ARTICLE

Received 16 Oct 2014 | Accepted 26 Mar 2015 | Published 7 May 2015 DOI: 10.1038/ncomms8034

Charge order from orbital-dependent coupling

evidenced by NbSe2
Felix Flicker1 & Jasper van Wezel2

Niobium diselenide has long served as a prototype of two-dimensional charge ordering,

believed to arise from an instability of the electronic structure analogous to the
one-dimensional Peierls mechanism. Despite this, various anomalous properties have
recently been identified experimentally, which cannot be explained by Peierls-like weak-
coupling theories. Here, we consider instead a model with strong electron–phonon coupling,
taking into account both the full momentum and orbital dependence of the coupling matrix
elements. We show that both are necessary for a consistent description of the full range of
experimental observations. We argue that NbSe2 is typical in this sense, and that any charge-
ordered material in more than one dimension will generically be shaped by the momentum
and orbital dependence of its electron–phonon coupling as well as its electronic structure. The
consequences will be observable in many charge-ordered materials, including cuprate
superconductors.

1 H. H. Wills Physics Laboratory, University of Bristol, Tyndall Avenue, Bristol BS8 1TL, UK. 2 Institute for Theoretical Physics, Institute of Physics, University of

Amsterdam, 1090 GL Amsterdam, The Netherlands. Correspondence and requests for materials should be addressed to F.F. (email: fl[email protected]) or
J.v.W. (email: [email protected]).

NATURE COMMUNICATIONS | 6:7034 | DOI: 10.1038/ncomms8034 | www.nature.com/naturecommunications 1

& 2015 Macmillan Publishers Limited. All rights reserved.
ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms8034

T
he charge density wave (CDW) order in 2H-NbSe2 has space. The second experimental result arguing for a strongly
been surrounded by controversy since its first observation coupled transition is the observation of a reduction of the density
nearly 40 years ago1. The most basic unresolved question of states persisting well above the transition temperature, in what
concerns the driving mechanism of the ordering transition. In has been termed a pseudogap regime4,23,24. In strongly coupled
quasi-one-dimensional (1D) materials, it is well known that CDW materials, a pseudogap typically appears through the
charge order arises from a Peierls instability: any interaction presence of locally fluctuating charge order without long-range
between states at the Fermi energy (EF) results in a periodic coherence23. These same charge fluctuations have recently been
modulation of both the electronic charge density and the atomic suggested to play a central role in the description of the pseudogap
lattice positions, which lowers the free energy through the phase of cuprate superconductors25–30. The precise nature of the
corresponding opening of a band gap2. Such a mechanism is pseudogap in these materials, however, is difficult to establish
termed weak coupling since an arbitrarily small interaction will owing to the presence of various nearby competing orders. NbSe2,
drive the system to order at low enough temperatures. In higher- being free of such complications, provides a test bed on which the
dimensional systems, a nesting of the Fermi surface is typically formation and characteristics of a pseudogap phase due to charge
required for such an instability to be effective. In NbSe2, however, fluctuations can be studied directly.
Fermi surface nesting does not occur1,3–5. The lack of a lock-in Here we present a theoretical analysis of the charge order in
transition and the coexistence of CDW order with NbSe2 based on a model of strong electron–phonon coupling.
superconductivity below 7.2 K nonetheless seem to suggest a Although the transition is phonon driven, we find that a full
weak-coupling origin of the charge order. As a result, a number of knowledge of the electron states scattered from and to, including
alternative weak-coupling mechanisms have been proposed, both wave vector and orbital dependence, is crucial in explaining
based on nested saddle-points in the electronic dispersion6, the experimental data. Despite a lack of nesting, the limited
local field effects5 or a combination of weak nesting with matching of states at EF acts in unison with the momentum
momentum-dependent electron–phonon coupling3,7. dependence of the electron–phonon coupling to select out the
There are a number of experimental observations, however, CDW-ordering vector, while the orbital characters of the bands
that are hard to reconcile with weak-coupling approaches in naturally explain why a CDW gap appears primarily in only one
NbSe2. The first concerns the size of the electronic gap in the of the Fermi surface pockets. The resulting momentum
charge-ordered state. Kinks in the density of states (DOS) dependence of the gap itself agrees with ARPES experiments as
observed by planar tunnelling experiments have been interpreted well as with the observed particle–hole asymmetry in the DOS.
to arise from a gap of D ¼ 35 meV8,9, while only much smaller Finally, we find that the phonon modes are softened over a broad
gaps between 2 and 5 meV were seen by high-precision angle- range of wave vectors, and that including the presence of strong
resolved photoemission spectroscopy (ARPES) experiments4,10, phonon fluctuations leads to a suppression of the transition
and older studies even reported no gap at all in ARPES and temperature, implying a range of intermediate temperatures
resistivity measurements11,12. Assuming a weakly coupled driving dominated by incoherent, fluctuating, charge order and a
mechanism for the charge order, the size of the gap is expected to corresponding pseudogap. In light of these findings, and the
be proportional to the transition temperature, DC1.76kBTC, in generic nature of the electronic structure of NbSe2, we argue that
direct analogy to the BCS theory of superconductivity13. A gap the momentum and orbital dependence of the electron–phonon
size of 35 meV would then be far too large to explain the observed coupling may be expected to play a similarly central part in the
TCDW of 33 K14. The 5 meV gap consistent with ARPES data is of description of other charge-ordered materials, especially in the
the expected size, but it occurs only in isolated points on the presence of relatively strong coupling as in the case of the layered
Fermi surface4,10, raising the question of how the correspondingly high-TC superconductors.
small gain in electronic energy can overcome the cost of
introducing the periodic lattice distortions associated with
CDW order. In addition, while the gaps seen in ARPES Results
measurements are connected by the CDW wave vector, other The extent of electronic nesting. NbSe2 is a layered material, in
equally well-connected points support either only a much smaller which hexagonal layers of niobium atoms are sandwiched
gap or no gap at all4,10. It was recently suggested, based on between similar layers of selenium atoms, displaced so that they
scanning tunnelling spectroscopy (STS) measurements15, that a lie above and below half of the Nb interstitial locations, as shown
particle–hole asymmetric gap of 12 meV, centred above EF, exists in the inset of Fig. 1. Consecutive sandwich layers are displaced to
in NbSe2. The offset in energy explains why ARPES only sees a have the complementary half of the interstices occupied, giving
smaller gap size, while the value of 12 meV is consistent with two formula units per unit cell. The Fermi surface consists of two
the order of magnitude expected for a strongly coupled CDW, concentric barrel-shaped pockets centred around both the G and
which may have DB4kBTC in analogy to strong-coupling the K points, as well as a very small pancake-shaped pocket
superconductors16–18. surrounding G, which we ignore in the analysis below. The inner
The idea that the CDW order in NbSe2 is of the strong coupling pockets at K are seen in ARPES to develop the largest CDW
variety, in which a sizable electron–phonon coupling is essential in gaps4,10. The CDW itself is of the 3Q type, with three equivalent
causing an instability of the electronic structure, is also consistent superposed modulations at 120° with respect to one another.
with two further recent experimental observations. First, it was To describe the electronic states, we employ a Slater–Koster
observed in X-ray scattering experiments that at the transition tight-binding fit to the band structure using a basis of Se-4p and
temperature a broad momentum-range of phonons simulta- Nb-4d orbitals31,32. For the two main bands crossing EF that give
neously softens towards zero energy7,19. This is in stark contrast to rise to the barrel-shaped pockets (which we will refer to as inner,
the sharp Kohn anomaly expected in a Peierls-like transition13 developing the CDW instability, and outer), the fit is constrained
and may be understood to be an effect of the importance of the by data from ARPES measurements10, while the remaining 20
entropy carried by phonon fluctuations in the strongly coupled bands are fit to recent LDA calculations5. In agreement with
scenario20–22. These phonon fluctuations are localized in real earlier reports32, we find that both the inner and outer bands at
space and therefore necessarily involve a broad range of momenta. EF are primarily composed of the d3z2  r2 orbitals centred on the
In an electronically nested system, on the other hand, the nesting two Nb atoms within a unit cell, in the form of bonding (inner)
vector localizes any fluctuations to a single point in momentum and antibonding (outer) combinations (as shown in Fig. 1).

2 NATURE COMMUNICATIONS | 6:7034 | DOI: 10.1038/ncomms8034 | www.nature.com/naturecommunications

& 2015 Macmillan Publishers Limited. All rights reserved.
NATURE COMMUNICATIONS | DOI: 10.1038/ncomms8034 ARTICLE

0.76
K
M

0(q)

QCDW
Γ

0 q 1
EF
0.56

g(q)
qg
0.56 qg 0.76

Figure 2 | Quantifying the extent of nesting in NbSe2. (a) Electronic

susceptibility as a function of momentum transfer along GM. Here q is
defined through q ¼ qGM, with qA[0,1]. Blue dotted line: in a perfectly
Γ M K Γ nested 1D band structure, the susceptibility is sharply peaked at q ¼ QCDW.
Red dashed curve: the result for the two bands in NbSe2 crossing EF, and an
Figure 1 | Model band structure of NbSe2. (a) The band structure of
orbital-independent electron–phonon coupling. The susceptibility is nearly
NbSe2 modelled by a tight-binding fit to both ARPES data and first-
flat, without any peaks. Black solid curve: the susceptibility of just the inner
principles band structure calculations. The two bands crossing EF (thick
band, in which a CDW gap is observed. Although a clear maximum can be
lines) are the ones involved in the formation of charge order. They consist of
seen at QCDW, it is not a sharp divergence indicative of perfect nesting.
bonding and antibonding combinations of the two niobium d3z2  r2 orbitals  2
within a unit cell, as indicated (with red and blue signifying postive and (b) The model electron–phonon coupling g ¼  aqg  q þ gmax as q is
negative parts of the orbital wave functions). (b) The Fermi surface varied along GM. (c) Evolution of the CDW vector QCDW with varying peak
resulting from the two bands crossing EF. (c) The layered atomic structure position qg of the model electron–phonon coupling. Blue dotted line: in a
of NbSe2, with Nb in red and Se in blue. The unit cell, containing two perfectly nested 1D band structure, the electronic structure determines the
formula units across consecutive sandwich layers, is indicated. CDW formation and QCDW is independent of qg. Red dashed curve: the
result for an orbital-independent electron–phonon coupling and the two
Focussing on the two non-interacting bands at EF, we bands crossing EF, which is close to a straight line at 45°, indicative of a
introduce the interaction between electrons c and phonons j: negligible role of the electronic structure. Black solid curve: the evolution
X X m;n relevant for NbSe2, given by just the inner band as imposed by the orbital-
^ int ¼
H ^ ym c
^ qc
gk;k þ q j ^n : ð1Þ
k kþq dependent electron–phonon coupling. This curve interpolates between the
m;n k;q flat plateau of a perfectly nested structure, and the 45° line dominated by
The bare phonon dispersion can be described by a Brillouin just the structure of the electron–phonon coupling.
function tending to a maximum of 11.2 meV at the zone
boundary19. The electron–phonon coupling g depends on the
momenta, as well as the band indices m and n, of the the ingoing the susceptibility to be entirely independent. This means that the
and outgoing electron states. Following Varma et al.33, we model lower value of the intraband coupling in the outer band leads to
both of these dependencies using knowledge purely of the an order of magnitude suppression of its contribution to w0 and
electronic structure, which has been shown to work well for a therefore to a correspondingly reduced gap in its Fermi surface
range of transition metals and their compounds. In terms of the pocket. It is thus the orbital character of the electronic bands, and
matrices of tight-binding overlap integrals Sk and generalized consequently the orbital dependence of the electron–phonon
eigenvectors Ak, the electron–phonon coupling matrix elements coupling, that causes a relative size difference between the CDW
are given by33: gaps in the pockets around the K-point, even if they are equally
h imn h imn well nested.
y y The expression for the susceptibility in equation (3) can further
gmn
kk 0 ¼ v m
k Ak Sk A k 0  Ak S k 0 Ak 0 v nk0 ; ð2Þ
be used to quantify both the amount of nesting present in NbSe2
where a constant prefactor has been omitted, and the band and to evaluate the relative contribution of the electron–phonon
velocity vn for electrons in band n is defined in terms of the band coupling matrix elements in selecting the observed value of the
energy xnk as vnk ¼ @xnk @k. In the interaction of equation (1) we (incommensurate) ordering wave vector QCDW ¼ ð1  dÞ 23 GM,
restrict attention to the longitudinal component of g, since only with dE0.021. In the left panel of Fig. 2, three susceptibilities are
the longitudinal acoustic phonons are observed to soften in compared. The dotted blue line shows the susceptibility for a
inelastic neutron and X-ray scattering experiments19. perfectly nested model with a single 1D band of cosine dispersion.
Inserting the tight-binding results into equation (2), we find The sharp peak characteristic of a nested Fermi surface is
that the interband electron–phonon coupling is strictly zero, apparent. The dashed red line is the bare susceptibility in NbSe2,
while the intraband coupling is about a factor of three stronger obtained by setting g ¼ 1 in equation (3). This does not include
for the inner band than for the outer. This result immediately any of the momentum and orbital dependence of the electron–
explains why the experimentally observed CDW gap is so much phonon coupling and is the form that has been widely used in
more pronounced in the inner band4,10, since the static electronic previous studies5,10,15. It is almost completely flat except for a
susceptibility (that is, the linear response to lattice distortions) very broad and low maximum far from QCDW, and can therefore
scales with the square of the electron–phonon coupling: not explain the observed ordering vector. The solid black line,
    on the other hand, displays the susceptibility of just the inner
X 2 f xm
 k  f xnk þ q band. A broad but distinct maximum develops at the
mn
w0 ðqÞ / gk;k þ q : ð3Þ experimentally observed CDW vector. The selection of the
k
xnk þ q  xm
k
ordering vector is thus shown to arise from the combined
Here f(x) is the Fermi–Dirac distribution function, and the static influence of both the electronic structure and the orbital and
Lindhard function is given by w0 when g ¼ 1. The absence of momentum-dependent electron–phonon coupling, with neither
interband coupling causes the contributions of the two bands to contribution being negligible4,10.

NATURE COMMUNICATIONS | 6:7034 | DOI: 10.1038/ncomms8034 | www.nature.com/naturecommunications 3

& 2015 Macmillan Publishers Limited. All rights reserved.
ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms8034

Restricting attention to the inner band, we can quantify the Figure 4 shows energy and momentum cuts through the
extent to which nesting contributes to the selection of QCDW. electronic spectral functions resulting from the solution to the gap
For this purpose we temporarily switch to a simplified form equations, as well as their corresponding DOS. A good match with
of g, which is a function only of the momentum transfer: the shape of the experimentally determined DOS in STS experiments
g ¼  a(qg–q)2 þ gmax. Even though in general the shape of the can be achieved by introducing a 4 meV shift in the chemical
susceptibility is determined by both the incoming momentum potential, which falls well within the range of experimental
and the momentum transfer, modelling the electron–phonon uncertainty15. The overall strength of the electron–phonon
coupling in this simplified form allows us to directly quantify the coupling, the constant of proportionality omitted in equation (2)
effect of varying the position of its peak, and thus its influence on and the only free parameter in the theory, is set to give a maximum
the position of the peak in w0 . The values of the parameters CDW gap magnitude of E12 meV. The Fermi surface on the right
a, qg and gmax can be estimated by fitting the experimentally of Fig. 4 then shows a CDW gap opening on the inner band only,
observed phonon softening7 to the expression for the and Fermi arcs forming at the same locations as observed in ARPES
renormalized phonon energy in the random phase measurements4,10,23. The extent of the gap on the Fermi surface, its
approximation (RPA), O2RPA ¼ O0 ðO0  w0 Þ, where we take O0 size (apparent from the dispersion’s back-folding in the centre of
to be the high-temperature (unrenormalized) dispersion. In the Fig. 4) and its restriction to one band are all in agreement with
right panel of Fig. 2 we plot the variation of the maximum in the experimental findings4,10,23. The plot of the DOS on the left
susceptibility (that is, the expected QCDW) as qg is varied along indicates that these results, which were previously interpreted to
GM. Again, the blue dotted line represents the situation of a imply a gap size of B5 meV4,10, are in fact consistent with the STS
perfectly nested band structure, in which the electron–phonon results indicating a particle–hole asymmetric gap of B12 meV15,23.
coupling never overcomes the peak in the susceptibility, and thus The asymmetry hides most of the gap structure above EF, where
has no effect on QCDW. The red dashed line has the same form of photoemission intensity is suppressed by the Fermi–Dirac
g for all bands, and is close to a straight line at 45° (grey dashed distribution23. The comparison between the experimental STS data
line). A true 45° line would indicate that QCDW ¼ qg, implying and the calculated RPA results shows that the particle–hole
that the CDW vector is completely determined by the electron– asymmetry arises naturally from the interplay between the
phonon coupling, with the electronic structure having no electronic structure and the momentum-dependent electron–
influence at all. The solid black line, which shows the behaviour phonon coupling.
of just the inner band, interpolates between these two extremes.
There is no plateau, but there is an indication of some flattening,
suggesting that while it is mainly the momentum dependence of The pseudogap. The central role of the orbital- and momentum-
the electron–phonon coupling that determines the value of dependent electron–phonon coupling in determining both the
CDW-ordering vector, the role of the electronic structure cannot CDW propagation vector and the structure of the gap suggests
be neglected entirely. that the charge order in NbSe2 falls in the strong-coupling regime,
in which the electronic structure alone is not sufficient to char-
acterize the emergent charge order. In the presence of strong
The CDW gap. Having identified the importance of both electron–phonon coupling, it is well known that the entropy
electron–phonon coupling and electronic structure, we now associated with localized lattice vibrations plays an essential part
revert to using the full expression of equation (2). To compare the in the system’s thermal evolution20. To understand the properties
size and shape of the CDW gap resulting from the orbital- and of NbSe2 above its CDW transition temperature, it is therefore
momentum-dependent g to the experimentally observed beha- necessary to consider the influence of phonon fluctuations
viour, the RPA gap equations shown diagrammatically in Fig. 3 beyond RPA. The so-called mode–mode coupling
are solved self-consistently. We simplify the calculation by using approximation (MMA) includes the lowest order terms of this
the 9  9 gap matrix which includes all higher harmonics type and has been shown to successfully describe the influence of
appropriate for a commensurate CDW wave vector at Q ¼ 23 GM, phonon fluctuations in some of the electronic properties of
as an approximation for the gap matrix of the incommensurate related dichalcogenide materials like TaSe2 and TiSe221,22,34. The
CDW in NbSe2 with QCDW ’ 0:98 23 GM. We further fix the additional terms included in the MMA fall into two categories: a
values of the electronic self-energies to 7 meV, approximating the vertex correction, which we neglect by appeal to Migdal’s
self-consistent result across k in the high-temperature phase, and theorem because of the large momentum transfer associated
constrain the momentum dependence of the gap function to a with QCDW, and self-energy corrections to the internal electron
five parameter tight-binding fit consistent with the symmetries of lines (indicated diagrammatically in Fig. 5), which we include.
the lattice. A self-consistent solution for the gap function to all In agreement with recent X-ray scattering observations on
orders in RPA can then be found by searching for a fixed point in NbSe27,19, the MMA-renormalized phonon spectrum shown in
the flow of consecutive iterations of the gap equation. Fig. 5b shows a broad range of phonon modes softening near the

0(k)
if Qi = 0 (k) if Qi = 0
k k+Qi 0 if Qi ≠ 0 k k+Qi (k,k+Qi) if Qi ≠ 0

Figure 3 | The RPA gap equation. Diagrammatic form of the self-consistent RPA gap equations. As indicated in the legend, the single arrows indicate bare
electronic Green’s functions G0 , and the double arrows the renormalized Green’s functions G and F. The anomalous functions Fðk; k þ Qi Þ ¼ hc ^y i
^ c
k k þ Qi
do not conserve crystal momentum, and correspond to the CDW order parameter. Notice that the momentum labels for the Green’s functions include
Matsubara frequencies, and the final diagram includes a summation over the anomalous momentum Qi of the internal electronic propagator.

4 NATURE COMMUNICATIONS | 6:7034 | DOI: 10.1038/ncomms8034 | www.nature.com/naturecommunications

& 2015 Macmillan Publishers Limited. All rights reserved.
NATURE COMMUNICATIONS | DOI: 10.1038/ncomms8034 ARTICLE

EF M K
10

Energy (meV)
DOS (a.u.)
0

−10
Γ
−20

−30 0 0
−80 80 M K
Energy (meV)

Figure 4 | Comparison to experimental results. (a) The calculated density of states for the bare dispersion (black line) and renormalized dispersion (red
line) compared with experimental results from STS measurements (blue points)15. The opening of the CDW gap leads to an asymmetric suppression of the
density of states, centred above EF. (b) The dispersion of the calculated spectral functions along MK, with the colour scale indicating intensity in arbitrary
units. The back-bending induced by the gap at EF is apparent in the band corresponding to the inner pocket about K. (c) Fermi energy cut of the spectral
function resulting from the self-consistent gap equations, showing the formation of Fermi arcs as a result of gaps opening along MK. Again the colour scale
indicates intensity in arbitrary units. Below the dashed line the bare bands are shown for comparison. Note that the outer band develops no discernible gap;
this can be compared directly with the results of recent ARPES measurements4,10.

the amplitude of the static order parameter vanishes, h^ jQ i ¼ 0,

T = TRPA T = TMMA
but
 dynamic fluctuations of the order parameter persist,
2
12 Bare 12
Bare h j
^ Q  i4020–22. The result is a locally fluctuating short-range
MMA
MMA
ordered state, characterized by the presence of a non-zero order
Energy (meV)

parameter amplitude, without any long-range phase coherence23.

Energy (meV)

8 8
Due to the non-zero amplitude, the gap in the electronic structure
will survive even above the transition temperature TCDW ¼ TMMA,
4
RPA
4 in the form of a pseudogap phase4,23. The dynamical fluctuations
in this regime may become locally pinned in the presence of
0
QCDW
0
QCDW defects, creating islands of order up to T ¼ TRPAE300 K,
0 q 1 0 q 1 in agreement with recent STM and X-ray scattering
experiments23,24. The properties of this regime are reminiscent
Figure 5 | Phonon dispersion in NbSe2. (a) The solid black line indicates
of the features that characterize the pseudogap phase of cuprate
the bare phonon dispersion as a function of momentum along GM, based
high-temperature superconductors35, which have recently been
on high-temperature X-ray scattering results19. Here q is defined through
shown typically to contain charge-ordered phases of their
q ¼ qGM, with qA[0,1]. The blue dotted curve shows the dispersion
own28–30. Specific similarities are the presence of a pseudogap,
resulting from the (mean field) RPA calculation at TRPA ¼ 303.5 K, while the
Fermi surface arcs, dynamical fluctuations and a corresponding
red dashed curve represents the MMA prediction at the same temperature
sensitivity to the timescales employed by experimental probes, as
(including the phonon fluctuations as shown in the diagrammatical
well as the existence of the crossover to the normal state at
representation of the renormalized phonon propagator). Both calculations
T* ¼ TRPA. Note that the existence of a crossover temperature
are based on the same tight-binding band structure. The fluctuations of the
T*E300 K, between the pseudogap and disordered regimes
phonon field can be seen to suppress the onset of charge order predicted by
predicted here, falls outside of the temperature range that has
the RPA theory. (b) The MMA dispersion at TMMA ¼ 33.5 K, which
been probed in NbSe2 to date4,24.
represents the true transition temperature. It can be seen that the
momentum of the softest phonon mode corresponds to the experimentally
observed CDW propagation vector. Discussion
Although various theories for the formation of charge order in
transition temperature, rather than the sharp Kohn anomaly NbSe2 have been able to explain many parts of its diverse
indicative of well-nested materials. The momentum of the softest experimental signatures, the material has so far evaded the
phonon mode at the transition temperature coincides with the simultaneous description of all of its properties by a single
experimentally observed CDW propagation vector. consistent theoretical account. The main features, which remain
The softening of phonon modes, as the charge-ordered state is difficult to reconcile with the weak-coupling theories considered
approached from high temperatures, is suppressed by the so far, include the emergence of CDW order in spite of a lack of
fluctuations beyond RPA. Since the temperature at which the nesting in the electronic structure, and an anomalously flat
first phonon mode reaches zero energy defines the transition electronic susceptibility, the softening of a broad range of phonon
temperature TCDW, this means that the charge-ordering transition modes rather than a sharp Kohn anomaly, a discrepancy in the
temperature in MMA, TMMA, is suppressed with respect to the sizes of the CDW gap observed in different electron pockets in
corresponding value in RPA, TRPA. Choosing the overall strength spite of them being similarly nested, disagreement about the
of the electron–phonon coupling such that TMMA matches the existence, size and particle–hole symmetry of the CDW gap as
experimentally observed value of TCDW ¼ 33.5 K, Fig. 5 shows seen by different experimental probes and the presence of a
that the corresponding RPA transition occurs already at pseudogap phase characterized by locally fluctuating charge order
TRPAE300 K. Physically, this suppression of the transition above the CDW transition temperature. We have shown here that
temperature can be ascribed to the presence of fluctuations in all of these observations can be consistently understood, within a
the phonon field. In the temperature range TMMAoToTRPA, the single theory, to result from the simultaneous influence of the
phonon fluctuations are strong enough to destroy the long-range presence of a strong electron–phonon coupling, its orbital
order predicted by (mean field) RPA theory. In that regime, dependence and its dependence on both incoming and outgoing

NATURE COMMUNICATIONS | 6:7034 | DOI: 10.1038/ncomms8034 | www.nature.com/naturecommunications 5

& 2015 Macmillan Publishers Limited. All rights reserved.
ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms8034

electron momentum. We have shown that none of these 10. Rahn, D. J. et al. Gaps and kinks in the electronic structure of the
ingredients can be neglected in the full description of superconductor 2H-NbSe2 from angle-resolved photoemission at 1 K. Phys.
NbSe25,21,22,34, suggesting their possible importance in other Rev. B 85, 224532 (2012).
11. Straub, T. et al. Charge-density-wave mechanism in 2H  NbSe2:
dichalcogenide materials as well.
photoemission results. Phys. Rev. Lett. 82, 4504–4507 (1999).
The combination of both the momentum-dependent coupling 12. Harper, J., Geballe, T. & Salvo, F. D. Heat capacity of 2H-NbSe2 at the charge
and the particular electronic structure of NbSe2 is found to be density wave transition. Phys. Lett. A 54, 27–28 (1975).
crucial in the selection of the charge-ordering vector, while the 13. Grüner, G. The dynamics of charge-density waves. Rev. Mod. Phys. 60,
orbital dependence naturally explains the different behaviours of 1129–1181 (1988).
14. Moncton, D. E., Axe, J. D. & DiSalvo, F. J. Study of superlattice formation in
different Fermi surface pockets. The same dependencies on
2H-NbSe2 and 2H-TaSe2 by neutron scattering. Phys. Rev. Lett. 34, 734–737
momentum and orbital character are also reflected in the (1975).
emergence of a particle–hole asymmetric gap, which has a size 15. Soumyanarayanan, A. et al. Quantum phase transition from triangular to stripe
of about 12 meV, centred above EF where it is largely inaccessible charge order in NbSe2. Proc. Natl Acad. Sci. USA 110, 1623–1627 (2013).
to observation by ARPES. These observations put NbSe2 in the 16. Rainer, D. & Sauls, J. A. inSuperconductivity: From Basic Physics to the Latest
Developments. (eds Butcher, P. N. & Lu, Y.) Ch. 2 45–78 (World Scientific, 1995).
regime of strongly coupled charge-ordered materials, in the sense
17. Wang, Y. et al. Weak-coupling d-wave BCS superconductivity and unpaired
that the specifics of the electron–phonon coupling qualitatively electrons in overdoped La2  xSrxCuO4 single crystals. Phys. Rev. B 76, 064512
affect the properties of the ordered state. (2007).
The pseudogap phase above TCDW may arise in such a scenario 18. Hashimoto, M., Vishik, I. M., He, R.-H., Devereaux, T. P. & Shen, Z.-X. Energy
from local fluctuations of the lattice, which destroy the long-range gaps in high-transition-temperature cuprate superconductors. Nat. Phys. 10,
483–495 (2014).
phase coherence of the CDW order parameter but not its
19. Weber, F. et al. Optical phonons and the soft mode in 2H-NbSe2. Phys. Rev. B
amplitude. Taking into account higher-order phonon fluctua- 87, 245111 (2013).
tions, we find that indeed the mean field transition temperature is 20. McMillan, W. L. Landau theory of charge-density waves in transition-metal
suppressed, leading to a regime above TCDW characterized by the dichalcogenides. Phys. Rev. B 12, 1187–1196 (1975).
absence of long-range charge order and the presence of phonon 21. Varma, C. M. & Simons, A. L. Strong-coupling theory of charge-density-wave
transitions. Phys. Rev. Lett. 51, 138–141 (1983).
fluctuations. This agrees with the experimental observation of
22. Yoshiyama, H., Takaoka, Y., Suzuki, N. & Motizuki, K. Effects on lattice
Fermi arcs and a pseudogap above the transition temperature, as fluctuations on the charge-density-wave transition in transition-metal
well as that of pinned charge order at high temperatures around dichalcogenides. J. Phys. C 19, 5591–5606 (1986).
local defects. All of these signatures of the pseudogap phase are 23. Chatterjee, U. et al. Emergence of coherence in the charge-density wave state of
reminiscent of similar features seen in the pseudogap phase of 2H-NbSe2. Nat. Commun. 6, 6313 (2015).
24. Arguello, C. J. et al. Visualizing the charge density wave transition in 2H-NbSe2
cuprate high-temperature superconductors.
in real space. Phys. Rev. B 89, 235115 (2014).
In conclusion, while the Peierls mechanism can explain charge 25. Emery, V. J., Kivelson, S. A. & Tranquada, J. M. Stripe phases in high-
ordering in 1D systems entirely in terms of the electronic temperature superconductors. Proc. Natl Acad. Sci. USA 96, 8814–8817 (1999).
structure, this is not the natural starting point for higher- 26. Wu, T. et al. Magnetic-field-induced charge-stripe order in the high-
dimensional materials. Instead, it is necessary to include full temperature superconductor YBa2Cu3Oy. Nature 477, 191–194 (2011).
information about the dependence of the electron–phonon 27. Walstedt, R. E., Mason, T. E., Aeppli, G., Hayden, S. M. & Mook, H. A.
Pseudogap term in the magnetic response of cuprate superconductors. Phys.
coupling on the momenta and orbital characters of the electronic Rev. B 84, 024530 (2011).
states involved in the CDW formation. This is evidenced by the 28. Chang, J. et al. Direct observation of competition between superconductivity
present case of NbSe2, a typical quasi-two-dimensional material and charge density wave order in YBa2Cu3O6.67. Nat. Phys. 8, 871–876 (2012).
without a strongly nested Fermi surface, in which knowledge 29. Ghiringhelli, G. et al. Long-range incommensurate charge fluctuations in
(Y,Nd)Ba2Cu3O(6 þ x). Science 337, 821–825 (2012).
about all aspects of the electron–phonon coupling proves to be
30. Wu, T. et al. Incipient charge order observed by NMR in the normal state of
essential in explaining the full range of experimental observations. YBa2Cu3Oy. preprint at http://arxiv.org/abs/1404.1617 (2014).
We suggest this consideration to be generically applicable across 31. Slater, J. C. & Koster, G. F. Simplified LCAO method for the periodic potential
charge-ordered quasi-two-dimensional and quasi-three-dimen- problem. Phys. Rev. 94, 1498–1524 (1954).
sional materials lacking strong nesting features. In particular, the 32. Doran, N. J., Titterington, D. J., Ricco, B. & Wexler, G. A tight binding fit to the
bandstructure of 2H-NbSe2 and NbS2. J. Phys. C 11, 685–698 (1978).
full momentum and orbital dependence of the electron–phonon
33. Varma, C. M., Blount, E. I., Vashishta, P. & Weber, W. Electron-phonon
coupling should be taken into account throughout the class of interactions in transition metals. Phys. Rev. B 19, 6130–6141 (1979).
transition metal dichalcogenides and related material classes, 34. Inglesfield, J. E. Bonding and phase transitions in transition metal
including high-temperature superconductors. dichalcogenide layer compounds. J. Phys. C 13, 17–36 (1980).
35. Timusk, T. & Statt, B. The pseudogap in high-temperature superconductors: an
experimental survey. Rep. Prog. Phys. 62, 61–122 (1999).
References
1. Wilson, J. A., Di Salvo, F. J. & Mahajan, S. Charge-density waves in metallic,
layered, transition-metal dichalcogenides. Phys. Rev. Lett. 32, 882–885 (1974).
2. Peierls, R. More Surprises in Theoretical Physics (Princeton University Press, Acknowledgements
1991). We thank M. R. Norman for suggesting the approach discussed here and for numerous
3. Doran, N. J. A calculation of the electronic response function in 2H-NbSe2 insightful discussions. J.v.W. acknowledges support from a VIDI grant financed by the
including electron-phonon matrix element effects. J. Phys. C 11, L959–L962 Netherlands Organisation for Scientific Research (NWO).
(1978).
4. Borisenko, S. V. et al. Two energy gaps and Fermi-surface "arcs" in NbSe2.
Phys. Rev. Lett. 102, 166402 (2009). Author contributions
5. Johannes, M. D., Mazin, I. I. & Howells, C. A. Fermi-surface nesting and the Both authors contributed equally to all aspects of the work.
origin of the charge-density wave in NbSe2. Phys. Rev. B73, 205102 (2006).
6. Rice, T. M. & Scott, G. K. New mechanism for a charge-density-wave
instability. Phys. Rev. Lett. 35, 120–123 (1975).
7. Weber, F. et al. Extended phonon collapse and the origin of the charge-density Additional information
wave in 2H-NbSe2. Phys. Rev. Lett. 107, 107403 (2011). Competing financial interests: The authors declare no competing financial interests.
8. Hess, H., Robinson, R. & Waszczak, J. STM spectroscopy of vortex cores and Reprints and permission information is available at http://npg.nature.com/
the flux lattice. Physica B 169, 422–431 (1991). reprintsandpermissions/
9. Wang, C., Giambattista, B., Slough, C. G., Coleman, R. V. & Subramanian, M.
A. Energy gaps measured by scanning tunneling microscopy. Phys. Rev. B 42, How to cite this article: Flicker F. et al. Charge order from orbital-dependent coupling
8890–8906 (1990). evidenced by NbSe2. Nat. Commun. 6:7034 doi: 10.1038/ncomms8034 (2015).

6 NATURE COMMUNICATIONS | 6:7034 | DOI: 10.1038/ncomms8034 | www.nature.com/naturecommunications

& 2015 Macmillan Publishers Limited. All rights reserved.