Zayid Ahmed

Department of Physics, Indian Institute

of Technology Gandhinagar

Master’s Thesis Supervised by Prof. Vinod

Chandra
Effects of Background magnetic Field on
Dirac Solutions and The Fermionic
Propagator

July 24, 2020

Abstract
Charged particles are very sensitive to the magnetic field as it changes
their properties and particles behave differently according to the field.
Here we have studied how the Dirac solutions and the energy eigenval-
ues change in presence of uniform background magnetic field and their
gauge dependence. After that we have studied the method given by Ju-
lian Schwinger to calculate the propagator. The fermionic propagator
changes in presence of magnetic field. We have discussed it’s gauge de-
pendence and how it changes according to the strength of background
magnetic field.

2
Contents
1 Introduction 4

2 Solutions of Dirac Equation 4

2.1 Choice of gauge . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.2 solutions and spinors . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.3 Orthonormality and Completeness and Spin Sum . . . . . . . . . 8
2.4 Landau levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.5 gauge dependence . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

3 Schwinger’s method 9
3.1 photon propagator . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3.2 fermionic propagator . . . . . . . . . . . . . . . . . . . . . . . . . 14

4 Gauge dependence, phase factor and field limits 15

5 Conclusion 16

3
1 Introduction
As a charged fermion, we consider a electron being the particle having the
largest specific charge i.e, being the most sensitive to the external field influence.
The large hadron collider (LHC) and the relativistic heavy-ion collider (RHIC)
can produce a very strong magnetic field at the very early stages of collision. A
very strong magnetic field exists at the core of magnetars, and the same may
have existed at the early stages of universe. In light of this we study different
types of elementary particle decays and scattering processes. Theoretically the
solutions of Dirac equation in vaccum are well known, with the Dirac spinors
for positive and negative energy along with there corresponding completeness
relations.
In presence of background magnetic field the Dirac spinors are going
to change as they depend on the spatial coordinates due the gauge used to
solve the Dirac equation and also depend on the momentum. As a result, their
corresponding completeness relation are also going change. These realistic fields
are very complicated but for simplicity many times we assume these fields are
uniform. The Dirac equation can be exactly solved in presence of uniform
magnetic field, then we can proceed to quantize these solutions and calculate
elementary particle decays and scattering cross-section. We use the uniform
background magnetic field along z-axis. The dispersion relation we get after
solving the Dirac equation is En2 = p2z − m2 + 2n|eB|, and n = 0 state is non-
degenerate for both positive and negative energy solutions. Thus the ground
state is spin polarised. In the case of strong magnetic field (|eB| >> m2 ),
electron prefer to lie in the lowest Landau level, and electron will have only
longitudinal momentum. In the weak magnetic field the higher Landau levels
contribute. The dispersion relation is gauge independent, but the spinors and
completeness relation are gauge dependent for both positive and negative energy.
We use the Schwinger’s method to calculate the propagator for charged
scalar field and fermionic field. This method has some nice properties, as we are
not fixing the gauge in the beginning of calculations. As a result we end up with
an expression for the propagator with an arbitrary gauge. In the case strong
magnetic field it can be shown that one part of the propagator corresponds to the
contribution of the ground Landau level and the other part corresponds to the
contribution of higher Landau levels. We will see that propagator has a phase
factor which plays an important role in transforming the two-point function.

2 Solutions of Dirac Equation

There are different methods to obtain the spinors in presence of constant
magnetic field, we have adopted to solve the Dirac equation in presence of
constant background magnetic field to get the spinors[1].

4
2.1 Choice of gauge
The background magnetic field of magnitude B along z-direction can be
fixed by choosing background gauge field AµB in many ways:

A0B = AxB = AzB = 0, AyB = xB + a (1)

or
A0B = AyB = AzB = 0, AxB = −yB + b (2)
or
1 1
A0B = AzB = 0, AyB = xB + c, AxB = − yB + d (3)
2 2
Where a, b, c, and d are constants. AµB is classical background field and not a
quantized dynamical field. Where x and y are coordinates not 4-vectors. We
assume gauge configuration as given in Eq.(1) with b = 0 in our calculation.
We will use Dirac-Pauli representation of Dirac matrices.

2.2 solutions and spinors

The Dirac equation for a particle of mass m and charge Q (where Q = −e
for electron and Q = +e for positron) in presence of background magnetic field
is given by:
dψ
i = HB ψ (4)
dt
where HB is Dirac Hamiltonian in presence of background magnetic field is
given by:
HB = α.Π + βm (5)
Πµ is kinematic momentum of the charged fermion. Πµ = (pµ + ec Aµ )
We can conclude from the Dirac equation that the stationary state, we can
write:  
φ
ψ = e−iEt (6)
χ
where φ and χ are two component objects. Using ψ in Eq.(4) we end up with
two coupled equation given by:

(E − m)φ = σ.(−i∇ + QA)χ (7)

(E + m)χ = σ.(−i∇ + QA)φ (8)

We eliminate χ from the above equation i.e. decoupling,

(E 2 − m2 )φ = (σ.(−i∇ + QA))2 φ (9)

Using the value of vector potential and evaluating RHS of the above equation,
∂
(E 2 − m2 )φ = [(−∇2 − (QB)2 y 2 − QB(2iy + σ3 )]φ (10)
∂x

5
As we can see that the variables x and z do not appear in the equation except
through the derivatives, by variable separable method we can write the solution
as,
φ = eip.Xy0 f (y) (11)
f (y) is a 2-component matrix which only depend on y, and possibly some mo-
mentum components. we have used p.Xy0 = px x + pz z, where px and Pz are the
eigenvalues of the momentum along x and z coordinates respectively.
σ3 appears in the equation which is a diagonal Pauli matrix. There will be two
independent solution of f (y), which can be taken to be the eigenstates of σ3
with eigenvalues s = −1, +1. We can choose two independent solution in the
form    
F+ (y) 0
f+ (y) = , f− (y) = (12)
0 F− (y)
where σ3 fs = sfs , using this in Eq.(10) equation satisfied by Fs is,

d2 FS
+ (QBy + px )2 FS + (E 2 − m2 − p2z + QBs)FS = 0 (13)
dy 2
We will use a dimensionless variable,
p px
ξ = |Q|B(y + ) (14)
QB
which transform the equation in the form,

d2
[ − ξ 2 + as ]Fs = 0 (15)
dξ 2
where,
E 2 − m2 − p2z + QBs
as = (16)
|Q|B
Eq.(15) is Hermite’s equation and solution exists provided as = 2ν + 1 where
ν = 0, 1, 2, ....This gives us the energy eigenvalues

E 2 = m2 + p2z + (2ν + 1)|Q|B − QBs (17)

and the solutions for Fs are

2
Nν e−ξ /2
Hν (ξ) = Iν (ξ) (18)

Nν are normalisation
 p|Q|B 1/2
Nν = √ (19)
ν!2ν π
The functions Iν satisfy the completeness relation
X p
Iν (ξ)Iν (ξ∗ ) = |Q|Bδ(ξ − ξ∗ ) = δ(y − y∗ ) (20)
ν

6
Q was arbitrary.we will discuss two cases one for Q = −e (electron case) and
second Q = +e (positron case).

Q = −e Case

The energy eigen values are,

En2 = m2 + p2z + 2neB (21)

which is the relativistic form of Landau energy levels. Where 2n = 2ν + 1 + s.

The solutions are two fold degenerate in general:
for s = 1, ν = n − 1 and
for s = −1, ν = n.
For n = 0 case when s = 1, ν becomes negative but ν cannot be negative, so for
n = 0 state only s = −1 state exists. Thus ground state is not degenerate.We
can say that ground state is spin polarised.
The solution(positive energy solutions) coressponding to n-th Landau level can
be written as,    
(n) I (ξ) (n) 0
f+ (y) = n−1 , f− (y) = (22)
0 In (ξ)
For n = 0 the solution f+ does not exist. Thus I−1 (y) = 0. Eq.(22) determine
the upper components of the spinors, lower components can be solved using
Eq.(8). Finaly, the positive energy solution for the Dirac equation can be written
as,
e−ip.Xy0 Us (y, n, py0 ) (23)
where X µ denotes the spacetime coordinate. And the spinors are given by,

In−1 (ξ) 0
   
 0   √ In (ξ) 
U+ (y, n, py0 ) = 
 Enp−m
z
In−1 (ξ) , U− (y, n, py0 ) =  2neB In−1 (ξ)
   (24)
√ En +m
− 2neB
En +m In (ξ) − Enp+m
z
In (ξ)

Q = e Case
The energy eigenvalues are,

En2 = m2 + p2z + 2neB (25)

which is the relativistic form of Landau energy levels. Where 2n = 2ν + 1 − s.

Again the solutions are two fold degenerate in general: for s = 1, ν = n and
for s = −1, ν = n − 1
. For n = 0 case, when s = −1, ν becomes negative. For n = 0 state only
s = 1 state exists. Thus ground state is not degenerate. Ground state is spin
polarised.

7
 With the same analogy as for the positive energy solution, the negative
energy solution for Dirac equation can be written as:

eip.Xy0 Vs (y, n, py0 ) (26)

where X µ denotes the spacetime coordinate. And the spinors are given by,
 pz  √2nB
In−1 (ξ˜)
 ˜

En
√ +m En +m In−1 (ξ )
 2nB In (ξ˜)  pz ˜ 
 , V+ (y, n, py0 ) =  − En +m In (ξ ) 

V− (y, n, py0 ) =  E +m
 nI ˜
(27)
(ξ )
n−1
  0 
0 In (ξ˜)

2.3 Orthonormality and Completeness and Spin Sum

Using the relation, Z ∞ √
In (a)Im (a)da = eBδnm (28)
∞
The ortho-normality of the spinors has to be modified as they have explicit
coordinaye dependence. Given by,
Z ∞ Z ∞
† 2En
dyUs (y, n, py0 )Us0 (y, m, py0 ) = dyVs† (y, n, py0 )Vs0 (y, m, py0 ) = δnm δss0
∞ ∞ E n+m
Z ∞ Z ∞ (29)
dyUs† (y, n, py0 )Vs0 Us (y, m, py0 ) = dyVs† (y, n, py0 )Us0 (y, m, py0 ) = 0
∞ ∞
(30)
They also satisfy the completeness relation.
The spin sum is
X 1
PU (y, y∗ , n, py0 ) = Us (y, n, py0 )Ūs (y∗ , n, py0 ) = SU (y, y∗ , n, py0 )
s
En + m
(31)
and
X 1
PV (y, y∗ , n, py0 ) = Vs (y, n, py0 )V̄s (y∗ , n, py0 ) = SV (y, y∗ , n, py0 )
s
En + m
(32)
One imortant property of spin sum is

PU (y, y∗ , n, py0 ) = −PV (y, y∗ , n, py0 ) (33)

2.4 Landau levels

We studied that the ground state (i.e. n = 0 solution) is non-degenerate.
In this case we have only one solution for the positive energy and one for the
negative energy, i.e. s = −1 for the positive energy and s = 1 for the negative
energy. We can also notice that the ground state wave function are the eigensa-
tates of σz the 3-D double spinor operator, where σz = ιγ1 γ2 , because for n = 0

8
solution s has a definite value and for all other higher Landau levels solution do
not have definite σz eigenvalue.
Considering the case of strong magnetic field we require to work n = 0
solutions and we can estimate the the strength of that field. Let say E is
the electron energy of the system and B is the strength of magnetic field. If
2neB > E 2 + m2 for a positive value of n then from the dispersion relation
E 2 = m2 + p2z + 2neB we can see that p2z becomes negative, that makes momen-
tum along z-axis imaginary, which is impossible. Thus for strong magnetic field
that makes 2eB greater than the square of electron energy minus the rest mass
square of the electron only n = 0 levels contributes to the energy levels and
only those corresponding wave functions should be used for the calculations. If
the electron energy of the system is of the order of 1M eV the for the magnetic
field magnitude greater than 1014 Gauss we should only have the n = 0 level
contributing towards the energy.
Thus, when the magnetic field magnitude is large only the lowest
Landau level will contribute to the electron energy but if the magnetic field
magnitude is low the higher Landau levels will start contributing.

2.5 gauge dependence

The Dirac spinors are gauge dependent but they are not physical observ-
able so that does not bother us. The orthonormality and completeness relations
are also gauge dependent as they includes the spinors. We see that the disper-
sion relation that we have in presence of uniform background magnetic field is
gauge independent. For any configuration of background gauge field that gives
us a uniform magnetic field along z-axis the dispersion relation will be gauge
independent as it includes only B, which is the magnitude of magnetic field.

3 Schwinger’s method
We will use the Schwinger’s method to calculate the propagator for scalar and
fermionic field in presence of uniform background magnetic filed.
We define the Green’s function as
H(x, ∂x )G(x, x0 ) = δ 4 (x − x0 ) (34)
The idea is to consider H as a hamiltonian that dscribes the proper time evolu-
tion of the system.

hx|x0 i = δ 4 (x − x0 )
[xα , pβ ] = −ιg αβ ,
pµ = ι∂µ
We introduce the unitary evolution operator which satisfy the Srondinger equa-
tion given by’
∂
i U (x, x0 : s) = H(x, p)U (x, x0 : s) (35)
∂x

9
where s is the proper time parameter. With boundary conditions;

U (x, x0 ; 0) = δ 4 (x − x0 )U (x, x0 ; −∞) = 0 (36)

Wew have
U (x, x0 ; s) = hx| e−ιHS |x0 i = hx| U |x0 i
(37)

Z 0
G(x, x0 ) = −ι dsU (x, x0 ; s) (38)
−∞

Equation () may be written as

ι∂s = hx| H(x, p)U |x0 i = hx| U (s)U (s)H(x, p)U (s) |x0 i

ι∂s hx(s)|x0 (s)i = hx(s)| H(x(s), p(s) |x0 (0)i (39)

we have used
U (s)H(x, p)U (s) = H(x(s), p(s))
, it’s like going from Schrondinger to Hiesenberg picture and hs| U (s) = hx(s)|
Our goal is to solve for x(s) and p(s) and express H(x(s), p(s)) in terms of x(s)
and x0 (0). Now we need to do the time ordering which is x(s) on the left and
x0 (0) on the right.That will give us:

hx(s)| H(x(s), p(s)) |x0 (0)i = F (x, x0 ; s) hx(s)|x0 (0)i

.
Then the equation for U (x, x0 ; s) becomes an ordinary linear differential equation
which may be integrated as
h Z s i
U (x, x0 ; s) = C(x, x0 )exp − ι ds0 F (x, x0 ; s) (40)

3.1 photon propagator

An important ingredient for the perturbative calculations is the
knowledge of propagator of charge particles in presence of magnetic field.
We are going to calculate the propagator for the case of a charged scalar field
in presence of a constant uniform magnetic field(constant in time and uniform
in space) by the method described by Julian Schwinger in 1951.
Following Schwinger’s method we have;

H(x, p)G(x, x0 ) = [Πµ Πµ − m2 ]G(x, x0 ) = δ 4 (x − x0 ) (41)

where
H(x, p) = Πµ Πµ − m2

10
is the Hamiltonian and Πµ = pµ − eAµ (x)
To compute hx(s)| H(x(s), p(s)) |x0 (0)i we require to solve the equations motion
for xµ and Πµ , i.e. Hiesenberg

dxµ (s)
= ι[H, xµ ] (42)
ds
dΠµ (s)
= ι[H, Πµ ] (43)
ds
Let us compute the commutators of the above equations,

[H, xµ ] = [Πµ Πµ − m2 , xµ ]
= 2ιΠµ

where [Aµ (x), xµ ] = 0 and,

[H, Πµ ] = [Πν Πν − m2 , Πµ ]
= Πν [Πν , Πµ ] + [Πν , Πµ ]Πν
= Πnν (−ιeFµν ) + (−ιeFνµ )Πν
= ιeΠν Fµν + ιeFµν Πµ

But P iµ Fµν = Fµν Πν + ι∂ ν Fµν , therefore,

[H, Πµ ] = ιe(2Fµν Πν + ι∂ ν Fµν )

As we are working witht a constant uniform magnetic field along z-axis, in this
case ∂ ν Fµν = 0, i.e., ∂ ν (∂µ Aν − ∂ν Amu ) = 0 Equations of motion,

dxµ
= −2Πµ (44)
ds
dΠµ
= −2eFµν Πµ (45)
ds
Let us introduce some notations, since the space separates naturally in a parellel
and a perpendicular direction with regards to the magnetic field, for vectors aµ ,
bµ ,

aµ|| = (a0 , 0, a3 )
aµ⊥ = (0, a1 , a2 , 0)

Such that

(a · b)|| = a0 b0 − a3 b3
(a · b)⊥ = a1 b1 + a2 b2
(a · b) = (a · b)|| − (a · b)⊥

11
The equations of motion separates as;
dx⊥
= −2Π⊥
ds
dΠ⊥
= −2eF ij Π⊥
ds
dx||
= −2Π||
ds
dΠ||
=0
ds
where i, j = 1, 2
Let us look at the equations in the perpendicular direction,
   
x1 (s) Π1 (s)
x⊥ = , Π⊥ = (46)
x2 (s) Π2 (s)

The energy-momentum tensor Fµν is;

 
0 −Ex −Ey −Ez
Ex 0 −Bz By 
Fµν =   (47)
Ey Bx 0 −Bx 
Ez −By Bx 0

Transverse part is,  

ij 0 −1
F =B (48)
1 0
we will take,  
0 −1
F = (49)
1 0
where F 2 = −I, F −1 = −F = F T Using these our equations of motion will be,

dx⊥ (s)
= −2Π⊥
ds
dΠ⊥
= −2eBF Π⊥
ds
The solutions of these two equations are

Π⊥ = exp{−2eBF s}Π⊥ (0)

2 sin eBs
x⊥ (s) − x⊥ (0) = − exp{−eBF s}Π⊥ (0)
eB
In this notation the tranverse part of the Hamiltonian is written as;

H⊥ = ΠT⊥ (s)Π⊥ (s) (50)

12
therefore the Hamiltonian after doing the normal ordering i.e, we need to place
x(s) to the left and x(0) to the right, for that we will use the commutator
relation.
H⊥ = ΠT⊥ (s)Π⊥ (s) (51)
Let us look at the parallel part of the equations of motion, which are,
dx|| (s)
= −2Π||
ds
dΠ||
=0
ds
Thus the parallel part of the Hamiltonian is,
H|| = ΠT|| (s)Π|| (s) (52)
The Hamiltonian is;
H = H⊥ − H|| (53)
Since all the operators in H are in the correct order we can write,
2
−(eB)
hx0 (s)| H |x00 (0)i = (4 sin 02 0 00 00 1 02 0 00 002 2 0 00
2 (eBs)(x⊥ − 2x⊥ x⊥ + x⊥ ) − ιeB cot eBs + 4s2 (x|| − 2x|| x|| + x|| ) − m ) hx (s)|x (0)i

(54)
Thus the evaluation operator is,
C(x0 , x00 )
 
−ιeB 0 ι
hx0 (s)|x00 (0)i = exp (x − x00 )2⊥ cot eBs + (x0 − x00 )2|| + ιm2 s
s sin eBs 4 4s
C(x0 , x00 ) is the phase factor, we can compute this function. For that we need
to compute hx0 (s)| Πµ (s) |x00 (0)i and hx0 (s)| Πµ (0) |x00 (0)i, these two will give us
set of equations,
∂ e
(ι 0µ + eA(x0 ) + Eµν (x0 − x00 )2 )C(x0 , x00 ) = 0 (55)
2
and
∂ e
(ι 0µ + eA(x0 ) + Eµν (x0 − x00 )2 )C(x0 , x00 ) = 0 (56)
2
After solving these we can find function C(x0 , x00 ) , which is,
C(x0 , x00 ) = ιφ(x0 , x00 )
x0
( Z )
1
whereφ(x0 , x00 ) = exp ιe dxµ (Aµ (x) + Fµν (x0 − x00 )ν
x00 2

Thus, the propagator in coordinate space is written as

Z ∞  
0 00 0 00 1 eB 0 00 2 1 0 00 2 2
G(x , x ) = iφ(x , x ) ds exp −i( (x − x )⊥ cot eBs − (x − x )|| − m s)
0 s sin eBs 4 4s
(57)

where φ(x0 , x00 ) is the phase factor and the other part is translationally invariant.

13
3.2 fermionic propagator
We will follow the same method here, the hamiltonian for a fermionic field is,
e
H = Πµ Πµ − m2 − σµν F µν (58)
2
The equations of motions are,
dxµ(s)
= −2Πµ
ds
dΠµ(s) e
= −2eFµν Πν − ι∂ ν Fµν − ∂µ Fµρ σ νρ
ds 2
For constant magnetic field, µ F µν = 0, we will end up with equations similar to
that for scalara field, which are;
dxµ(s)
= −2Πµ
ds
dΠµ(s)
= −2eFµν Πν
ds
Following the same notation as we did for scalar field propagator, we will sep-
arate the parallel and perpendicular parts of the equations of motion and the
calculate the Hamiltonian accordingly doing all the time ordering. As we can
see that this Hamiltonian has an extra term which is,
−e
σµν F µν = eBΣ3
2
where Σ3 = ιγ1 γ2 , B is the magnitude of magnetic field. The Hamiltonian is;

H = H|| − H⊥ − m2 + eBΣ3 (59)

Evolution operator for this Hamiltonian is;

C(x0 , x00 )
 
eB 0 ι
hx0 (s) |x00 (0)i| = exp −ι (x − x00 )2⊥ cot eBs + (x0 − x00 )2|| −2 s + ιeBΣ3
s sin eBs 4 4s
(60)
where the phase factor is C(x0 , x00 ) = ιφ(x0 , x00 ) and
( Z 0 )
x
0 00 µ 1 0
φ(x , x ) = exp ιe dx (Aµ (x) + Fµν (x − x )6ν
x00 2

Thus, the Green’s function is,

Z ∞
ds 1
G(x0 , x00 ) = ιφ(x0 , x00 ) exp (−ιm2 s + ιeBsΣ3 )

s sin eBs
0 
−ι 0 eBs
× exp ((x − x00 )2|| − (x0 − x00 )2⊥ )
4s tan eBs

14
Propagator is,

S(x0 , x00 ) = (ι∂x00 − qf Aµ (x00 ) + mf )G(x0 , x00 ) (61)

Z ∞
0 00 0 00 1
exp (−2 s + BsΣ3

S(x , x ) = φ(x , x ) ds
0 s sin eBs
(62)
(" !#)
ι eBs
× exp − (x0 − x00 )2|| − (x0 − x00 )2⊥
4s tanh eBs
(63)
" #
1 eBs 
× m+ (γ · x0|| − γ · x00|| ) − exp{−ιeBsΣ3 }(γ · x0⊥ − γ · x00⊥ )
2s sin eBs
(64)

In momentum space:
Z
ds exp{ιk · x}S(x0 − x00 )
S(k) =
Z ∞ (" #)

2 tanh eBs 2 2

= −ι ds exp ιs (γ · k|| ) − (γ · k⊥ ) − m
0 eBs
 1 
× (1 + γ1 γ2 tan eBs)(γ · k|| + m) − 2
γ · k⊥
cos (eBs)

4 Gauge dependence, phase factor and field lim-

its
The behaviour of frermionic two point function changes in presence
of magnetic field as we can see it contains a translationally invariant part and a
gauge dependent part. It includes a translationally non-variant part the phase
factor which is linked to it’s gauge transformation properties. We can write it
as;
dp4
Z
S(x0 , x00 ) = φ(x0 , x00 ) exp{−k · (x0 − x00 )}S(k) (65)
(2π)2
where S(k) is tanslationally invariant and gauge invariant part.
Gauge dependence is present in the phase factor. When we do the gauge trans-
form the field in presence of background magnetic field we get different fermionic
field at two different space-time points. Phase factor connects the fields at two
different space-time points, but when we transform fields at two different space-
time points the fields will have different phase factors. As we can see that the
phase factor breaks the translational invariance of the two-point function. An
interesting property of the phase factor is that it is path independent and when
we chose the path to be a straight line it becomes zero. The fields of charged

15
particles and their two-point functions become background dependent when we
work in background field. That is why phase factor arises in the two-point func-
tion.

When the magnetic field B = 0 and all the constants a = b = c = 0 only

then the propagator goes to the propagator in absence of background field, oth-
erwise it will have phase which is different from the phase factor. Schwinger’s
propagator is a function of gauge field and is defined for all gauge configurations
including the pure gauge. It can always be continuously transformed from the
case where it is a function gauge field to a case where it is a function of pure
gauge.
In the strong field limit, only contribution from the lowest Landau level needs
to be kept. In this limit the major contribution comes from the region where:
eB
s=
K|| − m2

we need to consider that part of the propagator, s is time parameter. For the
weak field limit, we can demonstrate propagator into an infinite series in power
of B, and terms with lowest power of B are more important. And as B goes to
zero, we get back to the normal fermion propagator in absence of background
field.

5 Conclusion
The presence of background magnetic field effects the fermionic fields and the
propagator, this in turn effect the different types of scattering and decaying
processes. Presence of the magnetic field also helps us to understand the pro-
cesses more clearly as it separates the different type of particles. The processes
like γ− > νν − , e− e+ − > µ− µ+ , e− µ− − > e− µ− etc. can be studied using
the Dirac spinors and the fermionic propagator. The scattering cross-section
and the decay rates should be gauge independent when we calculate using these
calculations.

16
References
(1) K. Bhattacharya, Solution of the Dirac equation in presence of an uniform
magnetic field, 2007, arXiv:0705.4275 [hep-th].

(2) J. Schwinger, Phys. Rev. 82, 664 (1951).

(3) A. Das, A. Bandyopadhyay, P. K. Roy, and M. G. Mustafa, General structure

of fermion two-point function and its spectral representation in a hot magne-
tized medium, Phys. Rev. D 97, 034024 (2018).

(4) C. Itzykson and J. Zuber, Quantum Field Theory, Mc Graw Hill press,
International student edition, Second edition.

(5) Alexander Kuznetsov, Nickolay Mikheev, Electroweak Processes in External

Electromagnetic Fields, springer(2003).

(6) Gauge invariance of elementary particle processes taking place in presence

of a background magnetic field Kaushik Bhattacharya(Ahmedabad, Phys. Res.
Lab) Oct 23, 2005

(7)K. Bhattacharya and P. B. Pal, Pramana 62, 1041 (2004) [arXiv:hep-ph/0209053].

17