A Related Representation of

The Dirac Equation and Maxwell Equations


                                 ZHANG Xia-fu

Room103, Building121of Teacher Apartments,

Kunming University of Science and Technology,

Kunming 650093, China   

Email:xiafu@ public.km.yn.cn


Abstract We suggest a special representation of the Dirac equation, in which two intrinsic spins

are containedso the representation involves more symmetries. We also suggest a formally

different expression of Maxwell equations, which also contain two intrinsic spins inside. The

Dirac equation and Maxwell equations are correlated by means of the matrices of their spins.


Key words: Theory of quantized fields, Dirac equation, Maxwell equations, spin

    Pacc Code: 0370



0.  Introduction

We suggest a special representation of the Dirac equationwhich can be derived by a unitary transformation from the acquainted representation of the Dirac equation. The feature of the representation is that, besides spin angular momentum, another intrinsic spin is included in the equation. The representation of the Dirac equation contains two intrinsic spins, thus it involves more symmetries.

Corresponding to the mentioned representation of the Dirac equation, we also rewrite Maxwell equations with the matrices that exist in the Dirac equation. As a result Maxwell equations show that they also contain two intrinsic spins, and are related to the Dirac equation with the common spin matrices.

1.       The Dirac equation containing two intrinsic spins

    When we make mention of spin we always remember 2×2 Pauli matrices. As we show in the following, there exist two sets of three 4×4 matrices, the properties of which are just the same as the 2×2 Pauli matrices. We will write the Dirac equation with them instead of Pauli matrices.

Set 1 of the three 4×4 matrices is written as:


Set 2 of the three 4×4 matrices is written as:


All the matrices are Hermitiean. We denote Set 1 with operators σ1, σ2 and σ 3, and Set 2 with operators τ1, τ 2 and τ 3 respectively. They can be represented as vector operators σ and τ, where each vector is composed of three components, i.e. σ = [σ1, σ2, σ 3] and τ = [τ1, τ2, τ 3]. The following operation rules for a pair of operators σ and τ are derivable from the matrix expressions (1.1) and (1.2):


Specifically, the matrices of Set 1 commute with the matrices of Set 2, that is


The form of the Dirac equation can be improved by means of σ and τ. We know that the Dirac equation generally is expressed as


Where the operators γμ satisfy anticommutive relations:


Our representation is to replace γμ in the above equation by


Considering rules (1.3) and (1.4), we can verify that expressions (1.7) obey anticommutive relations shown in expression (1.6):


The matrices of expressions (1.7) can be deduced from the conventional four Dirac matrices by means of a unitary transformation (see Appendix (1)). Substituting (1.7) into (1.5), consequently, a special representation of the Dirac equation is founded


Where contains a pair of vector operator σ and τ in the equation. We have seen that not only σ but also τ belong to the inherent properties of the particles described by the equation.

Under Lorentz transformations


Relativistic invariance of Eq. (1.5) requires that


For infinitesimal transformation, we have acquainted the following formulas


Corresponding to the representation of the Dirac equation given by us, in terms of rule (1.3) and (1.4) we can establish the expressions of Sμν as that


Hence the Lorentz transformation operators Sμν and Λ are distinct from the original expressions, and are perfect. Obviously the vector 


is just the spin angular momentum of the particles of spin S =1/2.

2.       Representing Maxwell Equations with The Above Matrices

We can also write Maxwell equations with the matrices (1.1) and (1.2) of σ and τ. The Maxwell equations containing the matrices (1.1) and (1.2) make the relations between Dirac equation and the equations of electromagnetic field more close.

We know that the tensor Fμν of electromagnetic field is represented by electric field E and magnetic field B as follows


We decompose Fμν  into the sum of six matrices, in which every matrix includes only one component of E or B and the other elements are zero, then replace the six matrices by using components of σ+τ and σ-τ. The tensor finally is written in the form:


Maxwell equations that are used to describe electromagnetic field are, in Caussian units,





The covariant form of Lorenz transformation of equations (2.4) and (2.5) can be written as


It is apparent that equation (2.7) contains the matrices of the intrinsic spins when we substitute expression (2.2) in the above equation.

We also define a matrix Gμν as follows to match Fμν, such that we can constitute a couple of equation (2.7) from equations (2.3) and (2.6).


By making the analogous analysis of establishing expression (2.2), the matrix Gμν also appears in such a form that contains σ and τ.


Base on the expressions (2.8) and (2.9), the Maxwell equations (2.3) and (2.6) can be simply taken in the form:


The pair of equations (2.7) and (2.10) is the complete Maxwell equations in which intrinsic spins are contained. The pair of the equations shows us to be more symmetries than the original forms. The equation (2.10) is only formally different but equivalent to the following covariant form of equations (2.3) and (2.6):


Under infinitesimal Lorentz transformation, four-vector Aμ are transformed according to


We still define transformation matrices Sκλ and operator Λ similar to (1.12):


We require that the Lorenz transformation of Aμ will be performed according to


Comparing the above formula with (2.12), we have


From equations (2.13) and (2.15) we derive


The Lorenz transformation of tensor Fμν will be performed by Λ and Λ-1 as follows


Calculating all the elements Ei and Bj of F’μν on the basis of the transformation formula and neglecting the terms of order εμκ ελν, we obtain


The transformation of Gμν is assumed just the same as the tensor Fμν :


We also calculate all the elements Ei and Bj of G’μν from (2.19), neglecting the terms of order εμκ ελν , the results coincide with (2.18) completely. It proved that formula (2.19) is correct, so Gμν is a tensor under Lorenz transformation. Equation (2.10) is a covariant form of Maxwell equations (2.3) and (2.6).

   According to the formulas (2.16), we concretely write each matrix of Sκλ as follows


comparing the above matrices with (1.1) and (1.2), we finally arrive at the formulas of Sμν of electromagnetic field.


It is clear for electromagnetic field the vector


is just the spin angular momentum of photon of spin S =1. The spin vector is contained in the forms (2.7) and (2.10) of Maxwell equations.

Equation (1.9), equations (2.7) and (2.10) disclose that, the Dirac equation and Maxwell equations bring two intrinsic spins with themselves. One is the vector – i[S23, S31, S12] and another one is the vector  – i[S14, S24, S34].

3. Runge-Lenz vector

    We know that there exists Runge-Lenz vector[5] when there exist Coulomb field. Runge-Lenz vector a satisfy


Here l is angular momentum.

It is easy to prove that the vector – i[S14, S24, S34 ] is Runge-lenz vector. For the Dirac equaton we have


Replace l withσ/2 and a with στ3/2 in formulas (3.1), we obtain


For Maxwell equations we have


Replace l with ( σ+τ)/2 and a with ( σ-τ)/2 in formulas (3.1), we obtain


(3.3) and (3.5) coincide with (3.1), thus the vector – i[S14, S24, S34 ] is Runge-lenz vector.

We have known that the vector – i[S23, S31, S12] is corresponding to the inherent magnetic property of particles and is called spin angular momentum. From (2.2) and (4.7) (see appendix (2)) we can also see that the vector – i[S14, S24, S34] should be corresponding to the inherent electric property of particles.


 (1): Transformation from ordinary representation of the Dirac equation to the above representation.

For Dirac matrices


there exist unitary matrix:


The Dirac matrices (3.1) can be transformed into the forms of expression (1.7) by the unitary matrix U:


  In addition, any 4x4 matrices A can be expressed by


where C0, Cij, Ck and Cl are arbitrary constants. It is easy to prove the above formula because the traces of matrices σi andτj and their productsσiτj are all equal to zero:


Hence we may construct a complete system of 4x4 matrices on the basis of the matrices of the operator σ and τ.

(2): Writing magnet field B and electric field E of Maxwell equations with potentials A and φ.

The magnetic field B and the electric field E generally are expressed by potentials A and φ


Here A is a vector potential and φ is a quantity called the scalar potential. We can rewrite (4.6) as


Where j=1,2,3. We assume E4=0, such that from (4.7) we get                                                                          


 It is just Lorenz condition of gauge transformation. In (4.7) we can see that the magnetic field is only associated with the spin angular momentum and the electric field is only associated with the Runge-Lenz vector.



[1] P. A. M. Dirac, The Principles of Quantum Mechanics, 2nd Ed, Oxford (1957)

[2] L. I. Schiff, Quantum Mechanics. 3rd Ed. McGraw-Hill Book Company (1963)

[3] L.D.Landau and E.M.Lifshits, Quantum Mechanics. Pergamon Press (1977)

[4] Albert Messiah, Quantum Mechanics. 2nd Ed, Unabridged Dover (1999)

[5] Thomas F.Jordan, Quantum Mechanics in Simple Matrix Form. New York (1986)