COMMENTS
Comments are short papers which criticize or correct papers of other authors previously published in Physical Review B. Each
Comment should state clearly to which paper it refers and must be accompanied by a brief abstract. The same publication schedule as
for regular articles is followed, and page proofs are sent to authors.
Comment on ‘‘Mechanism of the electric-field effect in the high-T
c
cuprates’’
N. Chandrasekhar
Department of Physics, Indian Institute of Science, Bangalore 560012, India
Oriol T. Valls and A. M. Goldman
School of Physics and Astronomy, University of Minnesota, Minneapolis, Minnesota 55455
Received 14 December 1995
The electric-field effect in the high-T
c
cuprates has been claimed to arise from the field-driven mobility of
free charges in the superconductor, a conclusion based on the observation of the field effect in
Bi
2
Sr
2
CaCu
2
O
8+y
at one sign of the bias. We show that this conclusion is unwarranted.
The electric-field effect in high-T
c
cuprates is a matter of
considerable scientific and technological importance. It has
sparked a certain amount of controversy. Two different, al-
though not necessarily mutually exclusive, mechanisms have
been proposed, based on early experiments restricted to in-
vestigations of the electric-field effect in
1
YBa
2
Cu
3
O
7 -x
.
The first is the conventional mechanism, and has its origin in
the Coulomb interaction of an applied field with free
charges.
1
This is fundamentally electronic, characterized by a
fast time constant, and results in the enhancement or deple-
tion of the number of carriers in a surface channel. This
should be a symmetric effect: changes in the number of car-
riers in the surface channel, and consequently changes in the
resistance observed, are equal in magnitude but not in sign
for a certain value of the bias, whether positive or negative.
The second explanation is based on the basal plane oxy-
gen diffusion process,
2
driven by the oxygen electric dipole
moments interacting with the applied external field. This pro-
cess is characterized by a slower time constant and, at least
in YBa
2
Cu
3
O
7 -x
, by asymmetry in the magnitude of the
effect: the changes in resistance and carrier concentration are
unequal in magnitude at positive and negative bias of iden-
tical absolute value. There are experimental data supporting
this mechanism in studies performed in geometries identical
to that discussed above.
3
The predicted asymmetry
2
has been
observed,
1,3
and is also found to depend on oxygen content.
The general picture of the basal plane oxygen diffusion pro-
cess is also supported by experimental and theoretical work
on persistent photoconductivity.
4,5
In their recent paper
6
Frey et al. claim to resolve the con-
troversy about the mechanism responsible by observing the
field effect in Bi
2
Sr
2
CaCu
2
O
8 +y
. They argued that, be-
cause they observed a field effect in this compound, which
lacks the CuO chains present in YBa
2
Cu
3
O
7 -x
, an expla-
nation in terms of the oxygen diffusion mechanism
2
was
ruled out.
This claim is erroneous. The presence of chains in
YBa
2
Cu
3
O
7 -x
makes it very convenient to develop a
simple model to describe the oxygen diffusion process in this
compound, and to make some quantitative predictions. But a
careful reading of Ref. 2 shows that the basic physics of the
process described there does not specifically require the pres-
ence of oxygen chains. All that is required is that there be
permanent oxygen dipole moments taking different values
depending on the chemical environment of the site. The pro-
cess in YBa
2
Cu
3
O
7 -x
can be qualitatively described with-
out explicit reference to chains: in the presence of an electric
field, oxygen migrates in order to lower the energy associ-
ated with its dipole moment. The migration of the oxygen
changes the valence state of some of the carrier donor atoms
Cu in YBa
2
Cu
3
O
7 -x
) and hence changes the carrier con-
centration. We will argue below that a very similar situation
should occur in Bi
2
Sr
2
CaCu
2
O
8 +y
, and therefore the ob-
servation of electric-field effects in this compound in no way
rules out oxygen diffusion as an explanation, even though
the quantitative conclusions based on a chain model
2
do not
necessarily apply.
Structure and superconducting properties are intimately
connected
7
in the compounds under study. The crystal struc-
ture of Bi
2
Sr
2
CaCu
2
O
8 +y
has been widely investigated by
several complementary techniques.
8
The unit cell has two
equivalent but shifted halves, each of which is 16 Å long in
the c direction. The crystal is composed of parallel planes.
Bismuth atoms are located at the centers of BiO
6
octahedra
sixfold coordinated with oxygen and Cu atoms are located
at the centers of square pyramids. Double Bi-O planes are
weakly held together by Van der Waals bonds. Both the
Bi-O and the Cu-O planes contribute to the electronic density
of states at the Fermi level. The Cu and the Bi atoms share a
common oxygen. The Cu-O bond is stronger than the Bi-O
bond, and the former determines the crystal lattice parameter.
The natural Bi-O bond is shorter than that dictated by the Cu
PHYSICAL REVIEW B 1 OCTOBER 1996-II VOLUME 54, NUMBER 14
54 0163-1829/96/5414/102183/$10.00 10 218 © 1996 The American Physical Society