Content uploaded by Geoffrey Torrington
Author content
All content in this area was uploaded by Geoffrey Torrington on Jan 08, 2015
Content may be subject to copyright.
Space radiation testing of radiation resistant glasses and crystals
Tammy D. Henson and Geoffrey K. Torrington
Sandia National Laboratories, P.O Box 5800, Albuquerque, NM 87185-0972
ABSTRACT
With the number of cerium doped radiation resistant glasses available to the designer of space optics rapidly decreasing, it is
critical to identify and characterize all potential sources of radiation resistant glasses and crystals. Unfortunately much of the
data on radiation testing of glasses is quite old and often not completed at very high dose rates as might be experienced by an
unshielded space optic in orbit for many years. In addition, many optical glasses and crystals are manufactured today with
much higher purity than in the past in order to increase their ultraviolet transmission properties. Consequently these glasses
are much more resistant to space radiation than in the past. In this paper we will present gamma radiation effects on the
transmission properties of today’s fused silica, sapphire, calcium fluoride, barium fluoride, Schott cerium doped radiation
resistant glasses, Schott colored glass filters, as well as some infrared glasses with up to a 10 Mrad dose.
Keywords: Space radiation, radiation resistant glasses, crystals, space optics
1. INTRODUCTION
Space optics must be able to survive without performance degradation in the space environment over the expected lifetime of
the instrument. Therefore, the optical designer must only use optical materials that will not degrade in the expected space
environment. Space radiation induces a loss in optical transmission in many optical materials. The loss in transmission can
be quite severe depending on the material and the radiation dose. In addition space radiation has been shown to change the
refractive index of several optical materials
1-4
. The dimensional stability of mirror substrates is also a concern. Several
researchers have found that some mirror substrates, such as Zerodur, are subject to compaction under the influence of space
radiation thus changing their radius of curvature
5-6
. This paper describes the results of transmission degradation in several
optical glasses and crystals after exposure to gamma radiation.
2. MATERIALS AND TESTING PROCEDURE
Several optical materials were selected for the gamma
radiation experiments as shown in Table 1. The Schott
colored filter glasses were chosen for testing because
these glasses are often needed in the design of bandpass
filters operating in the visible and near infrared portion of
the spectrum for space-based multispectral instruments.
There seems to be very little information in the literature
on the radiation resistance of these colored glasses. The
listing of the Schott cerium stabilized (or radiation
resistant) glasses includes all glass types that are
available today from Schott with the exception of
LF5G15. Infrared glasses should inherently be radiation
resistant due to their lack of transmission in the
ultraviolet and blue portion of the spectrum. However,
since no previous data on radiation effects of infrared
glasses could be found in the literature they were
included in these experiments. In addition we included
some optical materials with wide transmission ranges
which are often needed for space multispectral
instruments. All the materials were uncoated.
A Co
60
source was used for the gamma-ray irradiation.
The typical gamma dose rate was 218 rads/sec. The
Glass Type Glass
Schott Colored Filter Glasses BG38
GG495
OG590
RG715
RG850
Schott Radiation Resistant Glasses BK7G18
K5G20
F2G12
LaK9G15
SF8G07
SF6G05
Infrared Glasses Zinc Sulfide (Cleartran)
Zinc Selenide
Silicon
Germanium
Wide Transmission Range Glasses Synthetic Fused Silica
Sapphire
Barium Fluoride
Calcium Fluoride
Table 1. Optical materials used for gamma radiation experiments.
Inorganic Optical Materials III, Alexander J. Marker III, Mark J. Davis, Editors,
Proceedings of SPIE Vol. 4452 (2001) © 2001 SPIE · 0277-786X/01/$15.00
54
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 01/07/2015 Terms of Use: http://spiedl.org/terms
Schott colored filter glasses were irradiated at room temperature from 30 Krads up to 10 Mrads. All other optical materials
were irradiated at room temperature from 300 Krads up to 10 Mrads.
The optical transmission spectra of all the optical materials were measured prior to irradiation at room temperature using a
Carry 500 spectrophotometer. The spectral transmission range varied depending on the glass type. The spectral transmission
range for the Schott colored filter glasses was from their cut on wavelength up to 1100 nm. The spectral transmission range
for the Schott radiation hardened glasses was from 400 to 1100 nm. The spectral transmission range for the infrared glasses
was from their cut on wavelength up to 2700 nm. The spectral transmission range for synthetic fused silica was from 200 to
2000 nm, for sapphire was from 250 to 2700 nm, for calcium fluoride was 200 to 1100 nm for 3 windows and 200 to
2700 nm for one window, and for barium fluoride was from 200 to 1100 nm. The spectral transmission measurements were
repeated shortly after each radiation exposure. The nominal thickness for the Schott colored filter glasses and the fused silica
sample was 3 mm, for the Schott radiation hardened glasses and the infrared glasses was 5 mm, for sapphire was 6.35 mm,
for the calcium fluoride samples was between 5 and 10 mm, and for the barium fluoride samples was between 5 and 7 mm.
3. RESULTS AND DISCUSSION
3.1 SPECTRAL TRANSMISSION CHANGES
3.1.1 Schott Colored Filter Glasses
Schott colored filter glass is often used to provide out-of-band blocking for bandpass filters operating in the visible and near
infrared portion of the spectrum. The spectral properties of these glasses change with exposure to gamma radiation.
Specifically the short wavelength cut-on edge shifts to longer wavelengths and the overall transmission is reduced in the
transparent region. For multispectral space optical systems employing this type of bandpass filter, the radiation environment
may change the spectral characteristics of the filter. This could be very problematic, especially for systems that require a
high degree of spectral calibration. In this case some sort of radiation shielding would be required to protect the colored glass
from radiation, or colored glasses should not be used in the construction of the bandpass filters.
BG38 filter glass is an ionically colored bandpass filter with
transmission in the blue-green portion of the spectrum. The
spectral response of this filter glass before and after radiation is
shown in Figure 1. After exposure to 30 Krads of gamma
radiation, the blue wavelength cut-on shifted to longer
wavelengths by 3 nm and the peak transmittance was reduced by
about 0.5% at 500 nm wavelength. After exposure to 100 Krads,
the cut-on wavelength shifted about 8 nm and peak transmittance
was reduced by 1%. This trend continued, and after exposure to
10 Mrads, the cut-on wavelength shifted by 45 nm and the peak
transmittance was reduced by 9.5%. The long wavelength cut-off
wavelength remained unchanged with radiation exposure.
GG495, OG590, RG715, RG850 filter glasses are colloidally
colored longpass filters. The spectral response of these filter
glasses before and after radiation are shown in Figures 2 - 5.
After exposure to 30 Krads of gamma radiation, the blue
wavelength cut-on of GG495 and OG590 shifted to longer
wavelengths by about 1 nm and the peak transmittance was
reduced by 1%. For RG715 the blue edge was not shifted but the
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
300 400 500 600 700
Wavelength (nm)
External Transmission (Percent)
0 rads
30 Krad
100 Krad
300 Krad
1 Mrad
3 Mrad
10 Mrad
Figure 1. Transmission measurements of a Schott BG38
colored glass filter after exposure to gamma radiation
(t = 3.06 mm).
peak transmittance dropped by about 4%. For GG495, OG590, and RG715 the slope on the short wavelength edge was
reduced after exposure to radiation, therefore decreasing the transmittance of wavelengths near the edge by as much as 8%.
The RG850 was relatively unaffected by the 30 Krads radiation exposure. After exposure to 100 Krads, the cut-on
wavelength of the GG495 and OG590 shifted about 2 nm and peak transmittance was reduced by 1.5-2.5%. For the RG715
the cut-on wavelength again remained about the same but the peak transmittance was reduced by 10%. For GG495, OG590,
and RG715 the slope on the short wavelength edge was reduced, decreasing the transmittance of wavelengths near the edge
by as much as 19%. For RG850, the only effect was a 1.5% reduction in the peak transmittance. This trend continued, and
after exposure to 10 Mrads, the cut-on wavelength of the GG495 and OG590 shifted about 50-59 nm, the peak transmittance
was reduced by 6-8%, and the transmittance of wavelengths near the edge were decreased by as much as 54%. For the
Proc. SPIE Vol. 4452
55
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 01/07/2015 Terms of Use: http://spiedl.org/terms
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
400 500 600 700 800 900 1000 1100
Wavelength (nm)
External Transmission (Percent)
0 rads
30 Krad
100 Krad
300 Krad
1 Mrad
3 Mrad
10 Mrad
Figure 2. Transmission measurements of a Schott GG495 colored
glass filter after exposure to gamma radiation (t = 2.935
mm).
20%
30%
40%
50%
60%
70%
80%
90%
100%
700 800 900 1000 1100
Wavelength (nm)
External Transmission (Percent)
0 rads
30 Krad
100 Krad
300 Krad
1 Mrad
3 Mrad
10 Mrad
Figure 4. Transmission measurements of a Schott RG715 colored
glass filter after exposure to gamma radiation (t = 3.165
mm).
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
500 600 700 800 900 1000 1100
Wavelength (nm)
External Transmission (Percent)
0 rads
30 Krad
100 Krad
300 Krad
1 Mrad
3 Mrad
10 Mrad
Figure 3. Transmission measurements of a Schott OG590 colored
glass filter after exposure to gamma radiation (t = 3.05
mm).
30%
40%
50%
60%
70%
80%
90%
100%
800 850 900 950 1000 1050 1100
Wavelength (nm)
External Transmission (Percent)
0 rads
30 Krad
100 Krad
300 Krad
1 Mrad
3 Mrad
10 Mrad
Figure 5. Transmission measurements of a Schott RG850 colored
glass filter after exposure to gamma radiation (t = 2.885
mm).
RG715, the transmittance at wavelengths below 715 nm increased compared to earlier cases, the peak transmittance was
reduced by 20%, and the transmittance at wavelengths just above 715 nm was reduced by as much as 40%. For RG850, the
transmittance at wavelengths below 870 nm increased compared to earlier cases, and the peak transmittance was reduced by
7.5%.
3.1.2 Schott Radiation Resistant Glasses
Most optical glasses darken when exposed to radiation. The addition of cerium to the composition allows optical glasses to
be highly stabilized against discoloration
7
. However, the addition of cerium also causes a slight loss of transmission in the
range of short wave visible light. The degree of cerium stabilization varies from glass to glass. Optical glasses with the
addition of cerium are designated by the affix “G”, and by a code number which corresponds to the percent cerium oxide
content multiplied by ten. Schott glass technologies used to manufacturer over 30 different cerium stabilized glasses, but
today only manufacturers 7 cerium stabilized glasses. Six of those seven are discussed here.
All the glasses held up fairly well with radiation exposure, as was expected. The BK7G18 glass had only a very slight loss in
transmission of 0.7% at 400 nm wavelength after exposure to 300 Krads of gamma radiation as shown in Figure 6. The loss
Proc. SPIE Vol. 4452
56
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 01/07/2015 Terms of Use: http://spiedl.org/terms
increased very slightly to 1.1% at 400 nm after 1 Mrad exposure. After 3 Mrads exposure, the transmission losses increased
to 3.5% at 400 nm, 1% at 450 nm, and 0.5% at 600 nm. After 10 Mrads exposure, the losses increase to 5% at 400 nm and
remained the same at longer wavelengths. This glass held up the best to radiation exposure of all the Schott radiation
resistant glasses.
The K5G20 glass also had only a slight loss in transmission of 1.5% at 400 nm after exposure to 300 Krads of gamma
radiation as shown in Figure 7. The losses remained unchanged after exposure to 1 Mrad of radiation. After 3 Mrads of
exposure, the losses increased to 5% at 400 nm, 1.4% at 450 nm, 1% at 500 nm, 0.7% at 600 nm and 700 nm, and 0.5% at
800 nm. The losses increased to 8% at 400 nm, and 2% at 450 nm, and remained the same at longer wavelengths after 10
Mrad radiation exposure.
The F2G12 glass had a larger loss in transmission of 11% at 400 nm, 3.3% at 450 nm, and 0.5% at 500 nm after exposure to
300 Krads of radiation as shown in Figure 8. The loss at 400 nm increased slightly to 14.4% after 1 Mrad of radiation, but
remained the same at the longer wavelengths. After 3 Mrads of radiation exposure, the losses further increased to 22% at 400
nm, 6% at 450 nm, 1.5% at 500 nm, and 0.5% at 600 nm. The losses remained the same after 10 Mrad radiation exposure.
The LaK9G15 glass had a very large drop in transmission of 40.5% at 400 nm, 20.5% at 450 nm, 7% at 500 nm, 2% at 600
nm, 1% at 700 nm, and 0.5% at 800 nm after exposure to 300 Krads of radiation as shown in Figure 9. The loss increased to
46.5% at 400 nm, 22.5% at 450 nm, 8% at 500 nm, 3% at 600 nm, and remained unchanged at longer wavelengths after 1
Mrad radiation exposure. After 3 Mrads radiation exposure, the losses further increased to 54% at 400 nm, 38% at 450 nm,
11% at 500 nm, and remained about the same at higher wavelengths. The losses remained the same after exposure to 10
Mrads of gamma radiation. This glass had the largest loss in transmission, especially in the blue portion of the spectrum, of
all the Schott radiation resistant glasses.
The SF8G07 glass had a fairly large drop in transmission of 18.5% at 420 nm, 14.6% at 450 nm, 3.4% at 500 nm, and 0.5%
at 600 nm after exposure to 300 Krads of radiation as shown in Figure 10. The losses remained the same after 1 Mrad
radiation exposure. After 3 Mrads radiation exposure, the losses further increased to 28% at 420 nm, 24.5% at 450 nm, 6% at
500 nm, 1% at 600 nm, and 0.5% at 700 nm. The losses remained the same after exposure to 10 Mrads of radiation.
The SF6G05 glass had a drop in transmission of 9.1% at 460 nm, 9% at 500 nm, and 0.9% at 600 nm after exposure to 300
Krads of radiation as shown in Figure 11. The losses remained about the same after 1 Mrad radiation exposure. After 3
Mrads radiation exposure, the losses further increased to 12.3% at 460 nm, 12.6% at 500 nm, and remained the same at
longer wavelengths. The losses remained the same after exposure to 10 Mrads of radiation.
76%
78%
80%
82%
84%
86%
88%
90%
92%
400 500 600 700 800 900 1000 1100
Wavelength (nm)
External Transmission (Percent)
0 rads
300 Krad
1 Mrad
3 Mrad
10 Mrad
Figure 6. Transmission measurements of a Schott BK7G18 window
after exposure to gamma radiation (t = 5.095 mm).
76%
78%
80%
82%
84%
86%
88%
90%
92%
400 500 600 700 800 900 1000 1100
Wavelength (nm)
External Transmission (Percent)
0 rads
300 Krad
1 Mrad
3 Mrad
10 Mrad
Figure 7. Transmission measurements of a Schott K5G20 window
after exposure to gamma radiation (t = 5.03 mm).
Proc. SPIE Vol. 4452
57
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 01/07/2015 Terms of Use: http://spiedl.org/terms
20%
30%
40%
50%
60%
70%
80%
90%
400 500 600 700 800 900 1000 1100
Wavelength (nm)
External Transmission (Percent)
0 rads
300 Krad
1 Mrad
3 Mrad
10 Mrad
Figure 8. Transmission measurements of a Schott F2G12 window
after exposure to gamma radiation (t = 5.325 mm).
10%
20%
30%
40%
50%
60%
70%
80%
90%
400 500 600 700 800 900 1000 1100
Wavelength (nm)
External Transmission (Percent)
0 rads
300 Krad
1 Mrad
3 Mrad
10 Mrad
Figure 10. Transmission measurements of a Schott SF8G07 window
after exposure to gamma radiation (t = 4.965 mm).
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
400 500 600 700 800 900 1000 1100
Wavelength (nm)
External Transmission (Percent)
0 rads
300 Krad
1 Mrad
3 Mrad
10 Mrad
Figure 9. Transmission measurements of a Schott LaK9G15 window
after exposure to gamma radiation (t = 5.065 mm).
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
400 500 600 700 800 900 1000 1100
Wavelength (nm)
External Transmission (Percent)
0 rads
300 Krad
1 Mrad
3 Mrad
10 Mrad
Figure 11. Transmission measurements of a Schott SF6G05 window
after exposure to gamma radiation (t = 5.07 mm).
3.1.3 Infrared Glasses
Infrared windows made of four different materials were obtained from International Scientific Products (ISP). Cleartran
is a
water-clear form of Zinc Sulfide (ZnS) with a broad transmission range. The Cleartran
window material was obtained from
Rohm and Haas Advanced Materials. ZnS Cleartran
is a registered trademark of Rohm and Haas advanced materials,
formerly Morton Advanced Materials. The Zinc Selenide window material was obtained from II-VI. The Silicon window
material was obtained from Lattice materials, and the germanium window material was obtained from a Russian supplier.
The spectral transmission of these glasses before and after radiation exposure are shown in Figures 12 - 15. As expected, the
infrared glasses showed essentially no change in their transmission properties with exposure to gamma radiation within the
measurement uncertainty.
3.1.4 Synthetic Fused Silica
Fused silica is a synthetic form of SiO
2
. It has a very wide transmission range of 170 nm – 2200 nm for ultraviolet (UV)
grade material, has very low absorption due to it’s high purity, is not birefringent, and is naturally radiation resistant. These
properties make it an attractive material for space optics. A UV grade fused silica window was obtained from ISP. A
Russian company supplied the fused silica material. Figure 16 shows the spectral transmission of the fused silica window
before and after exposure to gamma radiation. The fused silica window had a slight loss transmission (0.5-1%) between
200nm and 320 nm wavelength after exposure to 300 Krads gamma radiation. Wavelengths longer than 320 nm had no loss
Proc. SPIE Vol. 4452
58
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 01/07/2015 Terms of Use: http://spiedl.org/terms
57%
59%
61%
63%
65%
67%
69%
71%
73%
75%
200 450 700 950 1200 1450 1700 1950 2200 2450 2700
Wavelength (nm)
External Transmission (Percent)
0 rads
300Krad
1 Mrad
3 Mrad
10 Mrad
Figure 12. Transmission measurements of a Cleartran
window after
exposure to gamma radiation (t = 5.07 mm).
52.0%
52.5%
53.0%
53.5%
54.0%
1500 1700 1900 2100 2300 2500 2700
Wavelength (nm)
External Transmission (Percent)
0 rads
300 Krad
1 Mrad
3 Mrad
10 Mrad
Figure 14. Transmission measurements of a Silicon window after
exposure to gamma radiation (t = 5.12 mm).
64%
65%
66%
67%
68%
69%
70%
71%
600 900 1200 1500 1800 2100 2400 2700
Wavelength (nm)
External Transmission (Percent)
0 rads
300 Krad
1 Mrad
3 Mrad
10 Mrad
Figure 13. Transmission measurements of a Zinc Selenide window
after exposure to gamma radiation (t = 5.07 mm).
45.0%
45.5%
46.0%
46.5%
47.0%
2000 2100 2200 2300 2400 2500 2600 2700
Wavelength (nm)
External Transmission (Percent)
0 rads
300 Krad
1 Mrad
3 Mrad
10 Mrad
Figure 15. Transmission measurements of a Germanium window
after exposure to gamma radiation (t = 5.045 mm).
in transmission within the measurement uncertainty discussed in section 3.2. The transmission properties remained about the
same after exposure to 1 Mrad radiation. After exposure to 3 Mrads, the transmission loss increased to 1.5-5.5% between
200 nm and 320 nm wavelengths with the largest loss being at the shortest wavelength. In addition there is a loss of between
0.5-1.5% between 320 nm and 400 nm wavelengths. After exposure to 10 Mrads radiation, the losses in the ultraviolet
increased to 1.65-9.5% from 200 nm to 320 nm, 0.5-1.65% between 320 nm and 400 nm, and 0.5% between 400 nm and 640
nm wavelength. The fused silica exhibited some losses in the ultraviolet with exposure to radiation, but only had very
minimal losses in the visible portion of the spectrum.
3.1.5 Sapphire
Sapphire is a very rugged radiation resistant optical material with a very broad transmission range (250 nm – 4500 nm). In
addition it has a high index of refraction and high dispersion, which make it an attractive material for the design of space
optics. However, it is a slightly birefringent material so its use may be limited to designs where low angle of incidence can
be maintained on the lens surface.
We obtained a sapphire sample from Crystal Systems Incorporated. It was their Hemlux grade with a 0001 orientation. The
spectral transmission of the Crystal Systems sample before and after exposure to gamma radiation is shown in Figure 17.
The sample held up fairly well with radiation exposure. It had a loss in transmission of about 0.5-1% in the 250 nm to 340
Proc. SPIE Vol. 4452
59
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 01/07/2015 Terms of Use: http://spiedl.org/terms
76%
78%
80%
82%
84%
86%
88%
90%
92%
94%
200 400 600 800 1000 1200 1400 1600 1800 2000
Wavelength (nm)
External Transmission (Percent)
0 rads
300Krad
1 Mrad
3 Mrad
10 Mrad
Figure 16. Transmission measurements of a UV grade fused
silica window after exposure to gamma radiation
(t = 3.09 mm).
80%
81%
82%
83%
84%
85%
86%
87%
88%
200 450 700 950 1200 1450 1700 1950 2200 2450 2700
Wavelength (nm)
External Transmission (Percent)
0 rads
300 Krad
1 Mrad
3 Mrad
10 Mrad
Figure 17. Transmission measurements of a Crystal Systems
0001 orientation sapphire window after exposure
to gamma radiation (t = 6.365 mm).
nm wavelength region after exposure to 300 Krads of gamma radiation. After 10 Mrads exposure, the transmission losses
only increased slightly to 1-1.5% in the 250 nm to 340 nm wavelength region. There was more variation in the transmission
measurements as a function of wavelength than was expected. It appears as though the measurement error may be larger for
the sapphire sample than for other materials. The rapid change in transmission around 800 nm is most likely due to the
detector change that occurs at this wavelength (see section 3.2 for a more detailed discussion of the measurement errors).
However, there are no detector, source, or grating changes for wavelengths between 800 nm and 2700 nm, so the additional
structure in the transmission measurements is not fully understood at this time.
3.1.6 Barium Fluoride
Barium fluoride (BaF
2
) is a naturally radiation resistant material with an extremely wide transmission range extending from
the visible to the long wavelength infrared. For this reason it is a highly desirable space optical material for multispectral
applications. We obtained BaF
2
windows from ISP. We had some interesting results with the radiation tests of the BaF
2
.
The first two windows sent to us were left over windows from a past order of many years ago and were somewhat unknown
in origin. Upon microscopic examination it was determined that these windows were polycrystalline. When exposed to
radiation these polycrystalline windows of BaF
2
had a major drop (10 - 30%) in transmission in the ultraviolet (UV) and in
the blue region after only 300 Krads of gamma radiation exposure as can be seen in Figure 18. They turned brown. The
transmission losses at 1 Mrad, 3 Mrad, and 10 Mrad only increased slightly in the UV (about 40% total) with slightly larger
increases in the visible (about 20% total). We then obtained 5 monocrystalline BaF
2
windows from ISP from their recent
stock where the polish and wedge were also known to be good. Three of these windows were tested under radiation and held
up well. They showed some slight losses in transmission of 1 - 2.5% in the UV and visible after 700 Krads of radiation. The
transmission only dropped slightly more with increased radiation with a total transmission loss of 2.5 - 3.5 % in the UV and
visible after 10 Mrads of radiation as can be seen in Figure 19. We are unsure why the latter windows behaved differently
than the first two. It could be because the first two windows were polycrystalline instead of monocrystalline, but that seems
unlikely. It may also be a purity issue. Looking at the transmission data before any radiation took place we saw that the first
two windows had poorer transmission in the deep UV (200-270 nm range). The point at which the transmission curve began
to change rapidly for the first two windows was 270 nm and for the latter windows was 230 nm. This was probably as a
result of higher purity in the latter windows.
An additional item that occurred during testing of the barium fluoride windows is that one of the polycrystalline windows
was accidentally dropped while trying to put it into an optical mount. The window was visually inspected an appeared to be
unharmed. However, when the transmission of the window was measured it was found to have quite different transmission
properties in the ultraviolet and blue portion of the spectrum compared to another identical window that was manufactured at
the same time. Upon microscopic examination of the window it was seen that the polycrystalline window that had been
dropped had microfractures running everywhere on its surface. We believe these microfractures are responsible for the
decreased transmission in the UV and blue portion of the spectrum due to increased light scattering.
Proc. SPIE Vol. 4452
60
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 01/07/2015 Terms of Use: http://spiedl.org/terms
35%
40%
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
200 300 400 500 600 700 800 900 1000 1100
Wavelength (nm)
External Transmission (Percent)
0 rad
300 Krad
1 Mrad
3 Mrad
10 Mrad
Figure 18. Transmission measurements of an ISP polycrystalline
BaF
2
window after exposure to gamma radiation (t =
6.85 mm).
86%
87%
88%
89%
90%
91%
92%
93%
200 300 400 500 600 700 800 900 1000 1100
Wavelength (nm)
External Transmission (Percent)
0 rad
700 Krad
2.7 Mrad
9.7 Mrad
Figure 19. Transmission measurements of an ISP monocrystalline
BaF
2
window after exposure to gamma radiation (t =
4.85 mm).
3.1.7 Calcium Fluoride
Calcium fluoride (CaF
2
) has traditionally not been as radiation resistant as barium fluoride. The Harshaw catalog
8
states that
calcium fluoride will develop significant color centers in the ultraviolet under long term exposure to space radiation.
However, the calcium fluoride that is being manufactured today is quite different from that of 10 and 20 years ago. Driven
by the photolithography industry, manufacturers are developing synthetic CaF
2
with extremely high purity to increase the
transmission of CaF
2
at eximer laser wavelengths of 157 nm and 193 nm, and to make CaF
2
less susceptible to developing
color centers under long term laser exposure. Fortunately these developments also make CaF
2
more resistant to transmission
losses under exposure to space radiation. We obtained CaF
2
windows from three vendors. Corning provided some of their
eximer grade synthetic CaF
2
windows, Schott provided a 157 nm eximer grade synthetic CaF
2
window and a 193 nm eximer
grade synthetic CaF
2
window, and ISP supplied “natural” CaF
2
windows. All the windows were single crystal materials.
The Corning, ISP, and Schott 193 nm grade CaF
2
windows were irradiated and tested at the same time over the wavelength
range of 200 nm to 1100 nm. However, the Schott 193 nm grade CaF
2
window arrived after the initial 300 Krads exposure
had been completed. Therefore, the lowest dose of radiation it received was 1 Mrad. The Schott 157 nm grade CaF
2
window
was obtained several months later and was tested at the same time as all the other glasses discussed in this paper. This
window was tested over the wavelength range of 200 nm to 2700 nm.
The transmission properties of all three vendors CaF
2
were
comparable over the wavelength range of 200-1100 nm before
radiation, with the exception of the Schott 193 nm grade synthetic
CaF
2
window. It had a major drop in its transmission around 203
nm as shown in Figure 20 indicating a lead contamination. The
natural CaF
2
window also had a slight absorption feature present
at 310 nm prior to any radiation. The natural CaF
2
window had a
major drop (20 - 30%) in transmission in the UV and blue region
after only 300 Krads of gamma radiation exposure. It turned
green. The transmission losses at 1 Mrad, 3 Mrad, and 10 Mrad
only increased very slightly from what was seen at 300 Krads as
shown in Figure 21. The Corning eximer grade synthetic CaF
2
windows held up exceptionally well to the radiation. Very slight
color centers were formed at 330 nm, 380 nm, and 510 nm
wavelengths with transmission losses of about 1 - 2% for 1 Mrad
exposure, and of about 2 - 3.5% at the 10 Mrad exposure at the
wavelengths stated as shown in Figure 22. All other wavelengths
45%
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
200 205 210 215 220
Wavelength (nm)
External Transmission (Percent)
0 rads
Figure 20. Transmission measurements of a Schott 193 nm
eximer grade synthetic monocrystalline CaF
2
window prior to any radiation (t = 9.94 mm).
had even less transmission loss. The Schott 157 nm eximer grade window also held up exceptionally well to the radiation. A
very slight color center was formed at 380 nm wavelength with transmission losses of about 1.7-2.3% after a 1 Mrad
exposure as shown in Figure 23. The transmission losses remained the same after 3 Mrad and 10 Mrad exposure. The Schott
Proc. SPIE Vol. 4452
61
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 01/07/2015 Terms of Use: http://spiedl.org/terms
55%
60%
65%
70%
75%
80%
85%
90%
95%
200 300 400 500 600 700 800 900 1000 1100
Wavelength (nm)
External Transmission (Percent)
0 rads
300 Krad
1 Mrad
3 Mrad
10 Mrad
Figure 21. Transmission measurements of an ISP “natural”
monocrystalline CaF
2
window after exposure to gamma
radiation (t = 6.40 mm).
87%
88%
89%
90%
91%
92%
93%
94%
95%
200 450 700 950 1200 1450 1700 1950 2200 2450 2700
Wavelength (nm)
External Transmission (Percent)
0 rads
300 KRad
1 Mrad
3 MRad
10 MRad
Figure 23. Transmission measurements of a Schott 157 nm eximer
grade synthetic monocrystalline CaF
2
window after
exposure to gamma radiation (t = 7.065 mm).
88.5%
89.5%
90.5%
91.5%
92.5%
93.5%
94.5%
200 300 400 500 600 700 800 900 1000 1100
Wavelength (nm)
External Transmission (Percent)
0 rad
300 Krad
1 Mrad
3 Mrad
10 Mrad
Figure 22. Transmission measurements of a Corning synthetic
monocrystalline CaF
2
window after exposure to gamma
radiation (t = 4.95 mm).
25%
35%
45%
55%
65%
75%
85%
95%
200 300 400 500 600 700 800 900 1000 1100
Wavelength (nm)
External Transmission (Percent)
0 rads
1 Mrad
3 Mrad
10 Mrad
Figure 24. Transmission measurements of a Schott 193 nm eximer
grade synthetic monocrystalline CaF
2
window after
exposure to gamma radiation (t = 9.94 mm).
193 nm eximer grade window with the lead contamination darkened significantly (20 - 60% loss in transmission) over the
UV, visible, and some of the near infrared (NIR) spectral region after the 1 Mrad dose as shown in Figure 24. It turned
brown. It remained about the same after the 3 Mrad and 10 Mrad exposure.
According to Robert Sparrow at Corning, “the very strong absorption feature at ~203 nm seen in one window prior to any
radiation, is well known to those growing calcium fluoride. It is caused by residual lead in the lattice. The lead
contamination could be from lead fluoride or lead oxide. If oxygen were present in the lattice, this would be expected to lead
to strong radiation induced absorption.”
All the other CaF
2
windows were free of lead contaminants. The “natural” CaF
2
window is likely to contain rare earths. The
rare earth contaminants create strong absorbers in the UV when exposed to radiation. This explains why the “natural” CaF
2
window darkened with radiation exposure. However, the synthetic eximer grade CaF
2
windows that were free from rare
earth or any other contaminants, had very little transmission loss after radiation exposure.
Proc. SPIE Vol. 4452
62
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 01/07/2015 Terms of Use: http://spiedl.org/terms
3.2 TRANSMISSION ERROR ANALYSIS
An empirical error analysis was performed on the Cary 500 spectrophotometer to estimate the error in measurement for this
study. The instrument is based upon the use of two sources and two detectors that divide the 200 nm to 2700 nm band into
three distinct wavelength bands. The near-ultraviolet (NUV) band uses a deuterium lamp and a silicon detector to cover the
range from 200 to 350 nm. The visible (VIS) band uses a quartz-tungsten-halogen incandescent lamp and the silicon detector
to cover the range from 360 to 800 nm. The near infrared/short-wave infrared (NIR/SWIR) band uses the quartz-tungsten-
halogen incandescent lamp and an InGaAs detector to provide coverage in the range from 810 to 2700 nm. Step changes
were observed at detector and source switchover points for some materials.
Since our interest is the broadband performance of the material, stray light and slit functions were not examined. We have
classified errors into two phenomenological types:
1. Repeatability. Shot noise, digitization, and other random noises.
2. Reproducibility. Sample mounting accuracy and instrument drift.
3.2.1 Method
Measurements for error analysis were performed using the CaF
2
sample, which is clear over the entire wavelength band of
interest. Repeatability was characterized by taking two consecutive measurements without disturbing the sample between
measurements. For the uncoated materials measured in this study, the dominant contributor to reproducibility was found to
be the mounting accuracy of both sample faces relative to the beam normal. We have also observed that mechanical axis of
the instrument is not necessarily aligned with the optical axis, implying that even a perfectly placed sample may still be
inclined relative to the beam.
Despite the fact that optical alignment can, in principle, be nulled by
fixturing or corrected mathematically, we shall treat it as a statistical
uncertainty of Type B.
9
With our available tooling, even using due
diligence, we could not assure alignment of the sample normal to the
beam to less than 2°. To determine the reproducibility of
measurement, we used mechanical shims to tilt our CaF
2
relative to
the instrument mechanical axis. Five measurements were taken at
relative horizontal angles of –1.5, -0.75, 0, 0.75, and 1.5 degrees
relative to the mechanical axis. These measurements were analyzed
statistically in each wavelength band to determine the instrument
reproducibility.
The results from all error analysis measurements are shown in Figure
25.
3.2.2 Repeatability
The expanded uncertainty with coverage factor k=2 arising from the
instrument repeatability is calculated as two times the standard
90%
91%
92%
93%
94%
95%
200 700 1200 1700 2200 2700
Wavelength (nm)
External Transmission (Percent)
-1.5 degrees
-0.75 degrees
-0.75 deg (Repeatability)
Mechanical Normal
+0.75 degrees
+1.5 degrees
NUV
VIS
NIR/SWIR
Figure 25. Error Analysis Measurements.
deviation of the pointwise difference between the repeated measurement (+0.75°) in percent. The repeatability of the
instrument is excellent. Results are shown in Table 2.
3.2.3 Reproducibility
For each of the five trial runs where the sample was inclined, we have calculated the pointwise difference between the run
and the composite average. The expanded uncertainty from repeatability and reproducibility (k=2), shown in Table 3, is
calculated as two times the standard deviation of these data.
Repeatability and reproducibility are uncorrelated, so they add in quadrature. Using the data for repeatability U
Repeat
and the
data for repeatability and reproducibility U
R+R
, we may infer the expanded uncertainty from reproducibility, U
Repro
, alone.
3.2.4 Inferred Uncertainty for Materials of Different Refractive Index
For clear materials (internal transmission >99%), the error in measurement due to optical alignment will be proportional, in
some way, to the index of refraction of the material. It is possible to use the existing data for CaF
2
to determine an
Proc. SPIE Vol. 4452
63
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 01/07/2015 Terms of Use: http://spiedl.org/terms
Band U
Repeat
NUV 0.06%
VIS 0.03%
NIR/SWIR 0.03%
Table 2. Uncertainty of Measurement Due to Repeatability
(k=2).
Band U
R+R
NUV 0.37%
VIS 0.38%
NIR/SWIR 0.30%
Table 3. Uncertainty in Measurement Due to Repeatability
and Reproducibility (k=2).
approximate functional relationship between the measurement error for each band and the refractive index of the material.
We base our argument upon the following assumptions:
1. The angle of incidence is sufficiently shallow to neglect polarization.
2. Reflection loss is characterized by the multiple reflection relationship
10
.
()
1
1
2
2
+
−
=
n
n
R
3. Uncertainty due to reproducibility is a linear function of reflection loss:
()
1
1
2
2
+
−
==
n
n
mmRU
Repro
,
where m is the scale factor as determined from empirical data.
4. The sample has zero internal absorption.
Given the refractive index of CaF
2
, we have calculated the scale factors m from the data in Tables 2 and 3. The total
uncertainty in measurement from repeatability and reproducibility is given by:
()
1
1
2
2
2
+
−
+=
n
n
mUU
RepeatC
Table 4 shows the calculated inferred uncertainty in measurement (k=2) for the materials in this study. Nominal indices of
refraction have been used to classify the materials. Uncertainty is expressed as a percent of measurement. Differences in
measurement less than the levels shown in Table 4 are not statistically significant.
Uncertainty of Measurement for Study Materials (k = 2)
Nominal Index NUV VIS NIR/SWIR Materials
n=1.43 0.38% 0.40% 0.31% Fused Silica, CaF
2
, BaF
2
n=1.52 0.51% 0.53% 0.42% BG38, GG495, OG590, RG715, RG850, BK7G18, K5G20
n=1.62 0.67% 0.69% 0.55% F2G12
n=1.75 0.87% 0.90% 0.71% LaK9G15, SF6G05, SF8G07, Sapphire
n=2.40 1.89% 1.49%
ZnSe, ZnS Cleartran
n=3.45 2.39% Silicon
n=4.06 2.75% Germanium
Table 4. Uncertainty in Measurement (k=2) for Study Materials.
4. CONCLUSIONS
The Schott colored filter glasses experienced fairly large changes in their cut-on wavelengths and transmission properties
with exposure to gamma radiation. If these filter glasses are to be used in the construction of space bandpass filters, they
should be adequately shielded to prevent spectral changes with radiation exposure or they must be frequently calibrated to
determine their spectral response function. The Schott cerium stabilized radiation resistant glasses held up well to radiation
exposure as expected. All of the glasses had some loss in transmission in blue portion of the spectrum after radiation
exposure. In addition, all the glasses except BK7G18 had some losses in the green and red portion of the spectrum as well.
However, the LaK9G15 glass had very large losses in the 400 nm to 500 nm spectral region after exposure to only 300 Krads
of gamma radiation. This is larger than what is stated in the Schott glass catalog
11
. The infrared glasses showed no change in
transmission from their cut-on wavelengths up to 2700 nm after exposure to gamma radiation within the transmission
Proc. SPIE Vol. 4452
64
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 01/07/2015 Terms of Use: http://spiedl.org/terms
measurement error. The fused silica exhibited some losses in the ultraviolet with exposure to radiation, but only had very
minimal losses in the visible portion of the spectrum. The sapphire material had some very slight losses in transmission in
the ultraviolet with exposure to radiation, and no losses in the visible region. The sapphire material showed larger than
expected spectral structure in the transmission measurements, which is not fully understood at this time. The BaF
2
and the
CaF
2
showed mixed results when exposed to radiation. Two samples of BaF
2
turned brown when exposed to 300 Krads of
gamma radiation. This is unusual for BaF
2
since it has a high radiation tolerance. We believe that some impurities must have
been present in the sample before radiation to cause this effect. The BaF
2
monocrystalline samples that were from recent
material stock held up exceptionally well to radiation with losses of only a few percent in the visible portion of the spectrum.
The “natural” CaF
2
sample turned green with exposure to radiation. This is most likely due to the rare earth contaminants
found in “natural” CaF
2
. The synthetic single crystal eximer grade CaF
2
samples that were free from contaminants only had
very minimal losses in transmission in the ultraviolet and visible portion of the spectrum after exposure to gamma radiation.
However, the one sample with the lead contamination, browned significantly with exposure to radiation. From this we can
conclude that it is very critical to use BaF
2
and CaF
2
material with the highest purity available and to test to be certain that the
material is free from contaminants if they are to be used in a space radiation environment.
ACKNOWLEDGMENTS
The authors wish to thank Craig Boney for his help in preparing the samples and coordinating the radiation testing and
transmission measurements, and Arnold Augustoni for completing all the transmission measurements done in this work.
Sandia is a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin Company, for the United States
Department of Energy under Contract DE-AC04-94-AL85000.
REFERENCES
1. M. Fruit, A. Gusarov, D. Doyle, G. Ulbrich, and A. Hermanne, “Space radiation sensitivity of glasses: first results
towards a comprehensive dose coefficients database,” SPIE Proc. Vol. 4134, pp. 261-267, 2000.
2. V. Arbuzov, A. Gusarov, and F. Ignatiev, “Radiation-induced absorption and refractive index difference in optical
glasses and their radiation resistant counterparts,” Glastech. Ber. Glass Sci. Technol. Vol 68 C1, pp. 501-506, 1995.
3. D. Doyle, and R. Czichy, “Influences of simulated space radiation on optical glasses,” SPIE Proc. Vol. 2210, pp. 434-
448, 1994.
4. P. Silverglate, E. Zalewski, and P. Petrone, “Proton induced radiation effects on optical glasses,” SPIE Proc. Vol. 1761,
pp. 46-57, 1992.
5. D. Doyle, J. Bourrieau, P. Bricard, B. Johlander, J. Juranek, A. Litzelmann, W. Pannhorst, and W. Rits, “The effect of
electron irradiation on the radius of curvature of a zerodur mirror,” SPIE Proc. Vol. 2775, pp. 166-188, 1996.
6. C. Merzbacher, E. Friebele, J. Ruller, and P. Matic, “Finite element analysis of deformation in large optics due to space
environment radiation,” SPIE Proc. Vol. 1533, pp. 222-228, 1991.
7. Schott Glass Technologies, Radiation Resistant Glasses Catalog, pp. 5.
8. The Harshaw Chemical Company, Harshaw Optical Crystals Catalog, pp. 23.
9. ANSI/NCSL Z540-2-1997. American National Standard for Expressing Uncertainty – U.S. Guide to the Expression of
Uncertainty in Measurement. October 9, 1997. pp. 6.
10. Schott Glass Technologies, Optical Glass Catalog 10022e. pp. 13, equation 6.
11. Schott Glass Technologies, Radiation Resistant Glasses Catalog, pp. 26.
Proc. SPIE Vol. 4452
65
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 01/07/2015 Terms of Use: http://spiedl.org/terms
