6 Optical Thin Films

Reflectance as Vector Addition

The case of the soap bubble is illustrated from the

viewpoint of vectors.

When φ = 0, r1 = –1/6, and r2 = +1/6.

When φ = 90°:

When φ = 180° (one QWOT):

When φ = 360° (two QWOTs or one half-wave OT):

Graphics for Visualization of Coating Behavior 7

Reflectance Amplitude Diagram

The reflectance amplitude diagram, often referred to

as a circle diagram, follows from the foregoing vector
diagrams (p.6).

The outermost circle represents r = 1.0 and also R = rr* =

1.0 or R = 100% reflectance. The origin of the real and
imaginary axes is where r = 0.0 or zero reflectance.

Any new layer starts from the point of the reflectance

amplitude of whatever lies beneath it, whether that is a
substrate or a stack of coating layers on a substrate. This
point is represented by point A in the figure, where the
starting reflectance amplitude is rA and its phase is φA.
After the addition of a given physical thickness (PT = d)
and optical thickness (OT) of a layer of given index, the
resulting reflectance reaches point B. The new reflectance
amplitude is rB and its phase is φB. Point B would then be
the starting reflectance for the next layer.
8 Optical Thin Films

Admittance Diagram

The admittance (Y) of a medium, as applied to optical

thin films, is an electrical quantity, which is normalized to
be equivalent to the index of refraction. When the effective
index of a thin-film stack is plotted as a function of
increasing thickness from the substrate to the end of the
last layer, an admittance diagram is produced. This is a
conformal mapping of the reflectance amplitude diagram.
For many cases the diagrams look quite similar, except
that they have been rotated 180° about a point. In the
locus of homogeneous layers, nonabsorbing layers are
circles.

The range of an admittance diagram is the semi-infinite

plane from 0 to ∞ to the right on the real axis and ±i∞ on
the imaginary axis. The two-layer case of a QWOT of M
and L is plotted here for comparison. Note that point S is
at 1.52 –i0.0, the index of the substrate; C is at 1.90132 –
i0.0; and Z is at Y = Ne = 1.0 –i0.0, where r = (1.0 – 1.0) /
(1.0 + 1.0 ) = 0.
Graphics for Visualization of Coating Behavior 9

Electric Field in a Coating

The electric field within a coating layer is of importance

when laser damage thresholds are considered and also
when working with absorbing layers. In the latter case, the
amount of energy absorbed in the layer depends upon the
relative value of the electric field within that layer. One
aspect of the laser damage issue seems to be that the
interface between real deposited layers has some
absorption and defects that are more vulnerable if the
electric field is high at that interface.

The relative volts/meter is a function of the real value of

the admittance and can be calculated as follows:

E = 27.46 / [Re(Y)]0.5 .

This is very convenient to visualize on an admittance

diagram. The closer the locus of the coating gets to Y = 0,
the higher the electric field, and therefore the more
vulnerable the coating is to high energy flux. A laser stack
is sometimes designed with non-QWOT layers that do not
terminate toward the left on the real axis. Instead, the
locus stops short of, or continues past, the crossing of the
real axis until it is further away from the risks of high
electric field at Y = 0 .
10 Optical Thin Films

Admittance versus Reflectance

Amplitude Diagrams

The admittance diagram is generally as useful as the

reflectance amplitude or circle diagram, particularly
in the realm of low reflectance (and thereby low
admittance). However, the admittance diagram is not as
useful for high reflectors such as a stack like (1H 1L)4, as
can be seen in the figure below. For high reflectors of many
layers, the admittance tends to go off of whatever scale is
chosen for the plot, whereas the reflectance amplitude
diagram is constrained to the unit circle.

Reflectance Amplitude Diagram

Graphics for Visualization of Coating Behavior 11

Triangle Diagram

The energy falling on an optical thin film will be either

reflected (R), transmitted (T), or absorbed (A). Scattering
is ignored for the purposes of this section. It can then be
stated that
R+T+A=1.

A convenient way to visualize the properties of materials

with absorption, such as metals, is using a triangle
diagram. When there is no absorption (A), the
transmittance (T) is simply 1 – R and the diagram is not
particularly useful.

This figure shows the path in R, T, and A for a coating of

aluminum on glass. It starts at 96% T, 4% R, and 0% A on
the bare glass surface. As the aluminum thickness
increases, the locus moves downward. This happens to pass
through a point of about 40% T, 43% R, and 17% A. When
the film becomes opaque, it is at a point of 0% T, 90% R,
and 10% A.

These triangle diagrams are useful when a material has a

significant value of k, the imaginary index, as in metals
and semiconductors.
12 Optical Thin Films

Single-Layer Antireflection Coating

The most common single-layer antireflection coating

(SLAR) on glass is a QWOT of MgF2. If the index of the
glass is 1.52 and that of the MgF2 is 1.38:
r1 = (1.0 − 1.38 ) / (1.0 + 1.38 ) = −0.1597 ,

r2 = (1.38 − 1.52 ) / (1.38 + 1.52 ) = −0.0483 .

When the film is infinitesimally thin, φ = 0, so that

r = ⎡⎣( r1 + r2 *1) / (1 + r1 r2 *1) ⎤⎦ = −0.2064 ;

R = 4.260% ( same as bare substrate ) .

When the thin film is one QWOT, at the wavelength under

consideration, then φ = 180°:

r = ⎡⎣r1 + r2 * ( −1) ⎤⎦ / ⎡⎣1 + r1r2 * ( −1) ⎤⎦ = −0.1123 ;

.
R = 1.260% .

Thus, this SLAR reduces the reflection of the glass surface

from 4.26% to 1.26% (only at the QWOT wavelength, and
0° AOI; at other wavelengths and angles, R is modified).
The locus of this coating as seen on a reflectance amplitude
or circle diagram as it grows from zero to one QWOT is a
semicircle moving clockwise from point S to L.
Behavior of Some Simple AR Coating Types 13

Two-Layer AR Amplitude Diagram Example

In this example (in air), the substrate is of index 1.52 and

its reflectance amplitude point is at S. The first QWOT
layer is of index M = 1.70. The second QWOT layer is of
index L = 1.38. This two-layer coating results in zero
reflectance at the origin point, Z, for this design
wavelength (only).

The supporting calculations are

rsubstrate = (1.0 − 1.52 ) / (1.0 + 1.52) = −0.20635; RS = 4.528% .
After the deposition of the first layer:

r1 = (1.0 − 1.70 ) / (1.0 + 1.70 ) = −0.25926

r2 = (1.70 − 1.52 ) / (1.70 + 1.52 ) = +0.05590; ϕ = 180°

{
rC = ⎡⎣ −0.25926 ( − )( +0.05590 ) ⎤⎦ / ⎡⎣1.0 + ( − ) − 0.25926 * ( +0.05590 ) ⎤⎦ }
= −0.310658; neffective = 1.90132; RC = 9.651%

After the deposition of the second layer:

r1 = (1.0 − 1.38 ) / (1.0 + 1.38 ) = −0.15966

r2 = (1.38 − 1.90132 ) / (1.38 + 1.90132 ) = −0.15888; ϕ = 180°

{ (⎣ )
rZ = ⎡⎣ −0.15966 ( − )( −0.15888 ) ⎤⎦ / ⎡ 1.0 + ( − ) − 0.15966 * ( −0.15888 ) ⎤
⎦}
= −0.00080; RC = 0.000064%!
14 Optical Thin Films

Wavelength Effects

If the SLAR were designed to be just one QWOT at a 510-

nm wavelength, that same physical thickness (PT) would
be 0.729 QWOT at a wavelength of 700 nm, and 1.275
QWOT at 400 nm. This means that the semicircle shown
previously (p. 12) would be less than a semicircle for 700
nm and more than one for 400 nm. The 400 and 700 nm
terminations of the layer would be further from the origin
than QWOT at 510 nm, and therefore would have a higher
reflectance. Specifically, the reflectance would be 1.799% at
400 nm and 1.786% at 700 nm, as compared to 1.260% for
510 nm.

The resulting reflectance versus wavelength would appear

as the figure below.
Behavior of Some Simple AR Coating Types 15

Wavelength Effects (cont.)

The two-layer AR design shown previously had one QWOT

of index 1.70 and one QWOT of index 1.38 to bring the
reflectance to essen-
tially zero at the
design wavelength.
If this was designed
for 510 nm, both
layers would be
1.275× thicker at
400 nm and 0.729×
thinner at 700 nm
than one QWOT. The wavelength effects would be even
more exaggerated than the SLAR case as shown.

Coatings such as these are referred to as “V-coatings”

because of the V
shape of the
reflectance spec-
trum. However, if a
layer of index 2.2
and thickness of two
QWOTs were added
between the two
layers, the result is
shown.
This additional half- wave
OT layer is referred to as
an achromatizing layer
because it minimizes the
changes in reflectance
with color or wavelength.
This might best be
understood in the frame-
work of the reflectance amplitude or circle diagrams at
different colors/wavelengths. At the design wavelength of
510 nm, the first QWOT of index 1.7 moves the locus from
the substrate to the left and on the negative real axis. The
first QWOT of index 2.2 moves the locus from this point
again to the left and on the negative real axis.
16 Optical Thin Films

Wavelength Effects (cont.)

The second QWOT of index 2.2 moves clockwise back to

the beginning of the 2.2 layer. The half-wave layer is called
an absentee layer because it does not change the
reflectance, and therefore acts as though it were absent.
The last layer of index 1.38 moves the locus from this point
to the origin of the coordinates, r = 0.

At a wavelength of 650 nm, the layers are not as thick as a

QWOT, and therefore are not full semicircles. All of the
layers are too short, but the two QWOTs of the 2.2 index
layer compensate for this by opening in such a way as to
push the end point of the last layer toward the origin and
closer to r = 0.

At a wavelength of 450 nm the layers are thicker than a

QWOT, and therefore are more than full semicircles on
such a diagram. All of the layers are too long, but the two
QWOTs of the 2.2 index layer compensate for this by
closing in such a way as to pull the end point of the last
layer toward the origin.