Optical constants of far infrared materials.

3: plastics

Donald R. Smith and Ernest V. Loewenstein

Room temperature optical constants of plastic materials have been measured over the 50-350-cm'1 spec-
tral range. The materials reported include high density polyethylene, TPX, Aclar, Kapton, Surlyn, and
Mylar. All except TPX are available in sheet form and exhibit birefringence as a consequence of stretch-
ing during the manufacturing process. Only the average of the two sets of optical constants is reported for
each material. The refractive index was calculated from the channeled spectrum as observed in reflection
from the sample, while the absorption coefficient was determined, in all cases but polyethylene, from a
transmission measurement.

1. Introduction set up to measure a reflection and a reference (alumi-

In two previous papers"12 we have discussed a
method for determining far ir optical constants of
solids from channeled spectra and presented data on
certain crystalline materials. In this paper we report
optical constants of a number of useful and readily
available plastic materials. Some of these materials
have been used for window, beam splitter, substrate,
or lens material, and those that have nbt been so used
may well find application. All 'the materials are
available as films except for TPX, which comes as
sheet, rod, or pellets for molding. The plastics most
widely used for far ir optical applications are polyeth-
ylene, Mylar (polyethylene terephthalate), and TPX
(4-methyl pentene-1). The latter is valuable as lens
material because of its unique property of being high-
ly transparent in both the visible and far ir region
with nearly the same refractive index in both spectral
regions.3 The optical constants of Mylar are tabulat-
ed in this paper, although they were already reported
graphically in Ref. 1. In addition to the above, this
publication reports the optical constants of polyeth-
ylene, Aclar, Kapton, and Surlyn.
A. Experimental Procedure
The procedure used in the measurements reported
here is the same as that reported in Ref. 1. All the
measurements were made in reflection because, as
was shown there, the contrast of the channeled spec-
trum fringes from low index materials is substantially
greater when viewed by reflection than by transmis-
sion. A special time shared double beam system was
The authors are with the Optical Physics Laboratory, Air Force
Cambridge Research Laboratories, Bedford, Massachusetts 01731.
Received 21 January 1975.
num mirror) interferogram simultaneously. A mir-
rored chopper with an open segment in it was placed
near a focal point in the optical system and the sam-
ple mounted behind the chopper so that the ir radia-
tion is alternately reflected from the sample and from
the mirror. The converging radiation was incident at
an angle of 6.5° to the normal in a cone of half angle
equal to approximately 6.5°. A correction was ap-
plied to the observed positions of the zero crossings
of the channeled spectra, as described in Ref. 1, to
compensate for both nonnormal incidence and the
beam convergence. As all the plastic samples
showed some degree of optical anisotropy, all the
measurements were carried out in polarized radia-
tion. A pile-of-polyethylene-sheets polarizer with an
observed polarizance exceeding 96% over the entire
spectral region was inserted into the beam to reject
the radiation polarized perpendicular to the plane of
incidence on the beam splitter. This arrangement
utilizes the stronger polarization in the interferome-
ter, 4 as the beam splitter introduces polarization into
the light beam. The 6-,um thick Mylar beam splitter
permits continuous coverage of the 50-350-cm-1
spectral region, which is the region for which optical
constants are reported in this paper.
B. Thickness Measurement
The thickness of the sample enters directly into
the determination of the refractive index when the
channeled spectrum method is used, and therefore
the need for precise thickness measurements is ap-
parent. Special problems were encountered in mak-
ing accurate measurements of these materials be-
cause of their softness. It was found that the probe
of the electronic comparator used would indent
slightly the surface of the sample being measured.
The procedure adopted was to place each sample be-

June 1975 / Vol. 14, No. 6 / APPLIED OPTICS 1335

 tween a pair of parallel-faced sapphire disks, measure
the thickness of the sandwich, then remove the sam-
ple and measure the sapphire only, and subtract the
thickness of the sapphire. The measurements, which
were all referred to a set of precision gauge blocks,
.yielded thickness values that are considered accurate
to about 1 im. All the samples were somewhat
wedged, with thickness variations across the surface
amounting in some cases to several micrometers, but
the average value is a sufficiently good representation
of the sample thickness for the precision of the values
reported herein.
C. Optical Properties of Polymer Films
Plastic films have complex structures that can be
treated as a mixture of crystalline and amorphous
phases. The picture generally accepted for the
structure of crystalline long-chain polymers is that
portions of many molecules are packed side by side in
a precise crystalline fashion, each molecule passing
through several crystalline regions, the molecules
being much longer than the crystalline regions. In
polyethylene, for example, the molecules are typical-
ly a few thousand angstroms in length, whereas the
crystalline regions are an order of magnitude smaller.
The amorphous regions consist of the portions of the
chain molecules that tie one crystalline region to the
next.
Commercial polymer films are oriented as a means
of improving their strength and durability. In gener-
al, commercial film is produced by quenching extrud-
ed film to the amorphous state and then orienting the
film by reheating and stretching the sheet approxi-
mately threefold in each direction. In order to stabi-
lize and reduce the tendency to shrink on heating,
the film is then usually annealed under restraint.
These operations, in addition to producing orienta-
tion of the molecules, also increase the crystallinity of
the film by producing small crystallites.
Very little work has been published on the optical
properties of polymers, probably because these prop-
erties are dependent upon many manufacturing pa-
rameters that are difficult to measure and are not
usually under the control of the investigator. Pa-
rameters such as the stretching temperature, stretch-
ing rate, amount of stretch, geometry, prior and post-
thermal treatments, polymer purity, and molecular
weight all contribute to the over-all orientation and
crystallinity of the finished film.
Examination of our samples under a polarizing mi-
croscope showed that all the materials except poly-
ethylene and TPX were optically biaxial. The two
exceptions were much thicker than the other films in-
vestigated, however, and thinner films would almost
certainly exhibit the birefringence typical of these or-
iented materials. In fact birefringences have been
observed in thinner films of polyethylene. In view of
the paucity of knowledge about the exact structure of
the bulk material and its relationship to the optical
properties we concluded that it would be most useful
to publish only an average value of the optical con-
1336 APPLIED OPTICS / Vol. 14, No. 6 / June 1975
stants, obtained by measuring at 450 to the principal
axes as observed in the visible. All samples were
measured in polarized light at the 450 position and in
a position perpendicular to that position. No signifi-
cant difference in the optical constants could be ob-
served for the samples between these two positions,
and the results were therefore averaged. This fact
seems strong evidence that the principal axes remain
unchanged in the far ir and the visible region.
II. Results
A. High Density Polyethylene
By far the commonest plastic material being used
for far ir optical components is polyethylene. It has
a nearly constant refractive index over a large wave-
length range and a reasonably low absorption coeffi-
cient. It is inexpensive, readily available in film,
sheet, and powder or pellet form, and is easy to work
into lenses or other shapes. Its principal drawback is
the fact that it is not transparent in the visible, which
makes it difficult to align optical systems incorporat-
ing thick polyethylene transmitting components.
This opacity is a consequence of the high degree of
crystallinity of the material.
The sample used for the measurements reported
here was approximately 1 mm thick and is believed
to be relatively unoriented because no birefringence
was observed under the polarizing microscope. The
optical constants are graphed in Fig. 1 and tabulated
in Table I. The absorption coefficient values are
somewhat more uncertain in the case of polyethylene
than in the other plastics reported herein, owing to
the fact that we had only reflection data available
1.53 r
POLYETHYLENE
1.52
w
z
1.51
10
Le
0
0
5
0
In
I I I
300
100 200
WAVENUMBER (cm1 )
Fig. 1. Optical constants of polyethylene.
 Table I. Optical Constants of High Density Polyethylenea B. Methylpentene
Refractive Absorption Coeff. Poly (4-methyl pentene-1), a polyolefin resin, is
Wavenumber Index (cm-l) marketed under the tradename TPX by the British
firm, Imperial Chemical Industries, and is not avail-
50 1.518 1.4
able from any American manufacturer. Its impor-
60 1.519 1.4
1.521 1.6
tance for the far ir is a consequence of the close
70
80 1.521 1.7 match between the visible and far ir refractive in-
90 1.521 1.8 dices, as well as its high transparency in both spectral
100 1.521 2.0 regions. 3 The material can be obtained in rods and
110 1.520 2.3 sheets as well as in pellet form, but is not available as
120 1.520 2.5 a film. It is readily molded into windows or other
130 1.520 2.9 shapes, but the heating must take place in a vacuum
140 1.521 3.2 oven to eliminate discoloring which seriously affects
150 1.521 3.6 the visible transparency. This discoloration may or
160. 1.521 3.9
may not affect the ir transparency.
170 1.521 4.3
4.8
The optical constants of TPX, shown in Fig. 2, in-
180 1.521
190 1.521 5.4 dicate that it is an ideal material for optical compo-
200 1.521 6.0 nents. The refractive index is nearly constant and is
210 1.521 7.0 coupled with a low, featureless absorption coefficient.
220 1.520 8.0 We have observed no birefringence in our samples.
230 1.519 7.8 The values of a reported in Ref. 3 are lower than ours
240 1.519 7.1 by about a factor of 2. The reason for this discrepan-
250 1.519 6.4 cy is not known.
a The refractive index is constant to within the experimental un-

certainty of 0.1%. The uncertainty in the absorption coefficient

is substantially larger, about 10%.

Table II. Optical Constants of TPXa

Wavenumber Refractive Absorption

(cm-I) Index Coefficient (cm-')
TPX
1.455 _ 50 1.447 6.0
60 1.446 6.3
70 1.446 6.4
80 1.446 6.4
x 90 1.446 6.2
w 1.450 _ 6.1
a 100 1.446
z 110 1.446 6.2
120 1.446 6.5
130 1.447 6.7
I I I I 140 1.447 6.7
1.445 .

150 1.447 6.8

160 1.447 6.7
170 1.448 6.8
10 1.448 6.8
180
IL 190 1.449 6.9
200 1.450 7.1
0 210 1.451 7.4
0 220 1.451 7.8
w
cn 5
m 230 1.451 8.2
240 1.450 8.3
0
C,) 250 1.449 8.3
260 1.449 8.2
I I I I I I I 270 1.449 8.3
0 100 200 300 280 1.449 8.7
290 1.449 8.9
WAVENUMBER (cMl) 300 1.449 8.9
310 1.449 8.7
Fig. 2. Optical constants of TPX.
320 1.448 8.5
330 1.448 8.6
340 1.450 8.8

a The refractive index is constant to within the experimental un-

and were forced to average over a large number of certainty of 0.1% . The uncertainty in the absorption coefficient
spectra. is 5%.

June 1975 / Vol. 14, No. 6 / APPLIED OPTICS 1337

 Table ll. Vibrational Frequencies of CTFE (Aclar) C. CTFE (Poly-chlorotrifluoroethylene)
Liang & Krim5 This worka Poly-chlorotrifluoroethylene [also known as poly-
102 cm -
chlorofluorethylene or (CFE)] is marketed under the
115 cm-'
185 200 tradenames Aclar (Allied Chemical) and Kel-F (Min-
235 240 nesota Mining and Manufacturing). It differs from
297 300 Teflon (poly-tetrafluoroethylene) in that a chlorine
335 340 atom is substituted for one of the fluorines in the mo-
nomer, which breaks up the high symmetry of the
a There is indication of an additional absorption peak below

50 cm', which was not observed in either work. Our absorption

molecule. The lowered symmetry results both in less
curve rises below 85 cm-', but the peak apparently lies outside close packing of the chains and a lower tendency to
our region of observation. crystallization which, in turn, permits the production
of transparent films. A study by Liang and Krim5
indicates that there are five vibrational frequencies
to be observed between 50 cm'1 and 350 cm-1 .
Table III summarizes the frequencies of absorption
maxima as observed by us and by Liang and Krim.
The optical constants are illustrated in Fig. 3 and 4
and tabulated in Table IV.
5 ACLAR
1. or

w 1.46
C
Z

.4 . ,

Cn
Table IV. Optical Constants of CTFE (Aclar)a
0 0.05
Z
Absorption
U- 0.04
0 Wavenumber Refractive coefficient
Lii
C.)
0.03 (cm-,) index (cm-)
z
0.02 50 1.471 12.2
U- 0.01 60 1.474 9.0
0 70 1.479 5.8
80 1.483 3.5
0 100 200 300
WAVENUMBER (crr')
90 1.487 3.9
100 1.488 7.0
Fig. 3. Refractive index of CTFE (Aclar) and the difference n - 110 1.486 10.3
n0 of the two accessible refractive indices. 120 1.484 10.3
130 1.482 7.1
140 1.484 3.5
150 1.488 2.0
160 1.490 5.9
170 1.492 13.5
180 1.493 24.0
190 1.491 32.5
200 1.489 34.7
I| ACLAR 210 1.488 33.4
T'E 60 220 1.488 35.8
230 1.488 42.2
U- 240 1.486 50.7
250 1.483 47.5
o 40 l 260 1.483 42.8
0
270 1.487 41.5
z 280 1.489 43.0
0
I- 290 1.491 45.0
0a- 20 H 300 1.492 46.0
310 1.493 45.9
U/)
m 320 1.494 46.8
330 1.495 48.9
I I I I
340 1.497 52.3
0 100 200 300 400 350 1.497 50.1
WAVENUMBER (cm-' )
aUncertainty in index is 0.3% (systematic) resulting from a
Fig. 4. Absorption coefficient of CTFE (midway between optic similar accuracy in the thickness of the sample. Uncertainty in
axes). the absorption coefficient is 50.

1338 APPLIED OPTICS / Vol. 14, No. 6 / June 1975

 Table V. Optical Constants of Kapton (Polyimide)a KAPTON
Refrac- Absorp- Refrac- Absorp-
Wave- tive tion coeffi- Wave- tive tion coeffi-
cient number index cient ^ 30(
number index E
a1 77 0 27 a

50.0 - - 205.0 1.726 102.2

210.0 1.725 88.9 , 201
55.0 1.769 - 0
60.0 1.770 3.1 215.0 1.726 72.9
1.781 7.5 220.0 1.729 70.4 z
65.0 0
70.0 1.784 7.5 225.0 1.723 61.4 10,
75.0 1.785 7.8 230.0 1.737 60.5 0
80.0 1.787 9.5 235.0 1.743 59.2 U,
o
85.0 1.793 8.3 240.0 1.749 66.1
90.0 1.797 9.5 245.0 1.758 80.1
95.0 1.806 16.0 250.0 1.758 96.1 WAVENUMBER (cm' )
100.0 1.804 23.4 255.0 1.745 155.9
105.0 1.808 32.2 260.0 1.733 372.3 Fig. 6. Absorption coefficient of Kapton.
110.0 1.805 42.2 265.0 1.721 126.0
115.0 1.804 49.3 270.0 1.712 113.9
120.0 1.806 51.6 275.0 1.716 85.2
125.0 1.805 59.1 280.0 1.723 65.3
130.0 1.815 71.3 285.0 1.728 57.8
135.0 1.811 84.8 290.0 1.730 51.1
140.0 1.820 108.8 295.0 1.733 44.5
145.0 1.808 146.3 300.0 1.734 39.5
150.0 1.793 153.5 305.0 1.736 38.1
155.0 1.792 171.1 310.0 1.739 39.2
219.6 315.0 1.740 42.4 Table Vl. Optical Constants of Surlyna
160.0 1.755
165.0 1.738 254.5 320.0 1.743 48.9 Refrac- Absorp-
Absorp-
170.0 1.723 202.1 325.0 1.742 54.6 Wave- tive tion co-
Wave- Refractive tion co-
175.0 1.709 128.2 330.0 1.745 53.6 efficient number index efficient
number index
180.0 1.716 102.3 335.0 1.746 48.4 a 27 a
a' 27 a
185.0 1.723 97.5 340.0 1.748 48.8
1.726 97.5 345.0 1.751 42.5 50.0 1.5192 9.8 205.0 1.5003 78.6
190.0
1.730 94.5 350.0 1.754 40.0 55.0 1.5197 10.2 210.0 1.4984 80.1
195.0
200.0 1.730 102.4 60.0 1.520-2 10.7 215.0 1.4968 81.3
65.0 1.5205 11.4 220.0 1.4956 82.2
The estimated uncertainty in a refractive index is 0.5%, in the 70.0 1.5207 12.0 225.0 1.4946 82.5
absorption coefficient it is 5%. 75.0 1.5203 12.8 230.0 1.4935 82.2
80.0 1.5200 13.6 235.0 1.4923 81.2
85.0 1.5195 14.6 240.0 1.4908 79.5
90.0 1.5190 15.6 245.0 1.4893 76.7
95.0 1.5183 16.8 250.0 1.4878 72.1
100.0 1.5177 18.2 255.0 1.4869 67.6
105.0 1.5171 19.5 260.0 1.4865 63.4
110.0 1.5167 21.1 265.0 1.4865 59.4
115.0 1.5162 23.0 270.0 1.4869 56.0
120.0 1.5157 25.0 275.0 1.4875 53.1
125.0 1.5153 27.0 280.0 1.4882 51.3
130.0 1.5149 29.4 285.0 1.4890 50.3
135.0 1.5145 31.8 290.0 1.4897 50.0
140.0 1.5140 34.3 295.0 1.4903 50.3
145.0 1.5136 37.1 300.0 1.4907 51.2
x 52.2
150.0 1.5132 40.0 305.0 1.4910
0
z 155.0 1.5129 43.0 310.0 1.4911 52.8
160.0 1.5124 46.2 315.0 1.4910 52.9
w
I-
165.0 1.5119 49.7 320.0 1.4909 52.6
4 170.0 1.5112 53.5 325.0 1.4908 51.9
t0i 175.0 1.5104 57.4 330.0 1.4907 50.9
W
C)
61.6 335.0 1.4909 49.6
180.0 1.5095
185.0 1.5083 65.8 340.0 1.4911 48.0
190.0 1.5066 69.5 345.0 1.4914
40 120 200 280 360 195.0 1.5046 73.3 350.0 1.4916
WAVENUMBER (cm' ) 200.0 1.5025 76.4

V. 5. Rfracft riv indpx of Kanton (nolvimide). This is the aver- Estimated uncertainty in refractive index is 0.1%, in absorp-
age of the two accessible indices; see text. tion coefficient 5%.

June 1975 / Vol. 14, No. 6 / APPLIED OPTICS 1339

 Table VII. Optical Constants of Mylar IOMIL SURLYN
(Polyethylene Terephthalate)a
Wave-
number Absorption
(cm-,) Refractive indices coefficients 1.55t
(7 ?II n2 2

50 1.717 1.752 13 13
60 1.715 1.756 28 21
70 1.712 1.762 36 27 X
80 1.710 1.761 45 40 .54
z
90 1.703 1.750 55 50
100 1.694 1.744 52 45
110 1.692 1.746 46 42
120 1.697 1.750 47 48
125 1.753 51 55 Li.
130 1.696 1.754 57 68 Id 1.531-
135 1.741 63 84
140 1.688 1.727 67 123
145 1.720 60 83
150 1.682 1.716 55 75
160 1.684 1.721 49 59
170 1.688 1.727 52 58 1.52
100 200 300
180 1.688 1.728 60 65
190 WAVENUMBER (cmiT')
1.687 1.729 64 69
200 1.688 1.729 62 66 Fig. 7. Refractive index of Surlyn.
210 1.691 1.732 -65 64
220 1.693 1.736 75 68
230 1.691 1.737 91 77
240 1.686 1.736 95 87
250 1.686 1.736 91 87
260 1.688 1.737 95 90
270 1.688 1.740 106 99
280 1.688 1.742 . 120 110
290 1.688 1.742 122 118
300 1.691 1.744 122 109
310 1.697 1.747 127 112 IOMIL SURLYN
320 1.705 1.750 133 119 E0 TRANSMISSION
330 1.713 1.752 138 132 80 x REFLECTION
340 1.721 1.753 144 151
a Estimated uncertainty in the index is 0.1%, in absorption
coefficient 3%. These data are shown graphically in Figs. 8 and 9 70 _
of Ref. 1. In this case both accessible optical constants are tab-
ulated to maintain consistency with the illustrations.
E 60 -

U
IL 50 _
D. Polyimide w
0
Kapton is -a transparent gold colored polyimide 0
film manufactured by DuPont, whose principal ap- z 40 _
plication is as a high temperature insulating material 0
in electrical equipment. A rather large observed bi-
refringence leads us to assume that this material is 30 -
0
highly oriented, but as with the other materials, our (n
measurements were made at 450 to the optic axis, M:
20 _
yielding only average values for the optical constants.
Identical values of optical constants were measured
when the sample was rotated by 90°. 101_
In Kapton we observed the largest values for the 0

refractive index, absorption coefficient, and disper-

sion of any material we studied. The refractive 0 100
, ,

200 300
index and the corresponding absorption coefficient
are illustrated in Figs. 5 and 6. Two strong absorp- WAVENUMBER (cmr')
tion bands occur at 164 cm-' and 259 cm-' with sev- Fig. 8. Absorption coefficient of Surlyn. The encircled points
eral weaker bands occurring at 110, 150, 200, and 330. were measured in transmission, the X's in reflection.

1340 APPLIED OPTICS / Vol. 14, No. 6 / June 1975

 The large absorption coefficient and high disper-
sion of Kapton make it unlikely that this material
will find wide use as an optical material in the far ir
region. One possible exception would be the use of a
thin film as a beam splitter material. Even this ap-
plication seems unlikely, however, since there is little
significant difference between the characteristics cal-
culated for this material and those for a Mylar beam
splitter of equal thickness.
E. Surlyn
Surlyn is the brand name for an ionomer polymer
manufactured by DuPont. The major constituent of
this film is ethylene, but the linking in Surlyn is af-
fected by the addition of sodium, potassium, or other
ions. Still Surlyn exhibits many properties of poly-
ethylene. The optical properties of Surlyn in the far
ir region, however, are different from those exhibited
by polyethylene. Whereas polyethylene exhibits
very little dispersion and low absorption, Surlyn on
the other hand exhibits a relatively large dispersion
and high absorption due to the medium strength ab-
sorption at 225 cm-'. Also, unlike polyethylene,
Surlyn is highly transparent in the visible region.
The refractive index determined from reflection mea-
surements on a 247.8-,um thick sample designated as
Surlyn A-1601 by the manufacturer is illustrated in
Fig. 7. As in the other cases the sample was oriented
at 45° to the optical axis. The absorption coefficient
was determined both from reflection and transmis-
sion measurements, and both these results are shown
in Fig. 8. The transmission measurements are
thought to be more accurate, but the difference be-
tween the two values is probably insignificant in this
case.
References
1. E. V. Loewenstein and D. R. Smith, Appl. Opt. 10, 577 (1971).
2. E. V. Loewenstein, D. R. Smith, and R. L. Morgan, Appl. Opt.
12, 398 (1973).
3. G. W. Chantry, H. M. Evans, J. W. Fleming, and H. A. Gebbie,
Infrared Phys. 9, 31 (1969).
4. E. V. Loewenstein and A. Engelsrath, J. Phys. (Paris) C-2,
Suppl. 3-4 28, 153 (1967).
5. C. Y. Liang and S. Krim, J. Chem. Phys. 25, 563 (1956).

The Society of Photographic Scientists and Engineers has

established a scholarship fund to be known as the Raymond Davis
Scholarship Fund. The proceeds of this fund will be used to
award scholarships to students of photographic science or
engineering.

Raymond Davis achieved international recognition for the

photographic research and standardization he carried on at
the National Bureau of Standards from 1911 to 1958. He was
one of the founders and the first president of the Society of
Photographic Engineers, which joined with the technical section
of the Photographic Society of America to form the Society of
Photographic Scientists and Engineers, in 1957.

He was a fellow of the Optical Society of America and represented

OSA on various standards committees from 1938 until recently.

Mr. Davis died in September 1974, and the members of the Society
feel that this scholarship program is a fitting memorial.

Contributions to this fund should be made out to the SPSE/Raymond

Davis Scholarship Fund, and should be sent to Robert Wood,
Executive Director, Society of Photographic Scientists and Engineers,
1330 Massachusetts Ave., N.W., 'Washington, D.C. 20005. Since the
Society is a non-profit scientific and educational institution,
contributions are tax deductible.

June 1975 / Vol. 14, No. 6 / APPLIED OPTICS 1341