Dynamics of the optical response of Nd:YAG to Ionizing radiation: Testing for radiation hardness using UV laser radiation

Abstract

The optical response of single-crystal Nd:YAG and Cr3+:YAG to
ionizing radiation has been previously studied using intense pulses of
gamma-rays at the HERMES III facility at Sandia National Laboratory,
where samples' transmission at 1064 nm was observed during exposure to
gamma radiation. A further study of similar samples when exposed to
10-ns UV laser pulses reveals nearly identical dynamics, with both tests
producing similar transient and permanent response in the medium. This
strongly suggests that the material response to UV radiation can be used
to gauge its gamma-radiation hardness, therefore yielding a material
testing technique that is much simpler and less costly than
gamma-radiation tests.

Full text

Dynamics of the Optical Response of Nd:YAG to Ionizing
Radiation: Testing for Radiation Hardness Using UV Laser
Radiation
B. L. Glebov1, K. Simmons-Potter1, and D. C. Meister2
1. University of Arizona
Tucson, AZ 85721
2. Sandia National Laboratories
Albuquerque, NM 87185
ABSTRACT
The optical response of single-crystal Nd:YAG and Cr3+:YAG to ionizing radiation has been previously studied
using intense pulses of gamma-rays at the HERMES III facility at Sandia National Laboratory, where samples’
transmission at 1064 nm was observed during exposure to gamma radiation. A further study of similar samples
when exposed to 10-ns UV laser pulses reveals nearly identical dynamics, with both tests producing similar transient
and permanent response in the medium. This strongly suggests that the material response to UV radiation can be
used to gauge its gamma-radiation hardness, therefore yielding a material testing technique that is much simpler and
less costly than gamma-radiation tests.
1. INTRODUCTION
One of the first materials to be used as a lasing gain material, Yttrium Aluminum Garnet (YAG) activated with
Neodymium [1,2] is now also one of the most popular and widely used. YAG activated with Cr3+ ions has also been
widely studied as lasing materials [3]. Furthermore, YAG activated with both Cr3+ and Nd ions has been extensively
studied because of an energy-transfer mechanism that exists between Cr3+ and Nd, which increases pumping
efficiency in the case where a broad-band pumping source, such as a flash lamp, is used to excite the laser
material[4]. Specifically, Cr3+ ions possess broad absorption in blue and UV parts of the spectrum, absorbing energy
more efficiently than Nd ions alone. Energy thus absorbed by Cr3+ is then transferred to Nd resulting in the observed
enhancement in laser performance.
The popularity and established production methodology of Nd:YAG has made it an attractive choice for
systems that are being developed with reliance on Commercial Off The Shelf (COTS) technologies. However,
Nd:YAG has been found to be unsuitable for harsh radiation environments. In particular, Nd:YAG been seen to
suffer from substantial transient and some permanent loss in optical transmittance in the presence of ionizing
(gamma) radiation [5-8], as ionizing radiation causes formation of color centers with absorption at the operating
wavelength, thus reducing lasing efficiency. Previous experiments have revealed that co-doped Cr3+:Nd:YAG, by
contrast, withstands gamma radiation without developing significant amounts of transmission loss that could impact
device operation [8].
However, testing with intense gamma-ray pulses can be costly and time-consuming, and so a less resource-
intensive alternative is presently sought. Previous experiments have revealed strong similarities between the
permanent color centers formed in YAG-based materials by gamma radiation exposure [6] and by UV irradiation in
the 200 – 400 nm range [9], with spectral changes observed in the absorption spectra of optically-irradiated YAG
attributed to valence changes of the iron impurity [10], where an electron trapped on a Fe3+ ion creates an
[
]
−
+3
Fe
color center. To date, no correlation has been shown between the transient, optical response of doped YAG to fast-
pulsed gamma radiation and the response of the materials to short-pulsed UV irradiation.
Optical Technologies for Arming, Safing, Fuzing, and Firing V, edited by Fred M. Dickey,
Richard A. Beyer, Proc. of SPIE Vol. 7434, 74340B · © 2009 SPIE
CCC code: 0277-786X/09/$18 · doi: 10.1117/12.828443
Proc. of SPIE Vol. 7434 74340B-1

2. EXPERIMENT
The present study employed a suite of single-crystal samples of YAG doped with various concentrations of Nd
and Cr3+. Singly doped Cr3+:YAG and Nd:YAG samples were acquired from United Crystals. Co-doped Cr3+,
Nd:YAG samples were obtained from Scientific Materials. Cr3+:YAG samples were specified as 0.5 mol% of Cr3+ in
the crystal. Nd:YAG samples were specified as 0.5 mol% and 1.2 mol% of Nd in the crystal. The co-doped samples
were specified as 0.5 mol% Cr3+ in the melt and 1.2 mol% Nd in the crystal.
Single-doped samples used in gamma-radiation tests were square rods, with 7 mm by 7 mm cross-sections and
averaging 12 mm in length. Co-doped samples used in gamma-radiation tests were discs, with 28 mm diameter clear
aperture and 9 mm in length. Samples used in UV-irradiation experiments were of the same cross-section as in
gamma-radiation experiments, but 2 mm in thickness. All samples had spectroscopic quality polish on the front and
back surface, and fine grind on the sides.
Gamma-radiation experiments consisted of measuring in real time sample transmission at 1064 nm in response
to intense pulses of highly energetic photons. These experiments were carried out at the HERMES III facility at
Sandia National Laboratory in Albuquerque, NM. This facility provides pulses with total doses averaging 60 krad
(Si), and pulse durations of ~30 ns. The exact setup and analysis of the resulting data have been discussed previously
[8].
UV-irradiation experiments were carried out using a KrF excimer laser operating at 248 nm at the Arizona
Materials Laboratory, a facility of The University of Arizona in Tucson, AZ. These experiments were set up to
measure the same quantities as the experiments at HERMES III [8], namely the samples’ transient transmittance at
1064 nm, in response to a fast UV pulse. The experimental layout is represented in Figure 1 below.
Measurements were performed using a dual-detector setup. Signals from the reference detector and the data
detector were combined in a custom analog signal mixer. The reference signal was passed through an inverting
amplifier with variable gain and then added to the data signal, which passed through the mixer with unity gain. The
variable gain was adjusted to provide a null rest signal. The mixer’s output was then fed into a recording
oscilloscope. In addition, a pulse detector registered light from the UV pulse scattered by the sample, and passed it
to the recording oscilloscope. This signal was used to trigger the data recording. The fast pulse detector gave a
timing resolution of better than 10 ns.
B/S
1064 nm
CW
248 nm
10 ns
sample
filter
filter
Analog
signal
mixer
Oscilloscope
Pulse detector
Reference
detector
Data
detector
UV laser
NIR laser
Proc. of SPIE Vol. 7434 74340B-2

Figure 1: UV-NIR pump-probe experimental setup.
The dielectric laser line mirrors that were used to guide the UV beam had an optical quality polish on the back
surface, which allowed for the probe NIR beam to be coupled to the same axis as the UV beam. This assured that the
probe beam was sampling the same volume as the pump beam was exciting. Filters were also placed in the test set-
up to prevent UV light from coupling into the NIR laser detector line.
Both the pump (UV) and the probe (NIR) beams passed through the samples unfocused. The excimer laser was
set to deliver 10.1 mJ per pulse. It irradiated an area approximately 4.8 mm x 7.9 mm. This translated into UV
pulse’s energy density of 26 mJ / cm2 per pulse, and a power density of 2.6 MW / cm2.
For the time-dependent data presented here, a trace was recorded for each single pulse from the excimer laser.
Twenty such traces were averaged together to produce the data presented the figures below. This equated to a total
dose of 0.52 J / cm2 for the samples used in the time-dependent measurements.
The UV-NIR pump-probe setup was used to observe sample response on the time scale of a few milliseconds.
Processes with longer lifetimes were studied using a Perkin-Elmer Lambda 950 UV-Vis spectrophotometer.
Differential absorption data, presented here, were obtained by subtracting the pre-irradiation spectrum from the post-
irradiation spectrum collected for the same sample.
3. RESULTS
Results from radiation testing at HERMES III have been described in detail and discussed in previous work [7-
8], but are presented here for clarity. Figure 2(a) shows the induced coefficient of absorption for gamma-irradiated
Nd-doped YAG laser materials. The y-axis in the figure has been scaled to facilitate comparison of the data with that
presented in Figure 2(b). The arrows in Figure 2(a) are indicative of the fact that the peak absorption coefficient in
gamma-irradiated 0.5 mol% Nd:YAG reached ~0.7 cm-1 immediately following the gamma pulse, while 1.2 mol%
Nd:YAG achieved a peak absorption coefficient of ~1 cm-1. Examination of the figure clearly shows a large,
transient induced loss in both the samples that scales inversely with Nd concentration, such that the lower Nd
concentration exhibits a much larger radiation-induced loss than the higher Nd-concentration material. The right-
hand figure, Figure 2(b), shows the similar transient response of identical materials to short-pulsed UV irradiation.
Evident in the figure is the comparable effect of UV irradiation on the samples as a function of dopant
concentration, with the high dopant concentration leading to an acceleration in annealing of the induced absorption
in the samples. Also evident is the strong similarity between the long-lifetime decay response of the induced optical
loss in the gamma and UV irradiated materials. While these general trends in radiation response of the materials are
consistent between the two irradiation sources, further examination of the data reveals differences in the time-
dependent decay response of the samples on the very short time scale. From the figures, it is clear that on the short
time scale, the gamma radiation process creates a substantially higher concentration of rapidly-decaying, less-stable,
absorbing centers which give rise to the initial rapid decay in the left-hand figure.
0.0
0.1
0.2
0.3
0.4
-1.0 0.0 1.0 2.0 3.0 4.0 5 .0
Time (ms)
Induced coefficient of absorption (1/cm)
0.5 mol% Nd:YAG
1.2 mol% Nd:YAG
0.000
0.005
0.010
0.015
0.020
-1.0 0.0 1.0 2.0 3.0 4.0 5.0
Time (m s)
Induced coefficient of absorption (1/cm)
0.5 mol% Nd:YAG
1.2 mol% Nd:YAG
(A) (B)
Proc. of SPIE Vol. 7434 74340B-3

Figure 2: Time-dependent induced absorption in Nd:YAG in response to gamma radiation (A) and UV radiation
(B).
Of importance in determining the role of color-center formation in the observed radiation-induced
photodarkening of the samples under investigation is the long-time-scale dynamics of optical absorption in the
samples across a broad wavelength spectrum. Impurities that are commonly associated with optical absorption, are
typically active at wavelengths that are distinct from the laser wavelength (at 1.06 μm), that is monitored in the
radiation experiments. It is therefore instructive to investigate steady-state optical absorption in the samples. Figure
3 below shows the differential absorption spectra taken from a 0.5 mol% Nd:YAG sample that was exposed to
gamma radiation, then subsequently (months later due to restrictions on sample release following HERMES III
testing) annealed to temperatures of ~500 °C in air, and then, finally, exposed to UV radiation. In the figure it can
be seen that a strong optical absorption feature across the UV portion of the spectrum is present in the gamma-
irradiated sample. Annealing the sample resulted in the removal of this feature and the sample’s recovery to its
original condition. Lastly, UV irradiation of the sample following the anneal (total dose of approximately 13 J/cm2)
resulted in the reappearance of the UV absorption. The particular impurity responsible for the observed absorption
can be attributed to trapped electrons on Fe3+ impurity ions [10]. These color centers are seen to be stable at room
temperature, remaining for several years after irradiation. Their reported annealing temperature is ~430 °C. Clearly,
then it can be said that this defect is responsible for at least some of the observed permanent photodarkening by
creating
[]
−
+3
Fe color centers.
-1.5
-1.0
-0.5
0.0
0.5
1.0
200 250 300 350 400 450 500 550 600
Wavelength (nm)
Differential absorption (1/cm)
gamma-irradiation
800 K anneal
UV irradiation
Figure 3: Steady-state differential absorption spectra showing creation and destruction of broad-band
absorption arising from [Fe3+]- color centers [10] due to gamma radiation, high temperature anneal, and UV
irradiation. Apparent sharp spectral features near 350 nm and 550 nm are instrument artifacts.
It is also instructive to compare the response of Cr3+:YAG and Cr3+,Nd:YAG to both gamma radiation and to
UV radiation. This data is shown below in Figure 4(a) and Figure 4(b) respectively. Examination of the figure
reveals similarities as well as some notable differences between the data. In both figures, the general radiation
resistance of Cr3+:YAG is evident as is the radiation-induced gain response of the Cr3+,Nd:YAG along with its long-
term radiation hardness. An important difference between the gamma-irradiation data and the UV-irradiation data is
the slight induced absorption in Cr3+:YAG that is visible in the UV-irradiation data. A small amount of
photodarkening can be seen in Cr3+,Nd:YAG as well, after the gain process has become extinct. It is important to
note that the magnitude of these effects is quite small making it well below the experimental sensitivity (noise limit)
Proc. of SPIE Vol. 7434 74340B-4

of the gamma radiation experiments. For this reason, such effects could not have been detected in previous
experiments and were only observed in the UV irradiations due to the greatly increased sensitivity (and very low
noise background) of the measurement. Thus, once again, the overall trend in radiation response of the materials to
gamma or UV radiation is remarkably similar, supporting the argument that UV irradiation can be useful as a
predictive tool for understanding the overall ionizing radiation behavior of these materials.
-0.08
-0.06
-0.04
-0.02
0.00
0.02
0.04
0.06
0.08
-1.0 0.0 1.0 2.0 3.0 4.0 5.0
Time (m s)
Induced coefficient of absorption (1/cm)
0.5 mol% Cr:YA G
0.5 mol% Cr, 1.2 mol%
Nd: Y A G
-0.012
-0.008
-0.004
0.000
0.004
-1.0 0.0 1.0 2.0 3.0 4.0 5.0
Time (m s)
Induced coefficient of absorption (1/cm)
0.5 mol% Cr, 1.2 mol%
Nd: Y A G
0.5 mol% Cr:YA G
2.0 mol% Cr:YA G
Figure 4: Induced gain and absorption in Cr3+:YAG and Cr3+, Nd:YAG in response to a fast gamma radiation pulse
(A) and UV pulse (B). In case of gamma radiation, data for 0.5 mol% and 2.0 mol% looks identical, within noise
limits, so data are presented for only one concentration.
4. DISCUSSION
Data from gamma- and UV-irradiation experiments for Nd:YAG were compared numerically using the double
exponential model below in Equation 1. This model was utilized because it provided efficient fits of the data and,
thus, gave a numerical description of the processes involved, not because it represents a functional description of
these processes.
()
C
t
A
t
At +
⎟
⎟
⎠
⎞
⎜
⎜
⎝
⎛−+
⎟
⎟
⎠
⎞
⎜
⎜
⎝
⎛−=Δ
2
2
1
1expexp
ττ
α
(1)
Amplitudes, time constants, and the constant C were all free for adjustment in the fitting procedures. In the
application of this model to the data, a correlation of better than 0.98 was achieved in all cases, which well exceeded
the quality of fits based on a single exponential. Fits were based on the data presented in Figure 2. Results of the fits
are presented in below, in Table 1.
Quantity 0.5 mol% Nd:YAG 1.2 mol% Nd:YAG
γ-radiation UV-irradiation γ-radiation UV-irradiation
A1 (cm-1) 0.45 0.004 0.83 0.006
τ1 (ms) 0.019 0.15 0.021 0.16
A2 (cm-1) 0.13 0.008 0.07 0.008
τ2 (ms) 3.16 1.52 1.55 1.44
C (cm-1) 0.13 0.006 0.08 0.0002
(A)
(B)
Proc. of SPIE Vol. 7434 74340B-5

Table 1 Fit results of relaxation of absorption at 1064 nm due exposure of Nd:YAG to fast pulses of either gamma
radiation or UV light at 248 nm.
The numerical fit data clarify what was seen in Figure 2. Specifically, when exposed to either gamma-radiation
or UV-radiation, absorption is induced in Nd:YAG at 1064 nm. This absorption is not stable at room temperature,
and recovery of the optical transmittance is observed. Numerically, the ensuing relaxation process can be split into
three parts – a fast process, a slow process, and some residual, permanent photodarkening. In the case of both
gamma- and UV-irradiation, the balance of these three processes is shifted toward the fast one for higher
concentrations of Nd.
While lifetimes of the faster process differ by an order of magnitude between UV- and gamma-irradiations, with
the gamma-radiation-induced lifetimes being faster, lifetimes of the slower processes are much closer together. This
indicates that both UV- and gamma-irradiation can access the physical processes responsible for the slower-
decaying part of the recovery curve. However, gamma-irradiation can also access faster-decaying processes,
attributable to less-stable absorption centers, that are not excited by the UV-irradiation in nearly the same
proportion. The fit data also clarifies the behavior of the permanent component of induced loss in the samples with
the magnitude of the fitting constant, C. The larger persistent loss of the 0.5 mol% Nd:YAG over the more highly
doped YAG is seen for both gamma and UV irradiations.
An analysis of this permanent photodarkening (stable over years after exposure) in Nd:YAG revealed formation
of color centers due to electrons trapped on Fe3+ ions (
[
]
−
+
3
Fe color center)[10], as is evidenced by the data in
Figure 3. These centers can be created by both gamma-radiation and UV-irradiation, though saturation concentration
of these color centers is lower in case of UV. This process can be described by the photochemical expression in
Equation 2.
[
]
−
−
+−+ +→++ OFehOFe 323
ν
(2)
Again, it is found that both UV- and gamma-irradiation cause similar spectroscopic changes in the Nd:YAG.
The difference is that gamma-irradiation can push the creation of these color centers to a greater concentration.
The results for both Cr3+:YAG and for Cr3+,Nd:YAG from gamma-radiation and UV-irradiation experiments
bore close similarity. In both cases, ionizing radiation produced some gain at 1064 nm. Double exponential fits were
performed on the data from the co-doped samples as well, again using Equation 1 as the model. Results were
dominated by a single exponential with a negative amplitude, corresponding to gain. In case of gamma-irradiation,
the gain lifetime was found to be 0.51 ms. In the case of UV-irradiation, the gain lifetime was found to be 0.67 ms.
Again, a strong correlation is observed between results of the two tests.
5. CONCLUSIONS
Experiments were performed in which changes in transmission at 1064 nm of YAG crystals activated with Cr3+
and Nd ions due to gamma-radiation and UV-irradiation was observed and analyzed. Dynamics of the processes
were compared between the two sources of ionizing radiation. Parallels and dissimilarities in how the two
environments affect these materials were outlined.
It was found that the photodarkening experienced by gamma- versus UV-irradiated Nd:YAG exhibited strongly
similar long-time scale and steady-state behavior. In addition, similar dependences on Nd-dopant concentration were
observed. Color centers responsible for the photodarkening were seen to be mostly unstable, decaying at room
temperature, with lifetimes of several involved processes varying from 20 μ to 4 ms. A proportionally small amount
residual absorption was associated with electrons trapped on Fe3+ ions, creating
[
]
−
+
3
Fe color centers. These
centers have annealing temperature of ~430 °C and broad absorption peaking at 310 nm and were seen to be created
by both gamma irradiation and, to a lesser extent, by UV irradiation.
In Cr3+:YAG and in co-doped YAG strong similarities were seen between the transient behaviors of gamma
irradiated and UV irradiated materials. Strong radiation resistance was observed in all samples doped with the Cr3+
ions, and in both kinds of tests. Though some photodarkening was detected in Cr3+:YAG and in Cr3+:Nd:YAG due
to pulsed UV-irradiation, this effect was small and well below the noise floor for the gamma irradiations.
Furthermore, this photodarkening was found to be unstable at room temperature. In addition, a net gain was
observed in Cr3+, Nd:YAG in both gamma-radiation and UV-irradiation experiments, likely due to an energy
Proc. of SPIE Vol. 7434 74340B-6

transfer mechanism between Cr3+ ions and Nd ions. Gain lifetime in 0.5 mol% Cr3+ : 1.2 mol% Nd:YAG due to a 10
ns UV pulse was found to be 0.67 ms, and 0.51 ms due to a 30 ns gamma pulse.
Overall, a close net correlation between results of testing doped YAG crystals in the two ionizing radiation
environments (UV and gamma radiation) was observed. General trends and overall material performance were
evaluated using the two radiation methods and similar material behaviors and radiation sensitivity was observed
regardless of radiation technique. The data presented provides support for the claim that pulsed UV-irradiation can
be used to predict the general behavior of materials in a pulsed-gamma-radiation environment.
6. ACKNOWLEDGEMENTS
This work was jointly supported by the University of Arizona, State of Arizona TRIF funds and by Sandia National
Laboratories. Sandia is a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin Company
for the United States Department of Energy’s National Nuclear Security Administration under contract DE-AC04-
94AL85000. The authors would like to additionally thank the Photoinduced Processing Laboratory, The College of
Optics and Photonics, University of Central Florida for sample sectioning and polishing, and Scientific Materials
Corporation for supplying many of the samples used in this study.
7. REFERENCES
[1] J.E. Geusic, H.M. Marcos, L.G. Van Uitert, “Laser oscillations in Nd-doped yttrium aluminum, yttrium
gallium and gadolinium garnets,” Applied Physics Letters 4, 182-184 (1964).
[2] L.F. Johnson, G.D. Boyd, K. Nassau, R.R. Soden, “Continuous operation of a solid-state optical maser,”
Physical Review 126, 1406-1409 (1962).
[3] Alphan Sennaroglu, Clifford R. Pollock, Howard Nathel, “Efficient continuous-wave chromium-doped YAG
laser,” Journal of the Optical Society of America B 12, 930-937 (1995).
[4] Z.J. Kiss, R.C. Duncan, “Cross-pumped Cr3+-Nd3+:YAG laser system,” Applied Physics Letters 5, 200-2002
(1964).
[5] T.S. Rose, M.S. Hopkins, R.A. Fields, “Characterization and control of gamma and proton radiation effects
on the performance of Nd:YAG and Nd:YLF lasers,” IEEE Journal of Quantum Electronics 31, 1593-1602
(1995).
[6] S.M. Kaczmarek, “Influence of ionizing radiation on performance of Nd:YAG lasers,” Crystal Research and
Technology 34, 1183-1190 (1999).
[7] A. Vaddigiri, K. Simmons-Potter, W. J. Thomes, Jr., and D. C. Meister, “Ionizing radiation effects in single-
crystal and polycrystalline YAG,” IEEE Transactions on Nuclear Science 53, 3882-3888 (2006).
[8] B.L. Glebov, K. Simmons-Potter, D.C. Meister, “Permanent and transient response of Nd:YAG and Cr:YAG
to ionizing radiation,” Proc. SPIE 7070, 70700F (2008).
[9] M. Bass, A.E. Paladino, “Color centers in Yttrium Galium Garnet and Yttrium Aluminum Garnet,” Journal
of Applied Physics 38, 2706-2707 (1967).
[10] I.Sh. Akhmadulin, S.A. Migachev, S.P. Mironov, “Thermo- and photoinduced defects in Y3Al5O12,” Nuclear
Instruments and Methods in Physics Research B 65, 270-274 (1992).
Proc. of SPIE Vol. 7434 74340B-7