Title: Organic Electronic Materials: Competing Phases,Spatial Order, and Device Applications
1Organic Electronic Materials Competing
Phases,Spatial Order, and Device Applications
- David K. Campbell
- Understanding Complex Systems UIUC
- May 2003
- with R.T. Clay, S. Mazumdar, S. Ramasesha, A.
Sandvik, and P. Sengupta - Supported by NSF and NCSA
2Outline
- Motivation and Bottom Line on Organic
Electronics - The Past Conducting Polymers
- The Present Organic Charge Transfer Solids
- (CTS)
- The Future Organic Molecular Crystals and
- self-assembled structures
- Conclusions
3Motivation and Bottom Line
- Organic Electronic Materials (OEM)conducting
polymers, organic charge transfer salts, and
organic molecular crystalsbring the richness of
organic chemistry to solid state physics and
electronic device applications - Like high Tc superconductors, OEM typically
have strong e-e and e-ph interactions. Can hope
for a unified theoretical approach to whole class
of materials via Peierls-extended Hubbard (PEH)
models. - OEM exhibit a large range of theoretically
interesting phenomena from the kink solitons
and bipolarons of conducting polymers through the
excitonic effects of molecular crystals to CTS
ground states that exhibit superconductivity (SC)
and antiferromagnetism (SDW) (as seen in high Tc)
but also charge density wave (CDW), bond-order
wave (BOW), and spin-Peierls (SP)phases. - OEM vary continuously in effective electronic
dimensionality from 1D (conducting polymers and
some CTS) through 2D (other CTS) to 3D (typical
molecular crystal).
4Motivation and Bottom Line
- Todays talk provides capsule summary of the
past (conducting polymers), present (charge
transfer salts), and future (organic molecular
crystals) of the field, illustrating some
possible device applications and offering a few
specific examples from our theoretical studies,
based on PEH models, that - Describe the nonlinear excitationskink solitons,
polarons, bipolaronsin conducting polymers - Predict novel coexisting BCDW and BCSDW
density wave ground states in CTS - Provide an explanation of previously puzzling
recent experiments in several organic CTS,
including recent observations of charge order
(CO) - Explain differences among 1D, weakly 2D, and
strongly 2D CTS - Set stage for a possible approach to the observed
organic superconductivity
5The Past Conducting Polymers
- 2000 Nobel Prize in Chemistry awarded for 1976
discovery of conducting films of (bromine) doped
poly-acetylene, the first synthetic metal
6The Past Conducting Polymers
- Polyacetylene, the hydrogen atom of conducting
polymers to chemists (CHCH)n, a
dimerized/bond alternating ground state to
physicists (CH)x, a Bond Order Wave (BOW). By
either name, a broken symmetry phase. (CH)x
exists in two isomers -
- trans-(CH)x
cis-(CH)x
7The Past Conducting Polymers
- Degenerate ground state of trans-(CH)x gt kink
solitons - Non-degenerate ground state of cis-(CH)x (and
most other conducting polymers!) gt no kinks
possiblewhat then?
8The Past Conducting Polymers
- Twenty-five year history of theoretical modeling
in one slide observed properties of conducting
polymers require inclusion of both
electron-phonon (e-ph) and electron-electron
(e-e) interactions. In idealized one-dimensional
(1D) situation, use Peierls-extended Hubbard
(PEH) model. - HPEH H0 Hee
- H0 -S j, s t0 - a(Dj)Bj,j1,s , Ka/2 S
j(Dj)2 - Hee U S jnj,nj, V S i,njnj1
- Here j is a site index, Bj,k,s cj,s ck,s
ck, s cj, s, is the kinetic energy/hopping
operator, uj intersite distortion (phonon),
Dj (uj uj1). - Parameters t0 2.5 eV, a ? 3.8 eV/Å, Ka ?
47eV/Å2, U 10 eV, V 2.5eV Initial studies had
U0V (SSH model), allowed analytic treatment,
via continuum limit and mapping to relativistic
quantum field theory (!). - Data on optics, as well as comparison to
oligomers, showed U and V both non-zero gt
accurate many-body methods needed gt Quantum
Monte Carlo
9The Past Conducting Polymers
- Pictorial descriptions of kink solitons
10The Past Conducting Polymers
- Pictorial descriptions of
- polarons
bipolarons
11The Past Conducting Polymers
- Excitons in conventional semiconductors, often
have bound particle-hole pairs excitons. Here
the strong e-ph coupling creates accompanying
lattice distortion implies neutral P-P- or
BP-BP- - pairs. In singlet state, such pairs
are unstable against radiative decay gt
light-emission gt Polymeric LEDs !! - First constructed by R. Friend and collaborators
(Cambridge) in February 1989 using
poly-paraphenylene vinylene (PPV)
12The Past Conducting Polymers
- Polymeric LEDs are now being commercially
exploited as the basis for large format,
inexpensive, flexible, multicolor displays. Such
displays can be created using inkjet technology
to polymeric components gt potentially dirt
cheap and disposable - Polymer LED
Organic LED Color TV
13The Present Organic CTS
- Organic Chemistry offers rich variety of CTS
- Principal molecules are flat
- p-conjugated organics
- Form 11 salts (TTF-TCNQ)
- 12 anionic salts (MEM(TCNQ))2
- 21 cationic salts (TMTSF)2PF6
- From T. Ishiguro and K. Yamaji, Organic
Superconductors, p. 2 - (Springer Series in Solid State Sciences Volume
88 (1990).
14The PresentOrganic CTS
- Crystal structures are stacks of the flat
- organic molecules, counterions lying
- between in parallel segregated stacks.
- In 0th approximation, treat crystals as
- orthorhombic. Largest conductivity
- along stack, called a. For (TMTSF)2PF6,
-
- ta ' 0.25 eV, tb ' eV, tc ' 0.0015 eV,
- so conductivities are in ratio
-
- sasbsc 1 (1/100) (1/25000)
- After N. Thorup et al, Acta Crys. B37 1236-1240
(1981).
15The Present Organic CTS
- Electronic anisotropies of material are such that
they range from quasi-1D to quasi-2D. This is
reflected in the bulk crystals quasi-1D
materials form needles, quasi-2 form pancakes. - Photos courtesy of James Brooks, NHMFL
The needle-like crystal of TTF-TCNQ, has a
transition to a CDW at about 100 K.
The flat, slab-like crystal of the k-phase of
(BEDT-TTF)Cu(NCS)2, which is a 10.4 K
superconductor.
16CTSExperimental Overview
- Generalized phase diagram summarizes properties
seen in a wide range of 21 cationic CTS,
including (TMTZF)2X, Z(SuT,SeS), XBr,PF6, . - Key observation Pressure is known to enhance
t?, thereby increasing effective electronic
dimensionality. Recent results on uniaxial
pressure confirm directly. - M. Maesato et al, Phys. Rev. B64 155104 (2001).
- 1DCTS ? Quasi-1D/Weakly 2D ? 2D
- Fig 1. Phase Diagram for (TMTZF)2X compounds.
- CL Charge localized state. Lower case letters
- designate locations of specific compounds at
- atmospheric pressure a) ZT, XPF6
- b)ZDTDS, XPF6 c) ZT, XBr
- d) ZS, XPF6 e) ZS, XClO4.
-
- D. Jerome, Science 252, 1509 (1991).
17CTS Experimental Overview
- In 1D CTS (TMTTF)2X, experiments establish CO
- NMR-line splitting below TCO shows two different
site charges. - D.S. Chow et al, PRL 85, 1698 (2000).
18CTS Experimental Overview
- In 1D CTS (TMTTF)2X, experiments establish CO
- B. Dielectric permittivity
-
- Monceau, Nad, Brazovskii, PRL 86, 4080 (2001).
19CTS Experimental Overview
- In Quasi-1D/Weakly 2D (TMTSF)2PF6 and (TMTTF)2Br,
X-Ray experiments show - a mixed CDW-SDW modulation
- develops below TSDW. To our knowledge,
- such a modulated state has never been
- observed before
- J.P. Pouget and S. Ravy, Synth. Metals 85, 1523
(1997). - X-Ray satellite intensities from (TMTSF)2PF6
below TSDW. - Satellites 1 and 2 are 2kF reflections satellite
3 is the 4kF - CDW reflections.
20CTS Experimental Overview
- In strongly 2D CTS a-(BEDT-TTF)2MHg(SCN)4,
M(K,Rb,Tl), analysis of cspin, Hall coefficient,
mSR, NMR, and ADMRO, leads to conclusion - We arrive at the possibility that the
- ground state below TA.. (' 10K)
- could be SDW (or CDW)
- accompanied by CDW (or SDW).
- T. Sasaki and N. Toyota, Mysterious ground
states - in the organic conductor a-(BEDT-TTF)2MHg(SCN)4
- Mixed SDW and CDW? Synth. Metals 70 849 (1995).
- Fig. 2 Temperature dependence of zero-field
- resistance, Hall coefficient, and spin
susceptibility.
21CTS Experimental Overview
- In strongly 2D a-(BEDT-TTF)2MHg(SCN)4, 13C NMR
data show - the whole NMR results seen above are in favor
of the picture of a nonmagnetic transition such
as charge density wave (CDW). However, we have no
idea which kind of CDW reconciles the
susceptibility anisotropy observed below 12 K and
other magnetic properties at present. - K. Miyagawa, A. Kawamoto, and K. Kanoda, 13C NMR
study of nesting instability in a
a-(BEDT-TTF)2RbHg(SCN)4 Physi. Rev. B. 56, R8487
(1997).
22CTS Theoretical Models and Concepts
- Strongly correlated electrons low
dimensionality ) weak-coupling Fermi
surface-based arguments are suspect (as in high
Tc materials). Use discrete 2D anisotropic
lattice model Hamiltonian with e-e and e-ph
interactions. The - (quasi-2D) Peierls-extended
Hubbard (PEH) model - HPEH H0 Hee Hinter
- H0 -S j,M s t0 - a(Dj,M)Bj,j1,M, M,s b S
j,Mvj,Mnj,M - K1/2 S j,M(Dj,M)2 K2/2 S j,Mvj,M2
- HeeU S j,Mnj,M, nj,M, V S i,Mnj,Mnj1,M
- Hinter -t? S j,M,s Bj,j,M,M1, s
23CTS Theoretical Models and Concepts
- As before, in PEH j is a site index along given
chain, M is a chain index. Bj,k,L,M,s cj,L,s
ck,M,s ck,M,s cj,L,s, - uj,M intersite phonon, Dj,M (uj,M uj1,M),
vj,M intrasite phonon, present because large
molecules are deformable - Parameters e-e are U gtgt V gt t0, e-ph are a, b
- Critical implicit parameter Band-filling, r
Ne/NL, where NL is the number of sites, and Ne is
the number of electrons.
24CTS Theoretical Models and Concepts
- Broken symmetries in 1D PEH at ½-filling
- 2kF BOW, modulation of bj
- (Bj, j1, M, M, s )
- 2kF CDW, modulation of charge nj
- 2kF SDW, modulation of spin nj,
- nj,
- No coexistence of two DWs!!
25CTS Theoretical Models and Concepts
- Phase Diagram in the U, V gt 0 quadrant
- Schematic ground state phase diagram
- of the pure EHM at ½-filling, showing
- familiar CDW (LRO) and SDW (algebraic
- decay) phases, as well as recently
- confirmed BOW (LRO) phase near
- U ' 2V line at intermediate coupling.
- M. Nakamura, Phys. Soc Jpn 68, 3123 (1999).
- P. Sengupta, A. W. Sandvik, D.K. Campbell,
- Bond-order-wave phase and quantum phase
- transitions I the one dimensional extended
- Hubbard model, Phys. Rev. B 65, 155113/1-18
- (2002).
26CTS Theoretical Models and Concepts
- Broken Symmetries in the 1D PEH at ¼-filling
- Organic stacks in 12 CTS are ¼-filled electron
bands ((MEM1)(TCNQ-1/2)2) and in 21 CTS are
¼-filled hold bands ((TMTTF1/2)2PF6-1 )
¼-filling is very important! - B1) 2kF BOW1CDW1 for U V 0 and a, b
- ! 0-Familiar Peierls state
- B2) 4kF CDW for a, b ! 0, U gt V gt Vc,
- where Vc 2t Vc 1 for U/t0 !
- 1 Vc gt Vc1 for finite U/t0-Familiar
- 1010 Wigner crystal.
- B3) New coexisting BCDW state-(2kF4kF)
- BOW 2kF CDW for U ¹ 0,V lt Vc-note
- 1100.. charge order
- B4) New 4kF 2kF CDW 2kF BOW for U gt
- V gt Vc and a gt ac-novel SP variant of
- ..1010.. Wigner crystal.
27CTS Theoretical Models and Concepts
- NB
- B1 and B2 are well-known B3 and B4 are new
results, to be proven below. Key issue is which
is the ground state for given parameters. - B3 has 2 different site charges, 3 different
bonds B4 has 3 different site charge, 2
different bonds. - Question of Charge Order in 1D materials is not
clear is it ..1010.. or ..1100..?
28CTS Theory Confronts Experiments
- 1D CTS Key Questions
- Quasi-1D ¼-filled, segregated stack organic CTS
21 cationic CTS (e.g. (TMTTF)2X), ¼-filled
holes 12 anionic CTS (e.g., (MEM)(TCNQ)2,
¼-filled electrons) - KQO What is Charge Order (CO)?
- Inhomogenous distribution of charge along organic
stacks-some sort of density wave. Two models
proposed for CO - 1010 CO, obtained by Hartree-Fock, 4kF CDW,
nearest-neighbor V dominates. - H. Seo and H. Fukuyama, J. Phys. Soc. Jpn. 66,
1249 (1997). - 1100 CO, obtained in (numerically) exact
many-body study, BCDW, cooperative e-e and e-ph
effect. - Mazumdar, Clay, Campbell, PRB 62, 13400 (2000).
29CTS Theory Confronts Experiments
- 1D CTS Key Questions
- Quasi-1D ¼-filled, segregated stack organic CTS
21 cationic CTS (e.g. (TMTTF)2X), ¼-filled
holes 12 anionic CTS (e.g., (MEM)(TCNQ)2,
¼-filled electrons - KQ1 Hartree-Fock vs. Exact many-body? Does
..1010.. CO occur for realistic parameters in
exact calculation? - KQ2 Combined e-e and e-ph interactions? Is there
always CO? Always spin-Peierls? - KQ3 Experimental signature of the two CO states?
Strong evidence for the one or the other? - KQ4 Are expected temperature scales in either
picture consistent with experiment?
30CTS Theory Confronts Experiments
- Recall PEH model in 1D version with intrasite
phonons - HPEH H0 Hee
- H0 -S j, s t0 - a(Dj)Bj,j1,s b S jvjnj,
- Ka/2 S j(Dj)2 KB/2 S jvj2
- HeeU S jnj,nj, V S i,njnj1
- Here j is a site index, Bj,k,s cj,s ck,s ck,
s cj, s, uj intersite phonon, - Dj (uj uj1), vj intrasite phonon.
- Parameters e-e are U gtgt V gt t0, e-ph are a, b
- Estimated Values ????
- TMTTF t0 0.1 0.15 eV, U 1.0 eV, V 0.35eV
(??) a, b ¹ 0 - TMTSF t0 0.2 0.25 eV, U 1.0 eV, ) U/t0
smaller than in TMTTF, V 0.4 0.5 eV (??) a,
b ¹ 0 - F. Mila, Phys. Rev. B 52, 4788 (1995).
- Methods applied include Hartree-Fock, ED, and
QMC Results?
31CTS Theory Confronts Experiments
- AKQ1 Hartree-Fock vs. Exact many-body? Does
..1010.. CO occur for realistic parameters in
exact calculation? For U ! 1, one finds ..1010..
only for - V gt Vc gt 2t0.
- Phase diagram in
(U,V) plane for a, b ! 0
32CTS Theory Confronts Experiments
- AKQ1 Hartree-Fock vs. Exact many-body? Does
..1010.. CO occur for realistic parameters in
exact calculation? - Vc(U) determined by exact methods (strong
coupling, QMC, and ED) vs. HF - Shibata, Nishimoto, Ohta, PRB 64, 235107 (2001).
- Clay, Mazumdar, Campbell, cond-mat/0112278 PRB
67, 115121/1-9 (2003) - H. Seo and H. Fukuyama, J. Phys. Soc. Jpn. 66,
1249 (1997). - For Vc(U) HF gets both magnitude and trend with U
wrong. For realistic U and V, ..1010.. can be
reached over only narrow range of parameters.
From experimental parameters conclude TMTSF is
strongly in ..1100.. Region, TMTTF may be close
to 1100/1010 boundary.
33CTS Theory Confronts Experiments
- AKQ2 Combined e-e and e-ph interactions? Is
there always CO? Always spin-Peierls? - Exact many body studies for U gt 6t0, V lt Vc(U)
show ground state is novel coexisting BCDW with
..1100.. CO and WSWS bond orders. - Data for U 6t0, V t0 on a 16 site open chain
for a b 0. CO and SP/BOW coexist in a BCDW
(2kF 4kF) BOW 2kF CDW, which occurs for a,
b ! 0 - Mazumdar, Clay, Campbell, PRB 62, 13400 (2000) -
Fig3a.
34CTS Theory Confronts Experiments
- Can SP/BOW coexist with ..1010.. CO for V gt
Vc(U)? - No previous studies of this!
- We have recently show answer is yes provided a
gt ac gt 0 state is 4kF CDW-SP - ..1010.. (4kF CDW-SP) bonds SSWW
- NB in 1010.. SP, 0 sites no longer equivalent
) slightly different charge on 0 s ) 3
different site charges! - Clay, Mazumdar, Campbell, cond-mat/0203266, PRB
67, 115121/1-9 (2003) - Bottom Line bond distortion can coexist with
both 1100 and 1010 states but with different
patterns - For V lt Vc(U), unconditional BCDW (1100) CO
WSWS BOW, 2 sites, 3 bonds - For V gt Vc(U), conditional 4kF CDW-SP (1010)
CO SSWW BOW, 3 sites, 2 bonds
35CTS Theory Confronts Experiments
- Detailed Results for Phase diagram
- with phonons
- Self-consistent Lanczös
- for N 12 and 16 sites
- Dimensions coupling constants
- la a2 / Kat, lb b2 / Kbt
- Results for N 16, U 8t
- NB Undistorted and 4kF BOW
- phases are finite-size artifacts.
36CTS Theory Confronts Experiments
- Expected Phase diagram for N ! 1
- The expected phase diagram in the la - lb plane
- for (a) V lt Vc and (b) V lt Vc.
- Thus SP behavior at low T is consistent
- with either 1100 or 1010 CO, but with
- key differences
- BCDW state 1100 charge order,
- two site charges, three bond orders
- in WSWS pattern small Dn, strong
- bond distortion.
- 4kF CDW-SP state 1010 charge order
(actually, three site charges in SP distorted
state), two bond orders in SSWW pattern large
Dn, weak bond distortion.
37CTS Theory Confronts Experiments
- AKQ3 Experimental signature of the two CO
states? Strong evidence for the one or the other? - Site Charges?
- NMR data show two different local charge
environments below TCO dont distinguish clearly
between 1010 and 1100. - Spectroscopic Studies of (EDO-TTF)2PF6 show 0110
CO - O. Drozdova et al., Yamada Conf LVI, Abstract
Number S72 (2001).
38CTS Theory Confronts Experiments
- Bond Orders?
- For 1010, no direct observation of yet of SSWW
pattern. - For 1100, considerable evidence for WSWS
pattern. - 12 TCNQ bond distortions measured by neutron
scattering - Dimerization followed by tetramerization seen
directly. - MEM(TCNQ)2 R.J.J. Visser et al., PRB 28, 2074
(1983). - TEA(TCNQ)2 Direct evidence of both WSWS and
1100 CO from X-ray and neutron diffraction H.
Kobayashi et al. Acta. Cryst. B 26, 459, (1970)
J. P. Farges, J. Physique 46, 465, (1985).
39CTS Theory Confronts Experiments
- AKQ4 Are expected temperature scales in either
picture consistent with experiment? - In a word NO! Experiments find CO at
intermediate temperature - TSP lt TCO lt TMI
- 1010
- T gt TMI 1010, 0101 equal weight ) metal
- T lt TMI 1010, 0101 symmetry broken ) insulator
but immediately with CO - Expect 1010 to give TCO TMI
- 1100
- Above TMI 1100, 0011, 0110, 1001, all equal
weight ) metal - TSP lt T lt TMI 1010, 0101 equal weight )
insulator but no CO - T lt TSP 1100, 0011 symmetry broken ) CO first
appears - Expect 1100 to give TCO TSP
- Bottom line Temperature dependence difficult to
understand in purely 1D model-Recent results
suggest considering 2D effects may resolve issue.
40CTS Conclusions and Future Directions
- We have a unified approach to 1D, quasi-1D/weakly
2D, and 2D organic CTS. Results are based on
exact many-body studies (strong coupling, QMC,
ED) of Peierls-extended Hubbard that
incorporates e-e and e-ph interactions. Focusing
on the 1D CTS, we have answered four key
questions (KQ1-KQ4) and can summarize our results
as follows - The presence of 1100 CO in 1D systems for most of
the expected range of parameters U, V. HF
prediction of large region of 1010 CO is wrong. - 1100 CO in 1D systems is one aspect of novel
BCDW (coexisting density wave) ground state. - Experiments confirm existence of BCDW and 100 in
12 TCNQ salts. (TMTTF)2 may exhibit 1010 CO.
41The Future Organic Electronics
- Idea/Dream Build (self-assemble?) structures
(dots, layers, multi-layers, crystals?) with
active elements (molecules with 2,3,states,
intrinsic rectifying properties?) linked by
network (how to make connections?).
42The Future Organic Electronics
- Sample idealized molecular devices
43The Future Organic Electronics
- Optical Micrograph of Actual
Molecular Circuit, fabricated by
soft-imprinting and chemical assembly
techniques
44The Future Organic Electronics
- How are these for
possibilities? - Efficient organic photovoltaic diodes based on
doped pentacene - Superconductivity in molecular crystals induced
by charge injection - Gate-induced superconductivity in a solution
processed organic polymer film - Self-assembled monolayer organic field-effect
transistors
JH Schön, Ch Kloc, E. Buchler, B. Battlogg JH
Schön, Ch Kloc, B. Battlogg JH Schön, A
Dodabalapur, Z. Bao, Ch Kloc, O. Schenker B.
Battlogg JH Schön, H Meng, Z. Bao,
In the words of Wolfgang Pauli, Schön ist sie
schon, aber leider falsch!
45Conclusion
- Despite that serious setback, the
future of organic electronic materials is
bright as we heard yesterday from John Rogers
from fundamental theoretical issues (the nature
of CO and the mechanism of SC in CTS, the
temperature dependence of mobility in organic
molecular crystals, the mechanism for
conductivity in molecules like DNA) through
commercial applications (cheap plastic active
displays, flexible organic thin-film transistors,
self-assembling novel logic and memory circuits),
this important class of complex materials
promises to provide challenges and opportunities
to chemists, physicists, materials scientists and
engineers for decades to come.
46Our References
- K. C. Ung, S. Mazumdar, and D. Toussaint,
Metal-Insulator and Insulator-Insulator
Transitions in Quarter-Filled Band Organic
Conductors, Phys. Rev. Lett. 73, 2603 (1994). - S. Mazumdar, S. Ramasesha, R. T. Clay, and D.K.
Campbell, Theory of Coexisting Charge and
Soin-Density Waves in (TMTTF)2Br, (TMTSF)2PF6,
and a-(BEDT-TTF2MHg(SCN)4, Phys. Rev. Lett. 82,
1522-1525 (1999). - S. Mazumdar, D. K. Campbell, R. T. Clay, S.
Ramasesha Comment on Wigner Crystal Type of
Charge Ordering in an Organic Conductor with a
Quarter-Filled Band (DI-DCNQI)2Ag, Phys. Rev.
Lett. 82, 2411 (1999). - S. Mazumdar, R. T. Clay, and D. K. Campbell
Bond and charge density waves in the isotropic
interacting two-dimensional quarter-filled band
and the insulating state proximate to organic
superconductivity,'' Phys. Rev. B 62, 13400-13425
(2000). - R. T. Clay, S. Mazumdar, and D. K. Campbell
Re-Integerization of Fractional Charges in the
Correlated Quarter-Filled - Band,'' Phys. Rev. Lett. 86, 4084-4087 (2001).
- R. T. Clay, S. Mazumdar, and D. K. Campbell,
Charge ordering and spin gap transitions in
q-(BEDT-TTF)2X - materials,'' J. Phys. Soc. Jpn 71, 1816-1819
(2002). - R. T. Clay, S. Mazumdar, and D. K. Campbell,
The pattern of charge ordering in
quasi-one-dimensional organic charge transfer
solids,'' Phys. Rev. B 67, 115121/1-9 (2003).
47(No Transcript)
48Theory Confronts Experiments
- Answers to Key Questions
- KQ1 Hartree-Fock vs. Exact many-body? Does
..1010.. CO occur for realistic parameters in
exact calculation? - Vc(U) determined by exact methods (strong
coupling, QMC, and ED) vs. HF - Shibata, Nishimoto, Ohta, PRB 64, 235107 (2001).
- Clay, Mazumdar, Campbell, cond-mat/0112278
- H. Seo and H. Fukuyama, J. Phys. Soc. Jpn. 66,
1249 (1997). - For Vc(U) HF gets both magnitude and trend with U
wrong. For realistic U and V, ..1010.. Be reached
over only narrow range of parameters. From
experimental parameters conclude TMTSF is
strongly in ..1100.. Region, TMTTF may be close
to 1100/1010 boundary.
49Theory Confronts Experiments
- Answers to Key Questions
- KQ2 Combined e-e and e-ph interactions? Is there
always CO? Always spin-Peierls? - Exact many body studies for U gt 6t0, V lt Vc(U)
show ground state is novel coexisting BCDW with
..1100.. CO and WSWS bond orders. - Data for U 6t0, V t0 on a 16 site open chain
for a b 0. CO and SP/BOW coexist in a BCDW
(2kF 4kF) BOW 2kF CDW, which occurs for a,
b ! 0 - Mazumdar, Clay, Campbell, PRB 62, 13400 (2000) -
Fig3a.
50Theory Confronts Experiments
- Can SP/BOW coexist with ..1010.. CO for V gt
Vc(U)? - No previous studies of this!
- We have recently show answer is yes provided a
gt ac gt 0 state is 4kF CDW-SP - ..1010.. (4kF CDW-SP) bonds SSWW
- NB in 1010.. SP, 0 sites no longer equivalent
) slightly different charge on 0 s ) 3
different site charges! - Clay, Mazumdar, Campbell, cond-mat/0203266
- Bottom Line bond distortion can coexist with
both 1100 and 1010 states but with different
patterns - For V lt Vc(U), unconditional BCDW (1100) CO
WSWS BOW, 2 sites, 3 bonds - For V gt Vc(U), conditional 4kF CDW-SP (1010)
CO SSWW BOW, 3 sites, 2 bonds
51Theory Confronts Experiments
- Detailed Results for Phase diagram
- with phonons
- Self-consistent Lanczös
- for N 12 and 16 sites
- Dimensions coupling constants
- la a2 / Kat, lb b2 / Kbt
- Results for N 16, U 8t
- NB Undistorted and 4kF BOW
- phases are finite-size artifacts.
52Theory Confronts Experiments
- Expected Phase diagram for N ! 1
- The expected phase diagram in the la - lb plane
- for (a) V lt Vc and (b) V lt Vc.
- Thus SP behavior at low is consistent
- with either 1100 or 1010 CO, but with
- key differences
- BCDW state 1100 charge order,
- two site charges, three bond orders
- in WSWS pattern small Dn, strong
- bond distortion.
- 4kF CDW-SP state 1010 charge order
(actually, three site charges in SP distorted
state), two bond orders in SSWW pattern large
Dn, weak bond distortion.
53Theory Confronts Experiments
- Answers to Key Questions
- KQ3 Experimental signature of the two CO states?
Strong evidence for the one or the other? - Site Charges?
- NMR data show two different local charge
environments below TCO dont distinguish clearly
between 1010 and 1100. - Spectroscopic Studies of (EDO-TTF)2PF6 show 0110
CO - O. Drozdova et al., Yamada Conf LVI, Abstract
Number S72 (2001).
54Theory Confronts Experiments
- Bond Orders?
- For 1010, no direct observation of yet of SSWW
pattern. - For 1100, considerable evidence for WSWS
pattern. - 12 TCNQ bond distortions measured by neutron
scattering - Dimerization followed by tetramerization seen
directly. - MEM(TCNQ)2 R.J.J. Visser et al., PRB 28, 2074
(1983). - TEA(TCNQ)2 Direct evidence of both WSWS and
1100 CO from X-ray and neutron diffraction H.
Kobayashi et al. Acta. Cryst. B 26, 459, (1970)
J. P. Farges, J. Physique 46, 465, (1985).
55Theory Confronts Experiments
- Answers to Key Questions
- KQ4 Are expected temperature scales in either
picture consistent with experiment? - In a word NO! Experiments find CO at
intermediate temperature TSP lt TCO lt TMI - 1010
- T gt TMI 1010, 0101 equal weight ) metal
- T lt TMI 1010, 0101 symmetry broken ) insulator
but immediately with CO - Expect 1010 to give TCO TMI
- 1100
- Above TMI 1100, 0011, 0110, 1001, all equal
weight ) metal - TSP lt T lt TMI 1010, 0101 equal weight )
insulator but no CO - T lt TSP 1100, 0011 symmetry broken ) CO first
appears - Expect 1100 to give TCO TSP
- Bottom line Temperature dependence difficult to
understand in purely 1D model-recent results
(c.f. S. Mazumdar, Talk ThuI2) suggest
considering 2D effects may resolve issue.
56Theory Confronts Experiments
- Answers to Key Questions
- Quasi-1D/Weakly 2D CTS
- KQ5 What is the alignment of the transverse
direction? - With 1100 CO along high conductivity (x) axis
can have 4 distinct transverse orders. - Shift by 0 sites between chains ) 0 phase shift
- 110011001100
- 110011001100
- 110011001100
- (Insulating!) stripes in y-direction , 1100, CO
in x (and diagonals) - Shift by 1 site between chains ) p / 2 phase
shift - 001100110011
- 011001100110
- 110011001100
- 1100 CO in x and y (diagonals (insulating) stripe
and 1010)
57Theory Confronts Experiments
- Shift by 2 sites between chains ) p phase shift
- 110011001100
- 001100110011
- 110011001100
- 1100 CO in x, 1010 CO in y (diagonals
(insulation) stripes) - Shift by 3 site between chains ) 3p/2 phase shift
- 001100110011
- 100110011001
- 110011001100
- 1100 CO in x and y (diagonals (insulating) stripe
and 1010) - Results?
-
58Theory Confronts Experiments
- For typical parameters, answer is 3), p phase
shift and we find coexisting SDW, so state
becomes BCSDW! - QMC results for site charges and interchain
- spin-spin correlations on a 12(x)4 lattice, t?
0.2t0
Schematic structure of the Coexisting BOW-CDW-SDW
BCSDW
59Theory Confronts Experiments
- Answers to Key Questions
- 2D CTS Consider q-(BEDT-TFF)2X
- KQ6 Crystal/lattice structure? Experimental
results? - q-(BEDT-TTF) lattice
- Triangular lattice 6 neighbors
- Tc ltlt tp, but Vc Vp
- q-(BEDT-TFF)2I3 superconductor,
- other X are semiconductors.
- Experiments in non-superconduction q-(ET) also
show two transitions as function to temperature
60Theory Confronts Experiments
- (1) High temperature metal-insulator (MI)
transition - Seen in all q-(ET) at T 200K
- 2D ¼-filled no MI transition in conventional
band theory - The Fermi surface predicted by bandstructure
calculation is a simple closed cylinder.
Nevertheless, most of them undergo transitions
into insulators - K. Miyagawa, A. Kawamoto, K. Kanoda, PRB 62,
R7679 (2000). - NB c-direction dimerization is seen at TMI
61Theory Confronts Experiments
- Low temperature transition to spin-gap state
(TSG) - Data for q-(ET)2RbZn(SCN)4, TSG 50K
- H. Mori, S. Tanaka, T. Mori, PRB 57, 12023
(1998). - Spin gap not consistent with 1D spin-Peierls
transition
62Theory Confronts Experiments
- (3) Charge ordering (CO) immediately for T lt TMI
- Splitting of NMR line seen
- K. Miyagawa, A. Kawamoto, K. Kanoda, PRB 62 R7679
(2000) and others. - No previous theory described all three effects.
TCO TMI suggests 1010 CO in 2D. Acutlaly, much
more subtle as we now show
63Theory Confronts Experiments
- Answers to Key Questions
- KQ7 What are possible CO arrangements in 2D q-ET
lattice? - H. Seo, J. Phys. Cos. Jpn. 69, 805 (2000).
- Vertical Diagonal Horizontal
Converted to square lattice - Stripes
added diagonal - Notice horizontal stripe has 1100 CO with p/2
phase shift - Compare Stripe Energies within 2D EH model
- H -t0 åltijgts (cyiscjs cyjscis) Uåi ni"ni
V åltijgt ni nj - Parameters for q-ET tp 0.14 eV, tc 0.01
eV, U 0.7 eV - Consider range of V 0.15 lt V lt 0.35
64Theory Confronts Experiments
- Answers to Key Questions
- KQ8 Which stripe CO dominates, and what are its
characteristics? - Results are 16-site exact diagonalization
- We find that the horizontal stripe
- has lowest energy for large range
- of V. Previous theoretical work
- incorrectly predicts vertical stripe
- H. Seo, J. Phys. Soc. Jpn. 69, 805 (2000).
- vertical strip favored within HF
- approximation
- R.H. McKenzie et al., PRB 61, 085109 (2001)
- ignores Vc (chooses pure square lattice)
65Theory Confronts Experiments
- Other characteristics of horizontal stripe CO
- We find strong bond distortions coexisting with
CO. These are not present in vertical/diagonal
states. - Bond distortions will lower energy further when
phonons included self-consistently
66Theory Confronts Experiments
- Answers to Key Questions
- KQ9 How do CO and BOW coexist in different spin
states? -
- S0 state (T lt TSG) FM state dominates
high T - - Tetramerization along p ! spin gap -
Dimerization along p - - Dimerization along c - Dimerizaiton along c
67Theory Confronts Experiments
- Exact diagonalization results show
- Charge order and c-direction bond dimerization
present in ferromagnetic state (i.e. above
spin-gap transition) - Spin gap transition involves tetramerization of
p-direction bonds. - Above spin-gap transition 1-1 and 0-0 bonds
equivalent - Below spin-gap transition 1-1 and 0-0 bonds
different
68Conclusions and Future Directions
- We have presented a unified approach to 1D,
quasi-1D/weakly 2D, and 2D organic CTS. Results
are based on exact many-body studies (strong
coupling, QMC, ED) of Peierls-extended Hubbard
that incorporates e-e and e-ph interactions. We
have answered nine key questions (KQ1-KQ9) and
can summarize our results as follows - The presence of 1100 CO in 1D systems for most of
the expected range of parameters U, V. HF
prediction of large region of 1010 CO is wrong. - 1100 CO in 1D systems is one aspect of novel
BCDW (coexisting density wave) ground state. - Experiments confirm existence of BCDW and 100 in
12 TCNQ salts. (TMTTF)2 may exhibit 1010 CO. - In quasi-1D/weakly 2D CTS, theory predicts novel
BCSDW in which 1100 CO along chains coexists
with 1010 transverse order and an SDW. Consistent
with data in (TMTSF)2PF6. - In strongly 2D CTS (q -ET)2X, find 1100 order in
both x and y directions (recall unusual
lattice). Theory explains two transition and spin
gap.
69Conclusions and Future Directions
- Challenges for the future include
- Resolving the T-dependent behavior of (TMTTF)2X
why is TMI gt TCO gtTSP, TSDW? (2D effects?) - Understanding the role crystal behavior in
(ET)-materials differences among q, a and k
phases, etc. Subtleties of structure in other
CTS? - Calculate spin excitation spectra in q-ET
- Establish relation of insulating phases to
(adjacent?) superconducting phases? Occurrence of
cooperation, rather than competition, between e-e
and e-ph interactions in BCDW/BCSDW may be the
key to understanding unconventional SC in
organics. Related to ideas in high TC SC but no
obvious analog of doping.
70References
- K. C. Ung, S. Mazumdar, and D. Toussaint,
Metal-Insulator and Unsulator-Insulatr
Transitions in Quarter-Filled Band Organic
Conductors, Phsy. Rev. Lett. 73, 2603 (1994). - S. Mazumdar, S. Ramasesha, R. T. Clay, and D.K.
Campbell, Theory of Coexisting Charge and
Soin-Density Waves in (TMTTF)2Br, (TMTSF)2PF6,
and a-(BEDT-TTF2MHg(SCN)4, Phys. Rev. Lett. 82,
1522-1525 (1999). - S. Mazumdar, D. K. Campbell, R. T. Clay, S.
Ramasesha Comment on Wigner Crystal Type of
Charge Ordering in an Organic Conductor with a
Quarter-Filled Band (DI-DCNQI)2Ag, Phys. Rev.
Lett. 82, 2411 (1999). - S. Mazumdar, R. T. Clay, and D. K. Campbell
Bond and charge density waves in the isotropic
interacting two-dimensional quarter-filled band
and the insulating state proximate to organic
superconductivity,'' Phys. Rev. B 62, 13400-13425
(2000). - R. T. Clay, S. Mazumdar, and D. K. Campbell
Re-Integerization of Fractional Charges in the
Correlated Quarter-Filled - Band,'' Phys. Rev. Lett. 86, 4084-4087 (2001).
- R. T. Clay, S. Mazumdar, and D. K. Campbell,
Charge ordering and spin gap transitions in
q-(BEDT-TTF)2X - materials,'' J. Phys. Soc. Jpn., in press.
- R. T. Clay, S. Mazumdar, and D. K. Campbell,
The pattern of charge ordering in
quasi-one-dimensional organic charge transfer
solids,'' submitted to Phys. Rev. B.
71(No Transcript)
72Broken Symmetry States and Charge Order in
Organic Superconductors
- Understanding Complex Systems UIUC
- May 2003
- with R.T. Clay, S. Mazumdar, and S. Ramasesha
- Supported by NSF DMR97-12765, NSF GRT, NCSA and
ERATO, Japan Science and Technology Corp.
73Outline
- Motivation and Bottom Line
- Materials The Organic CTS
- Experimental Overview evidence for change order
(CO) and novel coexisting density waves - Theoretical Models and Concepts Peierls
extended Hubbard (PEH) models, broken symmetry,
charge order - Theory confronts Experiment 1D CTS ?
- Quasi-1D/weakly 2D CTS ? 2D CTS
- Conclusions and Future Directions
74Motivation and Bottom Line
- Organic Charge Transfer Solids (CTS) offer
richness of organic chemistry in novel electronic
systems - Like high Tc superconductors, CTS have strong
e-e and e-ph interactions - CTS exhibit superconductivity (SC) and
antiferromagnetism (SDW) (as seen in high Tc) but
also charge density wave (CDW), bond-order wave
(BOW), and spin-Peierls (SP)phases. - Vary continuously in effective electronic
dimensionality from 1D to 2D.
75Motivation and Bottom Line
- Our theoretical studies, based on
Peierlsextended Hubbard (PEH) models - Predict novel coexisting BCDW and BCSDW
density wave ground states - Provide an explanation of previously puzzling
recent experiment in several organic CTS,
including recent observations of charge order
(CO) - Explain differences among 1D, weakly 2D, and
strongly 2D CTS - Set stage for a possible approach to the observed
organic superconductivity
76Materials The Organic CTS
- Organic Chemistry offers rich variety of CTS
- Principal molecules are flat
- p-conjugated organics
- Form 11 salts (TTF-TCNQ)
- 12 anionic salts (MEM(TCNQ))2
- 21 cationic salts (TMTSF)2PF6
- From T. Ishiguro and K. Yamaji, Organic
Superconductors, p. 2 - (Springer Series in Solid State Sciences Volume
88 (1990).
77Materials The Organic CTS
- Crystal structures are stacks of the flat
- organic molecules, counterions lying
- between in parallel segregated stacks.
- In 0th approximation, treat crystals as
- orthorhombic. Largest conductivity
- along stack, called a. For (TMTSF)2PF6,
-
- ta ' 0.25 eV, tb ' eV, tc ' 0.0015 eV,
- so conductivities are in ratio
-
- sasbsc 1 (1/100) (1/25000)
- After N. Thorup et al, Acta Crys. B37 1236-1240
(1981).
78Materials The Organic CTS
- Electronic anisotropies of material are such that
they range from quasi-1D to quasi-2D. This is
reflected in the bulk crystals quasi-1D
materials form needles, quasi-2 form pancakes. - Photos courtesy of James Brooks, NHMFL
The needle-like crystal of TTF-TCNQ, has a
transition to a CDW at about 100 K.
The flat, slab-like crystal of the k-phase of
(BEDT-TTF)Cu(NCS)2, which is a 10.4 K
superconductor.
79Experimental Overview
- Generalized phase diagram summarizes properties
seen in a wide range of 21 cationic CTS,
including (TMTZF)2X, Z(S,Se), XBr,PF6, . - Key observation Pressure is known to enhance
t?, thereby increasing effective electronic
dimensionality. Recent results on uniaxial
pressure confirm directly. - M. Maesato et al, Phys. Rev. B64 155104 (2001).
- 1DCTS ? Quasi-1D/Weakly 2D ? 2D
- Fig 1. Phase Diagram for (TMTZF)2X compounds.
- CL Charge localized state. Lower case letters
- designate locations of specific compounds at
- atmospheric pressure a) ZT, XPF6
- b)ZDTDS, XPF6 c) ZT, XBr
- d) ZS, XPF6 e) ZS, XClO4.
-
- D. Jerome, Science 252, 1509 (1991).
80Experimental Overview
- In 1D CTS (TMTTF)2X, experiments establish CO
- NMR-line splitting below TCO shows two different
site charges. - D.S. Chow et al, PRL 85, 1698 (2000).
81Experimental Overview
- In 1D CTS (TMTTF)2X, experiments establish CO
- B. Dielectric permittivity
-
- Monceau, Nad, Brazovskii, PRL 86, 4080 (2001).
82Experimental Overview
- In Quasi-1D/Weakly 2D (TMTSF)2PF6 and (TMTTF)2Br,
X-Ray experiments show - a mixed CDW-SDW modulation
- develops below TSDW. To our knowledge,
- such a modulated state has never been
- observed before
- J.P. Pouget and S. Ravy, Synth. Metals 85, 1523
(1997). - X-Ray satellite intensities from (TMTSF)2PF6
below TSDW. - Satellites 1 and 2 are 2kF reflections satellite
3 is the 4kF - CDW reflections.
83Experimental Overview
- In strongly 2D CTS a-(BEDT-TTF)2MHg(SCN)4,
M(K,Rb,Tl), analysis of cspin, Hall coefficient,
mSR, NMR, and ADMRO, leads to conclusion - We arrive at the possibility that the
- ground state below TZ.. (' 10K)
- could be SDW (or CDW)
- accompanied by CDW (or SDW).
- T. Sasaki and N. Toyota, Mysterious ground
states - in the organic conductor a-(BEDT-TTF)2MHg(SCN)4
- Mixed SDW and CDW? Synth. Metals 70 849 (1995).
- Fig. 2 Temperature dependence of zero-field
- resistance, Hall coefficient, and spin
susceptibility.
84Experimental Overview
- In strongly 2D a-(BEDT-TTF)2MHg(SCN)4, 13C NMR
data show - the whole NMR results seen above are in favor
of the picture of a nonmagnetic transition such
as charge density wave (CDW). However, we have no
idea which kind of CDW reconciles the
susceptibility anisotropy observed below 12 K and
other magnetic properties at present. - K. Miyagawa, A. Kawamoto, and K. Kanoda, 13C NMR
study of nesting instability in a
a-(BEDT-TTF)2RbHg(SCN)4 Physi. Rev. B. 56, R8487
(1997).
85Theoretical Models and Concepts
- Strongly correlated electrons low
dimensionality ) weak-coupling Fermi
surface-based arguments are suspect (as in high
Tc materials). Use discrete 2D anisotropic
lattice model Hamiltonian with e-e and e-ph
interactions. - (2D) Peierls-extended Hubbard (PEH) model
- HPEH H0 Hee Hinter
- H0 -S j,M s t0 - a(Dj,M)Bj,j1,M, M,s b S
j,Mvj,Mnj,M - K1/2 S j,M(Dj,M)2 K2/2 S j,Mvj,M2
- HeeU S j,Mnj,M, nj,M, V S i,Mnj,Mnj1,M
- Hinter -t? S j,M,s Bj,j,M,M1, s
86Theoretical Models and Concepts
- Here j is a site index along given chain, M is a
chain index. Bj,k,L,M,s cj,L,s ck,M,s ck,M,s
cj,L,s, uj,M intersite phonon, vj,M
intrasite phonon, Dj,M (uj,M uj1,M). - Parameters e-e are U gtgt V gt t, e-ph are a, b
- Critical implicit parameter Band-filling r
Ne/NL, where NL is the number of sites, and Ne is
the number of electrons.
87Theoretical Models and Concepts
- Broken symmetries in 1D PEH at ½-filling
- 2kF BOW, modulation of bj
- (Bj, j1, M, M, s )
- 2kF CDW, modulation of charge nj
- 2kF SDW, modulation of spin nj,
- nj,
- No coexistence of two DWs!!
88Theoretical Models and Concepts
- Phase Diagram in the U, V gt 0 quadrant
- Schematic ground state phase diagram
- of the pure EHM at ½-filling, showing
- familiar CDW (LRO) and SCW (algebraic
- decay) phases, as well as recently
- confirmed BOW (LRO) phase near
- U ' 2V line at intermediate coupling.
- M. Nakamura, Phys. Soc Jpn 68, 3123 (1999).
- P. Sengupta, A. W. Sandvik, D.K. Campbell,
- Bond-order-wave phase and quantum phase
- transitions I the one dimensional extended
- Hubbard model, Phys. Rev. B 65, 155113/1-18
- (2002).
89Theoretical Models and Concepts
- Broken Symmetries in the 1D PEH at ¼-filling
- Organic stacks in 12 CTS are ¼-filled electron
bands ((MEM1)(TCNQ-1/2)2) and in 21 CTS are
¼-filled hold bands ((TMTTF1/2)2PF6-1 )
¼-filling is very important! - B1) 2kF BOW1CDW1 for U V 0 and a, b
- ! 0-Familiar Peierls state
- B2) 4kF CDW for a, b ! 0, U gt V gt Vc,
- where Vc 2t Vc 1 for U/t0 !
- 1 Vc gt Vc1 for finite U/t0-Familiar
- 1010 Wigner crystal.
- B3) New coexisting BCDW state-(2kF4kF)
- BOW 2kF CDW for U ¹ 0,V lt Vc-note
- 1100.. charge order
- B4) New 4kF 2kF CDW 2kF BOW for U gt
- V gt Vc and a gt ac-novel SP variant of
- ..1010.. Wigner crystal.
90Theoretical Models and Concepts
- NB
- B1 and B2 are well-known B3 and B4 are new
results, to be proven below. Key issue is which
is the ground state for given parameters. - B3 has 2 different site charges, 3 different
bonds B4 has 3 different site charge, 2
different bonds. - Question of Charge Order in 1D materials is not
clear is it ..1010.. or ..1100..?
91Theory Confronts Experiments
- 1D CTS Key Questions
- Quasi-1D ¼-filled, segregated stack organic CTS
21 cationic CTS (e.g. (TMTTF)2X), ¼-filled
holes 12 anionic CTS (e.g., (MEM)(TCNQ)2,
¼-filled electrons) - KQO What is Charge Order (CO)?
- Inhomogenous distribution of charge along organic
stacks-some sort of density wave. Two models
proposed for CO - 1010 CO, obtained by Hartree-Fock, 4kF CDW,
nearest-neighbor V dominates. - H. Seo and H. Fukuyama, J. Phys. Soc. Jpn. 66,
1249 (1997). - 1100 CO, obtained in (numerically) exact
many-body study, BCDW, cooperative e-e and e-ph
effect. - Mazumdar, Clay, Campbell, PRB 62, 13400 (2000).
92Theory Confronts Experiments
- 1D CTS Key Questions
- Quasi-1D ¼-filled, segregated stack organic CTS
21 cationic CTS (e.g. (TMTTF)2X), ¼-filled
holes 12 anionic CTS (e.g., (MEM)(TCNQ)2,
¼-filled electrons - KQ1 Hartree-Fock vs. Exact many-body? Does
..1010.. CO occur for realistic parameters in
exact calculation? - KQ2 Combined e-e and e-ph interactions? Is there
always CO? Always spin-Peierls? - KQ3 Experimental signature of the two CO states?
Strong evidence for the one or the other? - KQ4 Are expected temperature scales in either
picture consistent with experiment?
93Theory Confronts Experiments
- Recall PEH model in 1D Version
- HPEH H0 Hee
- H0 -S j, s t0 - a(Dj)Bj,j1,s b S jvjnj,
- Ka/2 S j(Dj)2 KB/2 S jvj2
- HeeU S jnj,nj, V S i,njnj1
- Here j is a site index, Bj,k,s cj,s ck,s ck,
s cj, s, uj intersite phonon, vj intrasite
phonon, Dj (uj uj1). - Parameters e-e are U gtgt V gt t0, e-ph are a, b
- Estimated Values ????
- TMTTF t0 0.1 0.15 eV, U 1.0 eV, V 0.35eV
(??) a, b ¹ 0 - TMTTF t0 0.2 0.25 eV, U 1.0 eV, ) U/t0
smaller than in TMTTF, V 0.4 0.5 eV (??) a,
b ¹ 0 - F. Mila, Phys. Re. B 52, 4788 (1995).
- Methods applied include Hartree-Fock, ED, and
QMC Results?
94Theory Confronts Experiments
- Answers to Key Questions
- KQ1 Hartree-Fock vs. Exact many-body? Does
..1010.. CO occur for realistic parameters in
exact calculation? For U ! 1, finds ..1010.. Only
for V gt Vc gt 2t0. - Phase diagram in (U,V) plane for a, b ! 0
95Theory Confronts Experiments
- Answers to Key Questions
- KQ4 Are expected temperature scales in either
picture consistent with experiment? - In a word NO! Experiments find CO at
intermediate temperature TSP lt TCO lt TMI - 1010
- T gt TMI 1010, 0101 equal weight ) metal
- T lt TMI 1010, 0101 symmetry broken ) insulator
but immediately with CO - Expect 1010 to give TCO TMI
- 1100
- Above TMI 1100, 0011, 0110, 1001, all equal
weight ) metal - TSP lt T lt TMI 1010, 0101 equal weight )
insulator but no CO - T lt TSP 1100, 0011 symmetry broken ) CO first
appears - Expect 1100 to give TCO TSP
- Bottom line Temperature dependence difficult to
understand in purely 1D model-recent results
(c.f. S. Mazumdar, Talk ThuI2) suggest
considering 2D effects may resolve issue.
96Theoretical Models and Concepts
- Here j is a site index along given chain, M is a
chain index. Bj,k,L,M,s cj,L,s ck,M,s ck,M,s
cj,L,s, uj,M intersite phonon, vj,M
intrasite phonon, Dj,M (uj,M uj1,M). - Parameters e-e are U gtgt V gt t, e-ph are a, b
- Critical implicit parameter Band-filling r
Ne/NL, where NL is the number of sites, and Ne is
the number of electrons.
97Theoretical Models and Concepts
- Here j is a site index along given chain, M is a
chain index. Bj,k,L,M,s cj,L,s ck,M,s ck,M,s
cj,L,s, uj,M intersite phonon, vj,M
intrasite phonon, Dj,M (uj,M uj1,M). - Parameters e-e are U gtgt V gt t, e-ph are a, b
- Critical implicit parameter Band-filling r
Ne/NL, where NL is the number of sites, and Ne is
the number of electrons.