Raman spectral titration method: an informative technique for studying the complexation of uranyl with uranyl(VI)–DPA/oxalate systems as examples

Qian Liu a, Qianci Zhang a, Suliang Yang *a, Haiqiao Zhu a, Quanwei Liu b and Guoxin Tian *ac
aDepartment of Radiochemistry, China Institute of Atomic Energy, Beijing, 102413, China. E-mail: ysl79@hotmail.com; gtian@ciae.ac.cn
bChina Nuclear Power Engineering Co., Ltd, China
cNuclear Chemical Engineering Department, Harbin Engineering University, Harbin, Heilongjiang 150001, China

Received 4th May 2017 , Accepted 13th June 2017

First published on 13th June 2017

The Raman band at about 870 cm−1 originating from the symmetric stretch vibration (ν1) of uranyl, UO22+, has proven to be very informative for investigating the complexation of uranyl using perchlorate or nitrate of known concentration as internal standards. The concentration of uranyl can be conveniently calculated by using the ratio of the directly read band intensities of uranyl and the added reference, ClO4, with a factor of 1.72. While with NO3 of concentration lower than 1.8 M as the reference, a factor of 0.85 should be used. Furthermore, with added internal standards, the linear relationship between the Raman intensity and the concentration of the corresponding species is illustrated by the spectral titration of U(VI) with a very strong ligand, dipicolinic acid (DPA); and the application of a spectral titration method with Raman spectroscopy in studying the complexation of uranyl is demonstrated by the titration of U(VI) with oxalate. The stepwise changes in the Raman shift of 18, 17, and 6 cm−1, corresponding to the three oxalate anions successively bonding to UO22+, imply that the coordination modes are different. In the 1[thin space (1/6-em)]:[thin space (1/6-em)]1 and 1[thin space (1/6-em)]:[thin space (1/6-em)]2 ratios of metal to ligand complexes, the oxalate anions bond to the uranyl ion in side-on bidentate mode, but in the 1[thin space (1/6-em)]:[thin space (1/6-em)]3 complex the third oxalate bonds in head-on mode, which is much weaker than the first two.


To meet the rapidly growing demand for electricity, nuclear energy has been considered as one of the most important non-greenhouse-gas-emitting contributors to electricity generation to help cut pollutant and carbon dioxide emission that leads to air pollution in the short term and climate change in the long term. The primary fissile material used in current power-generating reactors is enriched uranium, and after the nuclear reaction in the reactors the spent nuclear fuel consists of about 95% of uranium, 1% of plutonium, and 3% of fission products. Uranium and plutonium in the spent fuel can be recovered by reprocessing to make new fuel, which is called the nuclear fuel cycle. The nuclear fuel cycle based on uranium as the major fissile material involves wet or moist conditions. The uranyl cation (UO22+) is known to be the most stable form of uranium under aerial wet conditions. Various methods have been developed for the analysis of uranium in solutions associated with the nuclear fuel cycle, such as spectrophotometry,1,2 fluorescence spectrometry,3,4 alpha spectrometry,5 electrochemistry,6 X-ray fluorescence spectroscopy,7 chromatography,8 inductively coupled plasma mass spectrometry,9 potentiometry and so on.10 Some of the methods require time-consuming pretreatment, and/or expensive and complicated instruments. On-line quick analytical methods are also urgently needed in the nuclear fuel cycle, especially for process-control in reprocessing, due to the strong radioactivity of some streams. The methods used to monitor these processes must be robust and capable of withstanding strong radiation and harsh chemical environments.

The distinct advantages of Raman spectroscopy to examine aqueous solutions or solid samples nondestructively and without tedious preparation have led to a wide application for analyzing some important components in the nuclear fuel cycle. Raman spectroscopy has been used to study the molecular structure of various uranyl minerals11 and to measure the concentration of metal-oxides, organics, inorganic metal oxo-anions and water in nuclear fuel reprocessing combined with a Coriolis meter.12 Quantitative analyses for uranyl speciation and relative abundance based on Raman spectroscopy have been carried out by measuring the absolute band intensity and analyzing the spectra.13–15 However, the absolute intensity of the Raman band is dependent on a number of factors, such as laser power, sample orientation, and other instrumental effects, which are easily disturbed and cannot be very well repeated. In addition, the spectral analysis focusing on assignments could identify the presence of uranyl complex species, but could not well calculate their real concentrations and then the related thermodynamic parameters, because of the absence of a standard band for the normalization of the band intensities. To overcome the drawbacks mentioned above, more repeatable analysis methods based on Raman spectroscopy by taking advantage of internal standards are essentially needed.

The linear uranyl cation unit consists of a uranium center and two oxygens oppositely coordinated to the uranium center through two very strong covalent bonds. On exposure to a laser beam, it gives a strong and sharp vibrational band in the range of 700–900 cm−1 due to the symmetric stretch (ν1) of O[double bond, length as m-dash]U[double bond, length as m-dash]O. The characteristics of the symmetric stretch vibration (ν1) of O[double bond, length as m-dash]U[double bond, length as m-dash]O make it possible to determine the concentration of uranyl by Raman spectroscopy. The uranyl ion in HNO3 or HClO4 solution shows the characteristic of the symmetric stretch (ν1) vibration at 870 cm−1. The insensitivity of the ν1 vibration of UO22+ towards the coexisting NO3 and ClO4 in the solutions may indicate that NO3 and ClO4 very weakly coordinate to or do not bond to the uranyl ion, so that they do not have detectable influence on the uranyl stretch vibration. In theory, due to the characteristics of the symmetric stretch vibration of nitrate at 1048 cm−1 and perchlorate at 934 cm−1 they could be conveniently used as internal standards to normalize the Raman band of the uranyl ion in HNO3 or HClO4 solutions.16 In addition, the symmetric stretch vibration (ν1) of uranyl is also apparently affected by the association between uranyl and the coordination ligands in the equatorial plane, and the exact change in ν1 depends on the ligands attached.14 If the intensities of the ν1 band of all uranyl complex species could be quantitatively correlated with their corresponding concentrations in a series of samples, spectral titration with Raman spectroscopy might be a very powerful method for investigating the complexation and speciation of uranyl in solutions containing a variety of ligands. Notably, a very detailed study on evaluating the application of Raman spectral analysis to uranium speciation and relative abundance was recently reported, but the assumed similar Raman cross-sections of each uranyl species are debatable.13

In the present work, we initiated a study on applying Raman spectrometry to the investigation of the uranyl ion in solutions containing very weak/non-coordinating ligands (NO3, ClO4), a moderate ligand (C2O42−), and a very strong ligand, dipicolinic acid (DPA).16 At first, the quantitative relationships were verified between the concentrations ([UO22+] and [NO3]) and the intensities of their Raman bands normalized to the band intensity of the symmetric stretch vibration of perchlorate at 934 cm−1. Then a spectral titration method was tried out with Raman measurement in the titrations of uranyl with a very strong ligand dipicolinic acid (DPA), illustrating that the method well-developed from other spectroscopic measurements is applicable in studying the reactions involving chemical species with Raman active vibrations. Finally, the extensively studied U(VI)-oxalate complexation was used as a typical example to demonstrate the great potentiality of Raman spectroscopy in exploring the coordination chemistry of uranyl. The results suggest that the linear relationships between the concentrations of the species and the corresponding Raman band intensities can be utilized to determine their concentrations, and further to study the complexation of uranyl with various ligands.

Results and discussion

Determination of UO22+ by using ClO4/NO3 as the internal standard

Fig. 1 shows the Raman spectrum of a uranyl nitrate solution in perchloric acid. Three distinctive bands were clearly observed at 870, 934 and 1048 cm−1, ascribed to the symmetric stretching band of UO22+, ClO4, and NO3, respectively. The spectral distinction among the Raman signals of UO22+, ClO4 and NO3 makes it possible to use any one of them as an internal standard for the determination of the other two.
image file: c7dt01631j-f1.tif
Fig. 1 Raman spectrum of 0.063 M UO2(NO3)2 in 0.200 M HClO4 collected with a laser power: 20 mW, exposure time: 20 s, and accumulation times: 5.

A set of samples containing uranyl nitrate of varying concentrations (0.0035–1.00 M) and perchlorate of a constant concentration (0.200 M) were prepared, and their Raman spectra were obtained to validate the application of added ClO4 as an internal standard for analyzing uranyl in nitrate/nitric acid solutions. As shown in Fig. 2, the Raman band intensity at 870 cm−1 increased proportionally with the concentration of uranyl, and a linear relationship between the relative Raman intensity (Iuranyl/Iperchlorate) and the concentration of uranyl was obtained in the range of 0.02 to 1.0 M. The Raman band at 870 cm−1 from uranyl could still be clearly observed with the concentration reaching as low as 0.0035 M, but the ratio of the signal to noise was too low to be used to calculate the concentration accurately.

image file: c7dt01631j-f2.tif
Fig. 2 The Raman spectra of uranyl-perchlorate solutions (upper) and the dependence of the relative intensity on the concentration of uranyl. Curanyl: 0.0035–0.100 M, Cperchlorate: 0.200 M.

In order to provide a general procedure to quickly calculate the concentration of uranyl nitrate in solutions, the relative intensities were further normalized to the corresponding concentrations. The normalized relative intensities ((Iuranyl/Curanyl)/(Iperchlorate/Cperchlorate)) stand at an average value of 1.72 ± 0.4. Then, for a sample with known perchlorate concentration, the uranyl concentration could be calculated from the Raman intensities by the following equation:

Curanyl = (Iuranyl/Iperchlorate) × (Cperchlorate/1.72)(1)

To verify eqn (1), Raman spectra of the samples containing varying uranyl nitrate and perchlorate were obtained, and the predicted values of uranyl concentrations were in good agreement with their known values (Table S1). The constant value of the normalized relative intensity of uranyl over perchlorate provides a very convenient method for determining the concentration of uranyl in nitric acid.

With an assumption that the method for determining uranyl could be adopted to the measurement of nitrate, the dependence of the relative Raman intensity (Initrate/Iperchlorate) on the concentration of nitrate was studied by following a similar procedure. For a set of samples containing varying nitric acid concentration but a constant perchloric acid concentration, within the range of 0.01–1.5 M nitrate, the relative Raman intensity of nitrate (Initrate/Iperchlorate) increases linearly with the increase of nitrate concentration, while the Raman signal is heavily disrupted by the noise baseline when the concentration is lower than 0.01 M. As the nitrate concentration is higher than 1.5 M, the measured relative Raman intensity goes astray below the values depicted by the linear relationship. Interestingly, the linear relationship between the relative integrated area, Anitrate/Aperchlorate, and the concentration of nitrate is well kept across the whole range of 0.01–3.48 M in the experiment (Fig. 3). The difference in the relative intensity of nitrate between low and high concentrations might have originated from the broadened Raman bands of nitrate in the solutions of higher concentration, as shown in Fig. 4. However, the mechanism responsible for broadening the Raman band of nitrate by its higher concentration is not clear. In contrast, no such phenomenon was ever observed for perchlorate with its concentration up to 2.5 M (Fig. S1).

image file: c7dt01631j-f3.tif
Fig. 3 The Raman spectra of varying nitric acid in 0.02 M perchloric acid (upper); and the dependence of the relative intensity/area on the concentration of nitrate. Cnitrate: 0.01–3.48 M (lower).

image file: c7dt01631j-f4.tif
Fig. 4 Raman spectra of nitric acid solutions (0.05–3.5 M) normalized by intensity.

Based on the observation about the dependence of the band intensity and the integrated area on the nitrate concentration, we suggest that the nitrate concentration should be calculated more accurately by using the relative integrated area, Anitrate/Aperchlorate, instead of the relative peak intensity (Initrate/Iperchlorate). However, for the samples of low nitrate concentration (<1.5 M), by directly reading the intensities of the nitrate concentration still could be accurately calculated with known perchlorate as the internal standard.

The relative intensities and integrated areas were further normalized to the corresponding concentrations of the samples in Fig. 3, and the average values of 0.85 and 0.83 were obtained for the normalized relative intensities (Initrate/Cnitrate)/(Iperchlorate/Cperchlorate) and the normalized relative integrated area ((Initrate/Cnitrate)/(Aperchlorate/Cperchlorate)), respectively. Hence, for a sample with known perchlorate, the nitrate concentration could be calculated from the Raman measurements by following equations:

Cnirate = (Initrate/Iperchlorate) × (Cperchlorate/0.85)(2)
Cnitrate = (Anitrate/Aperchlorate) × (Cperchlorate/0.83)(3)

The applicable ranges of nitrate concentration for eqn (2) and (3) are 0.01–1.5 M and 0.01–3.48 M, respectively.

Raman spectra of UO22+ titrated with DPA aqueous solution

DPA is a very strong ligand for actinide ions, and it can form very stable complexes with uranyl even in very acidic solutions. The stability constants (log[thin space (1/6-em)]β1 and log[thin space (1/6-em)]β2) for the 1[thin space (1/6-em)]:[thin space (1/6-em)]1 and 1[thin space (1/6-em)]:[thin space (1/6-em)]2 U(VI)/DPA complexes were determined to be 10.7 ± 0.1 and 16.3 ± 0.1, respectively, by spectrophotometry. The very high stability constant of 10.7 for the 1[thin space (1/6-em)]:[thin space (1/6-em)]1 U(VI)/DPA complex and the significant difference of about 5 orders of magnitude between the two stepwise stability constants (log[thin space (1/6-em)]K1 − log[thin space (1/6-em)]K2 = 10.7 − 5.6 = 5.1) indicate that in a very wide range of neutral to very acidic solutions, uranyl almost quantitatively forms a 1[thin space (1/6-em)]:[thin space (1/6-em)]1 complex with DPA. The 1[thin space (1/6-em)]:[thin space (1/6-em)]2 U(VI)/DPA complex only exists in a undetectable concentration before every uranyl ion is coordinated by one DPA ligand.17 The unique characteristic of the very strong complexation between uranyl and DPA provides an ideal example to let us look into the change in the Raman band of uranyl with the quantitative coordination in the equatorial plane.

There are two major bands in the range of 800–1100 cm−1 in the Raman spectra of DPA solutions differently neutralized with NaOH (Fig. S2). Two bands are observed at 1001 cm−1 and 1021 cm−1 for the solution from directly dissolving DPA into water, the intensity of the band at 1021 cm−1 weakens as the acid is neutralized. It is very clear that there is no significant interference between the Raman bands of DPA and UO22+, ClO4, NO3.

Fig. 5 shows the Raman spectra of a uranyl perchlorate solution titrated with DPA, the spectra are normalized to the Raman band of perchlorate. The band intensity of uranyl at 870 cm−1 decreased and the intensities of three new bands at 856, 826, and 1021 cm−1 increase as the concentration of DPA is increased. The band at 856 cm−1 is ascribed to the 1[thin space (1/6-em)]:[thin space (1/6-em)]1 U(VI)–DPA complex, and the other two bands at 826 and 1021 cm−1 belong to the DPA ligand bonding to uranyl. The linear dependence of the relative Raman intensities at 870 and 856 cm−1 on the addition of DPA confirms the strong complexation between uranyl and DPA.

image file: c7dt01631j-f5.tif
Fig. 5 Raman spectra of a uranyl perchlorate solution titrated with DPA, Curanyl = 0.036 M, CDPA = 0–0.0275 M (a); curve fitting data of overlapped peaks, Curanyl = 0.036 M, CDPA = 0.0118 M (b); linear correlation of relative intensity at 870/856 cm−1versus the addition of DPA (c).

Based on the linear relationship between the relative Raman intensities and the concentration of the corresponding species, the strong complexation between uranyl and DPA could also be used to determine the concentration of uranyl in solutions containing unknown nitrate and perchlorate. For uranyl solutions just containing either nitrate or perchlorate, the concentrations of uranyl and the only one anion, NO3 or ClO4, could be analyzed with Raman measurement by adding a known amount of the other anion as the internal standard using eqn (1), (2), or (3). For uranyl solutions containing both unknown nitrate and perchlorate anions, even though the concentrations of uranyl and the anions still could be determined by adding varying amounts of one of the anions as internal standards, the procedure becomes much more complicated and tedious. In contrast, by taking advantage of the strong complexation between uranyl and DPA, the concentration of uranyl in solution with unknown mixing nitrate and perchlorate could be easily determined by spectra titration with Raman measurement.

The overlapped bands of UO22+ and the 1[thin space (1/6-em)]:[thin space (1/6-em)]1 UO22+–DPA complex could be accurately deconvoluted with the equipped program of our Raman spectrometer or similar commercial programs, as shown in Fig. 5(b). The perfect straight lines of the plots of the relative intensities from the deconvoluted spectra versus the addition of DPA exhibit a linear relationship between the concentration of uranyl and added DPA (Fig. 5c):

IU(VI) = A × VDPA + B(4a)
where A is the slope of the straight line for UO22+, and B is the intercept on the X-axis, which is equal to the intensity of the uranyl in the first spectrum of the titration. Then from the point at which the intensity of free UO22+ is zero, the amount of DPA required to form the 1[thin space (1/6-em)]:[thin space (1/6-em)]1 UO22+–DPA complex could be deduced to be (VDPA × CDPA) = (−B/A) × CDPA, and the concentration of uranyl in the solution can be calculated by the following equation:
CU(VI) = (VDPA × CDPA)/Vt = (−B/A) × CDPA/Vt(4b)
where Vt represents the volume of the solution for the titration. For uranyl nitrate solutions of unknown nitrate, the titration of uranyl with DPA can be conducted in the same way except for the normalization of the spectra being done with the band of nitrate as the internal standard.

The complexation of UO22+ with oxalate

The result from the Raman measurements of uranyl in solutions with added nitrate or perchlorate as internal standards and from the titrations of uranyl with DPA clearly indicates that the Raman intensity (peak height or area) is linearly related to the concentration of the corresponding Raman active species. In addition, the Raman band of uranyl shifts with the coordination of ligands in the equatorial plane. These characteristics not only are the bases for the developed methods for determining the concentration of uranyl, but also are satisfied with the principles of the well-developed spectral titration method for studying the complexation of a variety of metal ions and ligands. To expand the application of Raman spectroscopy, spectral titrations of uranyl with oxalate were performed to confirm the great potential.

Fig. 6 shows the Raman spectra of a representative titration of U(VI) with oxalate. In all the samples, the same concentration of nitrate was maintained, so all the spectra were normalized to the intensity of nitrate. The weak band at 903 cm−1 that belongs to the oxalate ion was subtracted for better observation of U(VI)/oxalate complexes. With the addition of the oxalate titrant solution, the intensity of the Raman band at 870 cm−1 that belongs to the free UO22+ cation decreased and a new band appeared at 852 cm−1, which is ascribed to the formation of a 1[thin space (1/6-em)]:[thin space (1/6-em)]1 complex of U(VI) with oxalate, UO2(oxa). With more addition of the titrant solution, the intensity at 852 cm−1 reached maximum then decreased and a new band came out at 835 cm−1, which corresponds to the formation of the 1[thin space (1/6-em)]:[thin space (1/6-em)]2 complex, UO2(oxa)22−. As the concentration of oxalate was further increased, however, there is no observation of the obvious original band disappearing and a new band appearing, only continuous shifting was seen. The pattern of these changes in the Raman spectra is similar to those in the absorption or fluorescence spectral titrations. Usually during the titrations of the f-element ions with many other ligands, the decrease of the absorbance/fluorescence of the first complex is accompanied by the appearance of new absorption/fluorescence band(s) of successive complexes. Hence, the Raman spectra from the titrations could be used to calculate the formation constants of the complexes with the HypSpec program, which is widely used with other spectroscopy techniques.19

image file: c7dt01631j-f6.tif
Fig. 6 Representative Raman spectral titration of U(VI) with oxalate. Upper: Spectra collected during a titration (normalized to the intensity of 0.146 M nitrate). Lower: Raman spectra of U(VI)/oxolate complexes of relative intensity from deconvolution. Curanyl = 0.0706 M, Cnitrate = 0.146, V = 0.5 mL, CH+ = 0.0044–0.092 M, Coxalate = 0–0.44 M.

The variation of the spectra in Fig. 6 is interpreted with the assumption that three complexes of U(VI) with oxalate form successively in solution. The calculated formation constants of the 1[thin space (1/6-em)]:[thin space (1/6-em)]1, 1[thin space (1/6-em)]:[thin space (1/6-em)]2, and 1[thin space (1/6-em)]:[thin space (1/6-em)]3 complexes (log[thin space (1/6-em)]β1, log[thin space (1/6-em)]β2, log[thin space (1/6-em)]β3) are 6.0 ± 0.2, 10.1 ± 0.2, and 13.4 ± 0.4, respectively, which are in good agreement with the values in the literature.18 Görller-Walrand tentatively proposed a uranyl oxalate dimer (2[thin space (1/6-em)]:[thin space (1/6-em)]5) complex in acetone20 and Havel demonstrated that such dimeric species slightly improves the overall fitting of the formation constants in a spectrophotometric study of the aqueous system.21 However, the simplicity of Raman bands from the titrations (Gaussian or Lorentzian distribution) makes it easy to identify that only four species were involved in titration. The formation constant obtained here for the last species is more consistent with the values of the 1[thin space (1/6-em)]:[thin space (1/6-em)]3 complex than that of the 2[thin space (1/6-em)]:[thin space (1/6-em)]5 complex.21

It should be noted that, besides being used to calculate the thermodynamic parameters, Raman spectral titrations also provide structural information of the complexes. In general, for the same ligand, successive bonding with the same mode generates similar stepwise Raman shifts for uranyl complexes; and for different ligands, stronger bonding makes larger Raman shifts.14 For the complexation of U(VI) with oxalate, in the first two complexes, the oxalate anions bond to the uranyl ion in side-on bidentate mode, but the third one maybe bond in head-on or monodentate mode, which is much weaker than the first two. The changes in the bonding mode result in different stepwise Raman shifts for the first two and the third one (Scheme 1). We should be aware that even though the results from some calculations suggest that the third oxalate bonds to uranyl through a single carboxylate oxygen,22 the fairly strong coordination (log[thin space (1/6-em)]K3 of 3.3 herein, and 3.45 in the previous work18) indicates it might also form a chelate. Therefore, further investigation by other methods will be required to identify the actual structure of the 1[thin space (1/6-em)]:[thin space (1/6-em)]3 UO22+/oxalate complex.

image file: c7dt01631j-s1.tif
Scheme 1 Successive complexation of UO22+ with oxalate and the corresponding changes in the Raman shift and the coordination modes.

Change in Raman scattering cross-section with complexation

In previous studies of uranyl complexes, controversial assumptions on the Raman scattering cross-section were made by different researchers.15,16 A moderate increase of the relative integrated intensity of the ν1(O[double bond, length as m-dash]U[double bond, length as m-dash]O) band was observed in the complexation of uranyl with sulfate, which was ascribed to the corresponding small increase of Raman scattering cross-section caused by the complexation with sulfate. And a further hypothesis was assumed for quantitative analyses. It was described that the increase of Raman scattering cross-section was proportional to the number of sulfate groups in the complexes. In contrast, in a recent publication on evaluating the practices of Raman spectral analyses for uranium speciation and relative abundances, it was assumed that each uranyl species exhibited similar Raman cross-sections.13 However, based on our observation on the complexation of uranyl with DPA/oxalate, it seems to be very difficult to predict the change in the Raman cross-section caused by complexation. It clearly decreases with the bonding of DPA, while increases with the coordination with oxalates. But the enhancement is not proportional to the number of the bonded oxalates.

In all, speciation assignments and thermodynamic parameters can be easily determined by using the Raman spectral titration method. Because the Raman cross-section arbitrarily changes with various complex species of uranyl, special care should be taken to ensure appropriate relative abundance with traditional spectral analysis without an internal standard.



All chemicals except for uranium were reagent grade or higher. Deionized water from a Milli-Q system was used in the preparation of all solutions. UO2(NO3)2·nH2O was used as the starting material for all uranyl solutions. A stock solution of uranyl nitrate was prepared by directly dissolving UO2(NO3)2·nH2O in diluted nitric acid. For preparing a UO2(ClO4)2 stock solution, Na2U2O7 was precipitated from a UO2(NO3)2 solution by adding excess of NaOH and washed three times with deionized water, then the precipitate was dissolved with concentrated perchloric acid (70%) under stirring and diluted with water. Dipicolinic acid (pyridine-2,6-dicarboxylic acid, DPA, 98%) from Avocado Research Chemicals Ltd was used as received. Buffered DPA solutions were prepared by partially neutralizing DPA with a standard NaOH solution (1.022 M).


Raman spectra were recorded with a Renishaw inVia Raman microspectrometer at a nominal resolution of 1.4 cm−1 in the range between 800 and 1100 cm−1. The spectrophotometer employs a Leica imaging microscope with a 50× objective at a distance of 300 μm above the capillary tube containing the liquid sample. A charge-coupled device (CCD) array detector was used to achieve signal detection from an 1800 grooves per mm grating light path controlled by Renishaw WiRE software. A diode laser (532 nm line) was used as the excitation source. The excitation light with a maximum laser power of 20 mW was focused on the samples by using the microscope, and all the Raman spectra were recorded at 180° to the direction of the excitation beam. All the motors were controlled by Renishaw WiRE software. Glass capillaries were used to contain the liquid samples, and the two ends were sealed with Para-film. The exposure time was 20 s and the number of scans was 5 in all measurements.

DPA/oxalate titrations

A spectral titration method was tentatively used to determine the concentration of complex species of U(VI) by correlating the change in the Raman band of uranyl and the addition of DPA in solutions with nitrate or perchlorate as internal standards. Spectral titrations of U(VI) with oxalate were also conducted to demonstrate the applicability of Raman spectroscopy to quantitatively study the complexation of the uranyl ion. For a typical titration, 50 μL of the UO2(NO3)2 or UO2(ClO4)2 solutions were placed in each set of centrifuge tubes, then different aliquots of buffered DPA/oxalate solutions were added into each tube and diluted to an appropriate volume. The samples were mixed thoroughly (for 1–2 min) before the spectra were collected. Multiple titrations with different volumes of DPA/oxalate or U(VI) were performed. All the Raman spectra were normalized by the intensities of nitrate or perchlorate of exactly known concentration. The concentration of uranyl was calculated by interpreting the variation of the symmetric stretch vibration band (ν1) of uranyl at 870 cm−1 with the addition of DPA forming the U(VI)/DPA complex. The formation constants and the Raman spectra of the U(VI)/oxalate complexes were calculated, from the normalized Raman spectral data in the range 750–950 cm−1, with the HypSpec 2013 program.19


By monitoring the Raman band of uranyl with nitrate or perchlorate as the internal standard, a spectral titration method is used to determine the concentration of uranyl in nitric acid and perchloric acid solutions and to investigate the complexation of uranyl. For the first time, this work extends a spectral titration method to Raman spectroscopy. The results from the titrations of uranyl with DPA/oxalate demonstrate that the Raman spectral titration method not only provides thermodynamic information but also reveals structural details of the complexes.

The complexation of various ligands impacts the Raman scattering cross-section of uranyl; more attention should be paid to the speciation assignments and to the determination of relative abundance with traditional spectral analysis. These findings suggest that the Raman spectral titration method may be very powerful in the study of U(VI) complexation with various ligands in aqueous solutions with nitrate or perchlorate as a reference.


This work was supported by the National Natural Science Foundation of China (91426302).

Notes and references

  1. (a) J. Korkisch and G. E. Janauer, Anal. Chim. Acta, 1961, 25, 463–469 CrossRef CAS ; (b) M. N. A. Lutfullah, N. Rahman and S. N. H. Azmi, J. Hazard. Mater., 2008, 155, 261–268 CrossRef CAS PubMed .
  2. (a) L. Rao and G. Tian, J. Chem. Thermodyn., 2008, 40, 1001–1006 CrossRef CAS ; (b) G. Tian, L. Rao, S. J. Teat and G. Liu, Eur. J. Chem., 2009, 15(16), 4172–4181 CrossRef CAS PubMed .
  3. (a) A. A. Elabd and O. A. Elhefnawy, J. Fluoresc., 2016, 26, 271–276 CrossRef CAS PubMed ; (b) T. Saito, N. Aoyagi and T. Kimura, J. Radioanal. Nucl. Chem., 2015, 303, 1129–1132 CrossRef CAS .
  4. (a) C. Moulin, P. Decambox, P. Mauchien, D. Pouyat and L. Couston, Anal. Chem., 1996, 68, 3204–3209 CrossRef CAS ; (b) Z. Wang, J. Zachara, W. Yantasee, P. Gassman, C. Liu and A. Joly, Environ. Sci. Technol., 2004, 38, 5591–5597 CrossRef CAS PubMed ; (c) S. Maji, S. Kumar and K. Sankaran, J. Radioanal. Nucl. Chem., 2014, 302, 1277–1281 CrossRef CAS ; (d) M. Wu, L. Liao, M. Zhao, Y. Lin, X. Xiao and C. Nie, Anal. Chim. Acta, 2012, 729, 80–84 CrossRef CAS PubMed .
  5. (a) C. W. Sill, Anal. Chem., 1977, 49(4), 618–621 CrossRef CAS PubMed ; (b) G. Jia, M. Belli, U. Sansone, S. Rosamilia, R. Ocone and S. Gaudino, J. Radioanal. Nucl. Chem., 2002, 253, 395–406 CrossRef CAS .
  6. (a) A. Becker, H. Tobias and D. Mandler, Anal. Chem., 2009, 81, 8627–8631 CrossRef CAS PubMed ; (b) M. Shamsipur, F. Mizani, K. Alizadeh, M. F. Mousavi, V. Lippolis, A. Garau and C. Caltagirone, Sens. Actuators, B, 2008, 130, 300–309 CrossRef CAS .
  7. D. P. S. Rathore, Talanta, 2008, 77, 9–20 CrossRef CAS PubMed .
  8. M. L. Dietz, E. P. Horwitz, L. R. Sajdak and R. Chiarizia, Talanta, 2001, 54, 1173–1184 CrossRef CAS PubMed .
  9. A. Sabarudin, M. Oshima, T. Takayanagi, L. Hakim, K. Oshita, Y. Gao and S. Motomizu, Anal. Chim. Acta, 2007, 581, 214–220 CrossRef CAS PubMed .
  10. A. S. A. Ammar and H. M. Basheer, J. Radioanal. Nucl. Chem., 1993, 171, 435–441 CrossRef .
  11. (a) R. L. Frost, J. Cejka and M. L. Weier, J. Raman Spectrosc., 2007, 38(4), 460–466 CrossRef CAS ; (b) R. L. Frost and J. Čejka, J. Raman Spectrosc., 2007, 38(11), 1488–1493 CrossRef CAS ; (c) R. L. Frost, J. Čejka and M. J. Dickfos, J. Raman Spectrosc., 2008, 39(7), 779–785 CrossRef CAS .
  12. (a) S. Bryan, T. Levitskaia and S. Schlahta, WM2008 Conference, Pacific Northwest National Laboratory, Phoenix, AZ, February 24–28, 2008 ; (b) FY2009 Progress: Process Monitoring Technology Demonstration at PNNL, PNNL-19136.
  13. G. Lu, T. Z. Forbes and A. J. Haes, Anal. Chem., 2016, 88(1), 773–780 CrossRef CAS PubMed .
  14. C. N. Trung, G. M. Begun and D. A. Palmer, Inorg. Chem., 1992, 31, 5280–5287 CrossRef .
  15. A. Burneau, M. Tazi and G. Bouzat, Talanta, 1992, 39, 743–748 CrossRef CAS PubMed .
  16. S. P. Pasilis and J. E. Pemberton, Inorg. Chem., 2003, 42, 6793–6800 CrossRef CAS PubMed .
  17. C. Xu, G. Tian, S. J. Teat and L. Rao, Inorg. Chem., 2013, 52, 2750–2756 CrossRef CAS PubMed .
  18. (a) D. Ferri, M. Iuliano, C. Manfredi, E. Vasca, T. Caruso, M. Clemente and C. Fontanella, J. Chem. Soc., Dalton Trans., 2000, 3460–3466 RSC ; (b) P. D. Bernardo, P. L. Zanonato, G. Tian, M. Tolazzic and L. Rao, Dalton Trans., 2009, 4450–4457 RSC .
  19. P. Gans, A. Sabatini and A. Vacca, Talanta, 1996, 43, 1739–1753 CrossRef CAS PubMed .
  20. C. Görller-Walrand and K. Servaes, Helv. Chim. Acta, 2009, 92, 2304–2312 CrossRef .
  21. J. Havel, J. Suto-Guerrero and P. Lubal, Polyhedron, 2002, 21, 1411–1420 CrossRef CAS .
  22. (a) S. Tsushima, V. Brendlera and K. Fahmya, Dalton Trans., 2010, 39, 10953–10958 RSC ; (b) V. Vallet, H. Moll, U. Wahlgren, Z. Szabo and I. Grenthe, Inorg. Chem., 2003, 42, 1982–1993 CrossRef CAS PubMed .


Electronic supplementary information (ESI) available: Table S1, Fig. S1 and S2. See DOI: 10.1039/c7dt01631j

This journal is © The Royal Society of Chemistry 2017