We have measured vibronic emission spectra of an oxide of uranium formed after laser ablation of the metal in gaseous oxygen. Specifically, we have measured the time-dependent relative intensity of a band located at approximately 593.6 nm in 16O2. This band grew in intensity relative to neighboring atomic features as a function time in an oxygen environment but was relatively invariant with time in argon. In addition, we have measured the spectral shift of this band in an 18O2 atmosphere. Based on this shift, and by comparison with earlier results obtained from free-jet expansion and laser excitation, we can confirm that the oxide in question is UO, consistent with recent reports based on laser ablation in 16O2 only.

Laser ablation offers a simple and inexpensive means to produce plasma environments in which to study the high temperature emission spectra of atoms and molecules. In laser-induced breakdown spectroscopy (LIBS), a pulsed laser focused at or near the sample surface produces an ablation plasma, and the resultant emission spectrum of atomic de-excitation is typically used to determine the elemental composition of the ablated material.1–3 Emission spectra are also routinely used to study simple molecular species that form in the condensing plasma. The use of ultrafast methods to collect spectral data permits the study of spatially and temporally resolved in-growth of molecular species in the evolving ablation plume.4–9 Recent work in this area has stimulated the development of models to describe rovibronic transitions of diatomic species that can be used to evaluate molecular temperatures,4 local thermodynamic equilibrium conditions,7 and chemical kinetics pathways9 in the plasma plume.

Most of these recent molecular emission studies of laser-induced plasmas have focused on low-Z elements, where the discrimination of atomic and molecular emission lines is relatively straightforward. There is considerable interest in the application of LIBS technology to study uranium-bearing materials.10–15 A better understanding of uranium oxide formation during laser plasma condensation would be beneficial to both nuclear forensics and non-proliferation monitoring. However, comparatively little has been done to explore the molecular emission of uranium oxide species due to the complexity of the spectra. The use of molecular emission spectra to understand uranium oxidation chemistry requires the accurate assignment of uranium oxide emission bands.

Recently, laser ablation molecular isotopic spectrometry (LAMIS) has been proposed as a means by which the isotopic composition of samples can be determined, as the isotopic shifts in rovibrational transitions in diatomic molecules are larger and hence more easily resolved, in at least some cases, than isotope shifts in atomic emissions.16,17 Recent LAMIS studies have identified a uranium oxide line and measured the isotopic shift between the 235U and 238U substituted species.18–20 Mao et al.18 and Hartig et al.19 independently identify a uranium oxide emission peak at approximately 593.6 nm with an isotopic red shift of 35 pm and 50 pm, respectively, between the 235U and 238U substituted species.18,19 Their basis for identifying this peak as uranium oxide was its growth in intensity (relative to nearby atomic lines) as a function of time after laser impingement (thus suggesting molecular formation as temperature decreases in the expanding plasma). They further correlated the peak's spectral position with at least one known transition previously reported for UO identified in thermal emission,21 free-jet expansion, and laser excitation experiments.21–24 Hartig et al.19 did not positively identify the peak as due to UO but rather more generally as a uranium oxide species (UxOy) because of differences in position and shape from the previous reports. However, Hartig et al.20 and Mao et al.18 did positively identify the peak as due to UO, respectively, aided in this identification by the dependence of the relative atomic and molecular emission intensities on gas pressure,20 and the observation of two additional, although less distinct, bands at 594.57 nm and 595.22 nm.18 

Here we present molecular emission spectra of uranium oxide substituted with 16O and 18O to provide additional evidence that the spectral feature positioned at ∼593.6 nm is from UO emission. While free-jet expansion and laser excitation data for the isotopic shift between U16O and U18O have been shown,21,22 uranium emission spectra from laser ablation plumes have not been previously reported in an 18O2 environment. Our data show an isotopic shift between the 16O and 18O substituted isotopomers that is consistent with that observed in free-jet expansion and laser excitation experiments.

In our experiments, a New Wave Polaris II Nd:YAG (1064 nm wavelength, ∼5 ns pulse length, 2.75 mm output beam diameter) was used to generate 1.4 mJ pulses, which were focused onto the surface of a depleted uranium (<0.72% 235U) metal target using a planoconvex lens (6.3 cm focal length) to create an ablation plasma. Emission spectra were collected using a Shamrock SR-303i spectrometer equipped with an 1800 lines/mm grating and coupled to an Andor iSTAR intensified CCD (1024 × 1024 pixels, 13 μm pixel size). Spectra were collected at either 300 ns, 550 ns, 800 ns, 1050 ns, or 1300 ns after the arrival of each laser pulse. A 250 ns gate-width was selected in each case. Spectra were collected as the sum of 500 shots at 20 Hz, with a spectral window of 20 nm (580–600 nm for this study). The wavelength was calibrated using a Ne standard lamp, and the Ne atomic emission lines had a measured FWHM of 0.09 nm. The measured spectral intensities were corrected for the instrumental response using a standard lamp.

The atmosphere around the uranium metal target was controlled in an air-tight chamber with sapphire windows that are transparent to the incident ablation laser and emitted light. To achieve 16O2, 18O2, and Ar environments, the chamber was evacuated to ∼0.03 atm using a roughing pump and then filled with 1 atm of either isotopically purified 16O2 (99.98% purity), 18O2 (97.4% 18O2, 1.7% 17O2), or Ar (99.99%).

Figure 1 shows time-resolved spectra for uranium ablation in 1 atm 16O2 [Fig. 1(a)], 1 atm 18O2 [Fig. 1(b)] and 1 atm Ar [Fig. 1(c)] over the range of 590 nm–594 nm. The intensities are shown in log-scale, to emphasize the change in intensity of key features in the spectra. In Fig. 1(a), the evolution of three features as a function of delay time is highlighted in the spectra: the atomic emission lines at 591.54 nm and 593.38 nm and another prominent emission peak at 593.56 nm. In the 16O2 environment, the feature at 593.56 nm is difficult to discern at early delay times (300 ns), while at late times (1300 ns), it is more prominent than the atomic line at 593.38 nm. Notably, as delay time increases [as shown moving from the top spectrum to the bottom spectrum in Fig. 1(a), light to dark], the feature grows in intensity relative to its neighboring atomic lines even though there is an overall decrease in spectral intensity. In the 18O2 environment shown in Fig. 1(b), the peak at 593.56 nm is absent (or much reduced in intensity), and a prominent feature is seen at 593.74 nm that grows in relative intensity as delay time increases. The atomic lines at 591.54 nm and 593.38 nm are present at roughly the same intensities in both experiments. In the spectra obtained in an Ar environment, as shown in Fig. 1(c), the overall intensity is lower by at least a factor of 2 at all delay times. Further, in the Ar environment, the relative change in absolute intensity from 300 ns to 1300 ns is significantly less than in either O2 environment: in Ar, the signal decreases by a factor of 2, while in O2, the signal decreases by nearly an order of magnitude. Most notably, there is no prominent peak at either 593.56 nm or 593.74 nm in Ar, although there may be minimal contribution to the 593.56 nm peak due to residual 16O2 in the chamber, or the oxide layer on the uranium metal.

The spectral differences between the 16O2 and 18O2 ablation environments are emphasized in Fig. 2 for early [550 ns, Figs. 2(a) and 2(b)] and late [1300 ns, Figs. 2(c) and 2(d)] delay times. The most prominent peak in Fig. 2(a) is the neutral atomic emission line at ∼593.38 nm (also seen in Fig. 1), with two other neighboring atomic emission lines labeled at 592.55 nm and 592.93 nm. While the intensities of individual atomic lines are slightly different, the main features observed in the 16O2 and 18O2 environments are the same at the earlier delay time of 550 ns. The 16O2 and 18O2 spectra show a clear difference at the late delay time of 1300 ns with distinct features at 593.56 nm (in 16O2) and 593.74 nm (in 18O2), as shown in Fig. 2(d). Figures 2(b) and 2(d) show magnified views (from 593 nm to 594 nm) of the spectral features in Figs. 2(a) and 2(c), respectively. These respective features are only present at any appreciable intensity in their unique oxygen isotope environments. Though the features at 593.56 nm and 593.74 nm are observable at the 550 ns delay time [Fig. 2(b)], they become significantly more intense than the neighboring 593.38 nm atomic line at later delay [Fig. 2(d)]. Further, the relative intensity of the line at 593.56 nm (in 16O2) is roughly 2-fold higher in absolute intensity than the 593.74 nm line (in 18O2) at identical delay times.

In Fig. 3, the intensities of the 593.56 nm and 593.74 nm peaks are compared with the intensity of the neighboring atomic line at 592.55 nm [as shown in Fig. 2(a)], taken as ratios to the strong atomic line at 591.54 nm (labeled in Fig. 1) in oxygen [both 16O2 and 18O2, Fig. 3(a)] and argon [Fig. 3(b)]. In the 1 atm oxygen environment, the atomic line ratios (blue triangles) remain relatively time-invariant for delay times exceeding 500 ns (decreased by roughly 25% from the value at 300 ns delay), while the peaks at 593.56 nm (black circles) and 593.74 nm (red squares), taken as ratios to the atomic line at 591.54 nm, increase by more than 2-fold. However, in the 1 atm Ar environment, all of these ratios become relatively constant after 300 ns.

The time-dependent in-growth of the 593.56 nm peak relative to the neighboring atomic emission lines [Fig. 3(a)] in the 16O2 experiment suggests this is the same feature reported in recent studies by Mao et al.18 (593.55 nm) and Hartig et al.19,20 (593.57 nm), which were assigned as UxOy19 or UO.18,20 The absence of a peak at 593.56 nm in 18O2 is evidence that this is due to the emission from an oxide of uranium; in fact, there appears to be a local minimum in the spectrum at this position in the 18O2 ablation environment. The in-growth of the 593.74 nm peak in the 18O2 environment provides further evidence that these features are associated with uranium oxide emission, particularly as it has a similar time dependence to that of the 593.56 nm peak in 16O2 [Fig. 3(a)], and neither peak shows a relative increase in the Ar environment [Fig. 3(b)].

It is important to note that there are weak atomic emission lines reported at 593.55 nm and 593.74 nm.25 Further, the shoulder at 593.74 nm shown in Fig. 2(b) in the spectrum acquired in 16O2 could lead to the conclusion that the peak in 18O2 is due to atomic emission. However, the spectra shown in Fig. 2 and the high-resolution time-resolved spectra of Mao et al.18 both show there is never a time when the intensity at 593.74 nm exceeds the intensity of the atomic emission line identified at 593.38 nm. This is not the case in the 18O2 experiment, where the peak at 593.74 nm becomes the dominant feature in the spectrum [Fig. 2(d)]. Given the clear increase in intensity of the peaks at 593.56 nm and 593.74 nm at late delay times relative to the atomic line intensities, it is unlikely that either of these peaks are solely the result of atomic emission, especially considering their notable absence in Ar. Nevertheless, the presence of atomic emission interference could explain the aberrant increase in all the ratios shown in Fig. 3 at early delay (300 ns), when atomic emission has been shown to dominate molecular emission.7,17

As stated above, the relative intensity change in peaks alone is not sufficient to identify this emission as a UO band. Kaledin and Heaven21 reported molecular band centers for 238U16O by free-jet expansion and laser excitation at 593.52 nm and 593.66 nm at low-resolution and 593.48 nm at high-resolution [shown as filled black circles, Fig. 2(d)]—assigned to a 7sσ  → 7pπ metal-centered transition.26 That study also reports bandheads for 238U16O at 593.48 nm, 593.55 nm, and 593.68 nm in thermal emission spectra [shown as open black circles, Fig. 2(d)]. The peak we measured at 593.56 nm in 16O2 and the positions reported by Hartig et al.19,20 and Mao et al.,18 are approximately in the same location as those reported by Kaledin and Heaven and could potentially be the convolution of multiple bands.

As seen in Fig. 2(d), the apparent 593.56 nm peak width encompasses the positions of at least three peaks observed in the published free-jet expansion and emission data. Interestingly, the free-jet expansion data recorded at 130 K show the most intense peak at 593.48 nm, while the emission data (recorded at 2500 K) show the most intense peak at 593.52 nm, in contrast to the peak position of 593.56 nm measured in the high-resolution spectra shown in Ref. 18. Given that laser ablation plumes are initially much hotter (up to 10 000 K or more),1 it is possible that this observed shift in the 238U16O peak position is due to differences in the temperature at which the data were acquired. While a temperature-based attribution is speculative, it plausibly accounts for the 0.04–0.08 nm longer wavelength reported in the studies by Hartig et al.,19,20 Mao et al.,18 and this study.

For 238U18O, only free-jet expansion data are available in the study by Kaledin and Heaven, seen as red circles in Fig. 2(d). While an intense band center is reported at 593.74 nm (the same location as the peak measured in the present study), the free-jet expansion data show a very intense peak at 595.69 nm, a position which did not exhibit any notable shift or intensity change in our spectra (see supplementary material). We also note that in the free-jet expansion data of Kaledin and Heaven for 238U16O, a peak was observed at 581.38 nm that was slightly more intense than that at 593.48 nm. This peak was not apparent (or at least much less intense) in our spectra, and it was not reported in either of the two recent emission studies.18,19 It is possible that the intensity-ratios observed in free-jet expansion and laser induced fluorescence studies are not necessarily indicative of what might be observed in ablation spectra due to vastly different experimental conditions in temperature (130 K for free-jet expansion vs. thousands of kelvins for emission studies) and pressure. Particularly, the effect of pressure could affect the intensity ratios by kinetically favoring one vibronic transition over another.

While the Kaledin and Heaven study does not report emission data in an 18O2 environment, their free-jet expansion data signal a band center at 593.74 nm attributed to 238U18O using mass spectrometry results. This is in the same position where we observed the in-growth of a prominent emission feature in the 18O ablation environment, which was not observed in Ar. Further, they report an isotopic shift of 0.197 nm between the 238U16O and 238U18O in this position from the same assigned transition. In our study, the ∼0.18 nm shift between the 593.56 nm and 593.74 nm peaks in 16O2 and 18O2, respectively, is (given our spectral resolution) indistinguishable from the value reported by Kaledin and Heaven, acquired from free-jet expansion experiments. While there is an obvious difference in intensity between the peaks we measured at 593.56 nm in 16O2 and 593.74 nm in 18O2 (the 238U16O peak is roughly 2-fold higher in intensity), dramatic changes in relative intensity with isotopic substitution are shown in many of the transitions reported from free-jet expansion experiments by Kaledin and Heaven.

The growth in relative intensity we observed in oxygen ablation environments, and lack thereof in argon, are compelling evidence that the peak at 593.56 nm is a uranium oxide emission band. Further, the spectral shift in the peak from 593.56 nm in 16O2 to 593.74 nm in 18O2 corresponds to the shift shown by free-jet expansion/laser excitation between 238U16O and 238U18O. This affirms conclusions drawn from other recent laser ablation experiments. Considering the easily resolvable isotopic shift, this substitution technique may be useful for identifying (and assigning) future UxOy emission lines for exploring high-temperature uranium chemistry as well as for use in analytical techniques such as LAMIS.

See supplementary material for a comparison of the band positions published by Kaledin and Heaven (1997) and the full emission spectrum we recorded at 1300 ns in both 16O2 and 18O2.

We are grateful to M. Heaven for commenting on an early version of our manuscript and for numerous helpful suggestions. We thank R. Ryerson for generously providing 18O2 for this experiment, as well as E. Stavrou for helpful discussion. We also thank two anonymous reviewers for their comments. This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract No. DE-AC52-07NA27344. Funding was provided by a DTRA Basic Science Grant (HDTRA1-16-1-0020) and by DNDO's National Nuclear Forensics Expertise Development Program. The views and conclusions contained in this document are those of the authors and should not be interpreted as representing the official policies, either expressed or implied, of the U.S. Department of Homeland Security. LLNL-JRNL-729818.

1.
D. A.
Cremers
and
L. J.
Radziemski
,
Handbook of Laser‐Induced Breakdown Spectroscopy
, 2nd ed. (
John Wiley & Sons, Ltd
,
2013
).
2.
E.
Tognoni
,
V.
Palleschi
,
M.
Corsi
, and
G.
Cristoforetti
,
Spectrochim. Acta, Part B
57
,
1115
(
2002
).
3.
R. S.
Harmon
,
R. E.
Russo
, and
R. R.
Hark
,
Spectrochim. Acta, Part B
87
,
11
(
2013
).
4.
C. G.
Parigger
and
J. O.
Hornkohl
,
Spectrochim. Acta, Part A
81
,
404
(
2011
).
5.
C. G.
Parigger
,
A. C.
Woods
,
D. M.
Surmick
,
G.
Gautam
,
M. J.
Witte
, and
J. O.
Hornkohl
,
Spectrochim. Acta, Part B
107
,
132
(
2015
).
6.
A.
DeGiacomo
,
M.
Dell'Aglio
,
O.
De Pascale
,
R.
Gaudiuso
,
V.
Palleschi
,
C.
Parigger
, and
A.
Woods
,
Spectrochim. Acta, Part B
100
,
180
(
2014
).
7.
J.
Lam
,
V.
Motto-Ros
,
D.
Misiak
,
C.
Dujardin
,
G.
Ledoux
, and
D.
Amans
,
Spectrochim. Acta, Part B
101
,
86
(
2014
).
8.
J.
Lam
,
D.
Amans
,
C.
Dujardin
,
G.
Ledoux
, and
A.
Allouche
,
J. Phys. Chem. A
119
,
8944
(
2015
).
9.
M.
Gaft
,
L.
Nagli
,
N.
Eliezer
, and
Y.
Groisman
,
Spectrochim. Acta, Part B
110
,
56
(
2015
).
10.
F. R.
Doucet
,
G.
Lithgow
,
R.
Kosierb
,
P.
Bouchard
, and
M.
Sabsabi
,
J. Anal. At. Spectrom.
26
,
536
(
2011
).
11.
Y. S.
Kim
,
B. Y.
Han
,
H. S.
Shin
,
H. D.
Kim
,
E. C.
Jung
,
J. H.
Jung
, and
S. H.
Na
,
Spectrochim. Acta, Part B
74
,
190
(
2012
).
12.
S.
Chen
,
A.
El-Jaby
,
F.
Doucet
,
P.
Bouchard
, and
M.
Sabsabi
, “
Development of laser-induced breakdown spectroscopy technologies for nuclear safeguards and forensic applications
,”
Report No. IAEA-CN-220
,
2015
.
13.
E. J.
Judge
,
J. E.
Barefield
,
J. M.
Berg
,
S. M.
Clegg
,
G. J.
Havrilla
,
V. M.
Montoya
,
L. A.
Le
, and
L. N.
Lopez
,
Spectrochim. Acta, Part B
83
,
28
(
2013
).
14.
J. R.
Wachter
and
D. A.
Cremers
,
Appl. Spectrosc.
41
,
1042
(
1987
).
15.
I.
Choi
,
G.
Chan
,
X.
Mao
,
D. L.
Perry
, and
R. E.
Russo
,
Appl. Spectrosc.
67
,
1275
(
2013
).
16.
H.
Hou
,
G.
Chan
,
X.
Mao
,
V.
Zorba
,
R.
Zheng
, and
R.
Russo
,
Anal. Chem.
87
,
4788
(
2015
).
17.
X.
Mao
,
A.
Bol'shakov
,
I.
Choi
,
C. P.
McKay
,
D. L.
Perry
,
O.
Sorkhabi
, and
R. E.
Russo
,
Spectrochim. Acta, Part B
66
,
767
(
2011
).
18.
X.
Mao
,
G.
Chan
,
I.
Choi
,
V.
Zorba
, and
R. E.
Russo
,
J. Radioanal. Nucl. Chem.
312
,
121
(
2017
).
19.
K. C.
Hartig
,
I.
Ghebregziabher
, and
I.
Jovanovic
,
Sci. Rep.
7
,
43852
(
2017
).
20.
K. C.
Hartig
,
S. S.
Harilal
,
M. C.
Phillips
,
B. E.
Brumfield
, and
I.
Jovanovic
,
Opt. Express
25
,
11477
(
2017
).
21.
L. A.
Kaledin
and
M. C.
Heaven
,
J. Mol. Spectrosc.
185
,
1
(
1997
).
22.
M. C.
Heaven
,
J.
Nicolai
,
S. J.
Riley
, and
E. K.
Parks
,
Chem. Phys. Lett.
119
,
229
(
1985
).
23.
L. A.
Kaledin
,
A. N.
Kulikov
,
A. I.
Kobylyanskii
,
E. A.
Shenyavskaya
, and
L. V.
Gurvich
,
Russ. J. Phys. Chem.
61
,
1374
(
1987
).
24.
L. A.
Kaledin
,
A. N.
Kulikov
, and
L. V.
Gurvich
,
Russ. J. Phys. Chem.
63
,
801
(
1989
).
25.
B. A.
Palmer
,
R. A.
Keller
, and
R.
Engleman
,
An Atlas of Uranium Emission Intensities in a Hollow Cathode Discharge
(
LASL
,
1980
).
26.
R.
Tyagi
,
Z.
Zhang
, and
R. M.
Pitzer
,
J. Phys. Chem. A
118
,
11758
(
2014
).

Supplementary Material