Relationship between the parent charge transfer gap and maximum

transition temperature in cuprates

Wei Ruan1, Cheng Hu2, Jianfa Zhao2, Peng Cai1, Yingying Peng2, Cun Ye1, Runze Yu2,
Xintong Li1, Zhenqi Hao1, Changqing Jin2, 4, Xingjiang Zhou2,4, Zheng-Yu Weng3,4, Yayu
Wang1,4†

1
State Key Laboratory of Low Dimensional Quantum Physics, Department of Physics,
Tsinghua University, Beijing 100084, P. R. China
2
Beijing National Laboratory for Condensed Matter Physics, Institute of Physics, Chinese
Academy of Sciences, Beijing 100190, P. R. China.
3
Institute for Advanced Study, Tsinghua University, Beijing 100084, P. R. China.
4
Collaborative Innovation Center of Quantum Matter, Beijing, P. R. China.
†
Email: [email protected]

Abstract: One of the biggest puzzles concerning the cuprate high temperature
superconductors is what determines the maximum transition temperature (Tc,max), which
varies from less than 30 K to above 130 K in different compounds. Despite this dramatic
variation, a robust trend is that within each family, the double-layer compound always has
higher Tc,max than the single-layer counterpart. Here we use scanning tunneling microscopy to
investigate the electronic structure of four cuprate parent compounds belonging to two
different families. We find that within each family, the double layer compound has a much
smaller charge transfer gap size (∆CT), indicating a clear anticorrelation between ∆CT and
Tc,max. These results suggest that the charge transfer gap plays a key role in the
superconducting physics of cuprates, which shed important new light on the high Tc
mechanism from doped Mott insulator perspective.

Keywords: cuprates, Mott insulator, charge transfer gap, maximum transition temperature,
scanning tunneling microscopy

Introduction

Since the discovery of copper oxide (cuprate) high temperature superconductors about 30
years ago, understanding its microscopic origin has become one of the greatest challenges in
condensed matter physics. The problem is well-defined, i.e., to find the mechanism of Cooper
pairing for charge carriers residing in the two-dimensional CuO2 plane that is common to all
cuprates. However, after decades of experimental and theoretical investigations, very few
consensuses have been reached [1-5]. Figure 1A displays the schematic phase diagram of
hole-doped cuprates, in which only two well-defined phases are unambiguously determined.
First, the parent compound is an antiferromagnetic (AF) Mott insulator due to strong onsite
Coulomb repulsion, and the long-range AF order is suppressed by hole doping. Second, the
superconducting (SC) phase exists in a range of hole concentrations, and the Tc,max is reached
at optimal doping near the center of the dome. The definition and doping dependence of other
phases, such as the pseudogap phase [6] and charge orders in the underdoped regime [5], are
still under much debate.

One of the central issues about the cuprates is what determines the Tc,max of a specific
system, which varies dramatically from 25 K to 138 K in ambient pressure [7]. A highly
robust, but yet mysterious trend is that within each family the double-layer compound has
much higher Tc,max than the single-layer counterpart. Finding the physical parameter that
controls Tc,max is apparently a key step for solving the high Tc superconductivity puzzle. For
conventional BCS (Bardeen-Cooper-Schrieffer) superconductors, for example, a crucial
breakthrough for the final resolution of phonon-mediated mechanism is the discovery of the
isotope effect Tc ∼ m-1/2, where m is the isotopic mass [8, 9]. For the cuprates, there were
previous attempts to find the scaling relation between Tc and other physical parameters, such
as superfluid density [10] and metallic conductivity near Tc [11]. However, these are
macroscopic properties of the SC phase itself, which dictate the phase transition
phenomenology but provide little hint regarding the origin of Cooper pairing in the first place.
It is highly desirable to find a direct connection between Tc,max and the microscopic electronic
structure, especially that of the parent compound from where superconductivity emerges.
 The most relevant bands of the cuprates derive from the Cu 3𝑑 𝑥 2 −𝑦2 orbital and two O

2p orbitals in the CuO2 plane [12], as schematically drawn in Fig. 1B. Due to strong onsite
Coulomb repulsion and relatively small charge transfer energy, the half-filled parent
compound is a Mott insulator of the charge transfer type [13]. The hopping integral between
O and Cu sites, as well as the onsite and off-site Coulomb repulsions, may all affect the
behavior of doped charges hence Tc,max. Despite this complexity, we notice that the most
significant energy scale in the electronic structure of parent cuprate is the charge transfer gap
(CTG) between the upper Hubbard band (UHB) and the charge transfer band (CTB, or the
Zhang-Rice singlet band), which characterizes the energy needed for an O site electron to
migrate to the neighboring Cu sites [14]. If the doped Mott insulator picture is indeed the
valid theory for high Tc superconductivity, there might exist a direct link between the CTG
size ΔCT in the parent compound and Tc,max at the optimal doping. In a recent theoretical work,
the charge transfer energy εd - εp of various cuprates was calculated by first-principles
calculations and cluster dynamical mean-field theory [15], and was shown to anticorrelate
with Tc,max. However, so far the relationship between these two important quantities has not
been investigated experimentally.

In this paper we use scanning tunneling microscopy (STM) to investigate the CTG in
four parent cuprates belonging to the Can+1CunO2nCl2 and Bi2Sr2Can-1CunO2n+4 families
respectively. We found that within each family, the double-layer compound (n = 2) has much
smaller ΔCT than that of the single-layer (n = 1) compound. The anticorrelation between ∆CT
and Tc,max within each family suggests that the Mottness in parent cuprate plays a crucial role
in determining the SC properties upon charge doping. In particular, reducing the electron
correlation strength to the intermediate regime enhances superconductivity, which is
consistent with the pairing mechanism based on the doped Mott insulator picture.

Methods

For CCOC single crystals, polycrystalline samples are firstly synthesized by heating the
powder mixture of CaO and CuCl2 with a molar ratio of 2:1 at 1073 K for 24 h with
intermediate grindings. Then they are heated to 1203 K at a ramp rate of 60 K/h and are kept
there for 10 h. Finally, CCOC single crystals are obtained by cooling at a rate of 60 K/h down
to room temperature.

For 2-layer CCOC single crystals, the CaO, CuCl2 and CuO powders are firstly mixed
with a molar ratio of 3:1:1 and heated to 1223 K at a rate of 180 K/h and kept there for 24 h.
The polycrystalline samples are obtained by cooling down to room temperature at a rate of
240 K/h. The 2-layer CCOC single crystals are then obtained with a similar approach as the
CCOC, with the reaction temperature raised to 1373 K.

High quality Bi-2201 and Bi-2212 single crystals are grown by the traveling solvent
floating zone method [16, 17]. The typical size of the sample is about 2 × 2 × 0.1 mm3, which
can be obtained by cleaving the as-grown bulks. Various dopings can be obtained by
changing the chemical substitution and post-annealing under different temperatures and
atmospheres. For Bi-2212 samples, we obtain nearly insulating samples by annealing in
vacuum. All the samples are wrapped with Au foil during the annealing and quenched in
liquid N2 immediately after the annealing.

The details of the STM experiments were described elsewhere [18, 19]. All the STS
results reported here are obtained at liquid nitrogen temperature T = 77 K so that there is
finite tunneling current flowing through the highly insulating cuprate crystals. The samples
are precooled to 77 K and cleaved in the ultrahigh vacuum chamber with pressure ~ 10-11
mbar, and then transferred into the STM stage immediately. An electrochemically etched
tungsten tip is calibrated on a clean Au (111) surface before approaching the sample surface.
The differential conductance (dI/dV) curves are obtained by using standard ac lock-in
technique with a modulation frequency f = 423 Hz.

Experimental results

Figure 1C and 1D display the layered crystal structure of the four cuprates studied here.
The single-layer oxycloride Ca2CuO2Cl2 (CCOC) is a well-known Mott insulator, and the
Tc,max can reach 26 K by Na substation of Ca. Its double-layer counterpart is Ca3Cu2O4Cl2
(dubbed “2-layer CCOC” hereafter), in which the two CuO2 planes are separated by a Ca
layer, and Na doping leads to a Tc,max ~ 49 K [20]. Bi2(Sr,La)2CuO6+δ (Bi-2201) and
Bi2Sr2(Ca,Dy)Cu2O8+δ (Bi-2212) are the single-layer and double-layer compounds of the
widely-studied Bi family cuprates with Tc,max around 38 K and 95 K, respectively. There are
two reasons why we choose to study these four cuprates. First, they can be cleaved easily
between two van der Waals bonded neighboring layers, as indicated by the blue planes in Fig.
1C and 1D, resulting in ideal surfaces for STM studies. Second, for these systems high
quality single crystals with doping at or close to the parent Mott insulator phase are available.
For CCOC and 2-layer CCOC, pristine parent compounds can be grown by the flux method,
and the former one has been investigated extensively as a prototypical Mott insulator [18, 21,
22]. For Bi-2201, La replacement of Sr introduces electron doping that reduces the hole
density to the highly insulating regime close to the parent phase [16, 23]. For Bi-2212, Dy
substitution of Ca and further annealing in vacuum can also make it close to the parent Mott
insulator limit [17, 24].

Shown in Fig. 2A inset is the atomically resolved topographic image of CCOC taken
with bias voltage V = -3 V and tunneling current I = 8 pA, in which the exposed Cl atoms
form a square lattice. On the CCOC surface, the electron density of states (DOS) is highly
uniform in space, as reported previously [18]. Figure 2A displays a spatially-averaged
differential conductance (dI/dV) curve measured on a defect-free area of the CCOC surface.
The most prominent feature of the spectrum is a well-defined energy gap around the Fermi
level (EF), which is the CTG of this parent Mott insulator. To give a quantitative analysis of
the gap size, we take the logarithm of dI/dV signal as shown in Fig. 2B, which is a common
practice for extracting insulating energy gap from dI/dV spectroscopy [25, 26]. The two gap
edges show linear behavior in the logarithmic plot, which gives a more accurate fit for the
exponentially decayed gap edges. By using the intersection of the fitted dashed blue lines
illustrated in Fig. 2B, the CTG size of CCOC is determined to be ΔCT = 2.0 eV. This value is
also in quantitative agreement with that of a single CuO2 layer grown on Bi-2212 substrate by
molecular beam epitaxy technique [27].

We then performed the same measurements on the 2-layer CCOC Mott insulator. The
topographic image (taken with V = -3 V and I = 1 pA ) in Fig. 2C inset also shows a square
lattice of the surface Cl layer. Figure 2C displays a spatially-averaged dI/dV curve measured
on a defect-free area of 2-layer CCOC surface. The CTG size is extracted to be ΔCT = 1.4 eV,
significantly smaller than that of single-layer CCOC. In Fig. 2D we put the dI/dV curves of
the two samples together, which reveal striking similarities between them. They both show a
well-defined energy gap with steep gap edges. The EF lies close to the occupied states, or the
CTB, and is much further away from the unoccupied UHB. The negative half (occupied
states) of the two curves are quite similar, but the positive half shows a more pronounced
shift.

Figure 3A and 3B insets are the atomically resolved topographic images of Bi-2201 and
Bi-2212 scanned using bias voltages V = -1.2 V and -1.8 V, respectively. The exposed BiO
planes of both samples show stripe-like structural super-modulations. In addition to the
surface Bi-site vacancies, there are also bright and dark patches in both samples due to the
chemical and electronic inhomogeneities that are ubiquitous in underdoped cuprates [28, 29].
Because of the spatial inhomogeneities in these samples, the dI/dV curves exhibit more
complicated behavior. On the Bi-2201 surface (Fig. 3A), the red dI/dV curve is taken on the
spot marked by the red dot in the inset, which shows the CTG feature similar to those in
CCOC and 2-layer CCOC. The blue curve is taken on the spot marked by the blue dot in the
inset, which shows a broad electronic state within the CTG induced by local hole doping. The
dI/dV curves on the Bi-2212 surface (Fig. 3B) are very similar to those of Bi-2201, with some
locations showing the CTG feature (red curve) and some others showing the broad in-gap
states (blue curve). The evolution of the in-gap states and their spatial variations are highly
interesting in their own rights [30], but are not the focus of the current work. Plotting the
CTG-type dI/dV curves of the two samples together (Fig. 3C), we find that the CTB changes
little but the UHB shows a pronounced shift, in much the same way as that in Fig. 2D. The
CTG size also decreases significantly from ΔCT = 1.5 eV in Bi-2201 to ΔCT = 1.0 eV in
Bi-2212.

The results shown above clearly demonstrate that within each cuprate family, the
double-layer compound has much smaller CTG size than their single-layer counterpart. In Fig.
4 we summarize the ΔCT value of the four cuprates studied here as a function of their
respective Tc,max. There is an apparent anticorrelation between the two quantities within each
family: the smaller the ΔCT in the parent Mott insulator, the higher the Tc,max at the optimal
doping. The overall trend is consistent with the theoretically predicted anticorrelation
between the charge transfer energy εd - εp and Tc,max [15].

Discussion and conclusions

The experimental trend observed here implies that the Tc,max of a cuprate at optimal
doping is related to the CTG size in its parent compound. This is very surprising because the
latter is a much bigger characteristic energy as compared to the former, and the two regions
are well-separated in the phase diagram. We propose that the anticorrelation between ΔCT and
Tc,max actually sheds important new lights on the mechanism of superconductivity in the
cuprates. From the doped Mott insulator perspective, the superexchange coupling J between
local moments is responsible for the spin singlet pairing [31]. In the three-band model, J has a
complicated expression as a function of the bare charge transfer energy εd - εp and bare
Hubbard U [1, 14]. Our observation suggest the CTG size is the most significant single
parameter that determines the SC properties upon doping. In the scenario of single-band
Hubbard model (Fig. 1B), we may define an effective superexchange Jeff ~ 4teff2/ΔCT, where
teff is an effective hopping term characterizing the charge transfer process. Therefore the
smaller the ΔCT, the larger the Jeff and hence the stronger the pairing strength. In Fig. 4 inset
we plot Jeff ~ 1/ΔCT as a function of Tc,max, which has an approximately linear behavior (here
we ignore the variation of teff for different cuprates). From the doped Mott insulator
perspective, the CTG is necessary because it protects the local moments by putting a large
energy penalty on double occupancy of electrons, which is the origin of strong correlation.
However, if the CTG size is too large, it will weaken the pairing strength hence lowering the
Tc,max that can be achieved. We note that in Bi-2212, the highest Tc cuprate studied here, the
CTG size is merely 1.0 eV, putting it in the intermediately correlated regime.

The next question is what factors determine the CTG size in the parent compound, which
varies dramatically in the four samples studied here. Because the structure of a single CuO2
plane is nearly the same in all cuprates, it is quite obvious that the out-of-plane environment
is mainly responsible for the CTG size variations. As has been shown by first-principles
calculations, a key factor is the negatively charged apical ions [32]. Unlike the single-layer
compound, which has two apical ions locating symmetrically outside the CuO2 plane,
double-layer compound only has one apical ion outside each CuO2 plane, thus renders
smaller ΔCT. This scenario explains the experimental observation that within each family,
double-layer compound always has much smaller ΔCT than the single-layer one. Another
experimental trend is that the CCOC compound always has a larger ΔCT than the Bi-family
with the same number of CuO2 plane. The main difference between the two families is the
change of apical ion from Cl to O, as well as the distance between the apical ion and the
CuO2 plane. It is not straightforward why the former situation leads to a larger ΔCT than the
latter, but apparently it is consistent with the first-principles calculations based on the realistic
crystal structures [15].

The main message revealed by the observed relationship between ΔCT and Tc,max is that
the fate of a particular cuprate family is encoded in its parent compound. This conclusion
highlights the importance of Mottness in cuprate high temperature superconductors. Our
experiments indicate that the parent CTG size is the most relevant energy scale for the origin
of superconductivity. More importantly, ΔCT can be measured accurately by STM experiment,
thus providing a stringent constraint on available theories. Moreover, the sensitivity of ΔCT to
the out-of-plane environment suggests possible route to control ΔCT, and consequently Tc,max
[33-35]. Strategy for enhancing superconductivity along this direction has been proposed
theoretically [32], and is awaiting experimental confirmation.

Acknowledgement

This work is supported by NSFC and MOST of China. X. Z. thanks financial support
from the Strategic Priority Research Program (B) of the CAS (Grant No. XDB07020300).
References:

[1] P. A. Lee, N. Nagaosa, X.-G. Wen (2006) Doping a Mott insulator: Physics of high-temperature
superconductivity. Rev. Mod. Phys. 78:17-85
[2] P. W. Anderson, in Princeton Series in Physics (Princeton University Press, Princeton, New Jersey 08540,
1997).
[3] A. Damascelli, Z. Hussain, Z.-X. Shen (2003) Angle-resolved photoemission studies of the cuprate
superconductors. Rev. Mod. Phys. 75:473-541
[4] N. P. Armitage, P. Fournier, R. L. Greene (2010) Progress and perspectives on electron-doped cuprates.
Rev. Mod. Phys. 82:2421-2487
[5] E. Fradkin, S. A. Kivelson, J. M. Tranquada (2015) Colloquium: Theory of intertwined orders in high
temperature superconductors. Rev. Mod. Phys. 87:457-482
[6] T. Timusk, B. Statt (1999) The pseudogap in high-temperature superconductors: an experimental survey.
Rep. Prog. Phys. 62:61-122
[7] H. Eisaki, N. Kaneko, D. L. Feng et al (2004) Effect of chemical inhomogeneity in bismuth-based copper
oxide superconductors. Phys. Rev. B 69:064512
[8] C. A. Reynolds, B. Serin, W. H. Wright et al (1950) Superconductivity of Isotopes of Mercury. Phys. Rev.
78:487-487
[9] E. Maxwell (1950) Isotope Effect in the Superconductivity of Mercury. Phys. Rev. 78:477-477
[10] Y. J. Uemura, G. M. Luke, B. J. Sternlieb et al (1989) Universal correlations between Tc and ns/m* (carrier
density over effective mass) in high-Tc cuprate superconductors. Phys. Rev. Lett. 62:2317-2320
[11] C. C. Homes, S. V. Dordevic, M. Strongin et al (2004) A universal scaling relation in high-temperature
superconductors. Nature 430:539-541
[12] V. J. Emery (1987) Theory of high-Tc superconductivity in oxides. Phys. Rev. Lett. 58:2794-2797
[13] J. Zaanen, G. A. Sawatzky, J. W. Allen (1985) Band gaps and electronic structure of transition-metal
compounds. Phys. Rev. Lett. 55:418-421
[14] F. C. Zhang, T. M. Rice (1988) Effective Hamiltonian for the superconducting Cu oxides. Phys. Rev. B
37:3759-3761
[15] C. Weber, C. Yee, K. Haule et al (2012) Scaling of the transition temperature of hole-doped cuprate
superconductors with the charge-transfer energy. Europhys. Lett. 100:37001
[16] J. Meng, G. Liu, W. Zhang et al (2009) Growth, characterization and physical properties of high-quality
large single crystals of Bi2(Sr2−xLax)CuO6+δ high-temperature superconductors. Supercond. Sci. Technol.
22:045010
[17] X. F. Sun, S. Ono, X. Zhao et al (2008) Doping dependence of phonon and quasiparticle heat transport of
pure and Dy-doped Bi2Sr2CaCu2O8+δ single crystals. Phys. Rev. B 77:094515
[18] C. Ye, P. Cai, R. Yu et al (2013) Visualizing the atomic-scale electronic structure of the Ca2CuO2Cl2 Mott
insulator. Nat. Commun. 4:1365
[19] W. Ruan, P. Tang, A. Fang et al (2015) Structural phase transition and electronic structure evolution in
Ir1−xPtxTe2 studied by scanning tunneling microscopy. Science Bulletin 60:798-805
[20] Y. Zenitani, K. Inari, S. Sahoda et al (1995) Superconductivity in (Ca, Na)2CaCu2O4C12, The new simplest
double-layer cuprate with apical chlorine. Physica C 248:167-170
[21] J. J. M. Pothuizen, R. Eder, N. T. Hien et al (1997) Single Hole Dynamics in the CuO2 Plane at Half
Filling. Phys. Rev. Lett. 78:717-720
[22] F. Ronning, C. Kim, D. L. Feng et al (1998) Photoemission Evidence for a Remnant Fermi Surface and a
d-Wave-Like Dispersion in Insulating Ca2CuO2Cl2. Science 282:2067-2072
[23] Y. Peng, J. Meng, D. Mou et al (2013) Disappearance of nodal gap across the insulator-superconductor
transition in a copper-oxide superconductor. Nat. Commun. 4:2459
[24] Y. Zhang, C. Hu, Y. Hu et al (2016) In situ carrier tuning in high temperature superconductor
Bi2Sr2CaCu2O8+δ by potassium deposition. Science Bulletin 61:1037-1043
[25] R. M. Feenstra (1994) Tunneling spectroscopy of the (110) surface of direct-gap III-V semiconductors.
Phys. Rev. B 50:4561-4570
[26] R. M. Feenstra, J. A. Stroscio, edited by W. J. Kaiser and J. A. Stroscio (Academic Press, Inc., San Diego,
1993).
[27] Y. Zhong, Y. Wang, S. Han et al (2016) Nodeless pairing in superconducting copper-oxide monolayer
films on Bi2Sr2CaCu2O8+δ. Science Bulletin 61:1239-1247
[28] S. H. Pan, J. P. O'Neal, R. L. Badzey et al (2001) Microscopic electronic inhomogeneity in the high-Tc
superconductor Bi2Sr2CaCu2O8+x. Nature 413:282-285
[29] K. M. Lang, V. Madhavan, J. E. Hoffman et al (2002) Imaging the granular structure of high-Tc
superconductivity in underdoped Bi2Sr2CaCu2O8+δ. Nature 415:412-416
[30] P. Cai, W. Ruan, Y. Peng et al (2015) Visualizing the evolution from the Mott insulator to a charge ordered
insulator in lightly doped cuprates. Nature. Phy., DOI: 10.1038/NPHYS3840.
[31] P. W. Anderson (1987) The Resonating Valence Bond State in La2CuO4 and Superconductivity. Science
235:1196-1198
[32] C.-H. Yee, G. Kotliar (2014) Tuning the charge-transfer energy in hole-doped cuprates. Phys. Rev. B
89:094517
[33] I. Yamada, A. A. Belik, M. Azuma et al (2005) Single-layer oxychloride superconductor Ca2−xCuO2Cl2
with A-site cation deficiency. Phys. Rev. B 72:224503
[34] C.-Q. Jin, X.-J. Wu, P. Laffez et al (1995) Superconductivity at 80 K in (Sr,Ca)3Cu2O4+δCl2-y induced by
apical oxygen doping. Nature 375:301-303
[35] H. Hobou, S. Ishida, K. Fujita et al (2009) Enhancement of the superconducting critical temperature in
Bi2Sr2CaCu2O8+δ by controlling disorder outside CuO2 planes. Phys. Rev. B 79:064507
 Figure Captions:

Fig. 1. Phase diagram of hole-doped cuprates and crystal structure of the four materials

studied here. (A) Schematic electronic phase diagram of hole-doped cuprates, showing the

parent Mott insulator, the AF phase and the SC dome. The maximum SC transition

temperature is marked by the orange dot. (B) Schematic band structure of parent cuprate. The

charge transfer gap ΔCT is the energy distance between the UHB and the CTB. (C) The crystal

structure of Ca2CuO2Cl2 (CCOC) and Ca3Cu2O4Cl2 (2-layer CCOC), respectively. (D) The

crystal structure of Bi2(Sr,La)2CuO6+δ (Bi-2201) and Bi2Sr2(Ca,Dy)Cu2O8+δ (Bi-2212),

respectively.

Fig. 2. STM/STS results on pristine CCOC and 2-layer CCOC. (A) dI/dV spectrum

measured on the pristine CCOC surface, showing the well-defined charge transfer gap. The

inset shows the topographic image where dI/dV is taken. (B) The logarithm plot of the dI/dV

curve of CCOC, showing how the CTG is extracted. The gap edges and the floor are fitted

linearly (blue dashed lines), and the value of CTG is taken as the energy difference between

the two intersecting points. (C) dI/dV spectrum showing similar CTG measured on 2-layer

CCOC surface as shown in the inset topographic image. (D) Comparing the CTG in the two

materials, the occupied band edge shows little change, while the unoccupied band edge

displays more pronounced shift. The gap size shrinks significantly from CCOC to 2-layer

CCOC.

Fig. 3. STM/STS results on Bi-2201 and Bi-2212. (A) and (B) dI/dV spectra taken on

Bi-2201 and Bi-2212, respectively. The insulating charge transfer gap (red) and broad in-gap

states (blue) features show up at locations marked by the colored dots as shown in the insets,

which are atomically resolved topographic images of Bi-2201 and Bi-2212. (C) The

insulating dI/dV spectra of both Bi-2201 and Bi-2212 plotted in the same figure. The CTG
size of Bi-2212 is significantly smaller than that of Bi-2201.

Fig. 4. Anticorrelation between the CTG and Tc,max. The measured charge transfer gap

sizes in the above four materials as a function of their respective Tc,max. The experimental data

demonstrates an anticorrelation between the CTG size and Tc,max for each family, which is

illustrated by the two dashed lines. The inset shows the effective superexchange Jeff ~ 1/ΔCT

plotted as a function of Tc,max, assuming the same teff for different materials.
Figure 1
Figure 2
Figure 3
Figure 4