A novel time-resolved laser fluorescence spectroscopy system for research on complexation of uranium(IV)

Article abstract of DOI:10.1016/j.saa.2009.04.013

To date only a small number of studies have investigated the chemical speciation of complexes and the fluorescence properties of metal ions whose emitted fluorescence lifetime is in the range of only few nanoseconds. This is due to a lack of advanced meth

Full text of DOI:10.1016/j.saa.2009.04.013

Spectrochimica Acta Part A 73 (2009) 902–908
Contents lists available at ScienceDirect
Spectrochimica Acta Part A: Molecular and
Biomolecular Spectroscopy
journal homepage: www.elsevier.com/locate/saa
A novel time-resolved laser fluorescence spectroscopy system for research on
complexation of uranium(IV)
S. Lehmann∗, G. Geipel, G. Grambole, G. Bernhard
Forschungszentrum Dresden-Rossendorf, Institute of Radiochemistry, P.O. Box 51 01 19, D-01314 Dresden, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
To date only a small number of studies have investigated the chemical speciation of complexes and the
fluorescence properties of metal ions whose emitted fluorescence lifetime is in the range of only few
nanoseconds. This is due to a lack of advanced methods which allow the conduction of these measure-
ments. In the current study we set up a new time-resolved laser fluorescence spectroscopy system with
which the fluorescence properties of metal ions with very short fluorescence lifetimes such as uranium(IV)
and its compounds can be investigated. By studying the fluorescence properties of uranium(IV) in per-
chloric acid, we showed uranium(IV) to have a detection limit of 5 × 10−7 M and a fluorescence decay time
of 2.74 0.36 ns. We further investigated the fluorescence properties of uranium(IV) during the reaction
with fluoride and applied our novel laser system to study the complexation of uranium(IV) with fluoride.
Our data revealed the formation of a 1:1 complex of uranium(IV) and fluoride. The corresponding com-
plex formation constant of uranium(IV) fluoride UF3+ was found to be log ˇ0 = 9.43 1.94. Our results
demonstrate that our novel time-resolved laser fluorescence spectroscopy system can successfully con-
duct speciation measurements of metal ions and their compounds with very short-lived fluorescence
lifetimes. Using this laser system, it is possible to analytically investigate such elements and compounds
in environmentally relevant concentration ranges.
Received 13 November 2008
Received in revised form 23 March 2009
Accepted 15 April 2009
Keywords:
Time-resolved laser fluorescence
spectroscopy
Uranium(IV) fluoride
Detection limit
Fluorescence lifetime
Complex formation constant
© 2009 Elsevier B.V. All rights reserved.
1. Introduction
therein] described the detection limits of various actinide ions for
speciation analyses. Depending on the metal ion, these were at least
In order to predict the behaviour of hazardous metals such
as uranium in the environment, it is important to understand
their physical properties and the chemistry of these elements and
their compounds. Aqueous uranium species can be found in sub-
millimolar to micromolar concentrations in the environment [1,2].
Under anoxic and reducing redox conditions, uranium(IV) in par-
ticular should be considered in such concentration ranges, also as
previous studies have shown that the solubility of uranium(IV),
depending on the pH, is in this range [3,4]. Hence, research on
the speciation of uranium(IV) requires highly sensitive analytical
methods.
Various spectroscopic techniques have been used to study the
chemical structure of trace actinides such as uranium(IV) and to
determine complex formation constants. However, there are several
advantages of the time-resolved laser-induced fluorescence spec-
troscopy (TRLFS) over other spectroscopic analytical methods. First,
the high sensitivity of TRLFS allows the investigation of actinide
complexation in environmentally relevant concentrations in the
submicromolar range. Using TRLFS, Geipel et al. [5 and references
one order of magnitude lower compared with conventional UV–vis
spectroscopy or laser-induced photoacoustic spectroscopy (LIPAS)
[6]. Second, TRLFS is a fast analytical technique, making the mea-
surement of samples less time-consuming compared with LIPAS.
Furthermore, time-resolved fluorescence spectroscopy has been
shown to be a very powerful analytical method to study complex
matrices in the environment. Excitation wavelength of the laser
beam and the resulting fluorescence spectrum are element-specific.
This makes TRLFS a very selective and sensitive method for studying
fluorescent metals and their complexation reactions, i.e. their spe-
ciation, by using spectral and temporal features [7] without being
invasive and destroying the original sample. The interpretation of
the spectra produced by TRLFS provides structural information by
clarifying the chemical speciation.
A number of studies have used TRLFS to investigate actinides
and in particular the speciation of uranium(VI). While aqueous ura-
nium(IV) has been studied using other (spectroscopic) techniques
[6,8,9], to date no study has used TRLFS. This may be due to the
fact that uranium(IV) was considered lack luminescence with the
exception of uranium(IV) phosphate in solution [10 and reference
therein]. A first description of the fluorescence properties of ura-
nium(IV) in aqueous solution was published only few years ago. A
detailed description of the excitation of uranium(IV) in perchlorate
∗
Corresponding author. Tel.: +49 351 260 2438; fax: +49 351 260 3553.
E-mail address: s.lehmann@fzd.de (S. Lehmann).
1386-1425/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.saa.2009.04.013

S. Lehmann et al. / Spectrochimica Acta Part A 73 (2009) 902–908
903
medium, its transition energies and fluorescence spectra was pro-
vided by Kirishima et al. [10]. Other studies have focused on how
the luminescence of uranium(IV) originates [11–13], however, these
studies gave mixed results. The laser system used by Kirishima et
al. [10], however, did not allow the investigation of all aspects of
the fluorescence of uranium(IV). The excitation laser beam used
in their study had a minimum pulse width of 20 ns, which was
longer than the uranium(IV) luminescence lifetime and therefore
overlaid the U(IV) fluorescence. Furthermore, although they pre-
dicted a measurable uranium(IV) concentration of 1 × 10−6 M, the
technical possibilities at the time only allowed the detection of ura-
nium(IV) of a concentration of 1 × 10−4 M. In order to be of practical
relevance, however, TRLFS should be able to detect uranium(IV) in
environmentally relevant concentration ranges.
Only recently we have been able to set up a laser system
that provides the technical prerequisite to extend research on
very short-lived fluorescence emitting metals like uranium(IV)
and its compounds. We focused on the fluorescence properties
of uranium(IV) in an acidic medium. We aimed to determine the
minimum detectable uranium(IV) concentration as well as the fluo-
rescence lifetime of uranium(IV) as an important distinctive feature
in the characterization and speciation process.
We first applied our novel fluorescence spectroscopy system to
study uranium(IV) fluoride. Uranium(IV) fluorides – single com-
pounds and even mixtures of different uranium(IV) fluorides with
U(VI) fluorides) [14] – have been investigated intensively. The great
amount of data available on uranium(IV) fluorides together with the
easy handling make U(IV) fluorides an ideal compound for exper-
iments on the complexation of uranium(IV). By determining the
complex formation constant of the uranium(IV) fluoride complex
we confirmed the accurate functioning of the laser system.
300i, Acton, MA) and an intensified CCD camera (PicoStar HR,
LaVision, Inc., Göttingen, Germany). The laser system operates at
wavelengths between 220 and 1800 nm. A schematic representa-
tion of the complete fluorescence spectroscopy system is shown in
Fig. 1.
A flash lamp-pumped Nd:YAG laser {1} generates laser pulses
of 1064 nm wavelength. The output energy averages 1300 mJ per
pulse, the pulse width is ∼10 ns with a pulse-to-pulse stability
of 2%. By directing the laser through several non-linear optical
devices, the 2nd and the 3rd harmonics are generated and laser
pulses leave the Nd:YAG with a pulse width of 8 ns at a wavelength
of 355 nm. The outgoing laser beam pumps an optical paramet-
ric oscillator system {2}. Approximately 80% of the laser beam is
channelled directly into the power oscillator (PO). The remaining
20% are directed into the master oscillator (MO) where the required
wavelength is created by double-pumping a BBO crystal. By using
a grating spectrograph, only the desired wavelength with a full-
width at half maximum (FWHM) of approximately 0.2 cm−1 leaves
the laser cavity of the master oscillator. The master oscillator line
seeds the power oscillator, and the PO amplifies the pulse, which
then passes through the optional frequency doubler and is directed
into a glove box onto the sample filled in a quartz cuvette {3}.
For our experiments, we set the excitation wavelength to
245 nm. The corresponding pulse width shows an average of 2 ns
producing a pulse energy of 3–4 mJ at a repetition rate of 20 Hz. The
laser beam diameter is approximately 3.5 mm with a divergence of
less than 1 mrad. The pulse-to-pulse stability of the outgoing laser
beam at that wavelength is more than 8%. The room temperature
was kept constant at 21 1 ◦C to ensure stable working conditions
during the operation of the laser system.
The sample holder with the quartz cuvette {3} was placed in
the glove box under inert atmosphere (N2) with an oxygen con-
centration of less than 10 ppm. The temperature in the glove box
was 21.5 1 ◦C. The laser beam was directed into a quartz cuvette
with 1 cm pathlength. A delay generator (Stanford Research Sys-
tems, Inc., Sunnyvale, CA) {8} corrects the time difference between
the electrical trigger for the generation of the laser pulse and the
optical pathway of the laser pulse through the laser system. This
difference was 545 ns.
2. Experimental
2.1. Instrumentation
The novel laser system consists of a Nd:YAG laser combined
with a MOPO-SL (Spectra-Physics Laser, Inc., Mountain View, CA)
as excitation source, a spectrograph (Acton Research SpectraPro
Fig. 1. Setup of the laser system with fluorescence detection.

904
S. Lehmann et al. / Spectrochimica Acta Part A 73 (2009) 902–908
Fluorescence radiation emitted from the sample in the cuvette
was focused on an entrance slit of an optical fibre connected to a
spectrograph {4} set at 500 nm with a grating of 300 grooves/mm.
Detection of the fluorescence was performed by an intensified CCD
camera {5} with an optical gate width of down to 80 ps. For the
recording of the fluorescence spectra, the camera was internally
cooled down to −11 ◦C using the Peltier effect. A delay generator
(LaVision, Inc., Göttingen, Germany) {7} enables time delays of up to
20 ns with a minimum stepsize of 25 ps after application of the laser
pulse. An amplifier (Kentech Instruments Ltd., Wallingford, UK) {6}
is connected to the camera. All parameters of the spectrograph and
the camera as well as the recording and storage of the data were
computer-controlled using the software DaVis 6.1.1 (LaVision, Inc.,
Göttingen, Germany) {9}. All spectra were analysed using Origin 6.1
and 7.5 (OriginLab Corporation, Northampton, MA, USA). Spectra
were recorded between 265 and 465 nm.
Fig. 2. Deconvolution of 1 × 10−2 M U(IV) in 0.1 M HClO4.
All spectra were then analysed in terms of the ability to decon-
volute specific peaks or not. It was shown that a uranium(IV)
concentration of 5 × 10−7 M was detectable. Overlaying all the spec-
tra from the highest to the lowest uranium(IV) concentration, as
shown in (Fig. 3), resulted in a clear decrease in the fluorescence
intensity. However, not all uranium(IV)-specific peaks could be
deconvoluted to the minimal detectable uranium(IV) concentra-
tion due to different fluorescence intensities of the single peaks.
The minor peaks at 290, 318, 339, 344 and 350 nm are observable
down to 1 × 10−4 M, depending on the peak. The more prominent
peaks at 320, 336, 394 and 410 nm can be deconvoluted down
to a uranium(IV) concentration of 5 × 10−7 M. As a result, laser-
induced fluorescence spectroscopy provides a powerful tool for
examining aqueous uranium(IV) at very low, environmentally rel-
2.2. Reagents
Unless stated otherwise, all chemicals were of analytical reagent
grade. Uranium nitrate hexahydrate UO2(NO3)2·6H2O was used as
starting material for the preparation of uranium(IV) stock solu-
tion in perchloric acid medium (wHClO = 70 − 72%, Merck). It was
4
obtained from Chemapol/LaChema whereas sodium fluoride NaF
for the fluoride stock solution was obtained from Nünchritz. Deion-
ized water was used throughout the experiments. Preparation and
measurements of solutions and samples were done in a glove box
under nitrogen atmosphere with an oxygen concentration of less
than 10 ppm.
Uranium(IV) was prepared by electrolytic reduction of a 0.01 M
uranium(VI) in 0.1 M perchloric acid stock solution in a two-
compartment cell using platinum electrodes as anode and cathode.
The intensity of the current and the voltage is adjusted as necessary
until uranium(VI) was fully reduced to uranium(IV). The amount
of uranium(VI) in all samples was determined to be less than 5%,
which was controlled using a TRLFS system as described by Moll
et al. [15]. For the determination of the detection limit and the
fluorescence decay time of uranium(IV), dilution series of decreas-
ing U4+ concentrations in 0.1 M and 1 M perchloric acid medium
were prepared and samples were analysed. All recorded data were
evaluated using Origin. In order to study the complexation of ura-
nium(IV) with fluoride, a uranium stock solution was electrolyzed
as described before. Again, several sample series with constant
uranium(IV) concentrations were prepared. The concentration of
fluoride added to the samples varied between tenfold stoichiomet-
ric deficiency and eightfold stoichiometric excess with respect to
the uranium(IV) concentration. All samples were prepared in 0.1 M
perchloric acid medium.
3. Results and discussion
3.1. Determination of the detection limit of uranium(IV)
To determine the detection limit of uranium(IV), its concentra-
tion in the samples was decreased stepwise from 1 × 10−3 M to
1 × 10−8 M. The recorded spectra were deconvoluted (Fig. 2) using
Gaussian functions for peaks with fitting parameters such as peak
wavelength, peak height and full-width at half maximum. The num-
ber of peaks relevant for the identification of uranium(IV) and their
approximate positions was preset. All other parameters were not
restricted. With a correlation coefficient of R2 = 0.982, the follow-
ing peaks were identified: 290, 318, 320, 336, 339, 344, 350, 394 and
410 nm. These data are in accordance with previous studies [10,16]
Fig. 3. Fluorescence spectra of uranium(IV) at different concentrations in perchloric
acid.
and therefore, prove the presence of U4+
.

S. Lehmann et al. / Spectrochimica Acta Part A 73 (2009) 902–908
905
Table 1
Overview of the fluorescence lifetimes of U(IV) at different concentrations and pH
values.
c(U4+) [M]
pH
Fluorescence lifetime [ns]
pH
Fluorescence lifetime [ns]
1 × 10−3
1 × 10−4
1 × 10−5
1 × 10−6
5 × 10−7
1
1
1
1
–
2.74 0.13
2.88 0.43
2.49 0.31
2.60 0.11
–
–
0
–
0
0
–
3.05 0.49
–
2.55 0.22
2.61 0.27
evant concentrations. The amount of uranium(IV) measurable by
our fluorescence spectroscopy system is almost two orders of mag-
nitude below the detection limit of the LIPAS [6], and one and a
half orders of magnitude lower compared to UV–vis spectroscopy.
Furthermore, our data have confirmed the minimum detectable
uranium(IV) concentration predicted by Kirishima et al. [10] and
to even lower it by half an order of magnitude down to 5 × 10−7 M.
Fig. 5. Dependency of log(Ilum) on log([U(IV)])to determineconcentrational quench-
ing effects.
3.2. Determination of the uranium(IV) fluorescence lifetime
concentrations around 1 × 10−3 M. Therefore, the dependency of
the fluorescence intensity on the uranium(IV) concentration was
investigated. The results are shown in Fig. 5. The data of fluores-
cence measurements clearly indicate a linear correlation between
log Ilum and log([U(IV)]). Additionally, the fluorescence liftetimes
determined at varying uranium(IV) concentrations are all in the
acceptable error range. The sample containing 1 × 10−3 M ura-
nium(IV) does not display a significantly shortened fluorescence
decay time. As a result, no quenching due to concentrational effects
was observable.
A dilution series with uranium(IV) concentrations between
1 × 10−3 and 5 × 10−7 M was prepared from the uranium(IV) stock
solution. Time-resolved fluorescence spectra of these samples were
recorded in the decay time range of 0–20 ns. The increment was
varied between 25 and 1000 ps. All spectra were analysed using
Origin software. We were able to detect a monoexponential decay
of the fluorescence, which means that there was one uranium(IV)
species present in the samples under the applied conditions. Table 1
gives an overview of the averaged fluorescence lifetimes of different
uranium(IV) solutions. The fluorescence lifetime of aqueous ura-
nium(IV) in perchloric acid medium in a pH range between 0 and 1
was determined to be t = 2.74 0.36 ns (Fig. 4). As the fluorescence
lifetime and the average laser pulse width differ in less than 1 ns, the
decay of the fluorescence was tracked over several lifetimes. This
procedure allows to measure the uranium(IV) fluorescence lifetime
that is just a little longer than the applied laser pulse.
3.4. Complexation and fluorescence properties of uranium(IV)
fluoride
A number of studies have investigated uranium(IV) fluoride
solutions and uranium(IV) systems, providing useful insights into
the behaviour of uranium(IV) in (perchloric acid) solution, as well
as in the chemical species of solved uranium(IV). Some studies have
also focused on uranium(IV) fluoride complexation reactions with
varying U(IV): F ratios [9,17–21]. The amount of data obtained by
these studies together with the easy handling makes uranium(IV)
fluoride an ideal system for describing the complexation between
the reactants.
Uranium(IV) fluoride samples were prepared and measured.
Table 2 provides an example of a sample series. Fluorescence spec-
tra of the complexation between uranium(IV) and fluoride are given
in Fig. 6. Starting sample was pure U(IV) in 0.1 M HClO4, which
served as a reference point. Fluoride was added stepwise while the
uranium(IV) concentration was kept constant. Measurements com-
menced with a tenfold stoichiometric deficiency. The first spectrum
did not show a change in comparison with the pure U(IV) spectrum.
Neither a decrease in the fluorescence intensity nor a shift of peak
nor a general change in the spectrum shape was observed.
3.3. Investigation of concentrational quenching effects
Perfilev et al. [11] observed luminescence quenching due to
concentrational effects only at uranium concentrations greater
than 1 × 10−3 M, and described log Ilum to be linearly depen-
dent on log([U(IV)]) for [U(IV)] between 4 × 10−6 and 1 × 10−3 M.
However, this phenomenon cannot be excluded from the begin-
ning when investigating fluorescence properties of uranium(IV) in
A slow increase of the fluoride concentration up one-forth of
the uranium(IV) concentration resulted in a stepwise decrease of
the fluorescence intensity compared with the original U(IV) spec-
trum. This was shown clearly by the two prominent sharp peaks
at 320 and 410 nm. The minor peaks at 318, 339 and 344 nm
could not be deconvoluted due to the low fluorescence intensity.
At a fluoride concentration half the uranium(IV) concentration,
almost no fluorescence was measurable. Further addition of flu-
oride resulted in a slow change of the spectrum. A new broad peak
at 320 nm replaced the original U(IV) peak. In addition, a broad
peak appeared in the wavelength range between about 350 and
450 nm. This peak only showed one broad band with no detailed
structure. The newly formed spectrum had fully developed when
Fig. 4. Fluorescence lifetime of uranium(IV) in perchloric acid.

906
S. Lehmann et al. / Spectrochimica Acta Part A 73 (2009) 902–908
Table 2
Overview of a sample series for TRLFS measurements of the complexation of ura-
nium(IV) with fluoride at pH ∼ 1.
Sample no.
c(U4+) [M]
c(F−) [M]
0
1
2
3
4
5
6
7
8
1 × 10−3
1 × 10−3
1 × 10−3
1 × 10−3
1 × 10−3
1 × 10−3
1 × 10−3
1 × 10−3
1 × 10−3
1 × 10−3
1 × 10−3
1 × 10−3
1 × 10−3
1 × 10−3
1 × 10−3
1 × 10−3
1 × 10−3
1 × 10−3
1 × 10−3
0
0
1 × 10−4
2 × 10−4
3 × 10−4
3.4 × 10−4
4 × 10−4
4.4 × 10−4
5 × 10−4
6 × 10−4
6.4 × 10−4
6.8 × 10−4
7.2 × 10−4
7.6 × 10−4
8 × 10−4
8.8 × 10−4
1 × 10−3
2 × 10−3
5 × 10−3
8 × 10−3
1 × 10−3
9
10
11
12
13
14
15
16
17
18
19
Fig. 7. Slope analysis of the complex formation of uranium(IV) with fluoride to
determine the complex stoichiometry.
of the peaks could have been an indicator for the parallel existence
of two or more species that superimpose each other. The results
of the fluorescence lifetimes contradict this assumption. Hence, an
explanation for the broadening of the peak and the band cannot be
given so far.
Then, all spectra were deconvoluted. To determine the ele-
mental composition of the uranium(IV) fluoride complex(es), the
log([complex]/[U(IV)]free) against log([F−]free) of the samples 2–14
was plotted. The graph is shown in Fig. 7. The slope represents the
U(IV): F ratio. The complex formation constant can be derived from
the intersection of the linear fit with the y-axis. Further calculations
allow the determination of the complex formation constant at infi-
nite dilution. The analysis used in the following paragraph has been
described in previous publications [22,23].
a stoichiometrical uranium(IV): fluoride ratio was reached. A fur-
ther increase of the fluoride concentration resulted in an increase
in the fluorescence intensity until the fluoride concentration was
twice the uranium(IV) concentration. Further enhancement of the
fluoride–uranium(IV) ratio did not lead to a significant change in the
fluorescence intensity. Small changes in intensity may have been
the result of instrumental measurement error.
Both the broad peak at 320 nm as well as the band between 350
and 450 nm were evaluated separately with regard to their respec-
tive fluorescence decay time. They do not differ significantly from
the fluorescence lifetime of U4+ as well as from each other. For the
peak at 320 nm, the decay time average was 2.60 0.76 ns. The flu-
orescence lifetime of the broad band between 350 and 450 nm was
2.80 ns 1.08 ns. The overall fluorescence decay time of the whole
spectrum was determined to be 2.83 0.19 ns. These findings indi-
cate that there was one species present in the solution. Broadening
The plot displays a slope of 1.05 0.20. Thus, the U(IV): F ratio
is 1:1, i.e. the following reaction (1) has taken place:
U
4+ + F− → UF3+
(1)
The corresponding complex formation constant obtained from the
intersection was shown to be log K(UF3+) = 10.33 1.94 at a temper-
ature of 21.5 1 ◦C and at an ionic strength of I = 0.1 M. To extrapolate
the complex formation constant to infinite dilution, the Davis equa-
tion (2) was applied [23]:
ˇ(I) = ˇ0 · ꢀꢁꢂ
(2)
i
i
After transforming and logarithmizing equation (2), the complex
formation constant at infinite dilution, ˇ0, can be calculated as
follows:
log ˇ0 = log ˇ(I) − (log ꢂ1:1 − log ꢂ
− log ꢂF− )
(3)
4+
U
The activity coefficients ꢂ of UF3+, U4+ and F− were calculated using
Eq. (4):
√
I − B · I
log ꢂi = −z2 · A
(4)
√
i
1 +
I
z: ion charge, A: constant 0.5091, B: constant 0.2, and I: ionic
strength.
The calculated activity coefficients were: log ꢂ1:1 = −1.009,
log ꢂU = −1.794 and log ꢂF− = −0.112. From the complex forma-
4+
tion constant determined and the activity coefficients calculated,
the complex formation constant at infinite dilution was found to be
log ˇ0 = 9.43 1.94. This value is in good agreement with previous
studies reporting a value for UF3+ of log ˇ0 = 9.42 0.51 [21]. The
large error shows the complexity of working with an element emit-
Fig. 6. Fluorescence spectra of the uranium(IV) fluoride complex formation.

S. Lehmann et al. / Spectrochimica Acta Part A 73 (2009) 902–908
907
its maximum at the final excitation wavelength of 255 nm. In addi-
tion, the peak moved slowly from 272 to 280 nm. The reasons for
this phenomenon are currently unclear, however it is unlikely to
be caused by the presence of uranium(V). Although the excitation
wavelength of 255 nm indicates uranium(V), the deconvoluted flu-
orescence spectrum does not display a peak as given by Steudtner
et al. [24] who described one peak for uranium(V) with a maximum
at 440 nm in the range of 250–600 nm.
4. Conclusions
Research investigating the chemical properties of hazardous
metals such as the actinides is essential in order to predict their
physical behaviour in the environment and their interaction with
the geo- and biosphere. Time-resolved laser fluorescence spec-
troscopy is a powerful and highly sensitive analytical tool with
which metals such as uranium can be studied in environmentally
relevant concentration ranges. We set up a novel laser fluores-
cence spectroscopy system that provides very short laser pulses of
about 2 ns. This allows us to directly study complex formations of
aqueous metals such as uranium(IV) or americium(III) whose fluo-
rescence has a lifetime of only few nanoseconds. The fluorescence
was spatially dispersed by a grating spectrograph and measured in
a time-resolved manner by an intensified CCD camera.
Uranium(IV) was chosen as a representative example of an
actinide with a very short fluorescence lifetime. The detection
limit of uranium(IV) was 5 × 10−7 M. Furthermore, the fluorescence
decay time of uranium(IV) under certain conditions was measured.
Its fluorescence lifetime of 2.74 0.36 ns has not been reported
before.
Fig. 8. Dependency of the emitted fluorescence of uranium(IV) fluoride on the
applied excitation wavelength.
ting fluorescence that is as short-lived as the one of uranium(IV).
Furthermore, stability of the laser performance over time adds
to the challenge of investigating short-lived fluorescence. Despite
these difficulties, we were able to show the accurate performance
of our novel laser system.
3.5. Investigation of the dependency complex
formation–excitation wavelength
There are different explanations for the changes observed in the
fluorescence intensity and in the shape of the spectrum. They can
be due to changes in the concentration of free uranium(IV) and the
corresponding formation of uranium(IV) fluoride complexes. How-
ever, an alternative explanation would be that a different amount
of energy was required to excite the uranium(IV) fluoride complex,
which would lead to a shift in the excitation wavelength. To clarify
this issue, we further assessed the dependency of the complex for-
mation and excitation wavelength. As described by Kirishima et al.
[10], U4+ has one of the greatest absorptions at a wavenumber of
40820 cm−1, which corresponds to a wavelength of 245 nm. There-
fore, excitation of free uranium(IV) at a wavelength of 245 nm leads
to the emission of fluorescence. However, it may be possible that
there is a shift in energy for the excitation of bound uranium(IV)
or the whole uranium(IV) complex, respectively. Thus, one sample
consisting of 1 × 10−3 M U(IV) and 8 × 10−3 M F− was chosen. The
excitation wavelength applied to the sample was varied between
235 nm and 255 nm in a stepwise manner with 1 nm steps. Five
selected spectra are shown in Fig. 8. An overview of all spectra can
be found online as supplementary material.
The uranium(IV) fluoride system was shown to be ideal for
studying complexation between a metal ion and a reactant by
measuring the fluorescence properties. Our data revealed the for-
mation of a 1:1 uranium(IV) fluoride complex. The corresponding
complex formation constant of UF3+ measured was shown to be
log ˇ0 = 9.43 1.94, which is in accordance with previous studies
[21]. Thus, our results have demonstrated that our novel laser flu-
orescence spectroscopy system can successfully applied to study
the speciation and determine complex formation constants of com-
pounds, which emit very short-lived fluorescence of only few
nanoseconds.
5 Computer programs
TRLFS spectra were recorded using DaVis 6.1.1 (LaVision, Inc.,
Göttingen, Germany) software. All spectra were analysed using Ori-
gin 6.1 and 7.5 (OriginLab Corporation, Northampton, MA, USA).
Comparing the spectra recorded at different excitation wave-
lengths with those obtained from the uranium(IV) fluoride
complexation showed that the change in the intensity and in the
shape of the complexation spectra were caused by a decrease in
the concentration of free uranium(IV) and the correlating forma-
tion of UF3+. The changes were not, however, caused by a shift in
the excitation wavelength for newly formed complex. The emitted
fluorescence of the sample gets to a maximum at the excitation
wavelength of 244 nm, which correlates with the known excitation
wavelength of 245 nm for uranium(IV).
6 Supplementary material
Fluorescence spectra of a uranium(IV) fluoride sample, to
which excitation wavelengths between 235 and 255 nm in a step-
wise manner in 1 nm steps were applied, are available online at
http://www.sciencedirect.com
Acknowledgements
Moreover, the appearance of a new peak at 272 nm starting at
an excitation wavelength of 246 nm can be observed. This peak
reached a maximum at an excitation wavelength of 248 nm and
decreased to a minimum at the next excitation wavelength step.
Then, its luminescence intensity increased again until it reached
The authors would like to thank Mrs. U. Schaefer for doing ICP-
MS/AAS measurements, Mrs. C. Eckardt for ion chromatography
measurements, and the German Research Council (DFG) for finan-
cial support (contract no. GE 1011/4-1).

908
S. Lehmann et al. / Spectrochimica Acta Part A 73 (2009) 902–908
[12] S.V. Godbole, A.G. Page, Sangeeta, S.C. Sabharwal, J.Y. Gesland, M.D. Sastry, J.
Appendix A. Supplementary data
Lumines. 93 (2001) 213–221.
[13] A. Kirishima, T. Kimura, R. Nagaishi, O. Tochiyama, Radiochim. Acta 92 (2004)
705–710.
[14] T.A. O’Donnell, P.W. Wilson, Anal. Chem. 39 (1967) 246–248.
[15] H. Moll, G. Geipel, V. Brendler, G. Bernhard, H. Nitsche, J. Alloys Compd. 271–273
(1998) 765–768.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.saa.2009.04.013.
References
[16] G. Geipel, Annual Report FZR-400, 2003, p. 1.
[17] S. Bürger, Spurenanalyse von Uran und Plutonium sowie Speziationsunter-
suchungen an Plutonium mit massenspektrometrischen und kapillarelek-
trophoretischen Methoden, Ph.D. Thesis, Mainz, 2005.
[18] S.A. Fell, Analyst 114 (1989) 325–331.
[19] H. Moll, M.A. Denecke, F. Jalilehvand, M. Sandström, I. Grenthe, Inorg. Chem. 38
(1999) 1795–1799.
[1] G. Bernhard, G. Geipel, V. Brendler, H. Nitsche, Radiochim. Acta 74 (1996) 87–91.
[2] G. Bernhard, G. Geipel, V. Brendler, H. Nitsche, J. Alloys Compd. 271–273 (1998)
201–205.
[3] K. Opel, S. Weiß, S. Hübener, H. Zänker, G. Bernhard, Radiochim. Acta 95 (2007)
143–149.
[4] D. Rai, M. YYui, D.A. Moore, J. Solut. Chem. 32 (2003) 1–17.
[5] G. Geipel, M. Acker, D. Vulpius, G. Bernhard, H. Nitsche, Th. Fanghänel, Spec-
trochim. Acta A 60 (2004) 417–424.
[6] C. Mesmin, D. Roudil, A. Hanssens, C. Madic, Radiochim. Acta 91 (2003)
385–392.
[7] C. Moulin, P. Decambox, P. Mauchien, Anal. Chem. 68 (1996) 3204–3209.
[8] V. Neck, J.I. Kim, Radiochim. Acta 89 (2001) 1–16.
[9] G. Geipel, G. Bernhard, Annual Report FZR-285, 1999, p. 5.
[10] A. Kirishima, T. Kimura, O. Tochiyama, Z. Yoshida, Chem. Commun. 7 (2003)
910–911.
[20] R.A. Day Jr., R.N. Wilhite, F.D. Hamilton, J. Am. Chem. Soc. 77 (12) (1955)
3180–3182.
[21] R. Guillaumont, T. Fanghänel, V. Neck, J. Fuger, D.A. Palmer, I. Grenthe, M.H.
Rand, OECD-NEA Chemical Thermodynamics Series, Volume 5: Update on the
Chemical Thermodynamics of Uranium, Neptunium, Plutonium, Americium
and Technetium, Elsevier, Amsterdam, 2003.
[22] G. Geipel, G. Bernhard, V. Brendler, in: B.J. Merkel, B. Planer-Friedrich, C. Wolk-
ersdorfer (Eds.), Uranium in the Aquatic Environment, Springer, Berlin, 2002,
pp. 369–376.
[23] C.W. Davies, Ion Association, Butterworths, London, 1962.
[24] R. Steudtner, T. Arnold, K. Großmann, G. Geipel, V. Brendler, Inorg. Chem. Com-
mun. 9 (2006) 939–941.
[11] V.A. Perfilev, D.V. Demeshko, V.T. Mishchenko, Radiokhimiya 33 (1991)
133–136.