Adsorption of Plutonium Oxide Nanoparticles - Langmuir (ACS

Jan 4, 2012 - Moritz Schmidt , Sang Soo Lee , Richard E. Wilson , Karah E. Knope , Francesco Bellucci , Peter J. Eng , Joanne E. Stubbs , L. Soderholm...
0 downloads 0 Views 3MB Size
Download PDF
Article pubs.acs.org/Langmuir

Adsorption of Plutonium Oxide Nanoparticles Moritz Schmidt,† Richard E. Wilson,† Sang Soo Lee,† L. Soderholm,*,† and P. Fenter*,† †

Chemical Sciences and Engineering Division, Argonne National Laboratory, Argonne, Illinois 60439, United States S Supporting Information *

ABSTRACT: Adsorption of monodisperse cubic plutonium oxide nanoparticles (“Pu-NP”, [Pu38O56Clx(H2O)y](40‑x)+, with a fluoriterelated lattice, approximately 1 nm in edge size) to the muscovite (001) basal plane from aqueous solutions was observed in situ (in 100 mM NaCl background electrolyte at pH 2.6). Uptake capacity of the surface quantified by α-spectrometry was 0.92 μg Pu/cm2, corresponding to 10.8 Pu per unit cell area (AUC). This amount is significantly larger than that of Pu4+ needed for satisfying the negative surface charge (0.25 Pu4+ for 1 e−/AUC). The adsorbed Pu-NPs cover 17% of the surface area, determined by X-ray reflectivity (XR). This correlates to one Pu-NP for every 14 unit cells of muscovite, suggesting that each particle compensates the charge of the unit cells onto which it adsorbs as well as those in its direct proximity. Structural investigation by resonant anomalous X-ray reflectivity distinguished two different sorption states of Pu-NPs on the surface at two different regimes of distance from the surface. A fraction of Pu is distributed within 11 Å from the surface. The distribution width matches the Pu-NP size, indicating that this species represents Pu-NPs adsorbed directly on the surface. Beyond the first layer, an additional fraction of sorbed Pu was observed to extend more broadly up to more than 100 Å from the surface. This distribution is interpreted as resulting from “stacking” or aggregation of the nanoparticles driven by sorption and accumulation of Pu-NPs at the interface although these PuNPs do not aggregate in the solution. These results are the first in situ observation of the interaction of nanoparticles with a charged mineral−water interface yielding information important to understanding the environmental transport of Pu and other nanophase inorganic species.

1. INTRODUCTION In recent years nanoparticles have found widespread application in sunscreens and cosmetics1 as well as in catalysis,2−5 sensors,6 batteries,7 medicine,8 and electronics.9−11 As a consequence, large quantities of nanoparticles have been released into the environment, with the amount expected to increase.1 Of particular concern is that nanoparticles can have chemical properties very different from those of their molecular constituents, a factor that is clearly seen in the limited number of studies currently available on colloidal metal transport through the geosphere. For example, enhanced transport of Al3+ in natural surface waters12 has been associated with the formation of soluble aluminum-oxy-hydroxide aggregates, such as AlO4Al12(OH)24(H2O)127+ (Al13) cations. Similarly, plutonium transport through the geosphere proceeds at a rate much faster than predicted,13,14 a finding that has been attributed to colloidal transport. Distinguishing these two observations is further chemical detail; the intrinsic Al13 clusters can also transport heavy metals through surface complexation whereas the Pu transport appears, at least in part, to be associated with adsorption to existing Fe colloids.14 It is widely known that adsorption of dissolved metal species to noncolloidal mineral surfaces is a mechanism for impeding environmental transport. Sorption processes on charged interfaces are often described by models of the electrical double layer developed on the basis of the classical Gouy− Chapman−Stern theories.15,16 Few experimental studies have © 2012 American Chemical Society

shown how dissolved nanoparticles interact with solid−liquid interfaces at the molecular level. The use of synchrotron based analytical tools, including X-ray absorption spectroscopy and, most recently, surface X-ray scattering techniques at third generation synchrotron sources, has provided significant insights into the underlying interactions that control metalion adsorption onto surfaces.17−20 These advances have come through study of model systems of dissolved ions under welldefined solution conditions, often in contact with single crystal surfaces. This manuscript reports initial experiments designed to extend this understanding to nanoparticle−surface interactions through direct in situ observations of the adsorption of Pu-nanoclusters at the muscovite mica−aqueous interface. Plutonium is a man-made element, first synthesized by Seaborg and co-workers in 1940.21 It is a major radiotoxicity concern and as such is relevant to the safety assessment of nuclear waste disposal and legacy contamination sites. Pu readily forms stable and well-defined oxide nanoparticles in aqueous solution.22 In the tetravalent state plutonium forms oxo-bridged polymers by aggregation of hydrolysis products via oxolation reactions (eq 1), comparable to polyoxo-metalates known from groups V and VI transition metals.23 Received: September 22, 2011 Revised: December 9, 2011 Published: January 4, 2012 2620

dx.doi.org/10.1021/la2037247 | Langmuir 2012, 28, 2620−2627

Langmuir

Pu IV −OH + Pu IV −OH → Pu−O−Pu + H2O

Article

ing the Pu-NP surface loading, areal coverage and vertical distribution, as well as substrate relaxations, and coadsorption of Cl− ions.

(1)

There are several reasons for choosing Pu nanocluster adsorption onto muscovite mica as a model system for understanding nanoparticle adsorption in general. Pu nanoclusters, Pu38O5640+, have been synthesized and their structures are known.24 The core of the cluster is an oxide-deficient PuO2 (space group Fm3m) unit decorated on its outer surface with anions, which compensate the NP charge, and water molecules. Adsorbed anions, such as Cl−, have been shown to be labile, and therefore readily desorbed or exchanged.25 The effective charge of the cluster in solution is dependent on the number of surface-bound anions, which in turn depends on pH, Eh, and the nature and concentration of dissolved anions. In this manner the nanoparticles' overall charge is influenced by solution conditions, which will directly influence the nanoparticles′ reactivity toward charged mineral surfaces. Because of this, it is not immediately apparent if the Pu-NP will have a primarily cationic character (and be attracted to the negatively charged muscovite surface) or anionic (and be repelled). X-ray scattering data confirmed the persistence of the monodisperse cluster under a variety of solution conditions.24,25 From a technical perspective, the large atomic number of Pu (94), together with the availability of its LIII-absorption edge in the hard X-ray regime, provides an ideal X-ray contrast for in situ studies, simplifying data analysis and interpretation of its adsorption to surfaces. Muscovite was chosen for this study because its basal surface is representative of dominant phyllosilicate surfaces of many clay minerals. It cleaves to create a large, atomically flat (001) plane with a permanent negative surface charge (1 e− per unit cell area, AUC) so that it is amenable to high resolution structural studies. The choice of muscovite also simplifies the interpretation of the results because it avoids complications associated with pH-dependent surface charge found on most oxide surfaces and electron transfer reactions that might be observed on redox active minerals such as iron oxides. To our knowledge there are few published studies on the sorption of nanoparticles to mineral−water interfaces. A study on the sorption properties of plutonium nanoparticles by ex situ high-resolution transmission electron microscopy (HRTEM) reveals that Pu(IV) clusters are associated with quartz and goethite.26 A distortion of the Fm3m symmetry was observed when the particles grow on goethite, associated with an epitaxial relationship between the nanoparticle and substrate. The nature of the TEM experiment, however, requires the samples to be removed from the aqueous solution for study, leaving an open path for potential interference with the sorption reaction or precipitation event. A study investigated Al Keggin clusters adsorbed to Al2O3 single crystal surfaces by monitoring the change on the substrate’s isoelectric point, but without any structural insight into how the clusters adsorbed to the surface.27 Here, we investigated the adsorption of Pu-oxo nanoparticles on the muscovite (001) basal plane by in situ nonresonant and resonant X-ray reflectivity (crystal truncation rods, CTR, and resonant anomalous X-ray reflectivity, RAXR) at pH = 2.6. This is the first in situ investigation of the structure of nanoparticles sorbed at a charged mineral−water interface. Studying the sample in situ strongly reduces the possibility of perturbing the sorption reaction and other systematic errors. The results provide direct insight into nanoparticle adsorption in this complex, electrostatically controlled, interfacial system, includ-

2. MATERIALS AND METHODS 2.1. Sorption Experiments. CAUTION! 242Pu is a radionuclide with a half-life of 3.73 × 105 years. Its use requires the appropriate infrastructure and personnel trained in the handling of alpha-emitting isotopes. 2.1.1. Nanoparticle Preparation/Characterization. The plutonium oxide nanoparticles, [Pu38O56]40+, were prepared by the hydrolysis of 242 Pu(IV) in Li/HCl solution as described in the literature.25 Resulting crystals were dissolved in 100 mM NaCl and centrifuged several times to exchange Li+ by Na+. The solid, collected by centrifugation, was suspended in 10 mM HCl by gently warming the solution in a water bath at 90 °C. The final suspension was centrifuged to remove unsuspended solids and passed down a cation-exchange column to remove any remaining mononuclear plutonium ions. The total Pu concentration ([Pu]tot = 1 mM) of the final solution was determined by liquid scintillation counting. The presence of a pure solution of monodisperse Pu-NP was confirmed by UV−vis spectroscopy and high-energy X-ray scattering (HEXS).28,29 A pair distribution function (PDF) analysis of the HEXS data also confirmed the 11 Å particle size.24,25 The [Pu38O56]40+ core has a well-defined structure, closely based on the cubic PuO2 fluorite structure. The atomic pair-correlation function derived from the HEXS data does not show any correlation beyond 14 Å.24 Single-crystal structural analyses find variable numbers of Cl− ions bound to the NP surface.25 These Cl− ions are easily and reversibly exchanged by other anions without affecting the [Pu38O56]40+ core, rendering the overall cluster charge variable. 2.1.2. Muscovite Substrates. Muscovite (KAl3Si3O10(F,OH)2) is a natural phyllosilicate mineral from the mica group. It is an alumosilicate with a tetrahedron-octahedron-tetrahedron (TOT) layer structure, which is commonly found in many clay minerals, and charge compensating K+ ions in the interlayer space. Muscovite forms large single crystals and can be easily cleaved to yield an atomically flat (001) surface. Due to these favorable characteristics muscovite has been frequently used in structural studies and as an analogue for clay minerals, when single crystal substrates are required.20,30−33 The muscovite crystal used in the X-ray scattering experiment has a (001) spacing of 19.95 ± 0.01 Å and a unit cell area (AUC) of 46.72 Å2 (a = 5.187 Å, b = 9.007 Å).34 Cleaving parallel to the (001) plane exposes the TOT layer with a fixed lattice charge of 0.021e−/Å2 (1 e−/AUC), which does not depend on solution pH. Adsorption on the muscovite surface is commonly considered to be an electrostatically driven process, without significant contributions from chemical bonding of adsorbates to the surface.20,34−36 2.1.3. Sorption. A muscovite single crystal (12.7 × 12.7 × 0.2 mm3) was freshly cleaved and submerged in 2 mL of the Pu-NP solution with a total Pu concentration [Pu]tot = 1 mM and 100 mM NaCl as background electrolyte at pH = 2.6. After 12 h exposure time the crystal was taken out of the solution and residual liquid was removed before placing it on the cell pedestal and mounting in a doubly contained thin film X-ray reflectivity cell (see Figure S1 of the Supporting Information (SI)). Twenty microliters of the reaction solution was pipetted onto the single crystal surface and covered with a thin Kapton membrane (7.5 μm thick). The sample and the pedestal were then further enclosed in two independent layers of containment: (1) a 50-μm-thick Kapton dome and (2) a cylindrical Al cap with Kapton windows (75 μm thick). The encapsulated sample was then transferred to the Advanced Photon Source (APS). A second sample was prepared for α-spectrometry, as described previously:37 After reaction in the Pu-NP solution for two hours, the sample was rinsed with a solution containing only the background electrolyte and successively with deionized water. The sample was dried in air and fixed on a steel planchette with epoxy, its edges masked with a second steel planchette. An aperture in the second planchette limits α-detection to only the Pu adsorbed on the basal plane, thereby masking interference from Pu adsorbed to the crystal 2621

dx.doi.org/10.1021/la2037247 | Langmuir 2012, 28, 2620−2627

Langmuir

Article

edges, which does also not contribute to the XR-signal. The αspectrum was acquired using an AlphaAnalyst system from Canberra Inc. equipped with a passivated implanted planar silicon (PIPS) detector (450 mm2) and a 1024 channel multichannel analyzer. Energy calibrations were performed by linear regression analyses using a commercially prepared mixed α-source standard of 238U, 234U, 239Pu, and 241Am (Analytics, Atlanta, GA, SS: 59949−121). The same source was used to determine count rate efficiencies. 2.2. Crystal Truncation Rods and Resonant-Anomalous Xray Reflectivity. Detailed descriptions of both CTR and RAXR methodologies can be found in the literature.38−40 Data were measured at the Advanced Photon Source, beamline 6-ID-B (XOR, formerly MU-CAT). The incident X-ray beam with a typical flux of ∼1012 photons/s and a profile (collimated by slits and then focused vertically by a Kirkpatrick-Baez mirror) of 0.04 mm × 1.0 mm (V × H) was reflected from the sample and subsequently collected using an Xray CCD detector in a vertical scattering plane defined by the incident and reflected X-ray beams.41 CTR data (i.e., nonresonant XR data) were measured as a function of vertical momentum transfer qz (0.2− 5.3 Å−1) with a fixed incident X-ray energy (E = 14.0 keV) that was well separated from any Pu X-ray absorption lines. RAXR data were measured at fixed qz by scanning the incident energy around the Pu LIII absorption edge (18.061 keV) determined from X-ray fluorescence (XRF) on the same sample as described previously.42 RAXR spectra were measured at 17 selected values of qz. For both experiments, CTR and RAXR, system stability was controlled by repeated reference measurements at regular time intervals. The CTR measurement yields a total electron density, including possible contributions from water, adsorbed ions, and nanoparticles. CTR data were fit using a parametrized structural electron density model that has been described in detail elsewhere.34,43,44 The model comprises the ideal muscovite substrate, a two unit cell deep region of the substrate that is allowed to relax and adsorbed species that are modeled as a sum of Gaussian distributions as given in eq 2. The structure factor of each component is calculated as

⎡ q 2u 2 ⎤ j F = ∑ cj f j (qz) exp(iqzzj) exp⎢ − z ⎥ ⎥ ⎢ 2 ⎦ ⎣ j

⎡ ⎤ χ 2 = ⎢ ∑ (Ik − Icalc, k)2 /σk2⎥ /(N − Np) ⎢ ⎥ ⎣ k ⎦ R=

k

(2)

)

⎡ q 2u 2 ⎤ z j ⎥ 2 ⎥⎦ ⎣

j

(5)

3. RESULTS 3.1. Quantifying Sorption by α-Spectrometry. The total plutonium coverage was quantified by α-spectrometry after rinsing the sample in a 100 mM NaCl solution at pH = 2.6 and deionized water (DIW). Surface adsorption concentrations of 0.92 ± 0.05 μg/cm2 plutonium (or 10.8 Pu/AUC) were obtained, indicating that the adsorbed Pu-NPs were not removed from the surface by this rinsing procedure. When extrapolated to the full muscovite surface area (2 × 1.61 cm2), this corresponds to 2.98 μg Pu adsorbed on the faces of the crystal (not including the edges, which are masked in the αcounting experiment), equivalent to 3.1% of the total Pu in solution. Since the solutions were stable over the time of the measurement and a bulk precipitation event is expected to remove a much higher percentage of dissolved Pu, we conclude that the observed behavior is not controlled by bulk precipitation of PuO2 or other Pu solid phases. However, the amount of adsorption is substantially higher (by a factor of ∼40) than that expected for adsorption of Pu(IV) monomers whose expected coverage would be defined by charge compensation of the surface (0.25Pu/AUC). This reveals that the presence of Pu nanoparticles effectively increases the sorption capacity of the muscovite basal plane by more than an order of magnitude. This observation is particularly noteworthy, because the solution is at high ionic strength and acidic pH (100 mM NaCl at pH = 2.6), where Na+ and H3O+ ions compete with Pu-NPs for sorption to the muscovite surface.34,35,48 The large coverage of Pu observed from the α-counting and RAXR measurements indicates that Pu-NPs adsorb more strongly than Na+ or H3O+ whose adsorption free energies are reported to be ∼−11 kJ/ mol49 and −29.1 kJ/mol,48 respectively. Moreover, the bare charge of the Pu-NP is screened by surface-adsorbed Cl− anions (through either a Stern or diffuse layer of counterions around the nanoparticles). Adsorption of Cl− to the nanoparticles reduces their net positive charge and is expected to decrease the effective adsorption strength to the negatively charged surface. It also reduces the electrostatic repulsion between adsorbed PuNPs, potentially aiding the particle stacking or aggregation near the surface. Together with previous work,25 this apparently strong adsorption of the Pu-NPs to the surface suggests that the nanoparticles act as highly positively charged units, despite the large excess of Cl− in the solution. 3.2. Pu-NP Adsorption Structure. Figure 1 shows the measured CTR as a function of momentum transfer qz in the range 0.2−5.3 Å−1 (corresponding to Miller index L = 0.6−16.9 reciprocal lattice units (r.l.u.)) and the calculated values from the best-fit model (χ2 = 4.31, R-factor = 7.4%) used to derive the electron density profile (Figure 2). For comparison, data

FR(qz , E = [f ′ (E) + if ″ (E)]

∑ cj exp(iqzzj) exp⎢⎢−

(Ik − Icalc, k)/Ik /N

N and Np are numbers of data points and parameters used in the model fit, respectively, Ik and Icalc,k are the measured and calculated reflected intensities, respectively, and σk is the uncertainty in the kth data point.

where f j(qz) is the atomic scattering factor and cj, zj, and uj are the occupancy, height from the surface, and rms width of the jth atom. The quality of fit (i.e., a scaled χ2 and an R-factor; see below) was improved by including a broad distribution modeled as an extended range of electron density composed of multiple overlapping Gaussians, as described earlier.44,45 The structure of bulk water was expressed by a layered-water model.34,43 The Pu-specific electron density profile was determined by resonant anomalous X-ray reflectivity. The data were analyzed by applying a model in which the resonant structure factor is expressed as

×

∑

(4)

(3)

Here, f ′(E) and f″(E) are the anomalous dispersion terms of the resonant element, in this case Pu, that were determined by measuring the X-ray absorption near-edge structure (XANES) as X-ray fluorescence yield42 and applying a difference Kramers−Kronig transform.46 A model-independent analysis was used as a guide for modeling the plutonium electron density distribution.47 The total Pu coverage (=Σj cj) was constrained by the value determined by αspectrometry. This approach helps reduce the uncertainty of the Pu electron distribution. Fitting without this constraint yielded a comparable distribution and surface loading (11.1 Pu/AUC); however, the broad, more distant component (see section 3.2 for the detailed information about the structure) was ill-defined. For all fits the scaled χ2 and an R-factor quality of fit are reported. 2622

dx.doi.org/10.1021/la2037247 | Langmuir 2012, 28, 2620−2627

Langmuir

Article

Figure 1. Specular CTR. Specular X-ray reflectivity of the muscovite crystal (equilibrated with a solution of Pu nanoparticles with 1 mM [Pu]tot at pH 2.6 with a 100 mM NaCl background electrolyte) is shown as a function of momentum transfer, qz, measured with a photon energy at E = 14.0 keV. Measured reflectivity data are shown (red square with statistical error bars). The best fit is shown as blue line, and the CTR of muscovite in deionized water43 is shown as black symbols for comparison.

from the muscovite surface in deionized water (DIW) are shown.43 The substantial differences in the two data sets are associated with adsorption of the Pu-NPs and disruption of the hydration layer near the muscovite surface. These differences, most prominent at the midzones between Bragg-peaks (e.g., near qz ∼ 0.3, 1.5, and 4.1 Å−1), extend through the whole data set. Analysis of these data leads to the vertical electron density profile shown as a black line in Figure 2. This profile consists of the muscovite surface followed by a broad structured peak centered at about 6 Å above the muscovite surface with a total width of 11 Å, and a second weaker and broader profile whose electron density peaks at around 20 Å, and then declines slowly with increasing distance from the surface until it reaches the bulk water density more than 100 Å above the mineral surface. Such an electron distribution is atypical for the adsorption of a well-behaved cationic monomer to a mica surface under conditions similar to those used in this experiment.20,34,48 Most of the significant density enhancements near the interface are undoubtedly due to adsorbed Pu-NPs. The width of the first peak is in good agreement with adsorption of a Pu-NP with a size consistent with that determined in previous studies.24 However, the nonresonant data alone do not distinguish uniquely the distribution of adsorbed Pu nanoparticles from other possible structural features in the near-interface region (e.g., adsorbed water or ions from the background electrolyte). The muscovite substrate remains largely unaltered with no significant roughness (e.g., due to dissolution). Atomic structural displacements are limited to 100 Å from the interface. This distribution is in agreement with the findings from the CTR analysis. The upper part of Figure 2 shows a schematic representation of the species contributing to the observed electron density. The distribution width of the first peak in the Pu profile matches the size of the clusters in solution,24 suggesting that this fraction is composed primarily of Pu-NPs directly adsorbed to the surface. The peak at 20 Å suggests a second layer of 2623

dx.doi.org/10.1021/la2037247 | Langmuir 2012, 28, 2620−2627

Langmuir

Article

We estimate the average effective charge of Pu-NPs adsorbed directly on the surface by assuming that the Pu-NPs at the muscovite surface fully compensate surface charge. This estimation represents an upper limit of the nanoparticle charge because other Pu-NPs distributed above the first layer can also contribute to the surface charge compensation. The coverage measured by RAXR is 0.07 Pu-NP/AUC, suggesting that the average effective charge of one directly adsorbed Pu-NP is approximately +14. This result indicates that about 26 Cl− ions remain bound to the [Pu38O56]40+ core, significantly fewer than the 42 to 54 Cl− ions seen in the solid state.25 If all sorbed PuNPs (10.8 Pu/AUC or 0.28 Pu-NP/AUC) are considered to compensate the surface charge, the result would suggest ∼36 Cl− remain bound to each Pu-NP, still slightly less than the amounts observed in the original structures. This observation is consistent with the known lability of the Cl− ions bound to each Pu-NP.25 The size of the nanoparticles is similar in magnitude to the Debye length in a 0.1 molar solution (9.7 Å),51 which may lead to overcompensation of the surface charge.52 However, the dielectric constant of interfacial water close to the surface is likely to be reduced53 which would increase the Debye length beyond the size of the Pu-NPs in this regime. The apparent interfacial stacking or aggregation of nanoparticles is an indication that the uptake of nanoparticles is limited by steric constraints, and the observed surface loading represents an upper limit. The nanoparticles do not aggregate in solution in the absence of the muscovite interface, which was confirmed by HEXS and UV-vis spectroscopy. Hence, the observed aggregation near the interface must be a consequence of the increased local concentration of Pu-NP and possibly the release of Cl− counterions due to the negative surface charge (see below). Additional insight about the adsorption structure comes from examining the non-Pu electron density, which can be obtained from the difference between total electron density and Pu specific electron density (Figure 4). In the first 14 Å above the surfacethe size range in which correlations were observed in the HEXS analysis25where plutonium electron density due to the first nanoparticle fraction is observed, the density difference is significantly larger than the electron density of bulk water. We observe four distinct peaks, with a regular vertical spacing of ∼3 Å, at heights of 3.3 Å, 6.3 Å, 9.3 Å, and 12.2 Å from the surface. This observed excess electron density is too large to be only due to oxygen in adsorbed Pu-NPs (∼14% of the Pu-NP electron density) or liquid water filling the space between adsorbed Pu-NPs (∼40% of the total density excess). This suggests that there is an additional contribution to the density associated with the adsorbed Pu-NPs. The evenly spaced peaks can be explained by the more electron dense Cl− ions decorating the adsorbed nanoparticles′ surfaces. This attribution is supported by the observed areal coverage, which suggests that the Cl− decoration around the adsorbed nanoparticles is significantly smaller than what had been observed in solution.24,25 There are seven layers of Cl− anions around the Pu-NPs in solution, while we observe only four distinct peaks around the adsorbed Pu-NPs (schematically shown in Figure 2). The equidistant distribution with ∼3.0 Å spacing implies a layered structure, similar to what was found for the same nanoparticles in a bulk crystal.24,25 The distance of 3.0 Å lies within the range of the observed distances between two adjacent Cl− layers (∼2.5−3.5 Å) in the crystal, but does not exactly match the structures found earlier.24,25 It is not

Figure 3. RAXR spectra. Select resonant anomalous X-ray reflectivity data (black circles) measured from the muscovite (001) surface equilibrated with the same 1 mM Pu nanoparticle solution. Six of a total of 17 scans are shown. Each spectrum probes the variation of the specular reflectivity measured as a function of photon energy, E, at different fixed momentum transfer, qz (Å−1), as indicated. Spectra are plotted using the resonance amplitude normalization [(|Ftot(qz,E)|2 − | FNR(qz)|2)/(2|FNR(qz)|)], where Ftot and FNR are total and nonresonant structure factors, respectively,58 and are offset vertically for clarity (offsets are given in brackets for each spectrum). The blue lines show the calculated intensities from the best fits to the RAXR spectra.

adsorbed Pu-NPs, due to an aggregation or stacking process which may also be responsible for the extended broad distribution. No further distinct peaks or nodes are observed, implying that the stacking occurs in a randomly distributed fashion with respect to the muscovite surface and less welldefined layers of Pu-NPs. Moreover, the shape of the electron density profile suggests that the Pu nanoparticle density decreases with increasing distance from the muscovite surface. The broad, extended electron density profile could also arise from larger nanoparticles (up to ∼100 Å) adsorbed at the interface, which might themselves be a consequence of the described stacking process (there is, however, no evidence showing that such larger particles grow in the absence of the muscovite substrate). The relatively high ionic strength of the background electrolyte (100 mM NaCl) should facilitate aggregation of the Pu-NPs by partially screening the high charges of the [Pu38O56]40+ core. We can calculate the coverage of the adsorbed nanoparticles from the element-specific electron density distribution. By integrating the Pu electron density from the muscovite surface to the size of the Pu-NP (11 Å) we find the electron density ρ11 Å(Pu) = 5.5 e−/Å2, corresponding to 1 Pu-NP per 14.0 AUC, or 17% areal coverage (see SI for details on the calculation). We can also relate this coverage to the surface charge compensation process: Recognizing that the core of one PuNP has a nominal charge of +40, we would expect one nanoparticle to adsorb every 40 unit cells for charge compensation if there were no Cl− around the nanoparticle. The saturation coverage of mononuclear, aqueous ions on the muscovite (001) basal plane is normally limited by the compensation of the surface charge.20,34,35,48 2624

dx.doi.org/10.1021/la2037247 | Langmuir 2012, 28, 2620−2627

Langmuir

Article

aggregation of the particles, occur simultaneously at the interface. The particles show a stronger interaction with the surface than monovalent Na+ or H3O+ and act as expected for entities having a high positive charge despite their formal negative charge from previous structural characterization, consistent with the earlier finding that decorating Cl− ions are bound loosely to the particles.25 Nonetheless, the results suggest that some of the Cl− ions that decorate the Pu-NPs remain associated with the nanoparticles after adsorption. Additional electron density peaks were observed near the muscovite surface, which can most likely be attributed to Cl− coordination to the adsorbed nanoparticles, with contributions from water and Pu-NP-oxygen. These first in situ results on the interaction of oxide nanoparticles with a charged substrate provide new insights into nanoparticle−surface interactions in aqueous environments. Unlike traditional models of ion adsorption to mineral surfaces which assume distinct adsorption states of the ions in close contact with the surface, the present results show that these Pu-NPs adsorb in a combination of modes, including a first distinct adsorption layer followed by a broad distribution extending ≥100 Å from the surface and consistent with Pu-NP stacking or aggregation. The adsorption strength appears to be strong because the nanoparticles remain adsorbed even after multiple rinsing steps by NaCl solution and DIW (as done for α-spectrometry measurement). The measured sorption capacity of the Pu-NP is more than 40-times larger than that expected for the Pu4+ aqua-ion (10.8 Pu/AUC instead of 0.25 Pu/AUC). Such a strong interaction and increased sorption capacity of the plutonium nanoparticles to this phyllosilicate mineral provides a baseline for understanding, in general, the environmental transport of nanoparticles, and in particular, mobility of Pu in geological systems. Therefore while the adsorption of Pu to nanoparticulate mineral colloids is associated with increased transport in natural systems, the results presented here suggest that the formation of intrinsic Pu colloids would lead to increased retention of Pu with respect to the transport of simple ionic species. From a more general perspective, these results make clear that understanding the nanoparticle surface behavior is critical to understanding its adsorption behavior to charged surfaces. The lability of ions adsorbed onto nanoparticles with a charged core provides a variable that plays a role in the total adsorbate concentration and its surface distribution.

Figure 4. Plutonium and non-plutonium contributions to the electron density. The non-Pu electron density [i.e., the difference between the total and Pu electron densities (black line)] obtained from the best-fit model of the CTR data are shown with the Pu electron density obtained from the model-dependent (MD) fit of RAXR data (blue area), with each normalized to bulk water electron density (dashed black horizontal line).

surprising that the negatively charged interface should influence the distribution of the Cl− ions coordinating to the nanoparticles.54 It is possible that the negative charge of the muscovite surface will reduce the activity of Cl− around Pu-NPs close to the surface which leads to the release of Cl− ions to the solution.

4. DISCUSSION AND CONCLUSIONS The results show a novel and comprehensive picture of the interaction of plutonium oxide nanoparticles with a charged alumosilicate surface. The nanoparticles, with large, positively charged cores, spontaneously adsorb on muscovite due to its negative surface charge. These particles cover 17% of the whole surface, or approximately 14 unit cells per nanoparticle. This suggests that each particle compensates the charge of the unit cells it adsorbs on as well as that of adjacent unit cells. This appears reasonable when considering the Bjerrum length (the distance at which the electrostatic interaction between two charged particles is comparable in magnitude to the thermal energy kBT, where kB is the Boltzmann constant and T is the temperature), which is about 7 Å in bulk water55 but may be different for structured water near the surface because of a change in the dielectric constant.53,56 The specular reflectivity experiments presented in this work do not probe the lateral ordering of the nanoparticles on the muscovite surface, but the high coverage does make ordering effects seem plausible. A uniform distribution of the Pu-NPs on the surface would lead to an average separation of approximately 20 Å between neighboring adsorbed nanoparticles. Considering the high charge of each unit further adsorption onto those unoccupied spaces may be unfavorable. The broad distribution of Pu electron density distribution extending to more than 100 Å above the muscovite surface indicates that these Pu-NPs can agglomerate at the interface or sorb as larger aggregates. The results presented here demonstrate that these two processes, adsorption of the nanoparticles by electrostatic attraction and

■

ASSOCIATED CONTENT

S Supporting Information *

Details on the CTR and RAXR data analyses, calculation of the areal coverage based on the Pu-specific electron density profile, and details on the thin film cell used in the XR experiments. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected], [email protected].

■

ACKNOWLEDGMENTS This work conducted at Argonne National Laboratory, operated by UChicagoArgonne LLC for the United States Department of Energy under contract number DE-AC0206CH11357, is jointly supported by the United States 2625

dx.doi.org/10.1021/la2037247 | Langmuir 2012, 28, 2620−2627

Langmuir

Article

(19) Fuller, C. C.; Bargar, J. R.; Davis, J. A.; Piana, M. J. Mechanisms of Uranium Interactions with Hydroxyapatite: Implications for Groundwater Remediation. Environ. Sci. Technol. 2002, 36 (2), 158− 165. (20) Lee, S. S.; Fenter, P.; Park, C.; Sturchio, N. C.; Nagy, K. L. Hydrated Cation Speciation at the Muscovite (001)-Water Interface. Langmuir 2010, 26 (22), 16647−16651. (21) Seaborg, G. T.; Wahl, A. C.; Kennedy, J. W. Radioactive Element 94 from Deuterons on Uranium. Phys. Rev. 1946, 69 (7−8), 366−367. (22) Johnson, G. L.; Toth, L. M. Plutonium(IV) and Thorium(IV) Hydrous Polymer Chemistry; Oak Ridge National Laboratory: Oak Ridge, TN; May, 1978. (23) Mü l ler, A.; Peters, F.; Pope, M. T.; Gatteschi, D. Polyoxometalates: Very Large Clusters - Nanoscale Magnets. Chem. Rev. 1998, 98 (1), 239−272. (24) Soderholm, L.; Almond, P. M.; Skanthakumar, S.; Wilson, R. E.; Burns, P. C. The structure of the plutonium oxide nanocluster [Pu38O56Cl54(H2O)(8)](14‑). Angew. Chem., Int. Ed. 2008, 47 (2), 298− 302. (25) Wilson, R. E.; Skanthakumar, S.; Soderholm, L. Separation of plutonium oxide nanoparticles and colloids. Angew. Chem., Int. Ed. 2011, 50 (47), 11234−11237. (26) Powell, B. A.; Dai, Z.; Zavarin, M.; Zhao, P.; Kersting, A. B. Stabilization of Plutonium Nano-Colloids by Epitaxial Distortion on Mineral Surfaces. Environ. Sci. Technol. 2011, 45 (7), 2698−2703. (27) Lützenkirchen, J.; Kupcik, T.; Fuss, M.; Walther, C.; Sarpola, A.; Sundman, O. Adsorption of Al13-Keggin clusters to sapphire c-plane single crystals: Kinetic observations by streaming current measurements. Appl. Surf. Sci. 2010, 256 (17), 5406−5411. (28) Clark, D. L.; Hecker, S. S.; Jarvinen, G. D.; Neu, M. P. Plutonium. In The Chemistry of the Actinide and Transactinide Elements, Morss, L. R.; Edelstein, N. M.; Fuger, J.; Katz, J. J., Eds.; Springer: Berlin, 2011; Vol. 2, pp 813−1264. (29) Soderholm, L.; Skanthakumar, S.; Neuefeind, J. Determination of actinide speciation in solution using high-energy X-ray scattering. Anal. Bioanal. Chem. 2005, 383, 48−55. (30) Israelachvili, J.; Wennerstrom, H. Role of hydration and water structure in biological and colloidal interactions. Nature 1996, 379 (6562), 219−225. (31) Yang, H.; Kuperman, A.; Coombs, N.; Mamiche-Afara, S.; Ozin, G. A. Synthesis of oriented films of mesoporous silica on mica. Nature 1996, 379 (6567), 703−705. (32) Oelkers, E. H.; Schott, J.; Gauthier, J.-M.; Herrero-Roncal, T. An experimental study of the dissolution mechanism and rates of muscovite. Geochim. Cosmochim. Acta 2008, 72 (20), 4948−4961. (33) Sakuma, H.; Katsuyuki, K. Structure and dynamics of water on Li+-, Na+-, K+-, Cs+-, H3O+-exchanged muscovite surfaces: A molecular dynamics study. Geochim. Cosmochim. Acta 2011, 75, 63−81. (34) Schlegel, M. L.; Nagy, K. L.; Fenter, P.; Cheng, L.; Sturchio, N. C.; Jacobsen, S. D. Cation sorption on the muscovite (0 0 1) surface in chloride solutions using high-resolution X-ray reflectivity. Geochim. Cosmochim. Acta 2006, 70 (14), 3549−3565. (35) Pashley, R. M. DLVO and hydration forces between mica surfaces in Li+, Na+, K+, and Cs+ electrolyte solutions: A correlation of double-layer and hydration forces with surface cation exchange properties. J. Colloid Interface Sci. 1981, 83 (2), 531−546. (36) Pashley, R. M.; Israelachvili, J. N. DLVO and hydration forces between mica surfaces in Mg2+, Ca2+, Sr2+, and Ba2+ chloride solutions. J. Colloid Interface Sci. 1984, 97 (2), 446−455. (37) Wilson, R. E.; Schwindt, O.; Fenter, P.; Soderholm, L. Exploitation of the sorptive properties of mica for the preparation of higher-resolution alpha-spectroscopy samples. Radiochim. Acta 2010, 98 (7), 431−436. (38) Robinson, I. K.; Tweet, D. J. Surface X-ray-Diffraction. Rep. Prog. Phys. 1992, 55 (5), 599−651. (39) Fenter, P. X-ray Reflectivity as a Probe of Mineral-Fluid Interfaces: A User Guide. Rev. Mineral. Geochem. 2002, 49, 149−220.

Department of Energy Office of Science, BER, NSF, and the EPA. The X-ray reflectivity and resonant scattering data were collected at the X-ray Operations and Research beamline 6-IDB at the Advanced Photon Source (APS), Argonne National Laboratory.

■

REFERENCES

(1) Nel, A.; Xia, T.; Mädler, L.; Li, N. Toxic Potential of Materials at the Nanolevel. Science 2006, 311 (5761), 622−627. (2) Somorjai, G. A.; Borodko, Y. G. Research in nanosciences − Great opportunity for catalysis science. Catal. Lett. 2001, 76 (1−2), 1− 5. (3) Chen, M.; Goodman, D. W. Catalytically active gold on ordered titania supports. Chem. Soc. Rev. 2008, 37 (9), 1860−1870. (4) Vajda, S.; Pellin, M. J.; Greeley, J. P.; Marshall, C. L.; Curtiss, L. A.; Ballentine, G. A.; Elam, J.; Catillon-Mucherie, S.; Redfern, P. C.; Mehmood, F.; Zapol, P. Subnanometre platinum clusters as highly active and selective catalysts for the oxidative dehydrogenation of propane. Nat. Mater. 2009, 8, 213−216. (5) Li, J.; Ren, J.; Yang, G.; Wang, P.; Li, H.; Sun, X.; Chen, L.; Ma, J.-T.; Li, R. Simple and efficient deposition of Pd nanoparticles on Fe3O4 hollow nanospheres: A new catalytic system for methanol oxidation in alkaline media. Mater. Sci. Eng., B 2010, 172 (3), 207− 212. (6) Smulko, J. M.; Ederth, J.; Li, Y.; Kish, L. B.; Kennedy, M. K.; Kruis, F. E. Gas sensing by thermoelectric voltage fluctuations in SnO2 nanoparticle films. Sens. Actuators, B 2005, 106 (2), 708−712. (7) Deng, D.; Lee, J. Y. Hollow Core−Shell Mesospheres of Crystalline SnO2 Nanoparticle Aggregates for High Capacity Li+ Ion Storage. Chem. Mater. 2008, 20 (5), 1841−1846. (8) Lai, C.-Y.; Trewyn, B. G.; Jeftinija, D. M.; Jeftinija, K.; Xu, S.; Jeftinija, S.; Lin, V. S. Y. A Mesoporous Silica Nanosphere-Based Carrier System with Chemically Removable CdS Nanoparticle Caps for Stimuli-Responsive Controlled Release of Neurotransmitters and Drug Molecules. J. Am. Chem. Soc. 2003, 125 (15), 4451−4459. (9) Wang, F.; Fan, X.; Pi, D.; Wang, M. Synthesis and luminescence behavior of Eu3+-doped CaF2 nanoparticles. Solid State Commun. 2005, 133 (12), 775−779. (10) Trabattoni, S.; Moret, M.; Miozzo, L.; Campione, M. Selfassembly of gold nanoparticles on functional organic molecular crystals. J. Colloid Interface Sci. 2011, 360 (2), 422−429. (11) Peng, W.; Yanagida, M.; Han, L.; Ahmed, S. Controlled fabrication of TiO2 rutile nanorod/anatase nanoparticle composite photoanodes for dye-sensitized solar cell application. Nanotechnology 2011, 22 (27), 275709. (12) Furrer, G.; Phillips, B. L.; Ulrich, K.-U.; Pöthig, R.; Casey, W. H. The Origin of Aluminum Flocs in Polluted Streams. Science 2002, 297, 2245−2247. (13) Kersting, A. B.; Efurd, D. W.; Finnegan, D. L.; Rokop, D. J.; Smith, D. K.; Thompson, J. L. Migration of plutonium in ground water at the Nevada Test Site. Nature 1999, 397 (6714), 56−59. (14) Novikov, A. P.; Kalmykov, S. N.; Utsunomiya, S.; Ewing, R. C.; Horreard, F.; Merkulov, A.; Clark, S. B.; Tkachev, V. V.; Myasoedov, B. F. Colloid transport of plutonium in the far-field of the Mayak Production Association, Russia. Science 2006, 314 (5799), 638−641. (15) Chapman, D. L. LI. A contribution to the theory of electrocapillarity. Philos. Mag., Ser. 6 1913, 25 (148), 475−481. (16) Stern, O. Zur Theorie der Elektrolytischen Doppelschicht. Z. Elektrochem. 1924, 30 (21/22), 508−516. (17) Arai, Y.; Elzinga, E. J.; Sparks, D. L. X-ray Absorption Spectroscopic Investigation of Arsenite and Arsenate Adsorption at the Aluminum Oxide−Water Interface. J. Colloid Interface Sci. 2001, 235 (1), 80−88. (18) Fenter, P.; Rivers, M.; Sturchio, N. C.; Sutton, S. R. Applications of Synchrotron Radiation in Low-Temperature Geochemistry and Environmental Science; The Mineralogical Society of America: Washington, DC, 2002. 2626

dx.doi.org/10.1021/la2037247 | Langmuir 2012, 28, 2620−2627

Langmuir

Article

(40) Fenter, P. A.; Park, C.; Nagy, K. L.; Sturchio, N. C. Resonant anomalous X-ray reflectivity as a probe of ion adsorption at solid− liquid interfaces. Thin Solid Films 2007, 515, 5654−5659. (41) Fenter, P.; Catalano, J. G.; Park, C.; Zhang, Z. On the use of CCD area detectors for high-resolution specular X-ray reflectivity. J. Synchrotron Rad. 2006, 13 (4), 293−303. (42) Fenter, P.; Lee, S. S.; Park, C.; Soderholm, L.; Wilson, R. E.; Schwindt, O. Interaction of muscovite (0 0 1) with Pu3+ bearing solutions at pH 3 through ex-situ observations. Geochim. Cosmochim. Acta 2010, 74, 6984−6995. (43) Cheng, L.; Fenter, P.; Nagy, K. L.; Schlegel, M. L.; Sturchio, N. C. Molecular-Scale Density Oscillations in Water Adjacent to a Mica Surface. Phys. Rev. Lett. 2001, 87 (15), 156103. (44) Lee, S. S.; Nagy, K. L.; Fenter, P. Distribution of barium and fulvic acid at the mica−solution interface using in-situ X-ray reflectivity. Geochim. Cosmochim. Acta 2007, 71, 5763. (45) Lee, S. S.; Fenter, P.; Park, C.; Nagy, K. L. Fulvic Acid Sorption on Muscovite Mica as a Function of pH and Time Using In Situ X-ray Reflectivity. Langmuir 2008, 24 (15), 7817−7829. (46) Cross, J. O.; Newville, M.; Rehr, J. J.; Sorensen, L. B.; Bouldin, C. E.; Watson, G.; Gouder, T.; Lander, G. H.; Bell, M. I. Inclusion of local structure effects in theoretical x-ray resonant scattering amplitudes using ab initio x-ray-absorption spectra calculations. Phys. Rev. B 1998, 58 (17), 11215. (47) Park, C.; Fenter, P. Phasing of Resonant Anomalous X-ray Reflectivity Spectra and Direct Fourier Synthesis of Element-Specific Partial Structures at Buried Interfaces. J. Appl. Crystallogr. 2007, 40 (2), 290−301. (48) Park, C.; Fenter, P. A.; Sturchio, N. C.; Nagy, K. L. Thermodynamics, Interfacial Structure, and pH Hysteresis of Rb+ and Sr2+ Adsorption at the Muscovite (001)−Solution Interface. Langmuir 2008, 24 (24), 13993−14004. (49) Pashley, R. M. Hydration Forces between Mica Surfaces in Electrolyte-Solutions. Adv. Colloid Interface Sci. 1982, 16 (JUL), 57− 62. (50) Park, C.; Fenter, P. A.; Nagy, K. L.; Sturchio, N. C. Hydration and Distribution of Ions at the Mica-Water Interface. Phys. Rev. Lett. 2006, 97 (1), 016101. (51) Israelachvili, J. N. Intermolecular & Surface Forces, 2nd ed.; Academic Press: London, 1991. (52) Borukhov, I.; Andelman, D.; Orland, H. Effect of Polyelectrolyte Adsorption on Intercolloidal Forces. J. Phys. Chem. B 1999, 103 (24), 5042−5057. (53) Bonthuis, D. J.; Gekle, S.; Netz, R. R. Dielectric Profile of Interfacial Water and its Effect on Double-Layer Capacitance. Phys. Rev. Lett. 2011, 107 (16), 166102. (54) Saito, T.; Koopal, L. K.; Riemsdijk, W. H. v.; Nagasaki, S.; Tanaka, S. Adsorption of Humic Acid on Goethite: Isotherms, Charge Adjustments, and Potential Profiles. Langmuir 2004, 20 (3), 689−700. (55) Shklovskii, B. I. Screening of a macroion by multivalent ions: Correlation-induced inversion of charge. Phys. Rev. E 1999, 60 (5), 5802−5811. (56) Sverjensky, D. A. Interpretation and prediction of triple-layer model capacitances and the structure of the oxide-electrolyte-water interface. Geochim. Cosmochim. Acta 2001, 65 (21), 3643−3655. (57) Lee, S. S.; Park, C.; Fenter, P.; Sturchio, N. C.; Nagy, K. L. Competitive adsorption of strontium and fulvic acid at the muscovite− solution interface observed with resonant anomalous X-ray reflectivity. Geochim. Cosmochim. Acta 2010, 74 (6), 1762−1776. (58) Lee, S. S.; Nagy, K. L.; Park, C.; Fenter, P. Enhanced Uptake and Modified Distribution of Mercury(II) by Fulvic Acid on the Muscovite (001) Surface. Environ. Sci. Technol. 2009, 43, 5295−5300.

2627

dx.doi.org/10.1021/la2037247 | Langmuir 2012, 28, 2620−2627