| Operation of the TREAT reactor at the Idaho National Laboratory is being
resumed in order to facilitate transient testing of new fuels and materials. As part of the
restart effort, the possibility of converting the current HEU core to LEU is being studied.
As is the case for the current fuel, the replacement fuel would consist of fissile material
distributed throughout a carbon matrix material. As part of this study, Idaho State
University was given the task to develop methods that can be used to identify and
quantify non-graphitic carbon in the new LEU fuel matrix. This thesis reports the
investigation and application of methods that were deemed fit to meet the aforementioned
objectives of this project.
Transmission Electron Microscopy (TEM), Scanning Electron Microscopy
(SEM), and Dual Beam-Focused Ion Beam (DB-FIB) Microscopy were used as methods
for identification of non-graphitic carbon structures in the LEU fuel matrix. An image
analysis technique was also applied to the SEM images in an attempt to quantify the non-
graphitic content in the samples. X-Ray Diffraction (XRD) and Raman Spectroscopy
were used as methods for quantification through the application of the degree of
graphitization equation for XRD and the ratio of the disorder-induced D peak and order-
induced G peak intensities (I D /I G ) for Raman spectroscopy.
Amorphous carbon, turbostratic carbon, and quinolone insoluble (QI) particles
were identified at a nanoscopic level with TEM analysis, something the other microscopy
methods could not do. The DB-FIB instrument was useful in preparing good TEM
samples for analysis, but due to the heterogeneous nature of the samples it was difficult to
distinguish carbon phases using DB-FIB. The contrast between graphitic and non-graphitic carbon phases was also not visibly distinct in SEM images. Therefore, SEM
image analysis was considered too subjective and was not recommended as a technique
for quantification of the carbon phases in the TREAT fuel matrix.
Of the two methods for quantification, XRD was recommended for quantifying
graphitic content in the TREAT fuel matrix. It was found that adequate sample
preparation and careful placement of the sample in the diffractometer provided consistent
results. The estimated graphite quantity in the matrix was 88.7%, which is consistent with
a mass balance estimate of 88.9%. Raman spectroscopy, with emphasis on the value of
the ratio of D and G peak intensities, was not recommended as a method for
quantification of carbon phases in the TREAT fuel matrix. This technique produced
inconsistent results, from point to point within the same sample, likely due to graphene
armchair edges appearing as strong D peaks in the Raman spectra. Distinguishing
between actual disorders and armchair edges represented in the D peak would be difficult
to achieve. Also, unlike the XRD analysis, Raman spectroscopy is not a bulk method, for
which the results are an average of measurements across the entire surface of the sample.
Instead, Raman spectroscopy focuses on single points on the sample and to obtain a
representative average of the sample a significantly large number of point measurements
would be required |