Lecture notes: Superconductivity

Dr. Mohammad A Rashid

September 26, 2019
just.edu.bd/t/rashid

Contents

1 London’s Theory 2

1.1 Flux penetration from London equations . . . . . . . . . . . . . . . . . . 3

1.2 Coherence Length . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

2 Ginzburg-Landau Theory 8

2.1 The Ginzburg-Landau free energy . . . . . . . . . . . . . . . . . . . . . . 8

2.2 Equilibrium value of order parameter . . . . . . . . . . . . . . . . . . . . 10

2.3 The Ginzburg-Landau differential equations . . . . . . . . . . . . . . . . 12

2.4 Flux exclusion and zero electrical resistance . . . . . . . . . . . . . . . . 15

2.5 The Ginzburg-Landau coherence length . . . . . . . . . . . . . . . . . . . 16

2.6 Ginzburg-Landau parameter . . . . . . . . . . . . . . . . . . . . . . . . . 17

2.7 Flux quantization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

2.8 Duration of Persistent Currents . . . . . . . . . . . . . . . . . . . . . . . 19

2.9 Josephson effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

3 BCS Theory 23

3.1 BCS ground state . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

3.2 Cooper pairs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

3.3 Transition temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

3.4 The energy gap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

References 28

1
1 London’s Theory

The first theory to explain the occurrence of superconductivity in metallic superconduc-

tors was given by London brothers (Fritz London and Heinz London) in 1935. London
brothers started with the logic that if the electron in superconductor do not encounter
resistance they will continue to accelerate in an applied electric field E. Let ns and vs
be the number density (number/volume) and velocity of superconducting electrons in a
superconductor respectively. The equation of motion of electrons in the superconducting
state is given by
dvs
m = −eE (1)
dt
where m is the mass of the electron and e = 1.6 × 10−19 C. The current density (Am−2 )
is given by

Js = −ens vs . (2)

Differentiating it with respect to time we have,

dJs dvs
= −em . (3)
dt dt
Combining eq. (1) and (3) we can write

dJs ns e 2
= E. (4)
dt m

This is known as the first London equation:

A supercurrent is freely accelerated by an applied voltage, or, in a bulk

superconductor with no supercurrent or with a stationary supercurrent
there is no effective electric field.

Taking curl on both sides of eq. (4) we have

d ns e2
(∇ × Js ) = ∇ × E. (5)
dt m
From third Maxwell’s equation (Faradays law of electromagnetic induction) we know
that
∂B
∇×E=− (6)
∂t
and combining eq. (5) and (6) we can have

d ns e2 ∂B
(∇ × Js ) = − . (7)
dt m ∂t

2
Integrating both sides of above equation
ns e2
∇ × Js = − B + C. (8)
m
London brothers assumed the constant of integration, C to be zero so that eq. (8) takes
into account the fact of zero resistivity in superconductors. Hence with C = 0 we get
from above equation

ns e2
∇ × Js = − B. (9)
m
This is the second London equation. It yields the ideal diamagnetism, the Meissner
effect, and the flux quantization.

1.1 Flux penetration from London equations

The integral form of Amperes’ circuit law relates the magnetic field along a closed path
to the total current following through any surface bounded by the path. In mathematical
form:
I
B · dl = µ0 Iencl (10)
C

where C is the closed curve and Iencl is the total current flowing through any surface
bounded by c. Again, form Stokes’ theorem we can write
Z I
(∇ × B) · ds = B · dl (11)
S C

where S is any surface bounded by C and ds is the differential surface area. We may
express Iencl as the integral of the current density as follows:
Z
Iencl = Js · ds. (12)
S

Hence with help of eq. (11) and (12) we can rewrite eq. (10) as

∇ × B = µ 0 Js (13)

and taking curl on both sides yields

∇ × (∇ × B) = µ0 ∇ × Js . (14)

However,

∇ × (∇ × B) = ∇(∇ · B) − ∇2 B = −∇2 B (15)

since from second Maxwell’s equation (Gauss’s law for magnetism)

∇ · B = 0. (16)

3
Therefore, eq. (14) becomes
∇2 B = −µ0 ∇ × Js . (17)
Combining this with second London equation we get
 2
2 ns e 1
∇ B = µ0 B= 2B (18)
m λL
where
ns e2
 
1
= µ0 (19)
λ2L m
and λL has dimension of length and is known as Londons penetration depth.

This eq. (18) is seen to account for the Meissner effect because it does not allow a solution
uniform in space, so that a uniform magnetic field cannot exist in a superconductor. That
is, B(r) = B0 = constant is not a solution of eq. (18) unless the constant field B0 is
identically zero. The result follows because ∇2 B0 , is always zero, but B0 /λ2L is not zero
unless B0 is zero. Note further that eq. (13) ensures that Js = 0 in a region where
B = 0.

In the pure superconducting state the only field allowed is exponentially damped as we
go in from an external surface. Let a semi-infinite superconductor, placed in a magnetic
field, occupy the space on the positive side of the x axis, as shown in Figure 1, such
that B = ẑB, i.e., the field is only in the z-direction and can vary in space only in the
x-direction inside the superconductor.

Figure 1: A superconducting slab in an external field.

Since ∇ × B = µ0 Js , the current in the y-direction. Hence from eq. (18) we get
∂ 2B 1
2
= 2B (20)
∂x λL

4
which has a solution of the type

B(x) = B0 e−x/λL (21)

where B0 is the field at the surface and x is the depth inside the superconductor.
Any external field B0 is screened to zero exponentially inside a bulk
superconductor.

The eq. (21) shows that a uniform nonzero magnetic field can not exist in a supercon-
ductor, which is Meissner effect. In the pure superconducting state the only field allowed
is the exponentially decreasing field as shown in Figure 2.

Figure 2: Field penetration in a superconductor. The magnetic flux drops exponentially

inside the material. Penetration depth λL is defined as the depth at which the flux density
drops to its eth value. [R.G. Sharma]

An applied magnetic field B will penetrate a thin film fairly uniformly if the thickness
is much less than λL ; thus in a thin film the Meissner effect is not complete. In a thin
film the induced field is much less than B and there is little effect of B on the energy
density of the superconducting state, so that eq. (13) does not apply. It follows that the
critical field, that destroy the superconductivity, of thin films in parallel magnetic fields
will be very high.

Suppose x = λL then eq. (21) becomes

B0
B(x) = . (22)
e
From which we define London penetration depth:

The London penetration depth is the distance inside the surface of a

superconductor at which the magnetic field reduces to 1/e times its
value at the surface.

5
The equation for the London penetration depth is given by eq. (19) as:
r
m
λL = . (23)
µ0 ns e2
The penetration depth thus turns out to be inversely proportional to the square root
of ns , the superelectron density which changes with temperature. As the temperature
increases ns decreases and the flux penetration increases. At the critical temperature
Tc , ns decreases to zero and the whole material is penetrated with magnetic field thus
turning the superconductor into the normal state. The temperature dependent London
penetration depth is mathematically written as
"  4 #−1/2
T
λL (T ) = λ0 1 − (24)
Tc

where λL (T ) and λ0 are the London penetration depths at temperature T kelvin and 0
kelvin respectively.

Although Londons theory did explain the infinite electrical conductivity and the Meissner
effect, yet the calculated values of λL differ from the experimentally determined values.
This may be due to the uncertainty of the values of ns , e and m taken for free electrons,
which obviously cannot be justified. The parameters like the superelectron density, ns ,
their effective charge and effective mass have to be taken into account. After all, a
superconductor cannot be treated as a free electron metal. Instead, superelectrons in a
superconductor, interact coherently.

1.2 Coherence Length

The London penetration depth λL is a fundamental length that characterizes a super-

conductor. An independent length is the coherence length ξ. The coherence length is a
measure of the distance within which the superconducting electron concentration cannot
change drastically in a spatially-varying magnetic field.

The London equation is a local equation: it relates the current density at a point r to
the vector potential at the same point. So long as Js (r) is given as a constant time
A(r), the current is required to follow exactly any variation in the vector potential. But
the coherence length ξ is a measure of the range over which we should average A(r) to
obtain Js (r). We present a plausibility argument for the energy required to modulate
the superconducting electron concentration.

Any spatial variation in the state o r an electronic system requires extra kinetic energy.
A modulation of an eigenfunction increases the kinetic energy. It is reasonable to re-
strict the spatial variation of Js (r) in such a way that the extra energy is less than the
stabilization energy of the superconducting state.

We compare the plane wave ψ(x) = eikx with the strongly modulated wavefunction:
ϕ(x) = 2−1/2 (ei(k+q)x + eikx ). (25)

6
The probability density associated with the plane wave is uniform in space:

ψ ∗ ψ = e−ikx eikx = 1, (26)

whereas ϕ∗ ϕ is modulated with the wavevector q:

1
ϕ∗ ϕ = (e−i(k+q)x + e−ikx )(ei(k+q)x + eikx )
2
1
= (2 + eiqx + e−iqx )
2
= 1 + cos qx (27)

The kinetic energy of the wave ψ(x) is }2 k 2 /(2m); the kinctic energy of the modulated
density distribution is
1 }2 }2 2 }2
 
2 2
[(k + q) + k ] ≈ k + kq,
2 2m 2m 2m

where we neglected q 2 for q  k.

The increase of energy required to modulate is }2 kq/(2m). If this increase exceeds

the energy gap Eg , superconductivity will be destroyed. The critical value q0 of the
modulation wavevector is given by
}2
kF q0 = Eg , (28)
2m

We define an intrinsic coherence length ξ0 related to the critical modulation by ξ0 = 1/q0 .

We wave

}2 kF }vF
ξ0 = = (29)
2mEg 2Eg

where vF is the electron velocity at the Fermi surface. On the BCS theory a similar
result is found:
2}vF
ξ0 = (30)
πEg

The intrinsic coherence length ξ0 to is characteristic of a pure superconductor. In impure

materials and in alloys the coherence length ξ is shorter than ξ0 . This may be understood
qualitatively as in impure material the electron eigenfunctions already have wiggles in
them and we can construct a given localized variation of current density with less energy
from wavefunctions with wiggles than from smooth wavefunctions.

The coherence length ξ and the actual penetration depth λ depend on the mean free
path le of the electrons measured in the normal state. When the superconductor is very
impure, with a very small le , then λ ≈ λL (ξ0 /le )1/2 and ξ ≈ (ξ0 le )1/2 , so that λ/ξ ≈ λL /l.
This is the “dirty superconductor” limit.

7
2 Ginzburg-Landau Theory

The Ginzburg-Landau theory of superconductivity or called just GL theory is a phe-

nomenological theory and valid close to the transition temperature, Tc . Nevertheless,
it accounts well for the main characteristic properties of the superconductors. Since
superconductivity is caused by a second order phase transition, this theory draws an
analogy with the similar second order ferromagnetic phase transition in metals like iron
and nickel. The order parameter M, the magnetization in ferromagnets has now been
replaced by a macroscopic pseudowavefunction ψ(r), a complex order parameter for
superconductor such that

n∗s = |ψ(r)|2 = ψ ∗ (r)ψ(r) (31)

where n∗s is the effective number density of superelectrons in the superconductor. The
theory was developed by applying a variational method to an assumed expansion of the
free energy density in powers of |ψ(r)|2 and |∇ψ(r)|2 , leading to a pair of coupled differ-
ential equations for ψ(r) and the vector potential A(r). The result was a generalization
of the London theory to deal with situations in which ns varied in space, and also to
deal with the nonlinear response to fields that are strong enough to change ns .

The GL theory expresses the free energy of a superconductor in terms of the expansion
of ψ(r). Ginzburg and Landau treated the order parameter ψ(r) as the wave function
of the superconducting state which can vary with the location r. They expressed the
order parameter as

ψ(r) = |ψ(r)|eiφ(r) (32)

where φ(r) is the phase. The gradient of the phase at r is related to the momentum,
that is, the current flowing at the point r.

2.1 The Ginzburg-Landau free energy

The basic postulate of GL theory is that if ψ(r) is small and varies slowly in space, the
Helmholtz free energy density Fs close to the transition temperature, Tc can be expanded
in a series of the form
2
}2 ie∗ H2
 
2 β 4
Fs = Fn + α|ψ(r)| + |ψ(r)| + ∇ − A(r) ψ(r) + (33)
2 2m∗ }c 8π

where Fn is the Helmholtz free energy density in normal state; α and β are material
dependent phenomenological parameters to be determined experimentally (in the case
of classic BCS superconductors, α and β can also be calculated from the microscopic
theory); A(r) is the vector potential; H is the external magnetic field; e∗ and m∗ are
the effective charge and mass of superelectron, respectively.

In right side of eq. (33) the second and third terms correspond to the condensation free
energy density as superconducting state is more ordered than normal state. The fourth

8
term represents the kinetic energy of a charge particle in a magnetic field. Magnetic
field energy density is represented by the last term in eq. (33). The fourth term which
is proportional to the square of the gradient of ψ(r) indicates the uniform value of ψ(r)
at minimum energy. It is to be noticed that this square term also has the electromag-
netic potential A(r) because ∇ψ(r) is proportional to current which too depends upon
A(r). The free energy expression above assumes a configuration ψ(r), yielding minimum
free energy condition controlled by external parameters like temperature and magnetic
field. Quite a few solutions of eq. (33) are possible which yield various superconducting
parameters.

Let us now focus our attention on the term in the GL free energy which leads to super-
currents, the kinetic energy part:
2
}2 ie∗
 
Fkin = ∇ − A(r) ψ(r)
2m∗ }c
2
}2 ie∗
 
iφ(r)
= ∇− A(r) |ψ(r)|e
2m∗ }c
2
}2 iφ(r) iφ(r) ie∗
= ∗
(∇|ψ|)e + i|ψ|e ∇φ − A(r)|ψ(r)|eiφ(r)
2m }c
2
  ∗
 2
} ie iφ(r)
= ∇|ψ| + i|ψ| ∇φ − A(r) e
2m∗ }c
2
}2 ie∗

2
= ∗
∇|ψ| + i|ψ| ∇φ − A(r) × eiφ(r)
2m }c
( 2 )
∗
}2

e
= (∇|ψ|)2 + |ψ|2 ∇φ − A ×1
2m∗ }c
( 2 )
∗

1 e
= (}∇|ψ|)2 + |ψ|2 }∇φ − A (34)
2m∗ c

where in the second line eq. (32) is used. These expressions deserve several remarks.
Firstly, note that the free energy is gauge invariant, if we make the transformation
A → A + ∇Λ where Λ is any scalar function of position, while at the same time
changing ψ → ψ exp(−ie∗ Λ/c). Secondly, eq. (34) shows explicitly the contributions in
the kinetic energy density term. The first term gives the extra energy associated with
gradients in the magnitude of the order parameter and will introduce new physics that are
not included in the London theory. The second term gives the kinetic energy associated
with supercurrents in a gauge-invariant form. In the London gauge, φ is constant, and
this term is simply e∗2 A2 |ψ|2 /(2m∗ c2 ). Equating this to the kinetic energy density for a
London superconductor based on eq. (2), namely, A2 /(8πλ2eff ), we get
m∗ c2
λ2eff = . (35)
4π|ψ|2 e∗2
With the identification n∗s = |ψ(r)|2 , as in eq. (32), this agrees with the usual definition
of the London penetration depth, except for the presence of the starred effective number
density, mass, and charge values. As the order parameter decreases penetration length

9
increases. The kinetic energy density term can then be written as n∗s ( 12 m∗ vs2 ), where the
supercurrent velocity is given by
e∗ A e∗ A
m∗ vs = ps − = }∇φ − . (36)
c c
It should be noted that by writing the energy associated with the vector potential
in the simple form (34), we have restricted the theory to the approximation of local
electrodynamics.

In the original formulation of the theory, it was thought that e∗ and m∗ would be the
normal electronic values. However, experimental data turned out to be fitted better if
e∗ ≈ 2e. The microscopic pairing theory (BCS theory) of superconductivity makes it
unambiguous that e∗ = 2e exactly, the charge of a pair of electrons. In the free-electron
approximation, it would then be natural to take m∗ = 2m and n∗s = 12 ns , where ns
is the number density of single electrons in the condensate. With these conventions,
n∗s e∗2 /m∗ = ns e2 /m, so the London penetration depth is unchanged by the pairing.

2.2 Equilibrium value of order parameter

In the absence of fields and gradients (currents), e.g. deep inside of a bulk superconduc-
tor, we have from eq. (33)
β
Fs − Fn = α|ψ(r)|2 + |ψ(r)|4 (37)
2
which can be viewed as a series expansion in powers of |ψ(r)|2 , in which only the first
two terms are retained. An expansion in powers of ψ(r) itself is excluded since F must
be real. This difficulty cannot be avoided by taking the real part of ψ(r) since F should
not depend on the absolute phase of ψ(r). Odd powers of |ψ(r)| are excluded because
they are not analytic at ψ(r) = 0.

The two terms in the right side of eq. (37) should be adequate so long as one stays near the
second-order phase transition at Tc , where the order parameter |ψ(r)|2 → 0. Inspection
of eq. (37) shows that β must be positive if the theory is to be useful; otherwise the
lowest free energy would occur for arbitrarily large values of |ψ(r)|2 , where the expansion
is surely inadequate.

Minimizing Fs with respect to |ψ| we obtain

dFs
= 0 = 2α|ψ| + 2β|ψ|3 = 2(α + β|ψ|2 )|ψ|
d|ψ|
or
α
|ψ|2 = |ψ∞ |2 = − (38)
β
where the notation ψ∞ is conventionally used because ψ approaches this value infinitely
deep in the interior of the superconductor, where it is screened from any surface fields

10
or currents. When this value of ψ is substituted back into eq. (37), one finds

α2 H2
Fs − Fn = − =− c (39)
2β 8π
using the definition of the thermodynamic critical field Hc as the stabilization free energy
density of the superconducting state. Depending on whether α is positive or negative
two cases can arise as illustrated in Figure 3. If α is positive, the minimum free energy
occurs at |ψ|2 = 0, corresponding to the normal state. On the other hand, if α < 0, the
minimum occurs when |ψ|2 = |ψ∞ |2 as given by eq. (38).

Figure 3: Ginzburg-Landau free-energy functions for T > Tc (α > 0) and for T < Tc (α < 0).
Heavy dots indicate equilibrium positions. For simplicity, ψ has been taken to be real. [M.
Tinkham]

Evidently, α(T ) must change from positive to negative at Tc , since by definition Tc is the
highest temperature at which |ψ|2 6= 0 gives a lower free energy than |ψ|2 = 0. Making a
Taylor’s series expansion of α(T ) about Tc , and keeping only the leading term, we have

α(t) = α0 (t − 1) α0 > 0 (40)

where t = T /Tc . Putting these temperature variations of α and β into eq. (38), we see
that

|ψ|2 ∝ (1 − t) (41)

for T near, but below, Tc . This is consistent with correlating |ψ|2 with ns , the density
of superconducting electrons in the London theory since ns ∝ λ−2 L ∝ (1 − t) near Tc .

Having noted that e∗ = 2e, and taking the convention that m∗ = 2m, we can now
evaluate the parameters of the GL theory by solving eq. (35), (38), and (39). The

11
results are
ns m∗ c2 mc2
|ψ∞ |2 ≡ n∗s ≡ = = (42a)
2 4πe∗2 λ2eff 8πe2 λ2eff

e∗2 2 2 2e2 2
α(T ) = − H (T )λ = − H (T )λ2eff (42b)
m∗ c2 c eff
mc2 c

4πe∗4 2 16πe4 2
β(T ) = ∗2 4 Hc (T )λeff = 2 4 Hc (T )λ4eff
4
(42c)
m c mc

where e and m are now the usual free-electron values and λeff and Hc are measured
values, or those computed from the microscopic theory. Since the electrodynamics of
some superconductors are significantly nonlocal, it is evident that this prescription in
terms of an effective London penetration depth is straightforward only sufficiently near
Tc at which condition the GL theory is really exact.

It is worth noting that if we insert the empirical approximations Hc ∝ (1 − t2 ) and

λ−2 ∝ (1 − t4 ) into eq. (42), we find

|ψ∞ |2 ∝ 1 − t4 ≈ 4(1 − t) (43a)

1 − t2
α∝ ≈1−t (43b)
1 + t2
1
β∝ ≈ constant. (43c)
(1 + t2 )2

Since the theory is usually exactly valid only very near Tc , it is customary to carry only
the leading dependence on temperature; i.e., |ψ∞ |2 and α are usually taken to vary as
(1 − t) and β is taken to be constant, as anticipated in our preliminary discussion. Still,
the more complete forms in eq. (43) give some idea of how the theory can be extended
over a wider range of temperature, and they have a certain amount of experimental
support.

Finally, we recall that although our discussion of eq. (34) has centered on the kinetic
energy of the supercurrent, this term also describes the energy associated with gradients
in the magnitude of ψ(r). Moreover, no additional parameters are introduced since
gauge invariance requires a particular combination of ∇ and A in eq. (33). Thus, the
coefficients in the theory are completely determined by the values of λeff (T ) and Hc (T ).

2.3 The Ginzburg-Landau differential equations

In the absence of boundary conditions which impose fields, currents, or gradients, the
free energy is minimized by having ψ = ψ∞ everywhere. On the other hand, when
fields, currents, or gradients are imposed, ψ(r) = |ψ(r)|eiφ(r) adjusts itself to minimize
the overall free energy, given by the eq. (33).

12
 R
To derive the GL equations, let us first minimize the total free energy dV Fs (r) with
respect to variations in the function ψ ∗ (r). We have
Z Z 
dV δFs = dV αψδψ ∗ + β|ψ|2 ψδψ ∗

}2 ie∗ ie∗
   
∗ ∗
+ ∇ψ − Aψ · ∇δψ + Aδψ (44)
2m∗ }c }c
As all the terms are proportional to ψ, the integral may reduce to the volume of super-
conductivity only. In order to get rid of ∇δψ ∗ we integrate the term containing ∇δψ ∗
by parts and use Gauss’ divergence theorem. In such a case, we obtain
ie∗ ie∗ ie∗
Z   Z  
∗
dV ∇ψ − Aψ · ∇δψ + dV ∇ψ − Aψ · Aδψ ∗
}c }c }c
ie∗ ie∗
Z   Z  
∗ ∗
= dS n̂ · ∇ψ − Aψ δψ − dV δψ ∇ · ∇ψ − Aψ
}c }c
ie∗ ie∗
Z  
+ dV ∇ψ − Aψ · Aδψ ∗
}c }c
ie∗
Z  
= dS n̂ · ∇ψ − Aψ δψ ∗
}c
"  ∗ 2 #
Z ∗ ∗
ie ie ie
− dV δψ ∗ ∇2 ψ − ∇(Aψ) − A∇ψ + A
}c }c }c
2
ie∗ ie∗
Z   Z 
∗ ∗
= dS n̂ · ∇ψ − Aψ δψ − dV δψ ∇ − A ψ
}c }c
where the integration in the first term runs over the surface of the superconductor (n̂
is the unit vector of the normal). Now we substitute the result back into the previous
expression (44) and get
" 2 #
2 ∗
Z Z 
} ie
dV δFs = dV αψδψ ∗ + β|ψ|2 ψδψ ∗ − ∗
δψ ∗ ∇ − A ψ
2m }c
}2 ie∗
Z  
+ ∗
dS n̂ · ∇ψ − Aψ δψ ∗ . (45)
2m }c

There are two contributions to the variation, the bulk term and the surface one. Taking
into account that the superconducting specimens may have arbitrary sizes and shapes,
one can hardly expect that any compensation of these two terms may ever happen. So
each one must be set equal to zero independently of the other.

At the time being let us consider the volume integral. One can suppose it to be more
important for a macroscopic body. The corresponding variation must be zero, and
because δψ in the bulk is an arbitrary function, the following condition must take place
in the equilibrium state to minimize the volume contribution to the free energy:
2
}2 ie∗

2
αψ + β|ψ| ψ − ∇− A ψ=0 (46)
2m∗ }c

13
or
2
e∗

2 1
αψ + β|ψ| ψ + −i}∇ − A ψ = 0 (46a)
2m∗ c

This is known as the first Ginzburg-Landau differential equations. As one could see
from the derivation, the equation is an equilibrium condition for the system with the
free energy given by eq. (33).

Now let us continue the derivation of the second GinzburgLandau equation. A variation
of the joint free energy of the superconductor and magnetic field is taken with respect
to A. Please remember that the volume integral is taken over the whole infinite space:

(∇ × A)2 }2 ie∗ ie∗

Z Z     
∗ ∗
dV δFs = dV δ + ∇ψ − Aψ · ∇ψ + Aψ
8π 2m∗ }c }c
}2 ie∗ ie∗
   
(∇ × A) · (∇ × δA)
Z
∗ ∗
= dV + − ψ ∇ψ + Aψ
4π 2m∗ }c }c
ie∗ ∗ ie∗
  
+ ψ ∇ψ − Aψ δA
}c }c

Using the formula

∇ · (a × b) = b · (∇ × a) − a · (∇ × b)

with a = ∇ × A = H and b = δA we get

Z Z   
1
dV δFs = dV δA · (∇ × H) − ∇ · (H × δA)
4π
 ∗
e∗2
 
ie } ∗ ∗ 2
+ (ψ ∇ψ − ψ∇ψ ) + ∗ 2 |ψ| A δA (47)
2m∗ c mc

Here the integral of the second term in eq. (47) is transformed into a surface integral over
an infinite surface and disappears. This is a consequence of the fact that the integration
in the GinsburgLandau free energy is taken over the whole infinite space, and not over
the superconducting specimen only.
R
Now, setting dV δFs equal to zero within the superconductor, one obtains the second
GinsburgLandau equation:

c e∗ } ∗ ∗ e∗2 ∗
Js = ∇×H= (ψ ∇ψ − ψ∇ψ ) − ψ ψA (48)
4π 2im∗ m∗ c
or
e∗ e∗
 
Js = ∗ |ψ| }∇φ − A = e∗ |ψ|2 vs
2
(48a)
m c

where in the last step we have repeated the identification (36). Note that the current
expression (48) has exactly the form of the usual quantum-mechanical expression for

14
particles of mass m∗ , charge e∗ , and wavefunction ψ(r). Similarly, apart from the non-
linear term, the first equation has the form of Schrödinger equation for such particles,
with energy eigenvalue −α. The nonlinear term acts like a repulsive potential of ψ(r) on
itself, tending to favor wavefunctions ψ(r) which are spread out as uniformly as possible
in space.

In carrying through the variational procedure, one must provide boundary conditions.
A possible choice, which assures that no current passes through the superconductor-
vacuum interface, is
e∗
 
}
∇− A ψ =0 (49)
i c n̂
which ensures that Js · n̂ = 0 at the surface. This is the boundary condition used by
GL, and it is appropriate at an insulating surface. This can easily be derived from the
surface term in the free energy variation given by expression (45). If a media surrounding
the superconductor does not influence superelectrons, which is true for the vacuum or
a dielectric, then δψ may take arbitrary values at the surface. Then one can see that
to eliminate the surface integral, the condensate wave function at the surface of the
semiconductor must satisfy the boundary condition (49). Using the microscopic theory,
de Gennes has shown that for a metal-superconductor interface with no current, eq. (49)
must be generalized to
e∗
 
} i}
∇− A ψ = ψ (49a)
i c n̂ b
where b is a real constant. As shown in Figure 4, if An = 0, b is the extrapolation length
to the point outside the boundary at which ψ would go to zero if it maintained the slope
it had at the surface. The value of b will depend on the nature of the material to which
contact is made, approaching zero for a magnetic material and infinity for an insulator,
with normal metals lying in between.

Figure 4: Schematic diagram illustrating the boundary condition (49a) at an interface char-
acterized by an extrapolation length b. [M. Tinkham]

2.4 Flux exclusion and zero electrical resistance

The GL theory too leads to a London type relation between the electromagnetic vector
potential A(r) and the electric current density Js . Taking the curl of the eq. (48a), the

15
phase drops out, and we find the magnetic field:
e∗2 2 e∗2 n∗s
∇ × Js = − |ψ| ∇ × A = − H (50)
m∗ c m∗ c
which is the second London equation showing that in the absence of an external potential
the current will flow indefinitely and charge carriers will accelerate in the presence of the
potential. From the above expression it can easily shown that the London penetration
depth will be
1/2
m∗ c2

λeff = (CGS unit)
4πn∗s e∗2
1/2
m∗

= (SI unit)
µ0 n∗s e∗2
same as eq. (23) except the presence of star marks. This has the same consequence as the
London penetration depth i.e., the flux penetrates a superconductor a small depth λeff
only and so does flow the screening current. Deep inside the bulk of a superconductor
magnetic field can not penetrate unless the superconducting state is destroyed with a
high enough magnetic field.

2.5 The Ginzburg-Landau coherence length

To help get a feeling for the differential equation eq. (46), we first consider a simplified
case in which no fields are present. Then A = 0, and we can take ψ(r) to be real
since the differential equation has only real coefficients. If we introduce a normalized
wavefunction
 1/2
β
f= ψ (51)
|α|
where α = −|α|, the equation in one dimension becomes
}2 d2 f
∗ 2
+ f − f 3 = 0. (52)
2m |α| dx
This makes it natural to define the characteristic length ξ(T ) known as the Ginzburg-
Landau coherence length for the variation of ψ(x) by

}2 1
ξ 2 (T ) = ∗
∝ (53)
2m |α(T )| 1−t

In terms of ξ(T ), eq. (52) becomes

ξ 2 f 00 + f − f 3 = 0. (54)
df
Now multiplying both side of the above equation by f 0 = dx
we get
 
d 1 2 02 1 2 1 4
ξ f + f − f =0
dx 2 2 4

16
or
1
ξ 2 f 02 + f 2 − f 4 = C (55)
2
where C is the constant of integration. Far from the boundary in the superconducting
state, where ψ does not vary in space, we have that dx d
ψ = 0 i.e., f 0 = 0 and f 2 = 1
(following eq. (38)). With these we find the value of integration constant as C = 1/2.
Therefore, from eq. (55) we get
 2
2 df 1
ξ = (1 − f 2 )2 (56)
dx 2
which has a solution of the form
 
x
f (x) = tanh √ . (57)
2ξ
These lead us to the solution of ψ in one dimension as
 1/2  
|α| x
ψ(x) = tanh √ (58)
β 2ξ
which shows that a small disturbance of ψ from ψ∞ will decay in a characteristic length
of order ξ(T ).

Now that we have an idea of the significance of the length ξ(T ), let us see what its value
is. Substituting the value of α from eq. (42b) into the definition (53), we find
Φ0
ξ(T ) = √ (59)
2 2πHc (T )λeff
where
hc hc
Φ0 =∗
= = 2.0678 × 10−15 tesla m2 (60)
e 2e
is the flux quantum and known as fluxoid or fluxon.

2.6 Ginzburg-Landau parameter

The ratio of London’s penetration depth and the GL coherence length is called the GL
parameter represented by κ, that is
λeff (T )
κ= (61)
ξ(T )

Since both the parameters λ and ξ diverge as (1−t)1/2 , the dimensionless GL parameter,
κ remains temperature independent. For typical classical metallic superconductors λ is
much smaller than ξ, the ratio κ is smaller than 1. In fact, the value of the GL parameter
distinguishes between two
√ class of superconductors type I and type II. For most metallic
superconductor κ < 1/ 2 and for type II superconductors
√ (like alloy
√ superconductors
and high Tc superconductors) this ratio κ > 1/ 2. Thus κ = 1/ 2 ≈ 0.707 is the
dividing line between the two classes of superconductors.

17
2.7 Flux quantization

London brothers postulated that the flux inside a superconducting ring is quantized,
this also follows as a consequence of GL theory. We know that the flux penetrates a
superconductor a small depth λ only and so does flow the screening current. Let us
consider a ring made of a superconductor (toroid) as in Figure 5. The currents which
lead to flux quantization will only flow in a small part of the cross section, a layer of
thickness λ. Draw a contour C in the interior of the toroid, as shown in figure. Then
vs = 0 everywhere on C. It follows from eq. (48a) that

e∗
I I  
1
0= vs · dl = ∗ }∇φ − A · dl. (62)
C m C c

Figure 5: Quantization of flux in a superconducting toroid.

The order parameter ψ(r) should have a unique value with minimum energy at every
point along the circular path. In going round the circular path phase φ(r) should change
only by an integral multiple of 2π. Thus from eq. (62)
I
∇φ · dl = 2nπ (63)
C

where n is an integer. Having n 6= 0 requires that one not be able to shrink the contour
to a point, i.e., the sample have to has a hole as in our superconducting ring. The line
integral of the vector potential is
I Z
A · dl = ∇ × A · dS
C ZS
= B · dS
S
= Φ. (64)

18
Here S is a surface spanning the hole and Φ the flux through the hole. Combining these
results,
}c hc
Φ = 2nπ ∗
= n = nΦ0 (65)
e 2e
where Φ0 is the flux quantum as in eq. (60). Thus the flux through the ring is quantized
in integral multiples of Φ0 = hc/(2e). Flux quantization indeed follows from the fact
that the current is the result of a phase gradient. However, it is important to note that,
a phase gradient does not guarantee that a current is flowing.

The flux through the ring is the sum of the flux Φext from external sources and the flux
Φsc from the persistent superconducting currents which flow in the surface of the ring:
Φ = Φsc + Φext . The flux Φ is quantized. There is normally no quantization condition
on the flux from external sources, so that Φsc must adjust itself appropriately in order
that Φ assume a quantized value.

2.8 Duration of Persistent Currents

Consider a persistent current that flows in a ring of a type I superconductor of wire of

length L and cross-sectional area A. The persistent current maintains a flux through
the ring of some integral number of fluxoids as in eq. (60). A fluxoid can not leak out
of the ring and thereby reduce the persistent current unless by a thermal fluctuation a
minimum volume of the superconducting ring is momentarily in the normal state.

The probability per unit time that a fluxoid will leak out is the product
P = (attempt frequency) × (activation harrier factor)
 
∆F
= (attempt frequency) × exp − , (66)
kB T
where the free energy of the barrier is
∆F ≈ (minimum volume) × (excess free energy density of normal state). (67)
The minimum volume of the ring that must turn normal to allow a fluxoid to escape is
of the order of Rξ 2 , where ξ is the coherence length of the superconductor and R the
wire thickness. The excess free energy density of the normal state is Hc2 /(8π), whence
the barrier free energy is
Rξ 2 Hc2
∆F ≈ . (68)
8π

Let the wire thickness be 10−4 cm, the coherence length = 10−4 cm, and Hc = 103 G;
then ∆F ≈ 10−4 erg. As we approach the transition temperature from below, ∆F will
decrease toward zero, but the value given is a fair estimate between absolute zero and
0.8Tc . Thus the activation barrier factor is
 
∆F 7
≈ exp 108 ≈ 10−4.34×10 .


exp −
kB T

19
The characteristic frequency with which the minimum volume can attempt to change
its state must be of the order of Eg /}. If Eg ≈ 10−15 erg, the attempt frequency is
≈ 10−15 /10−27 ≈ 1012 s−1 . The leakage probability (66) becomes
7
P ≈ 1012 × 10−4.34×10 s−1
7
≈ 10−4.34×10 s−1 .

The reciprocal of this is a measure of the time required for a fluxoid to leak out, T =
7
1/P = 104.34×10 s. The age of the universe is only 108 s, so that a fluxoid will not leak
out in the age of the universe, under our assumed conditions. Accordingly, the current
is maintained.

There are two circumstances in which the activation energy is much lower and a fluxoid
can be observed to leak out of a ring; either very close to the critical temperature,
where Hc is very small, or when the material of the ring is a type II superconductor
and already has fluxoids embedded in it. These special situations are discussed in the
literature under the subject of fluctuations in superconductors.

2.9 Josephson effect

From the aforesaid discussion we understand now quite clearly that a superconductor
is defined by an order parameter ψ(r) with a phase which is constant within a super-
conductor. This phase can however be changed by an external electromagnetic field.
Variation of phase causes a current to flow.

Consider a very thin weak link between two halfs of a superconductor as shown in
Figure 6. The order parameter has its thermodynamic value on both sides x < x1 and
x > x2 , but is exponentially small at x = 0. Hence, any supercurrent through the weak
link is small, and ψ may be considered constant in both bulks of superconductor. In the
weak link, not only |ψ| is small, also its phase may change rapidly, e.g. from φ2 = φ1
to φ2 = φ1 + π by a very small perturbation. The current that tends to flow across the
weak link or the junction is dependent on the phase difference (φ2 − φ1 ).

Without the right half, the boundary condition (49) would hold at x1 :

e∗
 
} ∂
− Ax ψ = 0 (69)
i ∂x c x1

In the presence of the right half, this condition must be modified to slightly depending
on the value ψ2 :

e∗
 
} ∂
− Ax ψ = aψ2 , (70)
i ∂x c x1

where a is a small number depending on the properties of the weak link. Time inversion
symmetry demands that eq. (70) remains valid for ψ → ψ ∗ , A → −A, hence a must

20
Figure 6: A weak link between two halfs of a superconductor (top) and the behavior of their
order parameters (bottom).

be real as long as the phase of ψ does not depend on A. For the moment we choose a
gauge in which Ax = 0. Then, the supercurrent density at x1 is
" #
}e∗ ∂ψ ∂ψ ∗
Js,x = ∗
ψ1∗ − ψ1
2im ∂x x1 ∂x x1
a}e∗ ∗
= [ψ ψ2 − ψ1 ψ2∗ ]
2im∗ 1
a}e∗  i(φ2 −φ1 ) −i(φ2 −φ1 )

= e − e
2im∗
a}e∗
= sin(φ2 − φ1 ) (71)
m∗

Therefore, the expression for the current that flows through the weak link can be written
as
I = Ic sin(φ2 − φ1 ). (72)
No current will flow if the phase difference (φ2 − φ1 ) is zero. Ic in the above equation is

21
the critical current of the junction and depends upon the junction strength. This is DC
Josephson effect. This current flows without a potential difference. In a geometry, like
a superconducting ring, the phase difference will change with the flux in the ring which
is quantized. Equation (72) will now take the form
 
Φjunction
I = Ic sin (φ2 − φ1 ) + 2π . (73)
Φ0
In a Superconducting Quantum Interference Device (SQUID), shown in Figure 7 (left),
the two arms acquire different phases equal to 2π times the number of unit quantum
of flux depending on the enclosed flux. The current will be maximum when the phase
difference is an even multiple of π and minimum when the phase difference is an odd
multiple of π. The current pattern is thus oscillatory when plotted against the magnetic
field passing through the SQUID as shown in Figure 7 (right).

Figure 7: A dc SQUID with two Josephson-junctions (J-J1 and J-J2) mounted on a super-
conducting ring (left) and a typical Josephson current versus magnetic field pattern in a dc
SQUID (right). [R.G. Sharma]

An expression for the ac Josephson effect can be driven from the eq. (62) by differenti-
ating with respect to time, that is
∂ e∗ ∂A
} ∇φ = = −e∗ E. (74)
∂t c ∂t
Since the electric field E i.e., the electro motive force V is determined by the rate of
change of vector potential A. Thus the phase difference between the two superconductors
in a Josephson junction turns out to be,
 ∗   
eV 2eV
∆φ = t= t. (75)
} }
Therefore,
 
2eV Φjunction
I = Ic sin t + 2π . (76)
} Φ0
Hence if a voltage V is applied across the junction, an AC current with a frequency
2eV /} will flow.

22
3 BCS Theory

The basis of a quantum theory of superconductivity was laid by the classic 1957 papers
of Bardeen, Cooper, and Schrieffer. There is a “BCS theory of superconductivity” with
a very wide range of applicability, from He3 atoms in their condensed phase, to type I
and type I1 metallic superconductors, and to high-temperature supercondnctors hased
on planes of cuprate ions. Further, there is a “BCS wavefunction” composed of particle
pairs k ↑ and -k ↓, which, when treated by the BCS theory, gives the familiar electronic
superconductivity observed in metals and exhibits the energy gaps. This pairing is
known as s-wave pairing. Some of the accomplishments of BCS theory with a BCS
wavefunction are:

1. An attractive interaction between electrons can lead to a ground state separated from
excited states by an energy gap. The critical field, the thermal properties, and most of
the electromagnetic properties are consequences of the energy gap.

2. The electron-lattice-electron interaction leads to an energy gap of the observed mag-

nitude. The indirect interaction proceeds when one electron interacts with the lattice
and deforms it; a second electron sees the deformed lattice and adjusts itself to take
advantage of the deformation to lower its energy. Thus the second electron interacts
with the first electron via the lattice deformation.

3. The penetration depth and the coherence length emerge as natural consequences of
the BCS theory. The London equation is obtained for magnetic fields that vary slowly
in space. Thus the central phenomenon in superconductivity, the Meissner effect, is
obtained in a natural way.

4. The criterion for the transition temperature of an element or alloy involves the elec-
tron density of orbitals D(F ) of one spin at the Fermi level and the electron-lattice in-
teraction U , which can be estimated from the electrical resistivity because the resistivity
at room temperature is a measure of the electron-phonon interaction. For U D(F )  1
the BCS theory predicts
 
1
Tc = 1.14θ exp − (77)
U D(F )

where θ is the Debye temperature and U is an attractive interaction. The result for
Tc , is satisfied at least qualitatively by the experimental data. There is an interesting
apparent paradox: the higher the resistivity at room temperature the higher is U , and
thus the more likely it is that the metal will be a superconductor when cooled.

5. Magnetic flux through a superconducting ring is quantized and the effective unit of
charge is 2e rather than e. The BCS ground state involves pairs of electrons; thus flux
quantization in terms of the pair charge 2e is a consequence of the theory.

23
3.1 BCS ground state

The filled Fermi sea is the ground state of a Fermi gas of noninteracting electrons. This
state allows arbitrarily small excitations − we can form an excited state by taking an
electron from the Fermi surface and raising it just above the Fermi surface. The BCS
theory shows that with an appropriate attractive interaction between electrons the new
ground state is superconducting and is separated by a finite energy Eg from its lowest
excited state.

The formation of the BCS ground state is suggested by Figure 8. The BCS state in
(b) contains admixtures of one-electron orbitals from above the Fermi energy F . At
first sight the BCS state appears to have a higher energy than the Fermi state: the
comparison of (b) with (a) shows that the kinetic energy of the BCS state is higher than
that of the Fermi state. But the attractive potential energy of the BCS state, although
not represented in the figure, acts to lower the total energy of the BCS state with respect
to the Fermi state.

Figure 8: (a) Probability P () that an orbital of kinetic energy  is occupied in the ground
state of the noninteracting Fermi gas; (b) the BCS ground state differs from the Fermi state
in a region of width of the order of the energy gap Eg . Both curves are for absolute zero.
[C. Kittel]

When the BCS ground state of a many-electron system is described in terms of the
occupancy of one-particle orbitals, those near F are filled some what like a Fermi-Dirac
distribution for some finite temperature.

The central feature of the BCS state is that the one-particle orbitals are occupied in
pairs: if an orbital with wavevector k and spin up is occupied, then the orbital with
wavevector −k and spin down is also occupied. If k ↑ is vacant, then −k ↓ is also vacant.
The pairs are called Cooper pairs. The binding energy is strongest when the electrons
forming the pair have opposite moments and opposite spins. It follows, therefore, that
if there is any attraction between them, then all the electrons in the neighborhood of
the Fermi surface condense into a system of Cooper pairs. These pairs are, in fact, the
superelectrons. The Cooper pair have spin zero and have many attributes of bosons.

24
3.2 Cooper pairs

Consider two electrons of opposite spin that interact with each other via an attractive
potential, as shown in Figure 9. For a complete set of states of the two electron system
that satisfy periodic boundary conditions in a cube of unit volume, we take plane wave
product functions

Ψ(k1 , k2 ; r1 , r2 ) = exp[i(k1 · r1 + k2 · r2 )] (78)

Figure 9: Interaction between two electrons, 1 and 2, near the Fermi surface in a metal.

We change variables to the relative coordinates: r = r1 − r2 , k = 21 (k1 − k2 ) and to the

coordinates of the center of mass: R = 21 (r1 + r2 ), K = k1 + k2 so that

k1 · r1 + k2 · r2 = K · R + k · r (79)

Thus eq. (78) becomes

Ψ(K, k; R, r) = ei(K·R) × ei(k·r) , (80)

and the kinetic energy of the two-electron system is

}2 1 2
 
2
K + Ek = K +k (81)
m 4

We give special attention to the product functions for which the center of mass wavevec-
tor K = 0, so that k1 = −k2 . With an interaction H1 between the two electrons, we set
up the eigenvalue problem in terms of the expansion
X
χ(r) = gk ei(k·r) . (82)

25
The Schrödinger equation is
0
X
(H0 + H1 − ) = 0 = [(Ek0 − )gk0 + H1 gk0 ei(k ·r) (83)
k0

where  is the eigenvalue. We take the scalar product with exp(ik · r) to obtain
X
(Ek − )gk + gk0 hk|H1 |k0 i = 0, (84)
k

the sacular equation of the problem. Transforming the sum to an integral we have
Z
(E − )g(E) + dE 0 g(E 0 )H1 (E, E 0 )N (E 0 ) = 0, (85)

where N (E 0 ) is the number of two electron state with total momentum K = 0 and with
kinetic energy in dE 0 at E 0 .

Now consider the matrix elements H1 (E, E 0 ) = hk|H1 |k0 i. Studies of these by Bardeen
suggest that they are important when the two electrons are confined to a thin energy
shell near the Fermi surface within a shell of thickness }ωD above EF , where ωD is the
Debye phonon cutoff frequency. We assume that

H1 (E, E 0 ) = −V (86)

for E, E 0 within the shell and zero otherwise. Here V is assumed to be positive. Thus
eq. (85) becomes
Z 2m
(E − )g(E) = V dE 0 g(E 0 )N (E 0 ) = C, (87)
2F

where m = F + }ωD . Here C is a constant, independent of E. From eq. (87) we have

C
g(E) = (88)
E−
and
2m
N (E 0 )
Z
1=V dE 0 . (89)
2F E0 − 

With N (E 0 ) approximately constant and equal to NF , over the small energy range be-
tween 2m and 2F , we take it out of the integral to obtain
Z 2m
1 2m − 
1 = NF V dE 0 0 = NF V ln . (90)
2F E − 2F − 

Let the eigenvalue  of eq. (90) be written as

 = 2F − Eb , (91)

26
which defines the binding energy Eb of the electron pair, relative to the free electrons at
the Fermi surface. Then eq. (90) becomes
2m − 2F + Eb 2}ωD + Eb
1 = NF V ln = NF V ln (92)
Eb Eb
or
 
1 2}ωD
= ln 1 + . (93)
NF V Eb
Hence the binding energy of a Cooper pair may be written as
2}ω
Eb =  D . (94)
1
exp NF V
−1

For V positive (attractive interaction) the energy of the system is lowered by excitation
of a pair of electrons above the Fermi level. Therefore the Fermi gas is unstable in an
important way. The binding energy (94) is closely related to the superconducting energy
gap Eg . The BCS calculations show that a high density of Cooper pairs may form in a
metal.

3.3 Transition temperature

In the weak coupling limit kB Tc  }ωD the expression for the transition temperature has
been obtained by solving ground state energy equations and is given by the expression
 
1
kB Tc ≈ 1.14 }ωD exp − (95)
N (0)V
where N (0) is the density of electron states of one spin per unit energy at the Fermi
level, V the electron-phonon interaction parameter and kB is the Boltzmann constant.
Here we find that Tc is proportional to the phonon frequency ωD which is consistent
with the observation of the isotope effect in superconductors. We also notice in eq. (95)
that Tc is a strong function of the electron concentration as the density of state enters
in the exponential term. It is thus possible for one to make calculations for the change
in Tc as a result of alloying or applying pressure.

3.4 The energy gap

Another remarkable finding of the BCS theory is that it predicts a relationship of Tc

with the energy gap Eg , also denoted by 2∆ in some literature, which is given by

Eg (0) ≈ 3.5 kB Tc . (96)

It is interesting to note that the energy gap normalized to the zero temperature value
when plotted for various superconductors against the reduced temperature, T /Tc lie on a

27
Table 1: Measured energy gap ratio Eg (0)/(kB Tc ) for metal superconductors at zero temper-
ature.

Metal Eg (0)/(kB Tc ) Metal Eg (0)/(kB Tc )

Aluminum 3.3 Tin 3.5
Zinc 3.2 Mercury 4.6
Gallium 3.5 Vanadium 3.4
Cadmium 3.2 Molybdenum 3.4
Indium 3.6 Lanthanum 3.7

single universal curve. Experimental values of Eg (0) for different materials and different
directions in k space generally fall in the range from 3.0 kB Tc to 4.5 kB Tc with most
clustered near the BCS value 3.5 kB Tc (Figure 10). The energy gap at T = 0 is almost
independent of temperature. Close to Tc , gap versus temperature relationship is given
by
 1/2
T
Eg (T ) ≈ 3.2 kB Tc 1 − . (97)
Tc

Strictly speaking, this universal curve holds only in a weak coupling limit, but it is a
good approximation in most cases.

Figure 10: Reduced energy gap Eg (T )/Eg (0) plotted against the reduced temperature T /Tc
for most metal superconductors lie on a universal curve and fit well with BCS theory within
the width in the curve. [R.G. Sharma]

References
1. Introduction to Superconductivity by Michael Tinkham
2. Superconductivity Basics and Applications to Magnets by R.G. Sharma
3. Introduction to Solid State Physics by Charles Kittel

28