Optical field calculation in 1D geometries

In this exercise, we will compute reflection and transmission of light as well as optical fields
profiles in 1D geometries. This corresponds to light propagating along the z axis in layers which
are uniform along the x and y axes. This is a scalar problem, and it is sufficient to consider a
single electric field component Ey. For light polarized along the y axis, Maxwell’s equations
reduce to the 1D scalar Helmholtz equation

 ∂2 
2 2
0,
 ∂z 2 + k0 n ( z )  E y ( z ) = (1)
 

where n is the refractive index and were we have assumed a harmonic time dependence e − iωt
with=k0 ω= / c 2π / λ , where λ is the free space wavelength. In the following, we will skip
the subindex y for simplicity.

Reflection and transmission at an interface

We consider the situation depicted in Fig. 1 of a 1D plane wave propagating along the z axis
incident on an interface between layers 1 and 2 with refractive indices n1 and n2 respectively.
The unusual choice of the orientation of the z axis from right to left will be justified later.

2 1
A2 A1

B1

z
Fig. 1. An incoming wave is illuminating the interface between materials 1 and 2. The interface is placed
at the position z = 0.

The forward propagating incoming wave A1 in layer 1 reaches the interface and is partly
reflected resulting in a backward propagating wave B1 and partly transmitted into the material
2 resulting in the forward propagating wave A2. The three waves are of the form
= A1 ( z ) a1ein1k0 z ( z < 0) (2)
=B1 ( z ) b1e − in1k0 z ( z < 0) (3)
A2 ( z ) a2ein2k0 z
= ( z > 0) . (4)
For convenience, we have placed the interface at z = 0, but this position can be arbitrary. The
total field in region 1 is E1 = A1 + B1 and in region 2 it is E2 = A2. We now require that the field
and its derivative are continuous at the interface, leading to the two conditions
E1 ( z = 0) = E2 ( z = 0) (5)
 ∂E1 ∂E
= 2 . (6)
∂z z = 0 ∂z z = 0
Given an incoming wave A1, we can determine a reflection coefficient r12 and a transmission
coefficient t12 defined by
B1 = r12 A1 (7)
A2 = t12 A1 (8)
at z = 0. The 12 indices indicate that the incoming wave is travelling from material 1 towards
material 2. Insertion of (2)-(4) into Eqs. (5)-(6) straightforwardly leads to
n −n
r12 = 1 2 (9)
n1 + n2
2n1
t12 = . (10)
n1 + n2
Similarly, one may consider reflection and transmission at the interface of an incoming wave
incident from layer 2 instead. The corresponding reflection r21 and transmission t21 coefficients
are obtained by simply switching the subindices 1 and 2 in (9)-(10).

2 1
A2 A1

B2 B1

z
Fig. 2. Interface between materials 1 and 2 for two incoming waves. The interface is at the position z = 0.

We now consider incoming waves from both sides of the interface as illustrated in Fig. 2. The
field is now given by
= A1 ( z ) a1ein1k0 z ( z < 0) (11)
=B1 ( z ) b1e − in1k0 z ( z < 0) (12)
in2 k0 z
A2 ( z ) a2e
= ( z > 0) (13)
− in2 k0 z
B2 ( z ) b1e
= ( z > 0) . (14)
The incoming and outgoing fields are related as
=A2 t12 A1 + r21B2 (15)
B1 r12 A1 + t21B2 .
= (16)
The relationship between the fields on each side of an interface can be described using the
transfer matrix formalism, where the interface transfer matrix T is defined by
 A2   A1  T11 T12   A1 
= B  T=  B  T T   B  . (17)
 2  1   21 22   1 
By isolating A2 and B2 on the LHS of (15)-(16), we obtain that
T11 T12  1 t12t21 − r12 r21 r21 
= T =   , (18)
T21 T22  t21  − r12 1 
and furthermore insertion of the expressions (9)-(10) and the corresponding ones for r21 and t21
into (18) leads to an expression in terms of the refractive indices for the interface transfer matrix
given by
T11 T12  1  n2 + n1 n2 − n1 
= T =   . (19)
T21 T22  2n2  n2 − n1 n2 + n1 

The transmission line matrix

We have calculated the interface transfer matrix T = T12 at the interface between materials 1
and 2. We also need the transmission line matrix TL for a layer of length L of constant refractive
index n as sketched in Fig. 3.

L 0
z
Fig. 3. Transmission along a uniform layer of thickness L.

The transmission line matrix TL relating the components at z = 0 to those at z = L is defined by

 A( L)   A(0) 
 B( L)  = TL  B(0)  . (20)
   
Inspection of e.g. (2) and (3) reveals that the components of the transmission line matrix TL are
given by
 eink0 L 0 
TL =  − ink0 L 
. (21)
 0 e 
The multi-layer structure
3 2 1
A3 A2 A1

B3 B2 B1

L 0

z
Fig. 4. Structure with three layers and two interfaces positioned at z = 0 and z = L.

We now consider the three layer structure depicted in Fig. 4. The interface between layers 1
and 2 is positioned at z = 0 and the interface between layers 2 and 3 is at z = L. The total transfer
matrix T13 for the entire structure is
 A3 ( L)   A1 (0) 
 B ( L)  = T13  B (0)  . (22)
 3   1 
The total transmission matrix characterizing this structure is constructed from the interface
matrix between layers 1 and 2, the transmission line matrix for layer 2 and finally the interface
matrix between layers 2 and 3. It is thus given by
T13 = T23TL 2T12 , (23)
where T12 and T23 are the transfer matrices for the interfaces between layers 1 and 2 and
between layers 2 and 3 respectively, and TL2 is the transmission line matrix for layer 2.

We observe that the unusual choice of the orientation of the z axis means that the matrices in
(23) are ordered in the same way as the interfaces and layers in Fig. 4 leading to an easy intuitive
understanding of the construction of the total transfer matrix.

N 1
AN A1

BN B1

LN 0

z
Fig. 5. Structure with multiple material layers positioned between z = 0 and z = LN.
Generalizing this to a structure with N layers as shown in Fig. 5, we straightforwardly obtain
that the transmission matrix between layer 1 and layer N is given by
T1N = TN −1, N TLN −1 ... T45TL 4T34TL 3T23TL 2T12 , (24)
where
 AN ( LN )   A1 (0) 
 B ( L )  = T1N  B (0)  . (25)
 N N   1 

Reflection and transmission of a multilayer structure

We now consider illumination of the structure depicted in Fig. 5 by an incoming wave A1 from
the right only. The structure is not illuminated from the left, so we set BN = 0. Using (25), we
can now compute the total reflection r1N and transmission t1N coefficients for the entire
structure defined by
B1 (0) = r1N A1 (0) (26)
AN ( LN ) = t1N A1 (0) . (27)
Inserting the definitions (26)-(27) into (25) and solving for r1N and t1N , we obtain
T21
r1N = − (28)
T22
T12T21
t1=
N T11 − , (29)
T22
where the subindices Txy refer to the matrix elements 11, 12, 21 and 22 of the total transmission
matrix T1N computed using (24). Similarly, by illuminating from the left using BN and setting
A1 = 0, we obtain
T12
rN 1 = (30)
T22
1
tN 1 = . (31)
T22
When considering the reflection and transmission properties of a structure, we are usually
interested in the power reflection and transmission spectra. The power reflection R1N and
transmission T1N coefficients are related the reflection and transmission coefficients as
2
R1N = r1N (32)
2 nN
T1N = t1N , (33)
n1
where n1 and nN are the refractive indices of the first and N’th layers. The refractive indices
appear in (33) since the power P of a wave aeinkz propagating in a material with refractive index
n is given by
n 2
P= a . (34)
2µ0 c
Exercise 1A – Reflection and transmission for a DBR

Fig. 6. Grating structure consisting of NLP = 8 alternating layer pairs with refractive indices n1 = 3.5 and
n2 = 2.9.

We now consider a 1D distributed Bragg reflector (DBR). A DBR is a periodic structure

featuring alternating layers of high index n1 and low-index n2 as illustrated in Fig. 6. The
structure has NLP layer pairs, and the first and last layers are high-index layers.

Calculate and plot the power reflection and transmission of the DBR as function of
wavelength in the interval 800 to 1000 nm for NLP = 8, 16 and 32. What is the effect of
increasing the number NLP of layer pairs on the reflection and transmission?

Geometrical parameters: Use n1 = 3.5 and n2 = 2.9. The thicknesses of the DBR layers should
be chosen as L = λd/(4n), where n is the refractive index of the layer and λd is the design
wavelength chosen as λd = 900 nm. This results in L1 ≈ 64.3 nm and L2 ≈ 77.6 nm.

You may use the reflection spectrum in Fig. 7 to check your code.

Fig. 7. Reflection spectrum for the grating structure for NLP = 8 and n2 = 2.9.
Verify that the structure is lossless, i.e. show that T1N + R1N = 1.

We now fix NLP = 32. While keeping the other parameters fixed, calculate and plot the
power reflection of the DBR as function of wavelength in the interval 800 to 1000 nm for
n2 = 2.7, 2.9 and 3.1. What is the influence of the index contrast ∆n = n1 – n2 on the
reflection spectrum?

Field profile
In addition, it is often of interest to visualize the optical field profile E(z) as function of z in the
entire structure. To this purpose, we consider the geometry illustrated in Fig. 8, where the origin
has been moved to the right side of layer 1 and where the structure is illuminated by an incoming
wave A1 only such that BN = 0. We can set the amplitude of the transmitted wave AN to an
arbitrary number and we set AN = 1. In this case, the boundary condition at z = LN is
 AN ( LN )  1 
 B ( L )  = 0 . (35)
 N N   

N q 1
AN A1

BN = 0 B1

LN zq Lq L2 L1 = 0

z
Fig. 8. Structure with multiple material layers. The grey arrows are discussed in the text below.

Since the fields on each side of the structure are related using (25), the corresponding field at z
= 0 is computed by inverting the transfer matrix such that
 A1 (0)  A ( LN ) 
−1  N
 B (0)  = T1N  B ( L )  . (36)
 1   N N 
We have now computed the incoming wave A1 and the outgoing wave B1 amplitudes at the
position z = 0. The procedure for calculating the field at any position z inside the structure is
now as follows. In layer 1 the field at a position z ≤ L2 is computed simply using the transmission
line matrix for layer 1, where L in (21) is replaced by z such that
 A1 ( z )   ein1k0 z 0   A1 (0) 
 B ( z) =   . (37)
 1   0 e − in1k0 z   B1 (0) 
The field at the position z = zq in an arbitrary layer q inside the structure as shown in Fig. 8 is
computed by considering the transfer matrix connecting the position z = 0 and the position z =
zq given by
 T( zq ) = TLq ( zq )Tq −1,q TLq −1 ... T45TL 4T34TL 3T23TL 2T12TL1 , (38)
where the final transmission line matrix TLq ( zq ) describes the transmission from the right side
z = Lq of layer q to the position z = zq such that
 einqk0 ( zq − Lq ) 0 
TLq ( zq ) =  − in k ( z − L )
. (39)
 0 e q 0 q q 
We also remark the presence of a transmission line matrix TL1 to the far right of (38) due to the
new position of the origin at the right side of layer 1.

Here we are constructing the transfer matrix (38) of the geometry while “ignoring” the grey
region to the left of the point z = zq in Fig. 8. When using this procedure to compute the field,
it could appear that the influence of the structure to the left of the point z = zq in Fig. 8 on the
field is ignored. However, we should keep in mind that the influence of the left side of the
geometry is taken into account through the initial calculation of the A1(0) and B1(0) coefficients
in (36), which involves the total transfer matrix T1N for the entire structure.

Exercise 1B – Field plot for a DBR

Using the geometrical parameters from before and using n2 = 2.9 and NLP = 32, compute
the field profile along the entire structure defined using the boundary condition (35). Plot
the total field amplitude E= ( z ) A( z ) + B ( z ) for the two wavelengths λ = 900 nm and λ =
965 nm.

What is observed in the two plots? What happens to the incoming light in the two cases?

You may use the field profile in Fig. 9 to check your code.

Fig. 9. Electric field amplitude |E(z)| and refractive index profile for the DBR for NLP = 8 and λ = 900 nm.
Exercise 2A – Reflection and transmission for a cavity
We now consider a cavity surrounded by two DBRs with NLP layer pairs each as illustrated in
Fig. 10. The cavity is chosen as a high index layer and the thickness of the cavity layer LCav is
chosen as one material wavelength such that as LCav = λd / n1. The geometrical parameters of
the DBRs are the same as previously.

Fig. 10. Cavity sandwiched between two DBRs consisting of NLP = 8 alternating layer pairs each.

Calculate and plot the power reflection and transmission of the DBR as function of
wavelength in the interval 800 to 1000 nm for NLP = 16.

What happens to the reflection and transmission at the design wavelength λd?

Exercise 2B – Field plot for a cavity

Plot the total field amplitude E=
( z) A( z ) + B ( z ) for the cavity defined by the boundary
condition (35) for the wavelength λ = 900 nm. What is observed?