Journal of Nuclear Materials 424 (2012) 38–42

Contents lists available at SciVerse ScienceDirect

Journal of Nuclear Materials

journal homepage: www.elsevier.com/locate/jnucmat

Advantages and disadvantages of using a focused ion beam to prepare TEM

samples from irradiated U–10Mo monolithic nuclear fuel
B.D. Miller ⇑, J. Gan, J. Madden, J.F. Jue, A. Robinson, D.D. Keiser Jr.
Fundamental Fuel Properties, Idaho National Laboratory, P.O. Box 1625, Idaho Falls, ID 83415-6188, USA

a r t i c l e i n f o a b s t r a c t

Article history: Transmission electron microscopy (TEM), scanning electron microscopy (SEM), and focused ion beam
Received 14 November 2011 (FIB) milling were performed on an irradiated U–10Mo monolithic fuel to understand its irradiation
Accepted 19 January 2012 microstructure. This is the first reported TEM work of an irradiated fuel sample prepared using a FIB.
Available online 31 January 2012
Advantages and disadvantages of using the FIB to create TEM samples from this irradiated fuel will be
presented along with some results from the work. Sample preparation techniques used to create SEM
and FIB samples from the brittle irradiated monolithic sample will also be discussed.
Ó 2012 Elsevier B.V. All rights reserved.

1. Introduction
Understanding fuel performance under irradiation is a key
aspect to qualifying nuclear fuel. The Reduced Enrichment for Re-
search and Test Reactors (RERTR) program was tasked with devel-
oping a high-density, low-enriched uranium fuel to reduce the
demand of highly enriched uranium in research reactors [1]. The
primary fuel being studied is U–Mo based. The two major fuel
forms being studied are U–Mo/Al dispersion fuels and U–Mo
monolithic fuel in plate form [2,3]. Dispersion fuels contain
U–Mo fuel spheres imbedded in an Al-modified matrix. The Al ma-
trix is often Si enriched to suppress the fuel-matrix interaction
layer that forms during fuel fabrication. Monolithic fuels are U–
Mo foils rolled to a thickness of roughly 250 lm. This foil is secured
in Al cladding. The cladding is bonded to the fuel by either friction
stir welding or by hot isostatic pressing [3]. The cladding can be Si
modified and/or Zr foil barriers implemented between the fuel foil
and the cladding [4].
The use of monolithic fuels provides an overall higher density of
U in the fuel when compared to U–Mo dispersion fuels. For disper-
sion fuels, 1-mm-diameter punchings were obtained from the
plate and successfully created into SEM and TEM samples using
conventional TEM preparation techniques [5,6]. Performing
1-mm-diameter punchings on irradiated monolithic plates was
shown to be ineffective, as the punchings crumbled because the
irradiated fuel was brittle. Keiser was able to perform SEM analysis
on pieces of shattered U–Mo monolithic fuel that was punched
⇑ Corresponding author. Tel.: +1 208 351 2734; fax: +1 208 533 7863.
E-mail address: Brandon.Miller@inl.gov (B.D. Miller).
from the low-flux and high-flux ends of an RERTR plate [2]. Limited
data was obtained, which included local chemical composition and
some analysis of the intra- and intergranular bubbles throughout
the crystalline fuel regions. The main conclusion of this study
was that punching monolithic fuel is not a feasible option to create
SEM samples. This led to the need for a new method to create small
sections of irradiated monolithic fuel.
In addition to a new method to create monolithic sections for
microstructural analysis, a method to produce site-specific TEM
samples was needed. Typical TEM samples of the U–Mo dispersion
fuels are created using the electro-polish technique. This leads to
the creation of electron transparent areas randomly around the
U–Mo fuel particles. The need to obtain TEM samples in specific
areas led to the use of FIB to create TEM lamellas. Additionally,
with only two fuel-cladding interfaces in the irradiated monolithic
fuel samples, FIB would provide site-specific areas to better under-
stand the irradiated microstructure of the U–10Mo monolithic
fuel–cladding interaction layer.
Using FIB to create TEM lamellas out of non-irradiated materials
is well documented [7]. Samples created by FIB preparation of irra-
diated and fresh nuclear fuel TEM lamellas are less documented,
with most of the focus being on ion- and neutron-irradiated struc-
ture materials [8–14]. Some TEM characterization of as-fabricated
RERTR fuels has been performed using samples produced by FIB
[8,9]. For irradiated structural materials, the specific areas of inter-
est are ferritic steels and oxide dispersive strengthened (ODS)
steels [10–13]. Ion irradiations of uranium-based oxides have been
performed to understand the effect of radiation damage on the
microstructure [14]. FIB was used to create TEM lamellas of these
irradiated materials. This work will discuss the advantages/disad-
vantages of using the FIB to create TEM samples of neutron-irradi-
ated nuclear fuels.

0022-3115/$ - see front matter Ó 2012 Elsevier B.V. All rights reserved.
doi:10.1016/j.jnucmat.2012.01.022
 B.D. Miller et al. / Journal of Nuclear Materials 424 (2012) 38–42
2. Irradiated monolithic fuel sample preparation
Plate L1F140 from the RERTR-7 irradiation campaign was cho-
sen for sectioning. This monolithic plate has a 250-lm
U-10 wt%Mo foil sandwiched between Al-6061 cladding. The total
thickness of the fuel plate is 1.5 mm. The fuel and cladding were
bonded by friction stir welding [3]. The fuel plate was irradiated in
the Advanced Test Reactor (ATR) at Idaho National Laboratory
(INL). The calculated average fission density of this plate was
5.25  1021 fiss/cm3 [15].
Plate L1F140 was sectioned in half in the width direction for
optical microscopy. One sectioned half was loaded into a low-
speed saw, and a 1.25–1.5 mm cut was made in the width direc-
tion. This fuel strip was secured to an aluminum plate using epoxy.
Using the low-speed saw, 1-mm-thick cuts were made along the
width direction of the fuel strip, creating a small-sectioned,
1.5 mm  1.5 mm  1.0 mm fuel sample. Fig. 1 shows the cutting
process. The local average fission density, average fission rate, peak
heat flux, and average irradiation temperature of the fuel from the
sectioned monolithic fuel piece is 3.46  1021 fiss/cm3, 4.45 
1014 fiss/cm3 s, 195 W/cm2, and 112 °C, respectively [15].
At the Electron Microscopy Laboratory (EML), the fuel sample
was mounted in an epoxy metallography mount. The sample was
polished through 1200 grit. The met-mount was decontaminated
and gold-coated for SEM and FIB analysis. Using a JEOL 7000F
SEM equipped with a wavelength dispersive spectroscopy (WDS)
analyzer, SEM was performed to characterize the microstructure
of the monolithic fuel and identify areas were FIB TEM liftouts
were going to be obtained. Using a FEI Quanta3D Dualbeam FIB,
TEM lift-outs were obtained in the fuel regions and across the
fuel-cladding interaction layer of the fuel.
TEM lamellae were created by coarse trenching a 20 lm 
10 lm  1 lm sample using the FIB. The samples were welded to
a copper TEM grid for final TEM thinning. The lamellas were
thinned to a thickness of roughly 100 nm using 30 kV. The samples
were thinned to a final thickness of roughly 75 nm using 5 kV. Final
cleaning of the sample was conducted using 2 kV.
Fig. 2 shows the TEM lamella locations and areas where cross-
section analyzes were performed in the FIB. After the final FIB thin-
ning to make the sample transparent to a 200-keV electron beam,
TEM analysis was performed on the FIB-prepared samples using a
200-keV JEOL 2010 TEM equipped with energy dispersive spectros-
copy (EDS).
2.1. Key observations from using the FIB for creating TEM samples of
irradiated fuel samples
Advantages and disadvantages of using the FIB will be pre-
sented, including some of the results of the TEM and FIB analysis
Fig. 1. The sectioning process of monolithic fuel.
39
Fig. 2. Areas where TEM lamellas and cross sections were obtained using the FIB.
of the irradiated monolithic fuel. Further detailed analysis of the
SEM and TEM results, focusing on the microstructure of the high
and low burnup ends of the plate, will be provided in a later paper.
A major advantage of using the FIB to create TEM samples is the
reduced personal radiation exposure during TEM sample prepara-
tion. By using the fuel sample analyzed in the SEM, only one sam-
ple is needed to achieve both SEM and TEM results, thus removing
the extra dose received by creating a second sample for TEM anal-
ysis. The use of small FIB samples compared to conventional TEM
samples significantly cuts down the sample total activity and im-
proves the quality of EDS analysis for radioactive samples. Compar-
ing volume, the FIB sample can be over 10,00,000 times less in
volume compared to a conventional TEM sample. By reducing
the radiation background noise, more meaningful EDS measure-
ments can be obtained. Fig. 3 shows an EDS spectrum from a con-
ventionally prepared irradiated dispersion fuel and the EDS
spectrum from the monolithic fuel sample. Notice the ratio of the
main uranium peak to its background value. The conventional
TEM value has a ratio of 4 and the FIB sample has a ratio of 10. This
indicates a lower background in the FIB monolithic fuel sample.
Additionally, with the reduction in background, lower energy
X-rays (<5 keV) can be seen easier, providing better EDS data.
Another key advantage of using the FIB to create TEM samples is
the ability to create site-specific samples in key areas of the fuel
sample. This is especially true in a monolithic fuel sample. In
RERTR monolithic fuel samples, a fuel-cladding interaction layer
forms between the Al cladding and the U–Mo fuel. This region is
generally less than 10 lm thick. Using the FIB, TEM samples were
obtained across this interface to help understand its microstruc-
ture. Using conventional methods to create TEM samples, such as
jet-polishing, it would be difficult to create a TEM sample in this
specific region. Fig. 2 shows where cross-sectional TEM lamellas
were fabricated on the L1F140 monolithic fuel sample.
Another advantage of using the FIB is the ability to cut out re-
gions of the fuel to get cleaner cross sections for SEM imaging than
that of the mechanically polished surface, which tends to smear
out the fine details such as fine bubbles. For the monolithic fuel
sample, this ability provides easily visible sectioned surfaces to
better understand the fission gas bubble microstructure along
grain boundaries. Fig. 4a shows a SEM image of a cross section of
the fuel showing fission gas bubbles along the grain boundaries.
The cross-section technique acts like a fracture surface in the fuel,
40 B.D. Miller et al. / Journal of Nuclear Materials 424 (2012) 38–42
Fig. 3. EDS spectrums from (a) a conventionally prepared irradiated dispersion fuel, and (b) a TEM sample produced by the FIB.
providing the true size and shape of the fission gas bubbles and the
fuel grains. Fig. 4b shows a TEM image of the grain boundary
fission gas bubbles for comparison to the SEM image.
TEM analysis demonstrated that the TEM lamellas prepared
using FIB were useful and quality samples. Fig. 5 shows bright field
TEM micrographs of (a) the fuel-cladding interaction layer and (b)
the ordered fission gas bubble superlattice in the irradiated
U–10Mo fuel. A fission gas bubble superlattice has been found to
develop in U–Mo alloys irradiated at low temperatures, and this
feature allows for stable swelling behavior [5,6]. Comparing these
results to work performed on conventionally prepared irradiated
TEM dispersion fuels, the monolithic fuel microstructure was sim-
ilar to the microstructure seen in U–10Mo dispersion fuels, i.e., a
similar fission gas superlattice was observed, along with similar
bubbles on the U–10Mo grain boundaries. From the FIB-prepared
sample, bubble size distributions from the grain boundary bubbles
can be found. In conventionally prepared TEM samples, this task is
much harder. The ability to accurately measure the grain boundary
bubble sizes provides useful information about the percentage of
fission gas residing on grain boundaries and in the bulk of the
grains. This data can be employed in bubble models to improve
existing models. Additionally, the samples provided useful infor-
mation on understanding solid fission product behavior around
grain-boundary fission gas bubbles.
Using the FIB to create TEM samples does have disadvantages
compared to the more conventional techniques. One of the key
issues is the amount of TEM-visible area to analyze. Using conven-
tional TEM sample preparation methods, much more electron-
transparent area can be analyzed in the TEM, providing more
representative and complete pictures of the microstructure. Using
the FIB, one would have to make tens of samples to analyze the
same amount of area. This lack of visible area makes achieving sta-
tistically relevant information on defects and overall microstruc-
ture much more challenging.
Another key disadvantage of using FIB to produce TEM samples
is dealing with the high porosity within irradiated fuel samples.
The area around these pores will mill much faster than the bulk.
This leads to curtaining artifacts, which can be seen in Fig. 4b. They
appear as the streaking artifact below fission gas bubbles. This cur-
taining causes the area below a feature with respect to the ion
source to be polished slower than the non-curtained regions above
the feature [16]. This problem arises because the ion-beam polish-
ing in the FIB is along one direction, perpendicular to the sample
surface. Similarly, the multi-layers of differing material of the
monolithic fuel preferentially ion mill depending on the sputtering
behavior of the respective material. This leads to one layer thinning
much faster than the other. In the irradiated monolithic fuel, the Al
cladding polished much faster than the U–Mo fuel region. This
preferential polishing of the Al cladding at the interface dominates
during final thinning of the TEM lamella.
Another potential disadvantage of the FIB is the ion-induced
damage during milling. It is important to have pre-irradiated sam-
ples created by FIB and conventional preparation methods to com-
pare what is truly FIB damage or radiation damage. One other
potential disadvantage of using the FIB to create TEM samples is
the chance of contaminating the FIB sample chamber due to the
sputtered material having to settle somewhere after being
removed from the bulk. A detailed survey to identify any contam-
ination present in the FIB after TEM samples were created from the
irradiated fuel samples showed that no contamination was present
in the FIB. The fixture to hold the TEM copper grids was found to
have elevated contamination. This fixture is an inexpensive con-
sumable, and decontamination of the fixture was shown to reduce
contamination levels to pre-experimental values.
3. Summary
A sectioning method to create SEM and TEM samples from irra-
diated monolithic fuel samples was shown to be successful. FIB
was shown to be a useful technique to create high-quality TEM
samples from irradiated fuel sections. Additionally, the ability of
the FIB to create cross sections of the fuel that can be used to de-
velop better understanding of the bubble microstructure of the
U–Mo fuel has been demonstrated. Advantages and disadvantages
of using the FIB to create TEM samples were discussed. Preparing
TEM samples in the FIB reduces dose received to personnel, offers
the ability to create site-specific TEM samples, reduces the radia-
tion field interference when performing EDS in the TEM, and pro-
vides useful cross sections to better understand the fission gas
behavior along grain boundaries. Disadvantages include curtaining
from the milling process and the resulting small sample volume to
examine in the TEM from FIB-prepared samples. Many FIB samples
would be required to achieve the same electron-transparent area
from a conventionally prepared TEM sample. Elevated contamina-
tion in the FIB was not seen after TEM sample preparation.
4. US department of energy disclaimer
This information was prepared as an account of work sponsored
by an agency of the US Government. Neither the US Government
nor any agency thereof, nor any of their employees, makes any
warranty, express or implied, or assumes any legal liability or
responsibility for the accuracy, completeness, or usefulness of
 B.D. Miller et al. / Journal of Nuclear Materials 424 (2012) 38–42
Fig. 4. (a) SEM image of a FIB cross section showing fission gas bubbles populating
the grain boundaries of the U–10Mo fuel. (b) TEM image of the fission gas bubbles
along the grain boundaries.
any information, apparatus, product, or process disclosed, or repre-
sents that is use would not infringe privately owned rights.
References herein to any specific commercial product, process, or
service by trade name, trademark, manufacturer, or otherwise,
does not necessarily constitute or imply its endorsement, recom-
mendation, or favoring by the US Government or any agency there-
of. The views and opinions of authors expressed herein do not
necessarily state or reflect those of the US Government or any
agency thereof.
41
Fig. 5. TEM micrographs of the irradiated monolithic fuel at the (a) fuel-cladding
interaction layer and the (b) ordered bubble superlattice in a [1 1 0] orientation with
the U–10Mo diffraction pattern and the satellite spots from the bubble superlattice.
Acknowledgments
The authors would like to express their gratitude to the HFEF
staff at INL for assistance in sectioning the SEM/FIB sample. They
would also express their gratitude to the EML health physics
technicians and nuclear operators for their help and support. The
work was supported through funding provided by the US Depart-
ment of Energy (DOE) to the RERTR program at INL, operated by
Battelle Energy Alliance, LLC, under DOE Idaho Operations Office
Contract DE-AC07-051D14517.
References
[1] D.M. Wachs, D.D. Keiser, Jr., M.K. Meyer, D.E. Burkes, C.L. Clark, G.A. Moore, J.-F.
Jue, T. Totev, G.L. Hofman, T.C. Wiencek, Y.S. Kim, J.L. Snelgrove, GLOBAL 2007,
American Nuclear Society, Boise, ID, USA, September 9–13, 2007.
42 B.D. Miller et al. / Journal of Nuclear Materials 424 (2012) 38–42
[2] D.D. Keiser Jr., J.F. Jue, A.B. Robinson, M.R. Finlay, RRFM-2008, in: 14th
International Meeting on Research Reactor Fuel Management, European
Nuclear Society, Hamburg, Germany, March 2–5, 2008.
[3] C.L. Clark, J.-F. Jue, G.A. Moore, N.P. Hallinan, B.H. Park, D.E. Burkes, RERTR-
2006, in: 28th International Meeting on Reduced Enrichment For Research and
Test Reactors, Cape Town, South Africa, October 29–November 2, 2006.
[4] A.B. Robinson, G.S. Chang, D.D. Keiser Jr., D.M. Wachs, D.L. Porter, Irradiation
Performance of U–Mo Alloy based Monolithic Plate-type Fuel-Design Selection,
INL/EXT-09-16807, August 2009, <http://www.inl.gov/technicalpublications/
Documents/4363871.pdf>.
[5] S. Van den Berghe, W. Van Renterghem, A. Leenaers, J. Nucl. Mater. 375 (2008)
340–346.
[6] J. Gan, D.D. Keiser Jr., D.M. Wachs, A.B. Robinson, B.D. Miller, T.R. Allen, J. Nucl.
Mater. 396 (2010) 234–239.
[7] P.R. Munroe, Mater. Char. 60 (2009) 2–13.
[8] B. Yao, E. Perez, D.D. Keiser Jr., J.F. Jue, C.R. Clark, N. Woolstenhulme, Y. Sohn, J.
Alloy. Compd. 509 (2011) 9487–9496.
[9] E. Perez, B. Yao, D.D. Keiser Jr., Y.H. Sohn, J. Nucl. Mater. 402 (2010) 8–14.
[10] A.G. Certain, K.G. Field, T.R. Allen, M.K. Miller, J. Bentley, J.T. Busby, J. Nucl.
Mater. 407 (2010) 2–9.
[11] H. Kishimoto, K. Yutani, R. Kasada, O. Hashitomi, A. Kimura, J. Nucl. Mater.
367–370 (Part A) (2007) 423–427.
[12] M.A. Pouchon, J. Chen, M. Döbeli, W. Hoffelner, J. Nucl. Mater. 352 (2006) 57–61.
[13] K. Yabuuchi, M. Saito, R. Kasada, A. Kimura, J. Nucl. Mater. 414 (2011) 498–502.
[14] M. Tang, K.S. Holliday, J.A. Valdez, B.P. Uberuaga, P.O. Dickerson, R.M.
Dickerson, Y. Wang, K.R. Czerwinski, K.E. Sickafus, J. Nucl. Mater. 389 (2009)
497–499.
[15] D.D. Keiser Jr., J.F. Jue, B.D. Miller, J. Gan, A.B. Robinson, P. Medvedev, RERTR-
2011, in: 33rd International Meeting on Reduced Enrichment For Research and
Test Reactors, Santiago, Chile, October 23–27, 2011.
[16] L.A. Giannuzzi, F.A. Sevie, Introduction to Focused Ion Beams, Instrumentation,
Theory, Techniques, and Practice, Springer-Verlag, New York, 2005.