noc19-ph02 Lecture 87-Josephson's effect in superconductors and tunneling current across barriers

IIT Kanpur July 201
20 followers
Dec 3, 2023
From our earlier lectures, we have been introduced to two different length scales associated
with the superconductor. One is the penetration depth, which is the distance up to which the
magnetic field actually enters into the superconductor and the remaining part of the superconductor
has B is equal to 0. From the edges, how much of magnetic field penetrate, up to what distance
the magnetic field penetrates into the superconductor is given by the superconducting penetration
depth lambda and the coherence length is the size of the cooper pairs. The pair of electrons
which are paired up, the typical size of each cooper pair is the coherence length.
So these are two fundamental length scales associated with any superconductor, once it
gets into the superconducting state at T less than Tc. And one defines an important
parameter which is called kappa, which is the Ginzburg-Landau parameter, it is the ratio
of lambda to xi. And it was shown that if kappa is greater than 1 or more precisely
greater than 1 by root 2, then one has type 2 superconductors, namely that up to a lower
critical field, the you have the Meissner state that means for fields less than HC1,
the magnetic field is completely shielded from the interior of the sample, magnetization
is minus of H, here magnetization is minus of H, so the slope is minus 1.
And for fields above the critical field, the diamagnetism decreases very slowly. So in
this case, you have superconducting regions and the system breaks up into microscopic
normal regions, which are called as vortices. So the superconductor actually breaks up into
superconducting regions which are diamagnetic and there are normal regions which are of
course non-diamagnetic and so the diamagnetism decreases very gradually as you go up to HC2.
This sort of discussion and showing that these vortices actually organize themselves into
a very ordered arrangement was done by A. Abrikosov, for which he won a Nobel Prize
in 2000 and so these are type 2 superconductors which happen when kappa is greater than 1
by root 2. This was explicitly shown by Abrikosov and they and these vortices which are generated
above HC1 between HC1 and HC2, these normal regions actually organize themselves into
a very ordered arrangement, which is actually a hexagonal lattice of vortices.
And so this type 2 superconductor you will see a vortex state where the normal regions
are not randomly placed, but they are arranged in a very periodic ordered configuration.
So this was the vortex state and if kappa is less than 1 or kappa is less than 1 by
root 2, then you have the conventional type 1 superconductor where magnetization for fields
less than HC, your magnetization is minus 1 of H and at HC superconductivity is lost,
it becomes normal. Entire superconductor in one shot becomes normal. There are no vortices
generated in the case of type 1 superconductors.
Now before I get further, one very important point of these vortices is that if this is
the top view of the superconductor and I have applied a magnetic field to it, this is the
magnetic field applied to the superconductor, this is the top surface of the superconductor
and this is the material, then the vortices actually pass from end to end of the superconductor.
The vortices are actually passing from end to end of the superconductor. They are like
cylinders which are present inside the superconductor. These are normal cores, these are normal regions
of the superconductor.
And so if you apply a magnetic field where B is greater than BC1 and it is B should be
less than BC2, then you will form in type 2 superconductors these vortices. And what
is interesting is that these vortices, the diameter of these vortices is twice the coherence
length. Each vortex diameter is twice the coherence length and each of these vortices
being normal regions non-diamagnetic, the magnetic field is actually present inside
these vortices. So, there is magnetic field present inside the vortices and they also
leak a little out into the diamagnetic regions of the superconductor and the distance up
to which they leak out, the magnetic field inside the vortices leak out is over a distance
which is twice the penetration depth. And around each vortex there are shielding currents
which are circulating and these currents also extend up to a distance which is given by
the penetration depth.
Now, for each vortex you can find the integral of B bar dot ds which is the magnetic flux
in each vortex and one can show that each vortex carries a fixed amount of magnetic
flux, which is called as the magnetic flux quantum. So, each vortex carries a fixed amount
of magnetic flux which is the integral of the magnetic field enclosed by each vortex
over the area of this vortex and that turns out to be an integral multiple or it is basically
one flux quanta. Each typically most vortices in most superconductors carry only one flux
quanta and they are in integral multiples of this fundamental quantity H which is your
Planck's constant divided by twice of E. Why twice of E? This twice of E comes from
Cooper pairs which are the charge carriers in the superconductor and their total charge
is minus twice of E. And a sign that you have Cooper pairs inside superconductor, they are
not carrying a charge E, but the charge carriers in the superconducting state have a total
charge of twice of E, a proof of that came from measuring the magnetic flux of each vortex.
When they measure the magnetic flux of each vortex, it was shown to be H by 2E and this
2E proved that the Cooper pairs have a total charge of minus 2E or these are a pair of
electrons. And so, it is very important for you to remember that these vortices or these
normal regions which are being created in type 2 superconductors between the lower and
upper critical field carry a fixed quantum of magnetic flux and that magnetic flux is
in units of fundamental constants H divided by 2E. The value is about 2.07 into 10 raised
to minus 7 Gauss centimeter square.
And this discovery of the arrangement of vortices in type 2 superconductors, the theoretical
discussion and why they should be well ordered and so on, that discussion or that derivation
and that discovery, which was theoretical first, that work led to another Nobel Prize
in this field. Now, after having discussed the nature of superconductivity and all the
behavior of a superconductor in a magnetic field, what is the theory for the superconductors
and there is an opening up of a gap around the Fermi energy and no single electrons can
exist within plus minus delta of the Fermi energy.
Let us look at another interesting and a final aspect which I will discuss about superconductors
and that is junctions between a superconductor and a normal material. So, the first I will
discuss the behavior of a metal superconductor junction. So, I can think of a metal superconductor
junction as let us take a metal, a normal metal and put it and sandwich it with another
superconductor. So, you have a metal and you have another superconductor and let us apply
a voltage, measure the current which is going to flow through this system. Will it be unusual
or will it be ohmic or what will it be? Will you observe something like voltage by current?
Will you observe something as you see in Ohm's law or will it be nonlinear or will it be
like a semiconductor? What will it be?
So, let us understand if you actually measure the current versus voltage, you will find
that the current will be 0 up to a certain value of voltage and only above this value
which is proportional to delta by e charge of the electron, you will find that the current
through the system will appear. So, up to a certain value of voltage I is 0, over this
entire regime I is 0, whether you apply voltage in the positive direction or negative direction.
Up to a certain threshold voltage the current is going to be 0 through this and only beyond
a certain value current will appear and this is as far as a simple metal superconductor
junction is concerned. It is certainly not like an ohmic system where the moment you
apply a voltage you get some current and V is equal to I into R. It is certainly not
like this. It more looks somewhat like a forward bias diode. Of course, it is not like a reverse
bias diode, but it has some characteristics of a forward bias diode where beyond a certain
threshold voltage you start getting current through the system.
So, how does one understand it and there is a reasonable simple way to look at this
is that in a superconductor this end is the metal and this end is a superconductor and
you have an interface between them. This is the interface. So, on one end you have a metal,
you have a superconductor, you have single electrons on this end, you also have single
electrons on this end, but in the superconducting state there is some difference. This is the
Fermi level in the metallic state and the superconducting state, but in the superconducting
side in this region no single electron states are available because electrons pair up.
So, in this window of two delta around the Fermi energy you do not have a possibility
of having individual electrons. All the electrons around the Fermi energy start pairing up and
the system. You cannot have any single individual electron floating around with an energy which
is within plus minus EF of the Fermi energy. So, therefore, in this window these are all
occupied states. In the superconductor these are all vacant single electron states and
these are all occupied single electron states. So, you cannot send by applying a voltage
you cannot send an electron from this end to this end because these are all completely
occupied. So, the current will be 0. Now when you start applying a voltage you start shifting
the level on one side. However, here also no electron can flow through. This flow is
not allowed and only when your top of your occupied level in the metal side becomes equal
to the vacant state an electron can flow inside the system.
Within this gap region there is no single electron states available. So, no electron
can enter into the superconducting side and that is why current is going to be 0 up to
a certain voltage and only beyond a certain threshold voltage will you get a current in
the system. One can also think of it in the reverse way and you can also work it out that
how will you get a current in the reverse direction. It will be exactly symmetric.
And so this is how a metal superconducting junction works that you will have up to a
certain threshold no current through the system as you are increasing the voltage, but only
beyond a certain threshold voltage current appears inside the system and that is because
of the presence of a gap in the density of single electron states in the superconductor
which does not exist on the metal side, ok. And so that is why it is very non-ohmic.
Now and you can see that once the electrons go into the normal then you have single electrons
and then at higher voltages you start getting an ohmic like response because then you have
single electrons which are scattering because they are no longer paired. So, they start
scattering off and so you get resistance in this region. This is as far as a metal superconductor
junction is concerned. These are very useful. They are used for detecting photons and these
metal superconductor junctions have various applications. One example which I just said
is that they are used for detecting photons and so on, single photon detection applications.
Let us go on to another very important effect and this was another Nobel Prize which was
won for discovering this phenomena. And what is this phenomena? That is called as the phenomena
of Josephson junction. Now, I already showed you what happens when you have a metal and
you have a superconductor, ok. You have a metal on one side, you have a superconductor
on another side, you have an interface in between and we looked at the I-V characteristics
that the current is going to be 0 up to a certain voltage value and only beyond a certain
value there will be a current which will start flowing.
However, suppose you take a superconductor insulator superconductor junction and if the
width or let us say the thickness of this junction, this insulating junction is of the
order of the coherence length which is say 10 to 100s of nanometers. If this thickness
of this insulating layer which is present or sandwiched between two superconductors,
if that thickness is of the order of the superconducting coherence length, then if I do a similar experiment
namely I apply a voltage across the junction, I apply a voltage across these two superconducting
junctions and I vary the voltage and measure the current and I measure the voltage across
the junction, then I get something interesting. In this sort of a junction if the thickness
of the insulating layer is of the order of coherence length, what I will find is that
even at 0 voltage I will get a finite amount of current in the system which is absent in
the metal superconductor interface at 0 voltage there was no current.
The current was 0 and it remains 0 up to a certain threshold voltage. But here what you
see is that in this superconductor insulator superconductor junction even at 0 voltage
I have a finite current which can be sent through the system, no dissipation. Voltage
is 0, I get a finite non-zero current through the system which means resistance is equal
to 0 inside the system. So, I can send in dissipation less current from one end to the
other ok, current can flow through the system with 0 voltage across the system or the resistance
is actually 0 inside the system. And this current across this insulating junction
has a maximum value it is plus minus IC, this is called as the critical current of the junction
and the discovery of this phenomena and its explanation led to another Nobel Prize. Such
type of junctions where the thickness of the interface is of the order of coherence length
and you get this specific type of current voltage characteristics where you can send
in a current up to a threshold value without any voltage across this junction, this is
known as the Josephson junction phenomena. Very briefly this is what is the Josephson
junction phenomena where up to a finite value of current can be sent through the junction
without any resistance or dissipation across the junction and beyond this voltage develops
across the junction. What is the reason for this? The reason for
this is actually quite interesting, the way to think about these Josephson junction is
that you have two superconductors with an insulating barrier, this is the insulator
I, you have a superconductor 1 and you have a superconductor 2 on two ends, so this is
your SIS junction, superconductor insulator superconductor junction, so thickness of this
superconductor is of the order of coherence length.
Now what happens is that on this end you have a wave function and this end also you have
a wave function psi which corresponds to the superconducting state, this has a phase
phi 1, this wave function has a phase phi 2 and as I told you for in BCS theory because
of the pairing of electrons one with each other, any superconductor the electronic state
is not described by each electrons having an individual wave function, it is a macroscopic
quantum state, so the entire superconductor is described by one single wave function having
an amplitude and a phase. So this superconductor S1 has a phase phi 1 and of course amplitude
psi 0 and this is a wave function associated with S1 and this is the wave function associated
with S2.
Now if this is very thin what can happen is that this wave function leaks a little across
this boundary and it can enter into other region and similarly this wave function leaks
a little and it actually enters into the other region. As a result within this region there
is an interference, there is an interference of currents which are actually flowing across
the junction from either end superconductor or cooper pairs from this end are actually
tunneling, there is a tunneling across the junction.
Similarly cooper pairs from this end are tunneling across the junction and their wave
functions start interfering in this region and this gives rise to because cooper pairs
can tunnel across the junction you can get a finite current through the junction without
any dissipation across the junction because these are pairs which are just moving across
the junction and they are tunneling across the junction and the currents in between the
two junctions they start interfering, the two wave functions interfere if the thickness
of the junction is of the order of coherence length.
And this leads to this type of interesting current relationships where I is equal to
I0 sin delta, delta is the phase difference across the junction, phi 1 and phi 2 are the
phases on the two ends of the junction. If this is your superconductor and this is your
insulating barrier, phi 1 this is the superconductor S1, superconductor S2 this is your insulator,
phi 1 and phi 2 are the phases of the waves on the two ends of the on either sides of
the superconductor and the difference in the phase is phi 1 minus phi 2.
The amount of current that is going to flow through the junction is going to be given
by the interference of these two waves and the interference as you know from optics is
proportional to the difference in the phases. If the difference in phase is of some order
then you will get constructive interference, if the difference is of some other value then
there is a destructive interference. So, the Josephson junction phenomena is actually related
to interference of these Cooper pairs which are flowing from right and left side and their
wave functions are interfering leading to a net current which is I0 sin delta, delta
is the phase difference, the current is just given by the interference of the waves.
And similarly, the voltage which appears depends on the rate of change of the phase across
the junction. So, if you apply a finite voltage of the junction, if you apply a constant voltage
across the junction the delta is going to be proportional to the voltage applied V0
times t from this, if V0 is constant and then you will get I is equal to I0 sin plus some
constant V0 t and there is other factors, there are other factors which will sit out
here plus some other constant delta 2 and this gives rise to an AC effect.
The current across the junction whereas this is a DC current, if you apply a voltage across
the junction the phase starts evolving in this fashion and that evolution starts oscillating
the current across the junction and this gives rise to the AC Josephson effect, this is called
as the DC Josephson effect, I is equal to I0 sin delta, where delta is the phase difference
and this is called as the AC Josephson part. Because if you apply a voltage difference
across the junction, then the current tunneling through starts oscillating with as a function
of time, this is almost like sin omega t and this pre-factor gives you something called
as a Josephson frequency.
And a very important standard for frequencies, the Josephson frequency is used as a standard
for defining frequencies as a calibration standard. So, this Josephson phenomena have
different applications, if you take two such Josephson junctions in a loop and bias it
ok with a current and place it in a magnetic field where the magnetic field is actually
passing perpendicularly across this loop, then the currents from these two arms interfere
and lead to something called as a superconducting quantum interference device. The voltage of
the junction actually oscillates periodically depending on the flux that is present, the
magnetic flux which is present across the junction and the oscillation period is of
the order of phi0, it oscillates with a period of phi0 where phi0 is the flux quantum 2.07
into 10 raise to minus 7 gauss centimeter square.
So, if you see such oscillations, you know you are detecting fluxes which are so small,
this is the level of sensitivity that you can get with superconducting quantum interference
device because you will see voltage oscillations across the junction which depend on the flux
changing in the junction by these small amounts. And these type of Josephson junctions are
finding a lot of applications being used in quantum computation chips. This is an example
of a chip which I show here, this is taken from Wikipedia. These are a quantum computation
chips which has about 1 lakh Josephson junctions implemented on the chip and here there is
interferences of current and then you look at different behaviors of this.
And so, Josephson junction is also a very important effect, it is basically related
to Cooper-Peirce tunneling across an interface if that interface is thin enough and they
start interfering and this is the phenomena of Josephson junctions. So, superconductivity
as a phenomena has multiple possibilities of applications, they are used for MRI magnets,
for resistance loss making the losses power transmissions with zero losses, for levitated
magnetic trains, for circuits, they have multiple different applications, have a potential
for multiple different applications. I have just given you a survey of the phenomena
of superconductivity and there is research which continues in this direction and people
are still trying to understand what is the phenomena of superconductivity, how best to
understand it. BCS is only one particular aspect and a way to understand superconductivity.
People are trying to go beyond BCS and understand other superconductors and so, the field is
very rich and it is one of the areas at the forefront of solid state physics.
