Supplement 4.5: Polarisation of Electromagnetic Waves: Stokes Vectors and Müller Matrices (2/2)

Optical interactions as 4×4 matrix: Müller Matrices

The elements of a Stokes Vector S change as they are multiplied by a 4×4 matrix. This matrix indicates interactions of light with an optical element, or another optical effect which affects intensity and polarisation. When marking the Stokes vector after the interaction - just like the Jones Vector in the preceding section - as a dotted quantity, it may be written:

( I y ' I z ' U' V' )=( a 11 ... ... a 14 ... ... ... ... ... ... ... ... a 41 ... ... a 44 )( I y I z U V ) S '=M S

The intensity transformation matrix M is the Müller Matrix.

In order to find the Müller Matrix for an interaction, one may (if it is not stated in literature) go back to the easier accsssible transformation of the field strengths

E y ' = a 1 E y + a 4 E z E z ' = a 3 E y + a 2 E z

and apply the dotted field strength to the definition of the Stokes parameters. After sorting the arising terms, elements of the Müller-Matrix become identifiable.

Example 1: Matrix of a retarder, fast axis in y direction
Example 2: Matrix of the linear polariser with diagonal orientation

In the way it was performed in the examples it is possible to find the Müller Matrices for the former discussed rotation of the y,z coordinates by the angle α. This leads to the rotation matrix:

R(α)=( cos 2 α sin 2 α 1 2 sin2α 0 sin 2 α cos 2 α 1 2 sin2α 0 sin2α sin2α cos2α 0 0 0 0 1 )

The matrices of some components with their basic orientation given in the right column can be transferred to an orientation at the angle α:

M(α)=R(α)MR(α)

The effect of several components in a row would be calculated in the very same manner as the Jones Matrices using the non-commutative product of each Müller Matrix, for n components:

M= M n M n1 ... M 2 M 1

In the last position of an optical setup, there is often a photodetector for measuring the intensity of light. The detector can be depicted as a row vector D =( 1 1 0 0 ) by the help of which the sum of the first two elements of a Stokes Vector are calculated:

I y '+ I z '=( 1 1 0 0 )( I y ' I z ' U' V' )=( 1 1 0 0 )M( I y I z U V )

The row matrix of the detector can be further detailed by spectral factors in order to take into account its wavelength sensitivity and the quantum efficiency.

 

Müller Matrix Component or interaction
P=( 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ) Linear polariser transmitting along the y axis
Q φ =( 1 0 0 0 0 1 0 0 0 0 cosφ sinφ 0 0 sinφ cosφ ) Retarder with a phase difference φ of the wave parts, fast axis in y direction
Q λ/4 =( 1 0 0 0 0 1 0 0 0 0 0 1 0 0 1 0 ) λ/4 retarder (λ/4 waveplate) with φ=π/2, fast axis in y direction
Q λ/2 =( 1 0 0 0 0 1 0 0 0 0 1 0 0 0 0 1 ) λ/2 retarder (λ/2 waveplate) with φ=π, fast axis in y direction
T= n 2 cos δ t n 1 cosδ ( t 2 0 0 0 0 t 2 0 0 0 0 t t 0 0 0 0 t t ) Fresnel refraction at a dielectric interface.
The quantities n1 and n2 are the refractive indices of the incident and the refracted medium. Fresnel coefficients as given in the section about the Jones Matrices.
X=( r 2 0 0 0 0 r 2 0 0 0 0 r r 0 0 0 0 r r ) Fresnel reflection at a dielectric interface.
The negative signs of the mixed components result from the directional change of the light propagation due to reflection. Fresnel coefficients as given in the section about the Jones Matrices.
Y=( r 2 0 0 0 0 r 2 0 0 0 0 c s 0 0 s c )

with the abbreviations

c= r r cosΔ
s= r r sinΔ
Fresnel reflection at a metallic surface.
Angles of incidence from 0 to 90° cause phase angles Δ from 0 to 180° between the orthogonal partial waves. All quantities are functions of the complex refractive index m=n-in', where n is the real refractive index and n' is the extinction coefficient of the metal, both depending on the wavelength.

A more detailed presentation is given in e.g. David Clarke: Stellar Photometry (Wiley-VCH, 2010), Appendix A.
Task: Light through a λ/4 retarder