modulation, usually known
as a charge density wave, (CDW).
The excited states depend only on the amplitude of and not on
the phase. Due to the translational symmetry, the CDW can move along
the crystal, transporting a d.c. current, which ideally would be infinite,
but which is suppressed in real crystals due to several mechanisms such
as impurities and lattice locking.
The description of the Peierls ordered state for T = 0 is qualitatively
valid for temperatures near zero, while the long range order is not de-
stroyed. At sufficiently high temperatures, the complete destruction of
the order is expected, and the thermally excited electrons above the gap
have energies ⇡ | | and will gain energy if the gap decreases. A decrease
of | | enables the additional excitation through the gap, and the mech-
anism predicts the disappearance of the gap and of the long range order
above a certain temperature.
In these 1 D systems, a sharp dip in the phonon spectrum, at the
wave vector q = 2kF , known as the Kohn anomaly [15], is particularly
noticeable (Figure 12.3). It is due to the possibility of exciting all the
electrons from one side of the Fermi distribution into the other side with
the single wave vector and very little energy.
ωq
T3
T2
T1
Tc0
0 2kF q
Figure 12.3: Evolution of the Kohn anomaly, in the phonon spectrum,

with the temperature (T3 > T2 > T1 > Tc0 ).
The phonons with wave vector near q = 2kF involve displacements of

the lattice and are collective oscillations strongly coupled to the lattice,
as well as to the charge density. It is this coupling that is responsible for
the dip in the phonon spectrum. As the temperature is lowered, the dip
increases and, at the critical temperature Tc0 , one finally has !2kF = 0,
indicating the instability of the system relative to a modulation with
wave vector q = 2kF —the complete softening of the lattice and a static
distortion appears. This is one explanation for charge density waves in
solids. The wave vectors at which a Kohn anomaly is possible are the
nesting vectors of the Fermi surface, that is, vectors that connect most
of the points of the Fermi surface. For a one-dimensional chain of atoms,
this vector would be 2kF , connecting entirely the two flat Fermi Surfaces.
The description of the Peierls transition is very similar to the BCS
theory of superconductivity, namely:
• In both cases, the gap opens on the whole Fermi surface inducing
a rapid decrease in the energy of the electrons.
• In both cases, the instability is expressed by a correlation function

which diverges logarithmically.
• The transition temperature and the gap at T = 0 only differ by a

proportionality factor, following the equation 2| (T =0) |/Tc = 3.5.
• The effect of impurities in the Peierls transition is analogous to the

effect of magnetic impurities in a superconductor.
12.2.2 Competing Instabilities in a 1 D Electron Gas

In the framework of mean-field theory, the electron-electron interaction
near the Fermi level can be characterized by four coupling constants,
g1 , g2 , g3 and g4 , their values determining the most stable phases. g1
corresponds to the electron-electron interaction with momentum transfer
±2kF (backward scattering), g2 and g4 have small momentum transfer
(forward scattering) involving electrons at both kF and kF , (g2 ), or
only on one side, (g4 ). For g3 , the total electron momentum is changed
by 4kF , being only allowed if 4kF is a reciprocal lattice vector (umklapp
scattering), i.e. for a half-filled band.
Just to have a qualitative notion of the problem, one can, in the
mean-field approximation, look for the possibility of opening a gap at
the Fermi level, by evaluating the expectation values of some operators
which may originate stable ordered states. Let us then consider the
following operators
X
OCDW (q) = b+kF +k q,s akF +k,s (12.10)
ks
X
ss0
OSDW ↵ (q) = b+kF +k q,s ↵ akF +k,s0 (↵ = x, y, z) (12.11)
kss0
X
OSS (q) = b+ kF k, s akF +k+q,s (12.12)
ks
X
ss0
OT S↵ (q) = b+ kF k, s ↵ akF +k+q,s0 (↵ = x, y, z) (12.13)
kss0
where a and b are operators which describe electrons moving to the right
0
and to the left, respectively, and ↵ss are elements of the spin Pauli matri-
ces. The operators OCDW (q) and OSDW↵ (q) are the Fourier components
of the charge and spin densities, with wave vector 2kF + q, respectively.
The OSS (q) and the three possible OT S↵ (q) are the singlet and triplet
operators of the Cooper pairs of the superconducting state. To look
for ordered states, one would have to compute the expectation values
of these operators. If, for example, one considers the singlet supercon-
ducting state, the Hamiltonian leads to a solution which is similar to the
Peierls transition with a gap | S| in the Fermi surface which opens at

q = 0.
1
| SS |
= 2 EF e SS
g1 + g2 (12.14)
SS =
2⇡ vF
for g1 + g2 < 0. For g1 + g2 > 0, there is no self-consistent solution with
finite | |.
The transition temperature is given by
2C 1
Tc = EF e SS , (C = 1.781) (12.15)
⇡
The same argument can be applied to the other phase transitions, re-
placing SS by the respective parameters:
2g1 g2
CDW = (12.16)
2⇡ vF
g2
SDW ↵ = (12.17)
2⇡ vF
g1 g2
T S↵ = (12.18)
2⇡ vF
Figure 12.4 schematically represents the domains in the g1 g2 plane for
which the various O operators give stable phases.
g1
TS
SDW
(SS)
(CDW)
g2
SS CDW
Figure 12.4: Phase diagram in the g1 g2 plane showing the domains

for which the various operators give stable phases.
The more stable phase is the one which will have the highest transi-
tion temperature. Besides the CDW of Peierls type all other phases are
possible depending on the g1 , g2 , g3 , g4 coupling constants. The super-
conducting singlet (SS, CDW) and triplet (TS, SDW) occur for g1 < 0
and g1 > 0 respectively. The density wave and superconducting phases
are separated by the line g1 = 2g2 (Fig.12.4).
12.3 The Perylene-Metal Dithiolate Series

12.3.1 Crystal Structure, Energy Bands and Fermi Sur-
face
This family of charge-transfer salts appears in several phases with dif-
ferent crystal structures. There are also perylene-metal dithiolate salts
with other stoichiometries. In this paper, we will focus on the so called
↵-phases which are the most studied and, so far, the most interesting.
The ↵-(Per)2 [M(mnt)2 ], with M=Ni, Au, Cu Pt or Pd, form an
isostructural series, crystallizing in the monoclinic system (space group
P21 /c) [16, 17, 18]. The Fe and Co compounds are dimerized, the b
axis being doubled. The projection of the crystal structure of the plat-
inum compound along the b axis is illustrated in Fig.12.5. It shows that
Figure 12.5: Crystal Structure of the ↵-phase of the (Per)2 [Pt(mnt)2 ].

Projection along the b axis.
there are segregated regular stacks of [Pt(mnt)2 ] and perylene parallel to

b. Each column of [Pt(mnt)2 ] is surrounded by six columns of perylene
and each perylene stack has three stacks of perylene and three stacks
of anions as nearest neighbours. The dihedral angle between the two
mean planes of the molecules is 67 . Interchain coupling is weak, with
intermolecular distances only slightly smaller than the sum of the van
der Waals radii of the nitrogen and carbon atoms of the two molecules.
In the stack, adjacent overlapping [Pt(mnt)2 ] anions are separated by
3.65 Å, the distance between consecutive Pt atoms is 4.19 Å and the tilt
of the plane with respect to b is 29 . The perylene molecules present
a very good graphite-like mode of overlap at the distance of 3.32 Å,
shorter than in graphite (3.35 Å), the tilt of the plane with respect to
b being 37 . This mode of stacking allows very short distances between
carbon atoms (⇡ 3.3 Å) and therefore there is consistent evidence for
strong interactions. In table 12.1, the cell parameters for the different
(Per)2 [M(mnt)2 ] compounds are compaired [18, 19].
Table 12.1: Cell parameters for (Per)2 [M(mnt)2 ] [18, 19]. In the Co and
Fe salts, the b axis is doubled, when compared with the others [19].
M Au Pt Pd Ni Fe Co
a (Å) 16.602 16.612 16.469 17.44 17.600 17.75
b (Å) 4.191 4.194 4.189 4.176 8.136 8.22
c (Å) 30.164 30.211 30.057 25.18 30.088 30.88
( ) 118.69 118.70 118.01 91.57 123.72 123.0
For the ↵-phases of (Per)2 [M(mnt)2 ], the energy band in the perylene
chain is quarter filled and the band in the [M(mnt)2 ] chain is half filled.
The perylene chain is metallic at room temperature and undergoes a
metal-insulator transition at low temperature. The [M(mnt)2 ] chain, for
M=Pt, Ni, Pd (S=1/2), behaves as a Mott-insulator at room tempera-
ture [16] and undergoes a spin-Peierls transition, at low temperature.
The band structure of the perylene chain was first calculated, by
Veiros et al [20] and more recently by Canadell et al [21] (using the
extended Huckel approach with both single-⇣ and double-⇣ type atomic
basis sets). These calculations estimate the intrachain transfer integral
(tb ) as 148.6 meV (or 353.8 meV for the double-⇣ basis set), with the other
transfer integrals in the range of 0 - 2.4 meV (or 0 - 8 meV for the double-
⇣). This confirms that these materials are nearly perfect one-dimensional
conductors, and that the conduction is mainly along the stacking direc-
tion of the perylene molecules, which are partially oxidized, (Per)+1/2 .
These values of the transfer integrals are in agreement with the exper-
imental values of the bandwidth, W , which are in the range 0.4 - 0.6
eV (W = 4t) depending on the metal M, as estimated from both ther-
mopower and magnetic susceptibility measurements [22, 23, 24]. They
also agree with the experimental measurements of the anisotropy of the
electrical conductivity estimated as of b / a ⇡ 103 [25].
Since there are four perylene stacks per unit cell, there must be four
HOMO bands crossing the Fermi level. If we would neglect the transverse
interactions, the Fermi surface would be the superposition of four planes
at ±kF = ±0.375 b⇤ , (by convention, (b · b⇤ ) = 2⇡, b = 4.1891 Å;
kF = 14 b⇤ = 2b ⇡
= 0.375 Å 1 along b⇤ ), but because of these transverse
interactions, the Fermi surface splits into four sheets with some warping
as shown in Figure 12.6. The warping of the sheets is very small, of the
order of 2 meV along a⇤ and is even smaller along c⇤ .
With these relatively weak interchain interactions, the four Fermi
surface sheets may touch each other, and by hybridization between them
Figure 12.6: Estimated Fermi surface of the perylene chain in

(Per)2 [Pt(mnt)2 ]. Note that this is a largely enlarged picture around
kF = 0.375 b⇤ . (After reference [21]).
there could be regions with closed pockets. Although these effects usually
would be ignored, they may become important at very low temperatures,
as suggested by experimental results described below.
The geometry of the Fermi surface was also tested with detailed mea-
surements of angular effects in the magnetoresistance, but the results are
difficult to interpret because there are simultaneously angular effects and
orbital effects with quantum interference [26] (see below).
12.3.2 Transport and Magnetic Properties

The ↵-(Per)2 [M(mnt)2 ] (M=Pt, Au, Ni, Cu) compounds show room tem-
perature conductivities of the order of 700 S/cm along the high conduc-
tivity b axis and the anisotropy in the a, b plane is estimated to be of the
order of 900 - 1000 (see Figure 12.7).
10 -2
Resistivity/Ω m
10 -3
10 -4
10 -5
10 -6
3 10 30 100 300
Temperature/K
Figure 12.7: Resistivities as a function of temperature for the ↵-phases

of the (Per)2 [M(mnt)2 ] famility (after references [17, 25, 27, 23]).
The room temperature conductivity of the Pd compound is of the or-

der of 300 S/cm and the Co and Fe compounds (with a crystal structure
where the b axis is doubled) present a relatively lower conductivity. In all
members of the ↵ series, the resistivity, ⇢(T ), decreases with the tempe-
rature, with the conductivities following a metallic regime as ⇠ T ↵ ,
with ↵ = 1.5, reaches a minimum at a temperature T⇢ and goes through
a metal-insulator transition at Tc , corresponding to the maximum of
d ln ⇢/d(1/T ).
These results are compiled in Table 12.2.
Table 12.2: Metal-insulator transition temperatures of the ↵-

(Per)2 [M(mnt)2 ].
M Au Pt Pd Ni Cu Fe Co
Tc / K 12 8 28 25 33 58 73
The thermopower, S, in the metallic regime, is very similar for all the
↵-phases and in the range of 32 - 42 µV K 1 at room temperature, and
it decreases upon cooling, as expected for metals, and in agreement with
the resistivity data. At the metal-insulator transition temperature, Tc ,
dS/d(1/T ) also shows an anomaly and, at lower temperature, the ther-
mopower varies as 1/T , as expected for a semiconductor. The positive
sign of the thermopower in the metallic state is indicative of hole type
conduction, which is consistent with a 3/4 filled band of the perylene
chains. From the linear regime at high temperatures, it is possible to
estimate a bandwidth, W = 4t, of the order of 0.6 eV for the series, with
the exception of the Co and Fe compounds where it is of the order of
0.5 eV.
The magnetic properties, namely the magnetic susceptibility, , and
electron spin resonance, ESR, spectra have been studied in great detail in
all the members of the series [3, 4, 16, 25, 28]. It is important to mention
that, through the combination of the susceptibility data with the ESR
spectra, is was possible to separate the contributions to the susceptibility
from both the itinerant electrons in the perylene chain and the localised
electrons in the [M(mnt)2 ] chain, for the paramagnetic species where
M=Ni, Pd, Pt. The [Au(mnt)2 ], being diamagnetic, was used as a term
of comparison. Figure 12.8 schematically represents the general features
of the magnetic properties for the salts with paramagnetic anions (S =
1/2).
a) b) Perylene M(mnt) 2
M chain chain
Fast spin J
exchange
Perylene M(mnt) 2 Perylene Itinerant Localized

chain chain chain electrons Spins
Figure 12.8: a) In (Per)2 [M(mnt)2 ] (M=Ni, Pt, Pd), the crystal struc-
tures are formed by chains of [M(mnt)2 ] units surrounded by perylene
chains. b) Interactions along and between the chains: the perylene chain
is a 3/4 filled band metal, while the [M(mnt)2 ] chain is an insulating array
of localized spins, with fast spin exchange with the itinerant electrons.
As we will discuss below in some detail, it is the particular combi-

nation of the structural and the electronic and magnetic properties and,
in particular, the features of the Fermi surface, that make this series of
quasi one-dimensional organic conductors so full of surprises.
12.3.3 Charge Density Wave Formation: Precursor Ef-

fects
As discussed in section 2 and illustrated in Figure 12.2, it would be
expected to observe charge density wave formation and Peierls instabil-
ities in the ↵-phases of the (Per)2 [M(mnt)2 ] family, as a result of the
strong one-dimensionality evidenced by the crystal structure and high
anisotropy in the conductivity. A schematic representation of the pos-
sible charge density wave formation and lattice modulation is shown in
Figure 12.9 for the 3/4 (Per)0.5+ filled band, which is expected to be
unstable to tetramerization of the perylene chains (CDW).
In fact, metal-insulator transitions are clearly observed in the plots of
Figure 12.7 for all members of the family, and the transition temperatures
are given in Table 12.2, where Tc corresponds to the CDW formation
(M=Au: TCDW = 12 K; M=Pt: TCDW = 8 K); 2 0 = ⇣kB T , with
5 < ⇣ < 10. In the Pt compound, (and for the Ni and Pd), as the
tetramerization of the perylene chain occurs, there is a simultaneous
dimerization of the [Pt(mnt)2 ] chains, corresponding to a spin-Peierls
instability.
In order to study the expected structural instabilities of these 1D con-
ductors, (see section 2.1 and Figure 12.3), Henriques et al [17] performed
X-ray diffuse scattering measurements in (Per)2 [M(mnt)2 ] (M=Pt, Pd,

Au), at several temperatures until reaching below the observed metal-
insulator transitions, with the exception of the Pt compound.
a) E d) E
EF
k k
−π/b -kF 0 kF π/b −π/b −π/4b 0 π/4b
kx kx
b) e)
ky ky
q=2kF
λ=4b
c) b f) y
y
Figure 12.9: CDW formation for a 3/4 filled band (1/4 filled hole band)
in a regular (Per)0.5+ chain. a,b,c) Electron dispersion of the regular
chain, Brillouin zone showing the q = 2kF phonon wave vector, and the
regular lattice with cell parameter b, respectively. d,e,f) Opening of the
gap at kF = 4b ⇡
, CDW formation with = 4b, and tetramerization of
the perylene chain.
These structural fluctuations are observed in the metallic state of

(Per)2 [Pt(mnt)2 ], as diffuse lines, at the wave vector 12 b⇤ , as shown, at
10 K, in Figure 12.10, but start to appear at 25 K, well above the crit-
ical temperature Tc = TCDW = 8 K, which unfortunately could not
be reached with the available experimental set-up. These diffuse scat-
tering lines, attributed to the [M(mnt)2 ] stacks, are also observed in
(Per)2 [Pd(mnt)2 ], and, in this case, they condense in superstructure re-
flections below 28 K, characteristic of a spin-Peierls transition.
A puzzling feature shown in (Per)2 [Pt(mnt)2 ] (Figure 12.10), and not
shown in (Per)2 [Pd(mnt)2 ], is the presence of sharp diffuse lines passing
through the main Bragg reflections, belonging to diffuse sheets with the
reduced wave vector qb = 0b⇤ . Their origin could not be understood at
the time. Another unexpected result of this investigation is that no dif-
fuse lines were detected in (Per)2 [Au(mnt)2 ] down to 12 K, contrary to
(Per)2 [Pt(mnt)2 ] and (Per)2 [Pd(mnt)2 ] which show 1D structural fluc-
tuations below 25 K and 100 K respectively.
The instabilities in one-dimensional electronic systems may occur
at the qb = 2kF or the qb = 4kF wave vectors, where kF is the Fermi
Figure 12.10: X-ray pattern of (Per)2 [Pt(mnt)2 ] showing at 10 K (these

lines appear below 25 K), the broad 12 b⇤ diffuse scattering (white arrows)
and the sharp 0b⇤ diffuse scattering (black arrow). On this pattern, the
chain axis, b, is horizontal. This pattern is free from /2 contamination.
After reference [17].
wave vector. Through coupling with the phonons, these (CDW or SDW)
instabilities may induce structural instabilities at the same wave vector.
Considering that the conduction band (from the perylene chain) is
3/4 filled, corresponding to the formula (Per)+ 2 [M(mnt)2 ] with one elec-
tron per dithiolate, we will be in the presence of a qb = 2kF instability.
This corresponds to S = 1/2 for each [Pt(mnt)2 ] and [Pd(mnt)2 ] an-
ion, and a closed-shell ion, with spin pairing (S = 0) for the [Au(mnt)2 ]
anion. This is in agreement with the magnetic properties of these charge
transfer salts, showing a Curie-Weiss law (at high temperature) for the
Pt [29, 16] and Pd [4] species, and no such behaviour for the Au species,
which show only a small temperature independent susceptibility typical
of metals due to the perylene stacks [30].
Similar diffuse scattering studies were performed in (Per)2 [M(mnt)2 ]
(M=Cu, Ni, Co and Fe) [31]. The metal-insulator transitions observed
at 33, 25, 73 and 58 K for the Cu, Ni, Co and Fe salts, respectively, were
found to correspond to the tetramerization on the perylene chains in the
case of the Ni and Cu derivatives, and to the dimerization for the Fe and
Co derivatives (where the b axis was doubled). In all cases, the periodic
lattice distortion can be associated to a 2kF Peierls instability of the
perylene chains. The Ni compound shows, in addition, a 4kF distortion,
corresponding to the dimerization of the [Ni(mnt)2 ] chain of localized
1/2 spin, leading to a non magnetic spin-Peierls like ground state.
A remarkable phenomenon, which was observed, was the non-linear
electrical transport, both in broad brand and narrow band noise in the
CDW state of (Per)2 [M(mnt)2 ], (M=Au, Pt, Co), that provided, for
the first time in an organic conductor, clear evidence of coherent CDW
motion as the origin of non-linear transport. These non-linear effects
are denoted by a large increase of current above a critical threshold
field, associated with an increase of broad and narrow band noise. In
the Pt compound, the presence of narrow band noise with frequencies
proportional to the excess current was in excellent agreement with the
coherent motion of a 2kF CDW ( = 4b = 16.8 Å) [32, 33, 34, 35, 36].
12.3.4 Nature of the Phase Transition in (Per)2 [Pt(mnt)2 ]

The most remarkable and unique feature of the (Per)2 [M(mnt)2 ] 1D
solids is the fact that the perylene chain is a quarter (hole) filled band
metal, at room temperature, while the band in the [M(mnt)2 ] chain is
half filled, with localized spins for M=Ni, Pt and Pd (S=1/2), exhibit-
ing fast spin exchange with the itinerant electrons on the perylene chain
(see Figure 12.8). For M=Au, the chain is diamagnetic, making a good
reference to study the peculiar magnetic properties of the series [16]. In
this section, we will focus on the nature of the metal-insulator transi-
tion in the platinum compound, which has been the most studied (along
more than 40 years) and is still a matter of controversy. Figure 12.11
illustrates the cascade of phenomena observed in (Per)2 [Pt(mnt)2 ], when
going from room temperature to low temperatures.
SP-CDW Metal
cts
fe
ef
r
on to
q= rso nes
D
ch P
iti la
D S-
ain
u a
2k r 1
ns su
ec pl
tra al-in
on t of
Pr use
se
F
ff
et
On
Di
M
0 8 18 25 50 T
Tc Tρ
Tetramerization of P chain
Opening of gap
2Δ=8.6 meV
Peierls on P chain
Dimerization and S-P on D chain
Figure 12.11: Cascade of phenomena (precursor effects and transition)

for the (Per)2 [Pt(mnt)2 ] compound as the temperature is lowered. The
P chain is the perylene chain, while the D chain is the [Pt(mnt)2 ] chain.
The perylene chain is metallic at room temperature and, at T ⇠ 8 K,

undergoes a Peierls (CDW) transition to an insulating state, where it
tetramerizes with wave vector qP er = ⇡/2b and the opening of a gap of
8.6 meV. On the other hand, the [Pt(mnt)2 ], (S = 1/2), chain is a Mott
insulator that undergoes a spin-Peierls transition, dimerizing with wave
vector qP t = ⇡/b and forming a spin singlet. An interesting aspect is
that even though qP t = 2qP er , diffuse X-ray scattering, specific heat, and
electrical transport indicate that both the CDW and SP transitions occur
at the same, or at very close temperatures, (Tc = TSP -CDW ) [17, 31, 37].
This is an indication that the two chains are strongly coupled, even with
the mismatch in the q vectors.
The transition is associated to precursor effects when coming from
high to low temperatures: at T ⇠ 50 K, the magnetic susceptibility, ,
is affected by the onset of the spin-Peierls instability on the dithiolate
chain, while at T ⇡ 25 K, diffuse planes appearing on the X-ray diffrac-
tion pattern are a symptom of the displacive distortion at the phonon
wave vector q = 2kF , kF being the Fermi wave vector for the electron
dispersion (on the perylene chain).
The nature of the transition, and its precursor effects, has been stud-
ied again and again, using many experimental techniques, including elec-
tron spin resonance, magnetic susceptibility, transport properties (re-
sistivity and thermopower) and X-ray diffuse scattering, as previously
discribed, as well as specific heat [37] and nuclear magnetic resonance
(NMR), both 1 H [38, 39] and 195 Pt NMR [39].
To make the story short, the present understanding based on, par-
ticularly, the comparison of the NMR data with all other results, one
concludes that the tetramerization on the perylene chain, correspond-
ing to the Peierls transition, ’forces’ the dimerization on the [Pt(mnt)2 ]
chain, corresponding to a spin-Peierls transition (see Figure 12.12).
The details of the coupling between the Peierls transition in the pery-
lene chain and the spin-Peierls transition in the [Pt(mnt)2 ] is not yet
clear, but it appears that a unique combination of SP and CDW order
parameters arises. Based on NMR results, in particular, on the linear
scaling of 1/T1 vs. T (T ) [38], which show that the spin degrees of
freedom remain diffusive and weak until TSP -CDW , it is not likely that
the SP dimerization of [Pt(mnt)2 ] would be a result of quantum spin dy-
namics alone. However, electron paramagnetic resonance (EPR) studies
indicate that an exchange interaction between the itinerant and the lo-
calized spins, associated with the onset of the TSP -CDW transition, might
be involved in the cooperative SP-CDW transition. One [40] and two-
chain [38] 1D Kondo lattice models have been considered, but at present
a) Per Pt(mnt) 2 Per b) Per Pt(mnt) 2 Per
º
3.3 º 5A
A 3.6
4b
b 2b
Metal SP CDW
Figure 12.12: a) Configurations of the perylene chains (outer planes) and

[Pt(mnt)2 ] chain (inner plane) along b, in the metallic state (T > Tc ).
b) Configurations at T < Tc : Spin-Peierls (SP) in the [Pt(mnt)2 ] chain
(dimerized) and CDW in the perylene chain (tetramerized).
there is no complete theory that treats the two-chain (Per)2 [Pt(mnt)2 ]

case. It has been shown [41] that, even at high magnetic fields, this cou-
pling persists, and that the initial tetramerization in the perylene chains
may promote the stabilization of the SP dimerization and the resulting
spin-singlet ground state, in spite of the predictions of mean-field theory
[42] that place the spin-Peierls decoupling well below the CDW phase
boundary, as shown in Figure 12.13, which represents the temperature
vs. magnetic field phase diagram.
Three regions of the (Per)2 [Pt(mnt)2 ] phase diagram are shown: the
metallic phase (M) is generally above 8 K; the SP-CDW phase is below
8 K and below 20 - 25 T, and the high-field (HF) phase is below 8 K and
above 20 - 25 T. The SP and CDW order parameters remain coupled at
high fields. All NMR spectra and spin relaxation data indicate that the
full spin-Peierls state forms below the CDW transition seen in transport
measurements, at least at low magnetic fields. The CDW ground state
formation appears to be a necessary condition to drive the formation of
the spin-Peierls ground state, and the commensurability (2:1) of the two
order parameter periodicities is likely to favour this effect.
12.3.5 (Per)2 [M(mnt)2 ] under Extreme Conditions

The behaviour of matter under extreme conditions, meaning, in particu-
lar, very low temperatures, high pressures and high magnetic fields is of
the greatest importance, for both theoretical and technological develop-
ment. The history of superconductivity is a good example.
10
MFT:CDW
Temperature / K
8 Transport
6 M
4 MFT:
SP-decoupling 1H NMR
2 HF
SP-CDW
0
0 5 10 15 20 25 30
Magnetic field / T
Figure 12.13: Schematic temperature-magnetic field phase diagram of

(Per)2 [Pt(mnt)2 ] showing the spin-Peierls (SP), charge density wave
(CDW), M (paramagnetic-metallic), and HF (high field) phase bound-
aries, derived from the onset of splitting of NMR spectra, and metal-
insulator transitions from electrical transport measurements, for a field
applied along the a-axis. ‘MFT:CDW’ and ‘MFT:SP-decoupling’ are the
mean-field theoretical predictions. (After references [41, 39]).
The search for superconductivity in organic solids led to the study

of the transport properties under high pressure, based on the belief that
forcing the molecules in a chain to come closer would lead to higher
transfer integrals and eventually to a superconducting phase. This was
actually achieved with the observation of superconductivity, for the first
time in an organic solid, namely (TMTSF)2 PF6 , [11], the first of the
series of the so called Bechgaard superconducting salts.
Pressure effects on the transport properties of (Per)2 [M(mnt)2 ], for
M=Pt and Au, were first studied by Henriques et al, [43], at pressures
up to 27 kbar. It was observed that the room temperature conductivity
measured along the needle axis, (b-axis), increases with pressure and
tends to saturate above approximately 9 kbar. The results show that the
evolution of the electronic correlations taking place in the neighbouring
conducting chains was quite unusual, when compared with other 1D
organic conductors, and further studies were needed.
The low Tc values of the Peierls transitions in these compounds (8 K
in Pt and 12 K in Au) elect them as ideal systems to test the effect of
high magnetic fields on a CDW. Initial measurements up to 8 T [44])
detected, for the first time in a CDW system, a sensible decrease of the
transition temperature with magnetic field. Larger effects were latter
detected in measurements up to 18 T which, although not large enough
for a complete suppression of the CDW, put into evidence a larger and
anisotropic dependence in the Pt compound, which was ascribed to the

coupling to the magnetic chains undergoing a spin-Peierls transition [45].
The total suppression of the CDW was finally observed in 2004 [46] when
studying the high field behaviour of the CDW ground state in several of
the elements of the series. At very high magnetic fields (up to 45 T) a
cascade of transitions in the platinum compound, tentatively ascribed to
field induced charge density wave (FICDW) phases, was discovered [47].
Figure 12.14 is a schematic representation of the phase diagram of the
platinum compound, where field induced charge density wave (FICDW)
state phases are shown at high magnetic fields (up to 45 T).
T Metallic
Low R
Activated region
Low R Activated
Region Low R
CDW0
Region
FICDW
Activated
Low R
FICDW
Region 45 T B||c
|b
B| 45 T
Figure 12.14: Schematic representation of the phase diagram of

(Per)2 [Pt(mnt)2 ], where field induced charge density wave (FICDW)
phases are shown at magnetic fields up to 45 T. (After reference [47])
The theory of CDW under high magnetic fields is still incomplete,

but considerable achievements have been reached, particularly by Lebed
et al [48, 49, 50, 51], namely for the (Per)2 [Pt(mnt)2 ].
12.3.6 Quantum Interference Effects

The interference of electrons was first observed in 1927 by Davisson and
Germer [52] in the diffraction pattern of electrons incident on nickel,
and was the confirmation of the de Broglie hypothesis that material par-
ticles, such as electrons, behaved as waves with wave length = h/p
(p = mv). A double slit Young type experiment with electrons was first
reported by Jönsson in 1961 [53]. Quantum interference of electron waves
in a metal has been observed by Stark and Friedberg [54] and later by
Sandesara and Stark [55], and is known as the Stark interference effect.
The Stark interference effect is observed as large amplitude oscillations in
the transverse magnetoresistance signal, periodic in inverse field, which

result from the direct interference of normal-state electron waves, cor-
responding to two electron ”trajectories” on different Fermi-surface (FS)
sheets, with a phase difference of = 2⇡ / 0 , where / 0 is the ratio of
the magnetic flux enclosed between the electron paths to the magnetic
flux quantum 0 = h/e.
The Stark interference effect has been observed through magneto-
transport measurements in (Per)2 [Au(mnt)2 ], a one dimensional organic
conductor, by suppressing the charge density wave (CDW) state with
pressure (above 3 kbar), below the transition temperature TCDW = 12 K
[56], see Figure 12.15.
a) ka b) a
2 b
∼2δ ∼2δ kb
QIA 1 y QIB
2 1
kF B||c λ
Figure 12.15: a) Simplified Fermi surface for (Per)2 [Au(mnt)2 ] after Ref.
[21]. The lattice parameters are (a, b, c) = (16.6, 4.19, 26.6) Å and the
bandwidths, (ta , tb , tc ) ⇠ (2, 150, < 0.1) meV; kF = 0.56 Å 1 ± , ⇠
0.002 kb , and vF ⇠ 1.6 ⇥ 105 m/s. b) Real-space motion of the carriers
with wavelength = 2⇡~/eBa ⇠ 2.5/B (µm T) and amplitude y =
4ta /evF B ⇠ 50/B (nm T). Transmission or reflection of the carriers can
occur at the interference nodes QIA and QIB . (After reference [56]).
The quantum interference (QI) oscillation amplitude exhibits a tem-

perature dependent scattering rate, indicative of an inhomogeneous metal
CDW ground state. The QI oscillation frequency reveals a Fermi-surface
topology of the unnested Fermi surface, in agreement with that recently
computed by Canadell et al [21] and schematically shown in Figure 12.6.
12.3.7 Superconductivity at Last

Finally, there is the discovery of superconductivity in the Au compound,
with the peculiarity that it occurs in the vicinity of the CDW state
[57]. As shown in Figure 12.16a), the resistance of (Per)2 [Au(mnt)2 ],
under a pressure of 5.3 kbar, drops sharply below 0.35 K, indicative of
the onset of superconductivity, although a residual resistance remains
in the sample. In the inset of Figure 12.16 a), the magnetoresistance
at 25 mK is displayed showing, at low fields, the critical-field behaviour
characteristic of superconductivity, and at higher fields the oscillations

characteristic of the Stark interference.
a) 8 b)
6 10
Resistance / Ω
Temperature / K
3
2
Resistance / Ω
Metal
10
1 1
0.7
1 T~25 mK CDW
0.5 0.1
0.4 0 5 10 15
0.3 Magnetic Field / T SC
0 0.1 1 10 100 0 1 2 3 4 5 6 7
Temperature / Κ Pressure / kbar
Figure 12.16: a) Resistance vs. temperature for (Per)2 [Au(mnt)2 ] at

5.3 kbar. Inset: magnetoresistance, showing the SC critical field and
Stark effect. b) CDW-SC phase diagram (after reference [57]).
(Per)2 [Au(mnt)2 ] has a non-magnetic anion, and exhibits a Peierls

type transition to a CDW ground state below 12 K [22]. The resulting
insulating CDW state is suppressed in high magnetic fields [58, 47]. The
CDW state can also be suppressed under high pressure, but, under some
conditions, a residual ”quantum melted” CDW character remains [59].
Under pressures above 3 kbar, the magnetoresistance shows a metallic
state with long range order, exhibiting a quantum interference Stark
effect. The onset of superconductivity occurs at ⇡ 300 mK, and the up-
per critical field (perpendicular to the perylene stacking axis) is 50 mT.
Although to fully demonstrate superconductivity will require measure-
ments, such as heat capacity, or magnetic susceptibility, which are dif-
ficult to do under high pressure, the critical field phase diagram sug-
gests BCS-like superconducting behaviour, (see figure 3 of reference [57])
with the parameters B0 = 49.4 mT and Tc = 0.313 K. Figure 12.16 b)
schematically shows the proposed CDW-SC phase diagram. The metal-
CDW transitions were determined from the standard d ln(R)/d(1/T )
peaks on the resistance, and the metal-superconductor transition from
the onset of superconductivity. An interesting aspect in this compound is
that superconductivity emerges from a charge density wave state, at vari-
ance with most other superconductors, where superconductivity emerges
from antiferromagnetic states.
12.4 Concluding Remarks

The enthusiastic expectations of the early nineteen seventies did not
bring about room temperature superconductors, but led to new theore-
tical developments in the science of condensed matter and to promising
emergent technologies. On the scientific side, the results have shown how
impressive is the physics dealing with the motion of electrons in one di-
mension. In the particular case of the two-chain highly one-dimensional
conductors of the (Per)2 [M(mnt)2 ] family, the work done, along these
more than 40 years, has shown how rich in physical phenomena are these
amazing materials. The story told in this paper is far from exhaustive
and it is certainly unfinished—there will be more surprises as theore-
tical and experimental methods progress. There are still no theoretical
answers to many questions, such as, for example, the nature of the mech-
anism of coupling of the charge density wave (CDW) and the spin-Peierls
(SP) state in the perylene and in the [M(mnt)2 ] chains, respectively, and
the origin of the field induced spin density wave (FISDW) state under
high magnetic fields.
There were many actors in this story, which includes several PhD
thesis. The development of the research in organic conductors, in Portu-
gal, benefited from several close collaborations, mainly with the groups
of Denis Jérôme and Jean-Paul Poujet at the Université Paris-Sud at
Orsay and the group of the late James Brooks, from the National High
Magnetic Field Laboratory (NHMFL) in Tallahasse (USA). As for the
best part of the theoretical work, it is due to the precious contribution of
Claude Bourbonnais from the Université de Sherbrooke in Canada and
also to Enric Canadell from Institut de Ciència de Materials de Barcelona
(ICMAB-CSIC). A word of recognition is also due to the many co-authors
who contributed for the progress of research in this family of compounds.
Acknowledgements
This work was supported by FCT-Portugal under contracts UID/EEA/
50008/2013 and UID/Multi/04349/2013.
Bibliography
[1] W. A. Little. Possibility of synthesizing an organic superconductor.
Physical Review, 134:A1416–A1424, 1964.
[2] L. Alcácer. Highly conductive molecular complexes. PhD thesis,

University of California, Riverside, 1970.
[3] L. Alcácer and A. H. Maki. Electrically conducting metal dithiolate-

perylene complexes. Journal of Physical Chemistry, 78:215, 1974.
[4] L. Alcácer and A. H. Maki. Magnetic properties of some electrically

conducting perylene-metal dithiolate complexes. J. Phys. Chem.,
80(17):1912–1916, 1976.
[5] R. E. Peierls. Quantum Theory of Solids. Oxford University Press,

1955.
[6] H. Fröhlich. On the theory of superconductivity: The one-

dimensional case. Proc. R. Soc. A, 223:296, 1954.
[7] J. Bardeen, L. N. Cooper, and J. R. Schrieffer. Theory of supercon-

ductivity. Physical Review, 108:1175–1204, 1957.
[8] F. London. Superfluids, volume 1. John Wiley & Sons, Inc., New
York, 1950.
[9] H. Akamatu, H. Inokuchi, and Y. Matsunaga. Electrical conductiv-

ity of the perylene-bromine complex. Nature, 173:168–169, 1954.
[10] L. B. Coleman, M. J. Cohen, D. J. Sandman, F. J. Yamagishi, A. F.

Garito, and A. J. Heeger. Superconducting fluctuations and Peierls
instability in an organic solid. Solid State Communications, 12:1125,
1973.
[11] D. Jérôme, A. Mazaud, M. Ribault, and K. Bechgaard. Supercon-

ductivity in a synthetic organic conductor (TMTSF)2 PF6 . Journal
de Physique Lettres, 41(4):L95–98, 1980.
[12] H. Shirakawa and S. Ikedo. Infrared spectra of poly(acetylene).

Polym. J., 2:231–244, 1971.
[13] Hideki Shirakawa, Edwin J. Louis, Alan G. MacDiarmid, Chwan K.

Chiang, and Alan J. Heeger. Synthesis of electrically conducting
organic polymers: halogen derivatives of polyacetylene, (CH)x . J.
Chem. Soc., Chem. Commun., 16:578–580, 1977.
[14] D. Jérôme and H. J. Schulz. Organic conductors and superconduc-

tors. Advances in Physics, 51(1):293–479, 2002.
[15] W. Kohn. Image of the Fermi surface in the vibration spectrum of

a metal. Phys. Rev. Lett., 2:393, 1959.
[16] L. Alcácer, H. Novais, S. Flandrois, C. Coulon, D. Chasseau, , and

J. Gautier. Synthesis, structure and preliminary results on electri-
cal and magnetic properties of (Perylene)2 [Pt(mnt)2 ]. Solid State
Commun., 35:945, 1980.
[17] R. T. Henriques, L. Alcácer, J. P. Pouget, and D. Jérôme. Electri-
cal conductivity and x-ray diffuse scattering study of the family of
organic conductors (perylene)2 M(mnt)2 , (M=Pt, Pd, Au). Journal
of Physics C: Solid State Physics, 17(29):5197, 1984.
[18] A. Domingos, R. T. Henriques, V. Gama, M. Almeida, A. Lopes
Vieira, and L. Alcácer. Crystalline structure/transport properties
realtionship in the (Perylene)2 M(mnt)2 family (M=Au, Pd, Pt, Ni).
Synthetic Metals, pages B411–B416, 1988.
[19] V. Gama, R. T. Henriques, M. Almeida, and L. Alcácer. Transport
and magnetic properties of the new members from the series of
organic conductors (Per)2 M(mnt)2 , M=Cu, Ni, Co, Fe, Pd, Au, Pt.
Synthetic Metals, 42(3):2553–2556, 1991.
[20] L. F. Veiros, M. J. Calhorda, and E. Canadell. Electronic
structure of the 1/1 mixed molecular and polymeric conductor
(Perylene)2 Co(mnt)2 (CH2 Cl2 )0.5 and comparison with the 2/1 ↵-
(Perylene)2 M(mnt)2 phases. Inorg. Chem., 33(19):4290–4294, 1994.
[21] E. Canadell, M. Almeida, and J. S. Brooks. Electronic band struc-
ture of ↵-(Per)2 M(mnt)2 compounds. Eur. Phys. J. B, 42:R453,
2004.
[22] M. Almeida and R.T. Henriques. Organic Conductive Molecules and
Polymers, volume 1, chapter 2, pages 87–149. John Wiley and Sons,
Chichester, 1997.
[23] V. Gama, M. Almeida, R. T. Henriques, I. C. Santos, A. Domingos,
S. Ravy, and J. P. Pouget. Low dimensional molecular conductors
(Per)2 M(mnt)2 , M=Cu and Ni: Low and high conductivity phases.
J. Phys. Chem., 95:4263–4267, 1991.
[24] R. T. Henriques, M. Almeida, M. J. Matos, L. Alcácer, and C. Bour-
bonnais. Thermoelectric power of the (perylene)2 M(mnt)2 family
(M=Pt, Au, Pd). Synth. Met., 19:379, 1987.
[25] L. Alcácer. Metal-insulator transitions in the perylene-dithiolate
family. Molecular Crystals and Liquid Crystals, 120(1):221–228,
1985. A7.
[26] D. Graf, J. S. Brooks, E. S. Choi, M. Almeida, R. T. Henriques,

J. C. Dias, and S. Uji. Geometrical and orbital effects in a quasi-
one dimensional conductor. Physical Review B, 80:155104–155108,
2009.
[27] V. Gama, R. T. Henriques, and M. Almeida. New members of the

series of quasi-one-dimensional metals (perylene)2 [M(mnt)2 ], M=Ni
and Cu. Fizika, 21:19–22, 1989.
[28] R. T. Henriques, L. Alcácer, M. Almeida, and S. Tomic. Trans-

port and magnetic properties on the family of perylene-dithiolate
conductors. Molecular Crystals and Liquid Crystals, 120:237–241,
1985.
[29] L. Alcácer, H. Novais, and F. Pedroso. Molecular Metals, W. E.

Hatfield ed., chapter Perylene-Dithiolate Complexes; The platinum
salt, pages 415–418. Plenum, New York, 1979.
[30] S. Flandrois. Unpublished results.
[31] V. Gama, R. T. Henriques, M. Almeida, and J. P. Pouget. Diffuse

X-ray scatterinmg evidence for Peierls and ’spin-Peierls’ like tran-
sitions in the organic conductors (Per)2 [M(mnt)2 ] (M= Cu, Ni, Co
and Fe). Synthetic Metals, 55-57:1677–1682, 1993.
[32] E. B. Lopes, M. J. Matos, R. T. Henriques, M. Almeida, and J. Du-

mas. Charge density wave nonlinear transport in the molecular
conductor (Perylene)2 Au(mnt)2 , (mnt=maleonitriledithiolate). Eu-
rophysics Letters, 27:241–246, 1994.
[33] E. B. Lopes, M. Matos, R. T. Henriques, M. Almeida, and J. Dumas.

CDW nonlinear transport in the organic systems (Per)2 M(mnt)2 .
Synthetic Metals, 70:1267–1270, 1995.
[34] E. B. Lopes, M. J. Matos, R. T. Henriques, M. Almeida, and J. Du-

mas. Charge density wave dynamics in the molecular conductor
(Perylene)2 Pt(mnt)2 , (mnt = maleonitriledithiolate). Physical Re-
view B, 52(R2237-R2240), 1995.
[35] J. Dumas, N. Thirion, E. B. Lopes, M. J. Matos, R. T. Henriques,

and M. Almeida. Wavelet analysis of the charge density wave dy-
namics in the molecular conductor (Perylene)2 Pt(mnt)2 , (mnt =
maleonitriledithiolate). Journal de Physique I, France, 5:539–545,
1995.
[36] E. B. Lopes, M. J. Matos, R. T. Henriques, M: Almeida, and

J. Dumas. CDW dynamics in quasi-one-dimensional conductors
(Per)2 M(mnt)2 , M=Au and Pt. Synthetic Metals, 86:2163–2164,
1997.
[37] G. Bonfait, M. J. Matos, R. T. Henriques, and M. Almeida. Spin-

Peierls instability in (Per)2 [M(mnt)2 ] compounds probed by specific
heath. Journal de Physique IV, Colloque C2, 3, pages 251–254, 1993.
[38] C. Bourbonnais, R. T. Henriques, P. Wzieteck, D. Kongle, J. Vo-

iron, and D. Jérôme. Nuclear and electronic resonance approaches to
magnetic and lattice fluctuations in the two-chain family of organic
compounds (perylene)2 [Pt(S2 C2 (CN)2 )2 ] (M=Pt, Au). Physical Re-
view B, 44:641–651, 1991.
[39] E. L. Green, L. Lumata, J. S. Brooks, P. Kuhns, A. Reyes, S. E.

Brown, and M. Almeida. 1 H and 195 Pt NMR study of the parallel
two-chain compound Perylene2 [Pt(mnt)2 ]. Crystals, 2:1116–1135,
2012.
[40] J. C. Xavier, R. G. Pereira, E. Miranda, and I. Affleck. Dimeriza-

tion induced by the RKKY interaction. Physical Review Letters,
90(24):247204, 2003.
[41] E. L. Green, L. L. Lumata, J. S. Brooks S, P. L. Kuhns, A. P. Reyes,

M. Almeida, M. J. Matos, R. T. Henriques, J. A. Wright, and S. E.
Brown. Interaction of magnetic field-dependent Peierls and spin-
Peierls ground states in (Perylene)2 [Pt(mnt)2 ]. Physical Review B,
Rapid Communications, 84:121101 (R), 2011.
[42] J. W. Bray, L. V. Interrante, I. S. Jacobs, and J. C. Bonner. Ex-

tended Linear Chain Compounds, edited by T. S. Miller, chapter
The Spin-Peierls Transition, pages 353–415. Plenum Press, New
York and London, 1983.
[43] R. T. Henriques, L. Alcácer, D. Jérôme, C. Bourbonnais, and

C. Weyl. Electrical conductivity of (perylene)2 M(mnt)2 , (M=Pt,
Au) under pressure. Journal of Physics C: Solid State Physics,
19(24):4663, 1986.
[44] G. Bonfait, E. B. Lopes, M. J. Matos, R. T. Henriques, and

M. Almeida. Magnetic field dependence of the metal-insulator tran-
sition in (Per)2 Pt(mnt)2 and (Per)2 Au(mnt)2 . Solid State Commu-
nications, 80:391–394, 1991.
[45] M. Matos, G. Bonfait, R. T. Henriques, and M. Almeida. Modifi-

cation of the magnetic field dependence of the Peierls transition by
a magnetic chain. Physical Review B, 54:15307–15313, 1996.
[46] D. Graf, E. S. Choi, J. S. Brooks, M. Matos, R. T. Henriques,

and M. Almeida. High magnetic field induced charge density wave
state in a quasi-one-dimensional organic conductor. Phys. Rev. Lett.,
93(7):076406, 2004.
[47] J. S. Brooks, D. Graf, E. S. Choi, M. Almeida, J. C. Dias, R. T. Hen-

riques, and M. Matos. Magnetic field dependence of CDW phases in
Per2 M(mnt)2 , (M = Pt, Au). Journal of Low Temperature Physics,
142:787–803, 2006.
[48] A. G. Lebed. Theory of magnetic field-induced charge-density wave

phases. JETP Letters, 78:138, 2003.
[49] A. G. Lebed and S. Wu. Soliton wall superlattice charge-density-

wave phase in a magnetic field. JETP Letters, 86:135–138, 2007.
[50] A. G. Lebed and S. Wu. Soliton wall superlattice in quasi-one-

dimensional conductor (Per)2 Pt(mnt)2 . Physical Review Letters,
99:026402, 2007. Also: arXiv:cond-mat/0702500v2.
[51] A. G. Lebed. Universal theory of magnetic field-induced charge-

density wave phases: Theory versus experiment. Physical Review
Letters, 103:046401, 2009.
[52] C. Davisson and L. H. Germer. Diffraction of electrons by a crystal

of nickel. Phys. Rev., 30:705, 1927.
[53] C. Jönsson. Elektroneninterferenzen an mehreren künstlich

hergestellten feinspalten. Z. Phys., 161:454–474, 1961. English
translation: Am. J. Phys. 42, 4-11 (1974).
[54] R. W. Stark and C. B. Friedberg. Quantum interference of electron

waves in a normal metal. Phys. Rev. Lett., 26(10):556–559, 1971.
[55] N. B. Sandesara and R. W. Stark. Macroscopic quantum coherence

and localization for normal-state electrons in Mg. Phys. Rev. Lett.,
53(17):1681–1683, 1984.
[56] D. Graf, J. S. Brooks, E. S. Choi, M. Almeida, R. Henriques, J. C.

Dias, and S. Uji. Quantum interference in a quasi-one-dimensional
metal. Physical Review B, 75:245101–245108, 2007.
[57] D. Graf, J. S. Brooks, M. Almeida, J. C. Dias, S. Uji, T. Terashima,

and M. Kimata. Evolution of superconductivity from a charge den-
sity wave ground state in pressurized (Per)2 [Au(mnt)2 ]. Europhysics
Letters, 85:27009, 2009.
[58] D. Graf, J. S. Brooks, E. S. Choi, S. Uji, J. C. Dias, M. Matos, and
M. Almeida. Suppression of a charge density wave ground state in
high magnetic fields: spin and orbital mechanisms. Physical Review
B, 69:125113, 2004.
[59] N. Mitsu, K. Yamaya, M. Almeida, and R. T. Henriques. Pressure
effect on charge density wave in Per2 M(mnt)2 , (M=Pt, Au). Journal
of Physics and Chemistry of Solids, 66:1567–1570, 2005.

Electrons 1D

Uploaded by

Copyright:

Available Formats

Electrons 1D

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Electrons 1D

Uploaded by

Copyright:

Available Formats

12

Interacting electrons in one

Luís Alcácer1 , Rui T. Henriques1 and Manuel Almeida2

Corresponding author contact: [email protected]

In 1964, William Little, from Stanford University, proposed a model

Figure 12.1: The constituent molecules of the (Per)2 [M(mnt)2 ] charge

by Klaus Bechgaard and found to have a transition temperature of

12.2 The Interesting Physics: 1 D Instabilities

ducting type. The needed theoretical framework is identical to that of

2. In the presence of magnetic interactions, the ground state is unsta-

3. In a 1 D system with short range interactions, the thermal fluctu-

4. In a 1 D system, an arbitrarily small disorder induces electron

Figure 12.2: a) 1 D lattice, electron density n, corresponding to one

holes should be considered instead of electrons, and the degree of fill-

12.2.1 The Peierls Instability

A modulation of theplattice with wave vector q = 2kF is described by

where is the phase of (i.e., | |ei ): there is a modulation of the elec-

Figure 12.3: Evolution of the Kohn anomaly, in the phonon spectrum,

The phonons with wave vector near q = 2kF involve displacements of

• In both cases, the instability is expressed by a correlation function

which diverges logarithmically.

• The transition temperature and the gap at T = 0 only diﬀer by a

• The eﬀect of impurities in the Peierls transition is analogous to the

12.2.2 Competing Instabilities in a 1 D Electron Gas

Peierls transition with a gap | S| in the Fermi surface which opens at

Figure 12.4: Phase diagram in the g1 g2 plane showing the domains

12.3 The Perylene-Metal Dithiolate Series

Figure 12.5: Crystal Structure of the ↵-phase of the (Per)2 [Pt(mnt)2 ].

there are segregated regular stacks of [Pt(mnt)2 ] and perylene parallel to

Figure 12.6: Estimated Fermi surface of the perylene chain in

12.3.2 Transport and Magnetic Properties

Figure 12.7: Resistivities as a function of temperature for the ↵-phases

The room temperature conductivity of the Pd compound is of the or-

Table 12.2: Metal-insulator transition temperatures of the ↵-

Perylene M(mnt) 2 Perylene Itinerant Localized

As we will discuss below in some detail, it is the particular combi-

12.3.3 Charge Density Wave Formation: Precursor Ef-

X-ray diﬀuse scattering measurements in (Per)2 [M(mnt)2 ] (M=Pt, Pd,

These structural fluctuations are observed in the metallic state of

Figure 12.10: X-ray pattern of (Per)2 [Pt(mnt)2 ] showing at 10 K (these

12.3.4 Nature of the Phase Transition in (Per)2 [Pt(mnt)2 ]

Figure 12.11: Cascade of phenomena (precursor eﬀects and transition)

The perylene chain is metallic at room temperature and, at T ⇠ 8 K,

a) Per Pt(mnt) 2 Per b) Per Pt(mnt) 2 Per

Figure 12.12: a) Configurations of the perylene chains (outer planes) and

there is no complete theory that treats the two-chain (Per)2 [Pt(mnt)2 ]

12.3.5 (Per)2 [M(mnt)2 ] under Extreme Conditions

Figure 12.13: Schematic temperature-magnetic field phase diagram of

The search for superconductivity in organic solids led to the study

anisotropic dependence in the Pt compound, which was ascribed to the

Figure 12.14: Schematic representation of the phase diagram of

The theory of CDW under high magnetic fields is still incomplete,

12.3.6 Quantum Interference Eﬀects

the transverse magnetoresistance signal, periodic in inverse field, which

The quantum interference (QI) oscillation amplitude exhibits a tem-

12.3.7 Superconductivity at Last

characteristic of superconductivity, and at higher fields the oscillations

Figure 12.16: a) Resistance vs. temperature for (Per)2 [Au(mnt)2 ] at

(Per)2 [Au(mnt)2 ] has a non-magnetic anion, and exhibits a Peierls