ARTICLE pubs.acs.org/est
TRLFS Evidence for Precipitation of Uranyl Phosphate on the Surface of Alumina: Environmental Implications Mirella Del Nero,†,* Catherine Galindo,† Remi Barillon,† and Benoit Made‡ † ‡
Institut Pluridisciplinaire Hubert Curien, UMR 7178 CNRS/UdS, 23 rue du Loess, BP 28, 67037 Strasbourg Cedex 2, France Ecole Nationale Superieure des Mines, Mines ParisTech, Centre de Geosciences, 35 rue Saint Honore, 77305 Fontainebleau Cedex.
bS Supporting Information ABSTRACT: We studied the ligand-enhanced sorption of uranyl ions (112 μM) on R-alumina colloids suspended in (and pre-equilibrated with) solutions at various concentrations of phosphate ions (PT = 0900 μM). A highly sensitive technique, time resolved laser-induced fluorescence spectroscopy (TRLFS), was used to examine the chemical speciation of uranyl sorbed at trace concentrations (0.44 μmol U 3 g1). The suspensions with PT g 100 μM exhibited high uranyl adsorption, and a very high intensity of fluorescence that increased with the sorbed amounts of phosphate and uranyl. These samples exhibited similar spectral and temporal characteristics of fluorescence emission, evidencing a uniform speciation pattern and a single coordination environment for sorbed U, despite large variation in parameters such as aqueous uranyl speciation, U loading, and extent of coverage of alumina by secondary Al phosphates precipitating on the surface. The results pointed formation of surface precipitates of uranyl phosphates, which are characterized by high quantum yield, peak maxima at positions similar to those of U(VI) phosphate minerals and four lifetimes indicating distortions, in-homogeneities or varying number of water molecules in the lattice. The findings have major implications for our understanding of the mechanisms of immobilization of U at trace levels on surfaces of oxides submitted to phosphated solutions in soils with low pH.
’ INTRODUCTION Knowledge of the processes involved in the immobilization of uranyl by phosphate ligands in natural systems is a major challenge. Uranyl released during the early stages of the oxidative weathering of primary U deposits interact with phosphate ions (noted P) to form U(VI)-phosphate minerals1 which are poorly soluble at near neutral pH.2 Uranyl has been found accumulated in close association with P and Fe or Al in oxidizing layers or soils with low pH overlying primary ores, too. Many experimental studies have shown that phosphate or arsenate preadsorbed on minerals enhance uranyl sorption (e.g., refs 36). Studies of both experimental systems7 and natural samples8 have suggested that U and P are sorbed onto ferric (hydr)oxides in the form of ternary uranyl-phosphato surface complexes, or that uranyl is sorbed by Fe3þ-phosphate surface precipitates having a stronger affinity for U than the oxide surface. It was also proposed that the distribution of U in weathered zones involves the late desorption of uranyl surface complexes that favors the local precipitation of phosphates containing U, for example, nanoscale uranyl phosphates in iron nodules or U-bearing Al phosphates in soils overlying the Koongara9 or Core Hill1 deposits, respectively. It is clear from above-cited literature that a large variety of mechanisms is involved in the sorption of U in oxidizing horizons rich in P. r 2011 American Chemical Society
The reliable modeling of uranyl retention requires basic knowledge, which is lacking at present, on the structure of the uranyl phosphate sorption species forming in oxidesolution systems. Studies at the molecular scale to identify uranyl sorption species, and factors influencing their nature, are limited for oxidephosphate solution systems. Tang and Reeder10 have analyzed by extended X-ray absorption fine structure (EXAFS) spectroscopy the uranyl species sorbed on γ-alumina in presence of arsenate, taken as a chemical homologue for phosphate. They reported the existence of surface precipitates of U(VI)-arsenate for high solution concentrations of U (g50 μM). Galindo et al.6 have studied by in situ attenuated total reflectance fourier transform infrared (ATR-FTIR) spectroscopy the dynamics of phosphate ions at the R-aluminasolution interface, during the process of sorption at low pH of aqueous uranyl (110 μM) and phosphate ligands. They showed that surface precipitates of U(VI)-phosphates were forming for the weak concentration and the low surface area (SA < 1 m2.L1) of R-Al2O3 used in Received: January 6, 2011 Accepted: March 25, 2011 Revised: March 21, 2011 Published: April 06, 2011 3982
dx.doi.org/10.1021/es2000479 | Environ. Sci. Technol. 2011, 45, 3982–3988
Environmental Science & Technology
ARTICLE
Table 1. Effect of Phosphate Ligand on the Fluorescence Lifetimes τi (μs), on the Relative Contributions pi (%) of the U Emitting Components to Fluorescence Decay, and on the Position of the Emission Maxima, for 2.5 g 3 L1 Al2O3 Suspensions and for Related Solutions. PT = 0900 μM, UT = 0, 1, or 12 μM, pH 3.30 ( 0.05, tR = 19 h, I = 0.01M NaClO4 τi ((Δτi)
pi ((Δpi)
emission maxima (d = 1 μs, w = 1 μs, resolution = 1 nm)
Suspensiona b
AP0U0
9(2), 50(8)
AP0U12, AP0U1
1.5, 9, 52
92, 6, 2
488.9 510.1 533.5 559.7
AP10U1
1.7, 8, 44
78, 17, 5
489.8 510.3 533.6 559.2 585.5
AP100U12, AP400U12, AP900U12c AP100U1, AP400U1, AP900U1
3.6(1.2),17(2),56(4),160(9) Id.
35(4),34(2),24(2),8(2) Id.
483.8 498.4 519.0 541.8 567.3 592 Id.
BP0U12
1.6(0.1), 134(7)
g99, e1
474.8 489.2 509.9 533.3 558.1 582.2
BP10U12
1.6(0.1), 134(7)
g99, e1
473.6 489.0 509.9 533.4 558.0 583.7
BP100U12
1.5, 11, 145
97, 2, 1
473.8 489.6 510.4 533.6 558.7 585.8
BP400U12
1.5, 10, 136
89, 10, 1
476.8 492.3 513.4 536.6 562.1 589.1
BP900U12
1.6, 9, 136
75, 23, 1
483.9 495.9 515.6 538.6 563.9 586.8
Solution
U added to the (P-containing) suspensions pre-equilibrated for 3 h. b Due to impurities in alumina matrix showing very low fluorescence intensity. c Wet paste. a
the ATR-FTIR experiments, which favored a high coverage of the surface by phosphate and uranyl. Full identification of all mechanisms and species involved in the sorption of uranyl on alumina in the presence of phosphate ligands requires, however, spectroscopic investigations of the coordination environment(s) of U(VI) sorbed. Moreover, using techniques sensitive to U present at the trace levels on the Al-oxide is necessary, since it was suggested that the loading of the surface with P and U is the key parameter controlling the transition between ternary surface complex and U(VI)-phosphate surface precipitates.6 In this work, we examined the effect of phosphate ligands on the chemical speciation of uranyl sorbed at low pH (3.3) and at the trace levels (∼ 0.44 μmol 3 g1) on R-alumina. We used a spectroscopic technique that is highly sensitive to sorbed U, namely time resolved laser-induced fluorescence spectroscopy (TRLFS). The study is a continuation of previous work restricted to the speciation of phosphate (co)sorbed with uranyl on Al2O3, at high P and U surface coverage.6 TRLFS is a powerful tool to explore identity of species via (i) position and relative intensities of emission peaks, which depend on the U coordination environment, that is, here, on bonding modalities of phosphate and hydroxyl ligands in the equatorial plane of UO2, and (ii) fluorescence lifetimes, which are sensitive to the presence of quenchers around U. In recent years, the technique has been applied to studying uranyl surface species in situ, that is, in colloidal suspensions.11,12 The aim of the present work was to identify the uranyl species at the alumina colloidphosphate solution interface, and to determine their dependency on key factors of suspensions, that is, aqueous species (UO22þ and U(VI)-phosphato complexes) and phosphate surface species pre-existing onto Al2O3 before uranyl sorption. An exhaustive study of the respective roles of aqueous uranyl speciation and phosphate surface coverage was achieved here by analyzing a large amount of suspensions that were brought at different initial concentrations of aqueous phosphate (0900 μM) and alumina (0.72.5 g 3 L1) before addition of uranyl (1 or 12 μM) to the systems. Del Nero et al.13 have indeed shown that increasing the total concentration of phosphate in R-Al2O3 0.01 M NaClO4 suspensions resulted in an increase of both the amount of
phosphate sorbed and the concentration of dissolved phosphate (which may react with dissolved U to form stable aqueous complexes). The authors have evidenced that phosphate ions are sorbed at low pH on R-Al2O3 via a combination of reactions of formation of inner-sphere phosphate surface complex and surface precipitates of Al phosphates, with the relative abundances of species depending on P surface coverage. The effect of pre-existing phosphate sorption species on oxides, particularly of Al phosphate surface precipitates, is an issue that is central to the understanding of the mechanisms of uranyl sorption on naturally occurring surfaces of oxides modified by soil solutions rich in P. This is the first study providing TRLFS data on coordination environments of U sorbed in a uranyl-oxide-phosphate system.
’ EXPERIMENTAL SECTION Batch Experiments. The samples analyzed by TRLFS were obtained from experiments on the sorption of uranyl ions which were added to P-containing R-Al2O30.01 M NaClO4 suspensions pre-equilibrated during 3 h (tP = 3 h) at pH 3.30 ( 0.05 and 298 K. NaCl electrolyte was not used because chloride ions quench the fluorescence. Main parameters under investigation were the total concentrations of phosphate (PT = 0900 μM), R-Al2O3 (r = 2.5 or 0.7 g 3 L1), and U (UT = 1 or 12 μM), as controls of the extent of coverage of the surface by phosphate and/or uranyl. The time of reaction between U, P and Al2O3 was also varied (tR = 3 h14 days). The low pH (3.30 ( 0.05) was chosen (i) to ensure high coverage of alumina surface with phosphate13 and (i) to minimize direct precipitation from the solution of uranyl phosphate minerals (which are poorly soluble at pH near to the neutrality). The initial solutions of the suspensions analyzed by TRLFS (cf. Table 1) were undersaturated with respect to UP minerals, except for the most concentrated ones (with UT = 12 μM and PT g 400 μM) which were slightly oversaturated with respect to (UO2)3(PO4)2 3 4H2O(s). Moreover, uranyl has low affinity for alumina at low pH, which is the best scenario for assessing the effect of presorbed phosphate on the uranyl sorption.6 The amounts of sorbed P (PS) and sorbed U (US) were determined from solution 3983
dx.doi.org/10.1021/es2000479 |Environ. Sci. Technol. 2011, 45, 3982–3988
Environmental Science & Technology
ARTICLE
Figure 1. Phosphate ligand-promoted sorption of U(VI) on R-Al2O3: uranyl sorbed (a) and phosphate sorbed (b) after short (white symbols: 3 h3 days) or long (dark symbols: 14 days) UP-alumina reaction times, for suspensions with r = 2.5 g 3 L1, PT = 0900 μM, pH 3.30 ( 0.05, I = 0.01 M NaClO4 and UT = 1 or 12 μM. U was added after a 3 h pre-equilibration of the suspensions containing P. Triangle = U or P surface coverage, squares = percentage of P sorbed. Lines are guides for the eyes.
analysis after sample centrifugation. The experimental errors prevented any accurate determination for the suspensions with low r. The electrophoretic mobility (EM) of suspended alumina was measured and was converted into zeta (ξ) potential (Smolukowski equation). Details on procedures and materials, including characteristics of R-Al2O3 (size: ∼300 nm, SA: ∼7 m2 3 g1), are given elsewhere.6,13 TRLFS Measurements. The TRLFS equipments included a pulsed laser Nd:YAG (15 Hz, 7 ns pulse duration) as an excitation source (λex = 266 nm) and a monochromator coupled with a CCD camera (Princeton Instruments) for the detection. The applied laser energy was equal to ∼10 mJ. An aliquot of the batch suspension was transferred into a quartz cell and was stirred continuously at 293 K. Time-resolved spectra of the fluorescence emission of U were recorded between 400 and 750 nm for different gate delay times (d = 1300 μs). The uncertainty on the peak position was ∼1 nm. The decay profile of the fluorescence of U was recorded within the interval of time 0.12000 μs. Each sample was exposed to the laser light during 30 min during the recording of the emission spectra and decay profile. No sample degradation was observed for the duration of the exposure. The lifetime values (τi) of the fluorescent uranyl species were obtained by fitting the decay curve to multiexponential laws using the LevenbergMarquardt algorithm. For each suspension, the whole procedure for lifetime determination was repeated twice. The alumina colloids in 0.01 M NaClO4 with no phosphate and no uranyl emitted a weak fluorescence signal and a biexponential decay function due to possible impurities or defects in the alumina lattice. The contribution of this matrix was observable on the spectra of the low-UT and low-PT suspensions, but it was negligible for the other suspension spectra. The signalto-noise ratio was about 40 for suspensions at PT = 100 μM and at UT = 1 μM and it increased with PT or UT. TRLFS measurements were also made on solution samples, with pH 3.3, I = 0.01 M NaClO4, UT = 12 μM and PT e 900 μM. Spectroscopic Sample Characteristics. The spectroscopic solution samples are noted as BPTUT (PT and UT in μM). The spectroscopic suspension samples are noted as APTUT. The samples (Table 1) included suspensions with no phosphate (samples AP0U12 and AP0U1), or suspensions with low, moderate or high phosphate contents showing PS values varying from a few μmol 3 g1 (sample AP10U1) up to tenth of μmol 3 g1 (samples AP100U12AP900U12, AP100U1AP900U1). The
suspensions with high PS values included samples having surface concentrations of U in the range of a few μmol 3 g1 (samples AP100U12AP900U12) or lower than 0.5 μmol 3 g1 (samples AP100U1AP900U1).
’ RESULTS AND DISCUSSION Macroscopic Sorption and Zeta Potentials. Major trends on the macroscopic uptake of phosphate ions (blank experiments), and uranyl and phosphate ions (U sorption experiments) in 2.5 g 3 L1 R-Al2O30.01 M NaClO4 suspensions are as follows. The blank experiments showed that increasing PT led to a decrease in the percentage of P sorption (Supporting Information Table S1). The increase of PT led to an increase of the concentration of residual aqueous phosphate, as well as of phosphate sorbed. Increasing equilibration time, tP, from 3 h to 14 days had no effect on the sorption of P on R-Al2O3 within our experimental uncertainties. Thus, it is expected that increasing the value of PT of the alumina suspension would promote the complexation of uranyl by residual aqueous phosphate, in the U sorption experiments. Speciation calculations indicated that the chemical forms of uranyl added to the pre-equilibrated Al2O3P suspensions would include UO22þ and uranyl phosphate complexes (before sorption), with the percentage of complexes ranging from ∼2% (PT = 10 μM) to ∼62% (PT = 880 μM) of total U(VI). Thermodynamic calculations suggested that all the initial solutions of the suspensions were under-saturated with respect to uranyl phosphates, except for the most concentrated ones (with UT = 12 μM and PT g 200 μM), which were slightly oversaturated with respect to (UO2)3(PO4)2 3 4H2O(s). If occurring in these suspensions, direct precipitation from the solution is expected to contribute to the sorption of U less than 8% (PT = 200 μM), and up to 30% (PT = 880 μM). The experiments of uranyl sorption (Figure 1a) showed that increasing PT resulted in enhanced removal of aqueous uranyl, which was almost quantitative for PT > 100 μM whatever UT (1 or 12 μM). Increasing equilibration time (19 h e tR e 14 days) had no significant effect on uranyl uptake. Uranyl sorption had no measurable effect on the phosphate sorption isotherm for the low UT values investigated, that is, the percentage and the amount of phosphate sorbed (Figure 1b) were similar to those obtained in the absence of U (Supporting Information Table S1) within our experimental uncertainties. Supporting Information Figure S2 3984
dx.doi.org/10.1021/es2000479 |Environ. Sci. Technol. 2011, 45, 3982–3988
Environmental Science & Technology
ARTICLE
Figure 2. Effect of phosphate ligands on the fluorescence emission spectrum of uranyl in R-Al2O3 suspension samples (AP0U12, AP100U12, AP400U12) of Table 1 (d = 1 μs; w = 2 μs). Uranyl phosphate sorption species show shifts in peak positions relative to those of UO22þ and higher quantum yield than U species in related solutions (samples BP0U12 and BP400U12). Sample s-AP400U12 is the supernatant of the suspension AP400U12.
presents the results of the ξ potential measurements. The ξ potential data for the U-free suspensions showed a two-steps trend decrease of ξ with PS, that is, a steep decrease for PT < 100 μM and a slight decrease for higher PT values, suggesting that the mechanisms of P sorption were PT-dependent. Sorption of uranyl induced an additional decrease in the ξ potential value of the P-amended surfaces for the high-PT suspensions (at short reaction times only), confirming the sorption of U on the alumina surface. The U sorption had no measurable effect on the ξ potential for the low-PT suspensions. All the macroscopic sorption data and the ξ potential data obtained here for the NaClO4 medium are similar to those published for NaCl, indicating similar sorption mechanisms for both electrolytes. The readers can find a detailed discussion of the experimental data obtained for the Al2O3P systems and the Al2O3PU systems in the papers by Del Nero et al.13 and Galindo et al.,6 respectively. Briefly, these authors have concluded that (i) the sorption of P occurred mainly via formation of inner-sphere phosphate surface complex and via surface precipitation of Al phosphate whose amount increased with PS and tR, irrespective of the presence of U, (ii) U was sorbed in the high-PT suspensions mainly through the formation of surface species of high negative charges. TRLFS Results for Reference Samples. Solution samples with pH 3.3, I = 0.01 M NaClO4, UT = 12 μM and PT e 900 μM were taken as reference samples. Detailed TRLFS studies of the aqueous complexation of uranyl by phosphate can be found in literature.14,15 Increasing phosphate concentration of the reference solutions resulted in an increase of the overall intensity of fluorescence of the solutions, and in a shift of the band positions toward higher wavelengths, for d = 1 μs (Supporting Information Figure S2). This is consistent with the increasing formation of the U(VI)-phosphato complexes, UO2H2PO4þ and UO2HPO4, whose positions of fluorescence peak maxima are reported to be shifted toward higher wavelengths relative to that of UO22þ.15 The best approximation for fluorescence decay of U in the reference phosphate solutions was a triexponential function
suggesting the existence of three fluorescent components (Table 1), including UO22þ and uranyl phosphate species whose contribution increased with PT and whose lifetime (∼10 μs) was characteristic of UO2H2PO4þ and/or UO2HPO4.14 For all solutions, there was observed a small contribution to fluorescence of a long-lived species whose lifetime and peak maxima are consistent with those of uranyl oxide hydrates,16 indicating the presence of limited amounts of true colloids. This finding is in agreement with results of a study by vibrational spectroscopy showing the presence of U(VI) hydrolysis species in diluted uranyl solutions at pH 2.54.5.17 The TRLFS data are consistent with the increasing formation of U(VI)-phosphato complexes with PT. No spectroscopic evidence was found for direct precipitation of uranyl phosphates from the solution. Therefore, direct precipitation is not expected to contribute to the uptake of U in the short term sorption experiments. TRLFS Results for Suspension Samples. The fluorescence of uranyl was related univocally to formation of uranyl sorption species on the alumina surface, in the case of suspensions with high/medium phosphate concentration, which show a high percentage of uranyl sorption. As an example, Figure 2 shows that the fluorescence intensity recorded for the supernatant of the suspension AP400U12 was negligible compared to that of whole suspension sample. The most striking feature was that the sorption of uranyl on alumina, in the presence of phosphate ligands, resulted in a marked enhancement of fluorescence intensity (cf. spectra of BP400U12 and AP400U12). The enhancement due to sorption far exceeded that due to the formation of aqueous U(VI)-phosphato complexes (cf. spectra of BP0U12 and BP400U12), pointing to uranyl sorption species with a quantum yield higher than that of the aqueous UP species. This finding stressed that TRLFS is the proper technique to study in situ the surface speciation of uranyl present at trace levels at the interface between alumina and phosphate solution. We present and we discuss below the effects of suspension parameters on the position of emission peaks (Figures 24), the lifetime of the fluorescent components and 3985
dx.doi.org/10.1021/es2000479 |Environ. Sci. Technol. 2011, 45, 3982–3988
Environmental Science & Technology
ARTICLE
Figure 3. Effect of phosphate ligands on the fluorescence emission spectrum of U in R-Al2O3 suspension samples (AP0U1-AP900U1) of Table 1 (d = 1 μs; w = 2 μs).
Figure 4. Fluorescence emission spectra of U recorded at various delay time and gate width, for the high-PT suspension samples AP100U12 (a) and AP400U12 (b) of Table 1.
their relative contributions (pi) to the decay of fluorescence (Table 1). Emission Characteristics of U in Suspensions With PT < 100 μM. In the absence of P, the TR spectra of uranyl in the 2.5 g 3 L1 alumina suspensions AP0U12 and AP0U1 (which showed a percentage of U sorbed of ∼2%) did not provide any specific information on possible adsorbed U(VI) species. The spectra showed the emission peaks of the uranyl aquo ion in perchlorate solution, which were superimposed on the background signal from the solid matrix (Figure 2). Due to the screening effect of the matrix, the intensity of uranyl fluorescence was 6 times lower in the alumina suspension than in the corresponding aqueous solution, at d = 1 μs. The fluorescence decay function obtained by statistical analysis was triexponential, with the lifetime values of UO22þ and alumina matrix (Table 1). Main spectral features of U in the low-PT alumina suspensions (e.g., spectrum of AP10U1, Figure 3) were close to those of U in the related solutions. Although uranyl sorption species might have contributed to
fluorescence decay and TR spectra, the accurate determination of their emission characteristics was difficult because aqueous species were the predominant species contributing to the decay and/or the TR spectra at long delay time were too noisy. Emission Characteristics of U in Suspensions at Moderate/ High PT. Unlike for the low-PT suspensions, the spectral characteristics of uranyl (UT = 1 or 12 μM) in the 2.5 g 3 L1 alumina suspensions with moderate or high contents of phosphate (100 μM e PT e 900 μM) were specific to uranyl phosphate sorption species. Figures 2 and 3 illustrate that the fluorescence intensity of these suspensions increased with UT and PT, that is, with U and P surface coverage. In contrast, there was no change either in the shape of the emission spectra or in the position of the peaks. Reducing the ratio alumina/solution (to r = 0.7 g 3 L1) or increasing the UPAl2O3 reaction time (up to tR = 1 month) did not induce any significant change in the spectral characteristics of the suspensions (Supporting Information Figures S3 and S4). The peak maxima were shifted drastically to higher 3986
dx.doi.org/10.1021/es2000479 |Environ. Sci. Technol. 2011, 45, 3982–3988
Environmental Science & Technology wavelengths relative to those of the reference uranyl phosphate solutions. For all suspension samples, four lifetimes were resolved by the analysis of the fluorescence decay (Table 1). Assuming mono, bi or triexponential decays gave significantly worse fits. Increasing the concentrations of phosphate and uranyl, as well as the UPAl2O3 reaction time (up to tR = 1 month), had no significant effect either on the lifetimes or on the pi values. All these findings evidenced that the same four uranyl components were forming at a relatively constant ratio on the alumina surface, despite important changes in the aqueous speciation of U and in the surface coverage. Surprisingly, the four components exhibited the same spectral identity, as varying the delay time (d = 1300 μs) had no effect either on the position or on the relative intensities of the emission peaks (Figure 4). The position of the main emission peaks which is observed on the TR spectra of the alumina suspensions was shifted by ∼10 and ∼3 nm toward higher wavelength relative to those of the uranyl aquo ion and protonated uranyl phosphate complexes, respectively. The position was similar to that reported for uranyl sorbed at low pH on γ-alumina18 but the relative intensities of main peaks were different, indicating different U coordination environments. The position was also close to that reported for U(VI)-Al-phosphates such as phuralumite.19 Emission spectra provide also information on the frequency of the symmetric stretching vibration ν1 of the UO2 unit,19 which is empirically correlated with the UO axial bond length20 and is measured as an average of band spacing of fluorescence emission from the lower radiation emitting level. All the TR spectra recorded here for suspensions at moderate or high PT values showed a band spacing of 808 ( 11 cm1. This suggested that the UO axial bond lengths were equal to 1.80 ( 0.01 Å for the four lifetime components coexisting on the R-Al2O3 surface. The values reported here are consistent with those published for uranyl phosphate minerals.19 In contrast, the band spacing and the UO axial bond length for the free uranyl ion in NaClO4 are given at 859 ( 6 cm1 and 1.75 ( 0.01 Å, respectively.19 The relatively small ν1 value and the weakening of the OdUdO bond observed for our suspensions pointed the strong chemical interaction between equatorial oxygen atoms of sorbed uranyl and phosphate and/or hydroxyl ligands, as increased binding strength in the equatorial plane of the UO2 moiety withdraws electron density from the axial oxygen atoms.
’ DISCUSSION The TRLFS data presented here can be used to identify the uranyl surface species of high negative charges forming in RAl2O3 suspensions with low pH and with moderate or high contents of phosphate (PT g 100 μM). The data provided clear evidence that the P-enhanced sorption of uranyl was due to the formation of uranyl phosphate species on the R-Al2O3 surface. The suspensions showed indeed a spectral signature specific to the UP surface species, which exhibited a high quantum yield, four lifetime components and significant shifts in peak positions relative to those of the uranyl aquo ion and reference solutions. The most important feature was that the spectral and temporal characteristics of the fluorescence emission did not vary from one suspension sample to another, that is, by varying over ranges the concentrations of U, P, and alumina used in the sorption experiment. Thus, the same species of uranyl phosphate were forming at the surface of alumina for a variety of experimental conditions, that is, of aqueous UP complexes, of extent of
ARTICLE
formation of Al phosphate surface precipitates, of surface coverage by uranyl, etc. This result provides strong evidence for the formation of surface precipitates of uranyl phosphate on alumina (as the distribution and nature of surface complexes are expected to depend strongly on solution and surface parameters). Moreover, the four lifetime components coexisting on the surface of Al2O3 have similar emission spectra and show constant relative proportions for all the samples, further pointing the existence of uranyl phosphate surface precipitates with a single type of coordination environment for sorbed U atoms. In-homogeneities and disturbance in the precipitate lattice and/or variation in the number of water molecules linked loosely by hydrogen bonding to the U-coordinated water molecules might have affected the lifetime of sorbed U but not its emission spectrum, as shown previously for the hydrates of uranyl nitrate.21 The position of the peak maxima and the band spacing of the surface precipitates formed on the R-alumina are similar to those reported for the phuralumite, a mineral containing structural OH groups. Tang et al.5 examined the local structure of U sorbed on arsenate-pretreated alumina by using EXAFS. They suggested the formation of poorly crystalline and highly disordered uranyl arsenate surface precipitates with a structure similar to that of the layered mineral tr€ogerite. Possibly, the uranyl phosphate precipitates formed on the R-alumina exhibited a structure close to that of phuralumite, with varied number of layer stacking or interlayer water molecule/cation contents. Thus, the present TRLFS study leads to the significant conclusion that surface precipitation of uranyl phosphate is the key mechanism controlling the speciation of uranyl sorbed on Al oxide, for very low U surface coverage (US < 0.5 μmol 3 g1). Environmental Relevance. This paper shows that TRLFS is a powerful technique to explore the identities of uranyl species sorbed on colloids or on minerals, at trace surface concentrations that are relevant to the environment. The spectroscopic results reported here have implications for our understanding of the uranyl behavior in P-rich oxidizing soils or weathering profiles. They suggest that the immobilization of uranyl, present at trace levels in soil solutions, involves the precipitation of uranyl phosphate on the surfaces of natural oxides subjected to acidic solutions containing phosphate ligands. They also show that precipitates of uranyl phosphate can form on oxide surfaces whatever the extent of the transformation into Al phosphate of the original surface. This information is of help to elucidate the nature of the association between U(VI), P and Fe/Al in soils with low pH.
’ ASSOCIATED CONTENT
bS
Supporting Information. Table S1 gives experimental data of P sorption without U (blank experiments), and Figure S1 shows changes in alumina surface potentials induced by P sorption, in absence of U (blanks), and by sorption of P and U. Figures S2, S3, and S4 provide additional spectra of the fluorescence emission of uranyl in R-Al2O3/U/P systems and show effects of PT for solution samples BP0U12-BP900U12, and of r and tR for high-PT suspensions. This material is available free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION Corresponding Author
*E-mail:
[email protected]. 3987
dx.doi.org/10.1021/es2000479 |Environ. Sci. Technol. 2011, 45, 3982–3988
Environmental Science & Technology
’ ACKNOWLEDGMENT We are grateful to I. Fabing and C. Hoffmann for setting up the TRLFS equipments at IPHC. We thank E. Simoni to allow the use of the TRLFS equipment at IPNO for preliminary measurements, and G. Lagarde for his technical assistance. This work was supported financially by IPHC and the Alsace Region (REALISE network). We are grateful to the Associate Editor Professor R. Kretzschmar and to an anonymous reviewer for the improvement of the manuscript. ’ REFERENCES (1) Jerden, J. L. Jr.; Sinha, A. K. Geochemical coupling of uranium and phosphorous in soils overlying an unmined uranium deposit: Coles Hill, VA. J. Geochem. Explor. 2006, 91, 56 - 70. (2) Gorman-Lewis, D.; Burns, P. C.; Fein, J. B. Review of uranyl mineral solubility measurements. J. Chem. Thermodynamics 2008, 40, 335–352. (3) Chang, T; Barnett, M. O.; Roden, E. E.; Zhuang, J. Effects of solid-to-solution ratio on Uranium(VI) adsorption and its implications. Environ. Sci. Technol. 2006, 40, 3243–3247. (4) Tang, Y.; Reeder, R. J. Enhanced uranyl sorption on aluminium oxide pretreated with arsenate. Part I: batch uptake behaviour. Environ. Sci. Technol. 2009, 43, 4446–4451. (5) Tang, Y.; Reeder, R. J. Enhanced uranyl sorption on aluminium oxide pretreated with arsenate. Part II: spectroscopic studies. Environ. Sci. Technol. 2009, 43, 4452–4458. (6) Galindo, C.; Del Nero, M.; Barillon, R.; Halter, E.; Made, B. Mechanisms of uranyl and phosphate (co)sorption: Complexation and precipitation at R-Al2O3 surfaces. J. Colloid Interface Sci. 2010, 347 282–289. (7) Payne, T. E.; Davis, J. A.; Waite, T. D. Uranium adsorption on ferrihydrite—Effects of phosphate and humic acid. Radiochim. Acta 1996, 74, 239–243. (8) Del Nero, M.; et al. Mechanisms of uranyl sorption. In Energy, Waste and the Environment: A Geochemical Perspective; Giere, R., Stille, P., Eds.; Geological Society Special Publication 236: London, 2004, p 545. (9) Sato, T.; Murakami, T.; Yanase, N.; Isobe, H.; Payne, T. E.; Airey, P. L. Iron nodules scavenging uranium from groundwater. Environ. Sci. Technol. 1997, 31 (10), 2854–2858. (10) Tang, Y.; Reeder, R. J. Uranyl and arsenate cosorption on aluminum oxide surface. Geochim. Cosmochim. Acta 2009, 73, 2727–2743. (11) Gabriel, U.; Charlet, L.; Schl€apfer, C. W.; Vial, J. C.; Brachmann, A.; Geipel, G. Uranyl surface speciation on silica particles studied by timeresolved laser-induced fluorescence spectroscopy. J. Colloid Interface Sci. 2001, 239, 358–368. (12) Krepelova, A.; Brendler, V.; Sachs, S.; Baumann, N.; Bernhard, G. U(VI)-kaolinite surface complexation in absence and presence of humic acid studied by TRLFS. Environ. Sci. Technol. 2007, 41 6142–6147. (13) Del Nero, M.; Galindo, C.; Barillon, R.; Halter, E.; Made, B. Surface reactivity of R-Al2O3 and mechanisms of phosphate sorption: In situ ATR-FTIR spectroscopy and ξ potential studies. J. Colloid Interface Sci. 2010, 342, 437–444. (14) Brendler, V.; Geipel, G.; Bernhard, G.; Nitsche, H. Complexation in the system UO22þ/PO43-/OH(aq): potentiometric and spectroscopic investigations at very low ionic strengths. Radiochim. Acta 1996, 74, 75. (15) Bonhoure, I.; Meca, S.; Marti, V.; De Pablo, J.; Cortina, J.-L. A new time-resolved laser-induced fluorescence spectrometry (TRLFS) data acquisition procedure applied to the uranyl-phosphate system. Radiochim. Acta 2007, 95, 165–172. (16) Chisholm-Brause, C. J.; Berg, J. M.; Matzner, R. A.; Morris, D. E. Uranium(VI) sorption on montmorillonite as a function of solution chemistry. J. Colloid Interface Sci. 2001, 233, 38–49. (17) M€uller, K.; Brendler, V.; Foerstendorf, H. Aqueous uranium(VI) hydrolysis species characterized by attenuated total refection
ARTICLE
fourier-transform infrared spectroscopy. Inorg. Chem. 2008, 47 10127–10134. (18) Froideval, A.; Del Nero, M.; Gaillard, C; Barillon, R.; Rossini, I.; Hazemann, J. L. Uranyl sorption species at low coverage on Al-hydroxide: TRLFS and XAFS studies. Geochim. Cosmochim. Acta 2006, 70, 5270–5284. (19) Geipel, G.; Bernhard, G.; Rutsch, M.; Brendler, V.; Nitsche, H. Spectroscopic properties of uranium(VI) minerals studied by timeresolved laser-induced fluorescence spectroscopy (TRLFS). Radiochim. Acta 2000, 88, 757–762. (20) Barlett, J. R.; Conney, R. P. On the determination of uraniumoxygen bond length in dioxouranium(VI) compounds by Raman spectroscopy. J. Mol. Struct. 1989, 193, 295. (21) Leung, A. F.; Hayashibara, L.; Spadaro, J. Fluorescence properties of uranyl nitrates. J. Phys. Chem. Solids 1999, 60, 299–304.
3988
dx.doi.org/10.1021/es2000479 |Environ. Sci. Technol. 2011, 45, 3982–3988
