Molecular Scale Description of Anion Competition ... - ACS Publications

Viewer
Transcript

Article pubs.acs.org/Langmuir

Molecular Scale Description of Anion Competition on AmineFunctionalized Surfaces William Rock,† Muhammed E. Oruc,‡ Ross J. Ellis,† and Ahmet Uysal*,† †

Chemical Sciences and Engineering Division, Argonne National Laboratory, Argonne, Illinois 60439, United States Department of Chemical Engineering, Yildiz Technical University, Istanbul, Turkey

‡

S Supporting Information *

ABSTRACT: Many industrial and biological processes involve the competitive adsorption of ions with diﬀerent valencies and sizes at charged surfaces; heavy and precious metal ions are separated on the basis of their propensity to adsorb onto interfaces, often as anionic ion clusters (e.g., [MClx]n−). However, very little is known, both theoretically and experimentally, about the competition of factors that drive preferential adsorption, such as charge density or valence, at interfaces in technologically relevant systems. There are even contradictory pictures described by interfacial studies and real life applications, such as chlorometalate extractions, in which charge diﬀuse chlorometalate ions are extracted eﬃciently even though charge dense chloride ions present in the background are expected to occupy the interface. We studied the competition between divalent chlorometalate anions (PtCl62− and PdCl42−) and monovalent chloride anions on positively charged aminefunctionalized surfaces using in situ specular X-ray reﬂectivity. Chloride anions were present in vast excess to simulate the conditions used in the commercial separation of heavy and precious metal ions. Our results suggest that divalent chlorometalate adsorption is a two-step process and that the divalent anions preferentially adsorb at the interface despite having a charge/volume ratio lower than that of chloride. These results provide fundamental insight into the structural mechanisms that underpin transport in phases that are relevant to heavy and precious metal ion separations, explaining the high eﬃciency of low charge density ion transport processes in the presence of charge dense anions.

1. INTRODUCTION

structure of anionic species that interact electrostatically with positively charged surfaces that are relevant to separations. Unlike the adsorption of metal cations, which is driven by dative (i.e., coordinating) interactions with ligating groups on the surface, anions are extracted via a competitive “anion exchange” mechanism. For example, the production of Pt and Pd involves the preferential extraction of the PtCl62− and PdCl42− anions over Cl−, which is usually in excess by several orders of magnitude.9 Other heavy metal ions, such as the actinides, can also be separated as the chlorometalate species when in the presence of very high chloride concentrations.10 Extraction is performed using an anion exchanger, often a quaternary amine,11 which either is tethered to a solid support or is in a water-immiscible organic liquid phase. Thus, extraction pivots on the preferential adsorption of the chlorometalate anion onto a solid−liquid or liquid−liquid interface consisting of positively charged organic groups. The eﬃciency of this process, like in all anion exchange separations, depends on the ability of the chlorometalate anion to compete eﬀectively with chloride for adsorption onto the charged interface.

The advanced energy and electronic technologies of today rely on relatively rare elements, such as lanthanides, actinides, and platinum group metals (PGMs). Most of these elements are reﬁned, recycled, and reprocessed via solvent extraction and ion exchange processes.1 Although some of these separation technologies can be considered mature, they are not adequate for the exponentially growing volumes of heavy metal production and recycling around the globe. A breakthrough in these technologies will require a molecular scale understanding of the processes beyond the classical pictures assumed from the macroscopic properties. A fundamental aspect of most separation technologies is that they involve preferential ion adsorption on,2,3 or transfer through, interfaces.4 For example, during solvent extraction, hydrated ions in an aqueous phase are transferred into an organic phase, usually with the aid of amphiphilic extractant molecules that form reverse micellar structures around the target ions to solubilize them in the organic phase.1,5,6 Very little is known, however, about the molecular scale structures that form at interfaces during ion separations, with the exception of a few recent X-ray/neutron scattering and reﬂectivity studies of interfacially adsorbed lanthanide cations at complexing amphiphiles.6−8 Even less is known about the © 2016 American Chemical Society

Received: September 21, 2016 Published: October 7, 2016 11532

DOI: 10.1021/acs.langmuir.6b03479 Langmuir 2016, 32, 11532−11539

Article

Langmuir

system can be interpreted in the context of biological, geochemical, or other systems in which competitive ion adsorption plays an important role.19,20 In particular, experimental studies of the electric double layer (EDL) structure of concentrated electrolytes are scarce, and recent studies suggest that these systems might be more complicated than generally assumed.21 X-ray reﬂectivity (XR) probes the laterally averaged interfacial structure perpendicular to the surface normal with subnanometer spatial resolution.22−24 Although XR is not element speciﬁc, the strong electron density contrast between chloride (18 e− for Cl−) and chlorometalate (116 and 182 e− for PdCl42− and PtCl62−, respectively) anions allows for reasonably accurate determination of the surface chlorometalate concentrations. Despite the relatively low bulk concentrations and charge density of chlorometalate versus chloride, interfacial adsorption of PtCl62− and PdCl42− is highly favored, accounting for the high eﬃciency of ion transport processes that target these ions. We were also able to determine the maximal metalate coverage, providing important insights into the energy−entropy balance during the competition.

The very fast kinetics and high eﬃciencies of extraction of the chlorometalate from highly concentrated chloride aqueous phases imply that these charge diﬀuse ions are preferentially adsorbed at positively charged interfaces compared to chloride. Singly charged chlorometalate anions, such as AuCl4−, are even more charge diﬀuse and thus even more preferentially extracted.9 This relationship between charge density and extraction has been known in the separations community for decades and underpins many anion separations processes.12 However, from an interfacial perspective, this observation presents a dichotomy as ions with high charge density, such as chloride, are expected to interact more strongly with charged surfaces than ions with low charge density, such as chlorometalate.13−15 For instance, both theory and simulations suggest that the charge/volume ratio plays an important role in the competition of ions with diﬀerent valencies, particularly in the high surface charge regime. This ratio roughly expresses the combined eﬀects of electrostatic energy and entropy at the interface. Higher valence ions are expected to preferentially adsorb at the interface because of the electrostatic forces, and the entropic eﬀects favor the adsorption of smaller ions. Although in theoretical and computational studies (under ideal conditions) a larger charge/volume ratio leads to stronger adsorption,13,14 in reality, hydration, van der Waals interactions, and nonideal surface properties should also play signiﬁcant roles.16,17 For instance, the preferential extraction of PtCl62− (2/139 = 0.0144 e−/Å3) and PdCl42− (2/106 = 0.0189 e−/Å3) over Cl− (1/24 = 0.0417 e−/Å3),9 whose charge/volume ratio is more than twice those of the metalate anions, suggests that the charge/volume ratio does not always determine preferential adsorption (volumes are calculated using atomic radii). It is important to note that the extraction eﬃciency is ultimately related to the ﬁnal structures in the organic and aqueous phases, rather than the interfacial complexes.5,18 However, eﬃcient extraction is unlikely to occur if Cl− blocks the interfacial adsorption of chlorometalate anions. Clarifying this apparent contradiction between the interfacial picture of the competitive adsorption of ions at a charged interface and chlorometalate extraction improves our understanding of these complex interfaces signiﬁcantly, particularly those used in the separation of heavy and precious metal ions. With a goal of addressing this dichotomy, we probe the structure at the interface between an amine-functionalized silicon substrate and an aqueous solution having high concentrations of chloride ions and minute concentrations of platinum group chlorometalate ions (reﬂecting the conditions in separations processes). Studying a model system at the solid−liquid interface to understand some aspects of an extraction process at the liquid−liquid interface has two advantages. First, the kinetics of these separation processes can be very fast, making it very diﬃcult to catch observable structures at the liquid−liquid interface on reasonable time scales, unless some “arrested” metastable structures are formed.7 Second, a real extraction system involves too many complex interactions, such as ion−extractant interactions, extractant−extractant interactions, and interactions with organic and aqueous solvents. Therefore, it is usually not possible to completely isolate a certain eﬀect. Studying ﬁxed amine surfactants isolates the interfacial amine−anion interactions, allowing direct observation of the competition between the anions that is independent of other eﬀects. The results of this study provide a benchmark for improving our understanding of more complex systems. Also, this simple model

2. EXPERIMENTAL SECTION All chemicals were purchased from Sigma and used without further puriﬁcation. Silicon samples were cleaned in a piranha solution (7/3 sulfuric acid/hydrogen peroxide) at 120 °C for 30 min. Then they were rinsed with deionized (DI) water three times for 5 min each and ﬁnally dried with nitrogen. (3-Aminopropyl)triethoxysilane (APTES) solutions were freshly prepared in anhydrous toluene [∼1.0% (w/w)]. Clean silicon substrates were placed in these solutions for 1 h at room temperature. To prevent evaporation of the solvent during the reaction, the top of the Petri dish was closed. After surface functionalization, the samples were rinsed with excess toluene and dried with nitrogen. Chlorometalate solutions were prepared from palladium(II) chloride and hexachloroplatinic acid by dissolving appropriate amounts in background solutions (5 M LiCl or 3 mM NaCl at pH 1 as described in the text). The pH of the solutions was controlled with 1 M HCl. In addition to simulating industrial conditions,9 these concentrations also ensure that the chlorometalate speciation is ﬁxed at PtCl62− and PdCl42−. The XR measurements were taken at beamlines 33-IDD and 12IDD of the Advanced Photon Source (APS) at Argonne National Laboratory. A thin ﬁlm sample cell25 (Figure 1) and PILATUS detector were mounted on a six-circle diﬀractometer (a modiﬁed Huber psi-C diﬀractometer at 12-IDD and a Newport Kappa diﬀractometer at 33-IDD). The incident beam size was focused to

Figure 1. Depiction of the thin ﬁlm sample cell and X-ray scattering geometry. The Kel-f sample cell holds the silicon substrates with the help of a kapton ﬁlm (yellow lines) secured by an O-ring and compression from the sides (not shown) to prevent ﬂuid leakage. When the ﬂuid is injected, the kapton ﬁlm puﬀs up, allowing quick concentration equilibration. Immediately before the measurement, the kapton ﬁlm was collapsed to leave a liquid ﬁlm (a few micrometers thick) over the substrate. 11533

DOI: 10.1021/acs.langmuir.6b03479 Langmuir 2016, 32, 11532−11539

Article

Langmuir ∼80 μm vertically. The horizontal size was set to 0.5 mm by a pair of slits. The specular reﬂectivity signal was recorded as a function of the vertical momentum transfer |q ⃗| = (4π/λ) sin(2θ/2), where λ is the wavelength of the X-rays (0.60 Å at 33ID-D and 0.73 Å at 12ID-D) and 2θ is the scattering angle. The beam damage due to the X-rays was identiﬁed carefully by repetitive measurements on the same spot, and the ﬁnal measurements were limited to shorter time intervals. Samples were shifted perpendicular to the beam to a fresh spot between measurements. The X-ray data were analyzed using a box model, which assumes laterally homogeneous layers with interfaces in the form of error functions.23,26,27 Density, thickness, and roughness parameters of each layer in the model were obtained by least-squares ﬁtting of the calculated XR curves (according to the Parratt formalism)23,28 to the collected data.

solution (Figure 2c). However, a comparable measurement with PdCl4 was not taken because the speciation of PdCl4 cannot be ﬁxed to PdCl42− at low Cl− concentrations.29 The concentration-dependent measurements were taken on a single APTES ﬁlm for each metalate. All experiments begin with the background solution and progress from low to high chlorometalate concentrations. After each measurement, the sample chamber is thoroughly ﬂushed, by puﬃng up the kapton ﬁlm, with the new solution and allowed to equilibrate for at least 10 min, and the sample is translated to a fresh spot. After the solution is fully exchanged, interfacial equilibration is expected to occur in less than a few minutes.30 This provides a reliable background subtraction to identify the eﬀects of increasing the bulk metalate concentration on the interfacial concentration. The intensity of the XR signal (R) decreases in a manner proportional to q−4 even at an ideal ﬂat surface; this is known as the Fresnel reﬂectivity (RF).23,25 Normalizing R to RF enhances visualization of the Kiessig fringes (the minima and maxima in XR data due to the electron density gradients at the interface).23 Before discussing the results obtained from the modeldependent ﬁts, we investigate the qualitative behavior of the XR signal in panels a and b of Figure 2 as a function of increasing metalate concentration. Two major trends can be observed in both Pd and Pt data. The minima in the XR curves shift to higher q values in 0.5 mM metalate solutions (blue squares) compared to those in the background (black circles). However, as the concentration increases to 5 and 50 mM (green diamonds and red triangles, respectively), the minima move in the opposite direction (to smaller q values). Because q represents reciprocal space, these results suggest that the interfacial structure (APTES and ions) becomes thinner compared to the background solution at 0.5 mM metalate, counterintuitively implying desorption of ions. However, as the metalate concentration increases further, the thickness increases as expected. Also, the amplitude of the Kiessig fringes increases signiﬁcantly in the 5 and 50 mM solutions, implying that the density contrast at the interface is also increasing. 3.1. APTES in 5 M LiCl. We ﬁrst discuss APTES ﬁlms in background solutions without any metalate ions. Figure 3 shows the EDP of the APTES ﬁlm structure obtained from ﬁtting the top reﬂectivity data set in Figure 2a. The ﬁlms used

3. RESULTS AND DISCUSSION We measured the XR at the APTES−electrolyte interface for concentrations of 0.5, 5, and 50 mM PdCl42− and PtCl62− in 5 M LiCl as shown in panels a and b of Figure 2, respectively. Before injecting the metalate anions, we measured the XR with the 5 M LiCl background solution (black circles in each plot). We also studied 0.5 mM PtCl62− in a 3 mM NaCl background

Figure 2. Specular XR intensity normalized to Fresnel reﬂectivity as a function of the vertical momentum transfer (q) from aqueous−APTES interfaces at diﬀerent bulk metalate concentrations in (a and b) 5 M LiCl and (c) 3 mM NaCl solutions. The solid black lines show the ﬁts to the data (symbols) as discussed in the text.

Figure 3. Electron density proﬁle (EDP) of APTES on silicon in 5 M LiCl (solid red line). The green solid lines show the same EDP without interfacial roughness. The bottom left inset is the molecular model of APTES. The superimposed cartoon depiction of the APTES ﬁlm structure with Cl− adsorption is deduced from the EDP. 11534

DOI: 10.1021/acs.langmuir.6b03479 Langmuir 2016, 32, 11532−11539

Article

Langmuir Table 1. Best Fit Parameters for the XR Data and Fits Shown in Figure 2a ﬁrst APTES layer

silicon oxide PdCl4

2−

−

in 5 M CI

PtCl62− in 5 M CI−

PtCl62− in 3 mM CI− error (±)

0 0.5 mM 5 mM 50 raM 0 0.5 mM 5 raM 50 mM 0 0.5 mM

second APTES layer

adsorption layer

ρ (e−/Å3)

L (Å)

σ (Å)

ρ (e−/Å3)

L (Å)

σ (Å)

ρ (e−/Å3)

L (Å)

σ (Å)

ρ (e−/Å3)

L (Å)

σ (Å)

0.68 0.68a 0.68a 0.68a 0.68 0.68a 0.68a 0.68a 0.69 0.69a 0.001

12.9 12.9a 12.9a 12.9a 12.4 12.4a 12.4a 12.4a 10.3 10.3a 0.04

2.2 2.2a 2.2a 2.2a 1.8 1.8a 1.8a 1.8a 2.2 2.2a 0.4

0.48 0.48 0.48 0.51 0.57 0.57 0.57 0.58 0.49 0.49 0.01

6.4 6.4a 6.4a 6.4a 6.0 6.0a 6.0a 6.0a 5.0 5.0a 0.1

1.9 1.9 1.8 3.5 2.7 2.7 2.6 2.6 3.6 3.6 0.7

0.42 0.42 0.43 0.44 0.46 0.45 0.45 0.47 0.41 0.54 0.01

6.1 6.1a 6.1a 6.1a 6.0 6.0a 6.0a 6.0a 5.6 5.6a 0.9

1.7 1.7 1.7 2.6 4.3 4.4 4.7 4.9 3.6 1.8 0.7

0.43 0.46 0.46 0.41 0.46 0.45 0.55 0.58 0.46 0.57 0.01

3.0 1.0 1.3 4.3 3.1 2.7 1.9 2.5 1.6 1.3 0.9

3.4 2.9 3.0 1.8 4.3 3.5 4.1 4.5 3.4 4.7 0.8

The silicon electron density and the Si−SiO interface roughness are ﬁxed to 0.71 e−/Å3 and 1.5 Å, respectively, for all samples. The superphase density is ﬁxed to 0.36 e−/Å3 for the 5 M LiCl background solutions and 0.33 e−/Å3 for the 3 mM NaCl background solution. All the parameters were allowed to ﬂoat within reasonable limits for zero metalate concentration background samples. Then, the silicon oxide parameters and APTES layer thicknesses are ﬁxed (marked with a superscript a) for the measurements with metalate ions. The uncertainties shown in the bottom row are estimated from the least-squares ﬁtting of the background solutions; they are very similar and usually smaller for other ﬁts in which some of the parameters are ﬁxed. a

quasi-bilayer formation.34 We modeled the ﬁlm with two separate boxes and allowed the parameters to vary freely. The ﬁts show that the second layer is ∼10−20% less dense than the ﬁrst. A multilayer surfactant is representative of interfaces in commercial liquid−liquid extraction systems,8 where atomically ﬂat monolayers are unusual. A separate layer was assigned for the adsorbed Cl− ions. As shown in Figure 3, this layer has almost the same density as the previous layer and is only ∼3 Å thick. Indeed, it is possible to obtain an identical EDP using a single box (with an ∼8 Å thickness) for the top APTES layer and the Cl−.35 However, this extra slab allows an easier comparison between diﬀerent samples with varying metalate concentrations. Nevertheless, all the results presented here can be reproduced with three-slab models with no qualitative diﬀerences (see the Supporting Information for details). The APTES coverage, calculated from the EDP in Figure 3, is ∼24 Å2/molecule in the ﬁrst layer and ∼29 Å2/molecule in the top layer. The same analysis of the APTES ﬁlm used for PtCl62− adsorption (Figure 2b) gives values of ∼22 and ∼27 Å2/molecule for the ﬁrst and second APTES layer, respectively. The APTES ﬁlm used in the NaCl background solution, which was prepared on a silicon substrate from a diﬀerent batch during an earlier beamtime, has coverages of ∼30 and ∼32 Å2/ molecule. Because the APTES ﬁlms are prepared under ambient conditions, diﬀerent relative humidity levels during the preparation are likely responsible for the slight diﬀerences in the ﬁlms.36,37 These coverages are used to estimate the surface charge, assuming that all the amine groups at the top surface are protonated, and therefore positively charged in the highly acidic (pH 1) solution. 3.2. Chloride Desorption from the Surface with Metalate Injection. When 0.5 mM metalate is injected into the 5 M LiCl background solution, the minima in the reﬂectivity signal (Figure 2, blue squares) shift toward higher values of q, suggesting a decrease in the thickness of the interfacial layer. A comparison of the EDPs calculated by ﬁtting the XR data (Figure 4) shows that this is due to a decrease in the electron density of the slab assigned to the adsorbed anions. This decrease in the EDP must be caused by a net decrease in the number of adsorbed ions or by thinning of the APTES

with other samples also have the same qualitative EDP with only slight diﬀerences in the densities and thicknesses (Table 1 and Figure 4). The ∼6 Å lengths of each APTES layer match

Figure 4. Comparison of EDPs of APTES on silicon in 5 M LiCl before (black dashed lines) and after (blue solid lines) addition of 0.5 mM metalate calculated by ﬁtting the XR data in Figure 2. The PdCl42− curve is moved up by 0.1 for the sake of clarity. The top right inset shows the diﬀerential EDPs for both plots generated by the diﬀerence between the dashed black and solid blue curves. They are almost identical for both metalates, within the errors of the measurement (Figure SI-4). The inset schematic describes a possible explanation for the decrease in the EDP due to the metalate anions adsorbed in the diﬀuse layer disturbing the chloride anions in the Stern layer.

well with the size of the molecule calculated by the MMFF94 force ﬁeld (Figure 3, inset).31 Contrary to most well-known self-assembly processes, such as octadecyltrichlorosilane (OTS),32,33 APTES self-assembly is not self-limiting after the formation of the ﬁrst layer; the APTES Si has three Si−O groups that can attach to the surface or self-polymerize, and the hydrocarbon chain is too short to block the formation of APTES−APTES bonds. Therefore, the thickness of the ﬁlm increases with reaction time (Figure 3).34 Our synthesis parameters matched well with previous reports that show 11535

DOI: 10.1021/acs.langmuir.6b03479 Langmuir 2016, 32, 11532−11539

Article

Langmuir ﬁlm. The injection could dislodge physisorbed APTES molecules; this is unlikely because the ﬁlm is thoroughly washed during preparation and ﬁrst injected with the background solution. The presence of metalate ions could rearrange the APTES layer. This could change the shape of the EDP but would conserve the total amount of interfacial electrons. The data show a net decrease in the number of interfacial electrons, making this scenario unlikely. Finally, the metalate ions could break some of the Si−O bonds, thinning the ﬁlm; this is also unlikely, especially because this strong chemical interaction would need to occur without enough chlorometalate adsorption to increase the electron density. Therefore, it is most likely that these changes arise from desorption of adsorbed ions. Interestingly, the subtraction of the background EDP from the 0.5 mM EDP gives a very similar diﬀerential EDP for both metalate anions (Figure 4, inset), suggesting that the eﬀect is approximately the same for both PdCl4 and PtCl6: Cl− anions desorb from the surface and are not replaced by metalate anions. This behavior is reﬂective of the identical bulk phase liquid−liquid extraction equilibria for PdCl42− and PtCl62−:9 PdCl4 2 − + 2[Cl.HA] = [PdCl4.HA 2] + 2Cl−

PtCl 6 2 − + 2[Cl.HA] = [PtCl 6.HA 2] + 2Cl− Figure 5. (a) EDPs of chlorometalate-adsorbed APTES ﬁlms derived from the ﬁts to the XR data in panels a−c Figure 2 from top to bottom, respectively. The dashed lines correspond to the EDPs in the appropriate background solutions. The top and bottom curves are shifted +0.1 and −0.1, respectively, for the sake of clarity. (b) Diﬀerential EDPs for [PdCl42−] adsorption, calculated by subtracting the dashed blue line from the green and red curves for PdCl42−/Cl− relative concentrations of 10−3 and 10−2, respectively. (c) Diﬀerential EDPs for PtCl62− adsorption, calculated by subtracting the dashed lines in the bottom two proﬁles from the green, red, and magenta curves for PtCl62−/Cl− relative concentrations of 10−3, 10−2, and 1.7 × 10−1, respectively.

where A is the amine extractant. Why do Cl− ions leave the surface after the introduction of the metalate? Most likely, metalate anions are attracted to the surface and disturb the Cl− ions in the Stern layer. However, metalate anions cannot be adsorbed, possibly because of their hydration (Figure 4, inset schematic). Because XR is sensitive to the gradient in the EDP, the few metalates that are not adsorbed exactly at the surface (and therefore cannot form a well-deﬁned layer) do not create enough contrast to be directly measured. Assuming charge compensation, this is the simplest explanation. Also, the almost identical diﬀerential EDPs for both metalates (Figure 4, inset) suggest that there is no signiﬁcant metalate adsorption on the Stern layer, which would result in observable diﬀerences due to their dissimilar electron densities. It is known that, on negatively charged mineral surfaces, divalent cations adsorb as outer sphere and extended outer sphere complexes with little eﬀect on the total electron density proﬁle.38 Hence, it is not surprising that divalent anions behave similarly at a positive interface. In this picture, ∼4% of the Cl− ions leave the surface at a bulk metalate concentration of 0.5 mM; it is also assumed that for every two chloride ions leaving the Stern layer, one metalate is adsorbed in the diﬀuse layer. However, in liquid−liquid anion extraction systems using amine/ammonium extractants, partial or complete dehydration of the anion is necessary to complete transport from one phase to another via the interface. Therefore, under conditions where ion extraction is known to proceed rapidly (i.e., at higher metalate concentrations that reﬂect process conditions), there should be evidence of penetration of the anion from the diﬀuse layer into the Stern layer. 3.3. Chlorometalate Adsorption at the Stern Layer. Increasing the metalate concentration in the bulk to 5 and 50 mM causes the minima in the XR data to shift to smaller values of q and increases the Kiessig fringe amplitudes (Figure 2, green diamonds and red triangles); this is direct evidence of the adsorption of metalate anions on the Stern layer. Figure 5 shows the electron density proﬁles derived for these higher relative metalate concentrations, including the measurement in

a 3 mM NaCl background (Figure 2c), which has a relative metalate concentration of 0.17. The diﬀerential EDPs (Figure 5b,c) provide important insights into the adsorption of metalate anions as a function of the relative metalate concentration. The integration of each curve gives the change in the electron density per unit area at the interface; this can be converted to the coverage of metalate anions by using the number of electrons they have. We can normalize each of the coverages to the available number of sites obtained from the APTES molecular areas, assuming full coverage (θ = 1) when two APTES molecules are occupied by one metalate (Figure 6). It is important to note that XR provides EDPs that are convoluted with the experimental resolution (π/qmax ∼ 4 Å in these experiments).35 Therefore, the diﬀerential EDPs are too broad to determine the exact position of the ions. However, the area under the EDP is a more robust parameter.27 Figure 6 summarizes all the results obtained from the XR experiments. The ﬁrst remarkable result is that, after calculating the number densities and normalizing with respect to the surface charge, there is no signiﬁcant diﬀerence between PdCl42− and PtCl62− adsorption at the interface over a relative concentration range of three decades. The charge/volume ratios of PdCl42− and PtCl62− are 45 and 35% Cl−, respectively. PdCl42− has a planar shape, while PtCl62− is octahedral. Apparently, these factors do not create a signiﬁcant diﬀerence 11536

DOI: 10.1021/acs.langmuir.6b03479 Langmuir 2016, 32, 11532−11539

Article

Langmuir

diﬃcult to deconvolute from diﬀusion-controlled processes. However, with the advent of new microﬂuidic techniques that couple rapid mixing with well-deﬁned interfacial areas,44 extraction studies should now be able to access absolute interfacial kinetic values in these rapidly evolving systems. Such studies would be invaluable for comparison with the surface structures presented herein.

4. CONCLUSIONS We studied the competitive adsorption of platinum and palladium chlorometalate anions on amine-functionalized surfaces in highly concentrated chloride solutions; the conditions were reﬂective of commercial heavy and precious metal ion extractions. XR reﬂectivity data clearly show that the metalate anions are able to adsorb to the surface even at relative metalate/chloride concentrations of 10−3 in the bulk, suggesting that energetics due to Coulombic interactions play an important role in this regime and favor divalent over monovalent anions regardless of the charge/volume ratio; this is contrary to the expectations from MC simulations and mean ﬁeld theories.13,14 Hydration of the ions may be mainly responsible for this diﬀerence between experiment and theory. Nevertheless, this is consistent with the well-known very rapid extraction kinetics for these metalate species. At higher relative concentrations, metalate anion coverage levels oﬀ after occupying one-third of the available sites, indicating that entropic eﬀects (mainly driven by the smaller volume of chloride ions)14 play a critical role in determining the maximal coverage. These results show that at low relative concentrations, a low charge/volume ratio does not prevent metalates from being adsorbed to the interface, explaining their eﬃcient liquid/liquid extraction. On the other hand, entropic eﬀects keep the maximal possible coverage low at high metalate concentrations. We can speculate that in a solvent extraction process, where ions are transferred into an organic phase from the aqueous phase by forming reverse micellar structures (contrary to forming a static interfacial structure as in our experiment),18 limited surface coverage may help prevent third-phase formation that occurs after high loading of metals at the interface. Such a relation could explain the inversely proportional relation between the charge/volume ratio and the extraction eﬃciency in chlorometalate anions. Both platinum and palladium chlorometalate adsorption have two regimes. At very low metalate concentrations, they are adsorbed on the diﬀuse layer without forming a well-deﬁned structure and cause an apparent decrease in the interfacial electron density. As the concentration increases, metalate ions are adsorbed at the interface preferentially over chloride, suggesting that the electrostatic interactions dominate for most technologically relevant concentrations. At very high metalate/ chloride relative concentrations, entropic eﬀects become important and limit the maximal coverage of the metalates. These observations will help to develop a better understanding of interfacial processes that include the competitive adsorption of ions with diﬀerent charges and sizes.

Figure 6. Plot of metalate coverage at the interface as a function of the relative metalate concentration in the bulk. The data points for the lowest bulk relative concentration (10−4) are estimated from Figure 4 according to the model described in the text. All the other data points are calculated by integrating the diﬀerential EDPs in panels b and c of Figure 5. The error bars do not represent the error in the integration (much smaller) but show the approximate range of surface concentrations based on the model used to estimate the metalate coverages. The black solid line represents the metalate coverage corresponding to the same relative concentration on the surface and in the bulk. The overlap of [PtCl62−] and [PdCl42−] results suggests that the diﬀerences in their geometries do not play a signiﬁcant role in the competition with chloride.

in their competition against chloride at the APTES−water interface. This is a surprising result because computer simulations suggest that PdCl42− and PtCl62− have diﬀerent hydration shells,39,40 which are expected to aﬀect interfacial interactions,41−43 due to the diﬀerences in their molecular geometries. This multitude of seemingly contradictory phenomena underscores the importance of further research into the molecular-level interactions of highly concentrated electrolyte solutions at charged interfaces. The black solid line in Figure 6 represents the case in which the relative metalate concentrations at the surface and in the bulk are the same. The results suggest that metalate adsorption at the interface is signiﬁcantly favored at a metalate/chloride ratio of ≤0.1. At a concentration ratio of 0.17, the measured surface concentration is almost the same as the bulk concentration. This strongly suggests that the metalate coverage saturates signiﬁcantly below the theoretical limit of 2NH3+ + M2−, which assumes the complexation of two amine groups with one metalate. The data were ﬁt with a Langmuir isotherm (blue dashed lines), which shows that only approximately one-third of the available sites are occupied by metalate anions at high metalate concentrations. The ﬁt does not provide absolute adsorption free energy because we use relative concentrations without any activity coeﬃcients. However, it helps to qualitatively visualize the fact that the interfacial metalate concentration starts to become saturated at high relative concentrations. The preferential adsorption of the metalate anions onto the ammonium surface is qualitatively reﬂective of their rapid kinetics and high extraction eﬃciency seen in analogous liquid− liquid extraction systems.9 However, there is a great paucity of quantitative kinetic data for these systems that can provide insight into the absolute interfacial mechanisms associated with the extraction of Pt and Pd under these conditions. This is in part due to the rapid kinetics, making interfacial reactions

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.langmuir.6b03479. 11537

DOI: 10.1021/acs.langmuir.6b03479 Langmuir 2016, 32, 11532−11539

Article

Langmuir

■

Details of the models used to ﬁt XR data and atomic force microscopy (AFM) characterization of APTES ﬁlms (PDF)

A.; Kang, S. O.; Bowman-James, K. Supramolecular chemistry of environmentally relevant anions. Adv. Inorg. Chem. 2006, 59, 175. (13) Zhou, S.; Wang, Z.; Li, B. Mean-field description of ionic size effects with nonuniform ionic sizes: A numerical approach. Phys. Rev. E 2011, 84 (2), 021901. (14) Wen, J.; Zhou, S.; Xu, Z.; Li, B. Competitive adsorption and ordered packing of counterions near highly charged surfaces: From mean-field theory to Monte Carlo simulations. Phys. Rev. E 2012, 85, 041406. (15) Taboada-Serrano, P.; Yiacoumi, S.; Tsouris, C. Behavior of mixtures of symmetric and asymmetric electrolytes near discretely charged planar surfaces: A Monte Carlo study. J. Chem. Phys. 2005, 123 (5), 054703. (16) Lee, S. S.; Schmidt, M.; Laanait, N.; Sturchio, N. C.; Fenter, P. Investigation of Structure, Adsorption Free Energy, and Overcharging Behavior of Trivalent Yttrium Adsorbed at the Muscovite (001)− Water Interface. J. Phys. Chem. C 2013, 117 (45), 23738−23749. (17) Lee, S. S.; Fenter, P.; Nagy, K. L.; Sturchio, N. C. Changes in adsorption free energy and speciation during competitive adsorption between monovalent cations at the muscovite (001)-water interface. Geochim. Cosmochim. Acta 2013, 123, 416−426. (18) Ellis, R. J.; Meridiano, Y.; Muller, J.; Berthon, L.; Guilbaud, P.; Zorz, N.; Antonio, M. R.; Demars, T.; Zemb, T. ComplexationInduced Supramolecular Assembly Drives Metal-Ion Extraction. Chem. - Eur. J. 2014, 20 (40), 12796−12807. (19) Benjamin, I. Reaction Dynamics at Liquid Interfaces. Annu. Rev. Phys. Chem. 2015, 66, 165−188. (20) Fenter, P.; Lee, S. S. Hydration layer structure at solid−water interfaces. MRS Bull. 2014, 39 (12), 1056−1061. (21) Smith, A. M.; Lee, A. A.; Perkin, S. The Electrostatic Screening Length in Concentrated Electrolytes Increases with Concentration. J. Phys. Chem. Lett. 2016, 7 (12), 2157−2163. (22) Als-Nielsen, J.; Jacquemain, D.; Kjaer, K.; Leveiller, F.; Lahav, M.; Leiserowitz, L. Principles and applications of grazing incidence xray and neutron scattering from ordered molecular monolayers at the air-water interface. Phys. Rep. 1994, 246 (5), 251−313. (23) Pershan, P. S.; Schlossman, M. Liquid Surfaces and Interfaces: Synchrotron X-ray Methods; Cambridge University Press: Cambridge, U.K., 2012. (24) Vaknin, D. X-ray Diﬀraction and Spectroscopic Techniques for Liquid Surfaces and Interfaces; Wiley Online Library: New York, 2012. (25) Fenter, P. A. X-ray reflectivity as a probe of mineral-fluid interfaces: A user guide. Rev. Mineral. Geochem. 2002, 49 (1), 149− 221. (26) Danauskas, S. M.; Li, D.; Meron, M.; Lin, B.; Lee, K. Y. C. Stochastic fitting of specular X-ray reflectivity data using StochFit. J. Appl. Crystallogr. 2008, 41 (6), 1187−1193. (27) Uysal, A.; Chu, M.; Stripe, B.; Timalsina, A.; Chattopadhyay, S.; Schlepütz, C. M.; Marks, T. J.; Dutta, P. What x rays can tell us about the interfacial profile of water near hydrophobic surfaces. Phys. Rev. B: Condens. Matter Mater. Phys. 2013, 88 (3), 035431. (28) Parratt, L. G. Surface Studies of Solids by Total Reflection of Xrays. Phys. Rev. 1954, 95 (2), 359−369. (29) Tait, C. D.; Janecky, D. R.; Rogers, P. S. Z. Speciation of aqueous palladium(II) chloride solutions using optical spectroscopies. Geochim. Cosmochim. Acta 1991, 55 (5), 1253−1264. (30) Lee, S. S.; Fenter, P.; Park, C. Optimizing a flow-through X-ray transmission cell for studies of temporal and spatial variations of ion distributions at mineral-water interfaces. J. Synchrotron Radiat. 2013, 20 (1), 125−136. (31) Hanwell, M. D.; Curtis, D. E.; Lonie, D. C.; Vandermeersch, T.; Zurek, E.; Hutchison, G. R. Avogadro: an advanced semantic chemical editor, visualization, and analysis platform. J. Cheminf. 2012, 4 (1), 17. (32) Steinrück, H.-G.; Will, J.; Magerl, A.; Ocko, B. M. Structure of nAlkyltrichlorosilane Monolayers on Si (100)/SiO2. Langmuir 2015, 31 (43), 11774−11780. (33) Richter, A.; Durbin, M.; Yu, C.-J.; Dutta, P. In situ time-resolved X-ray reflectivity study of self-assembly from solution. Langmuir 1998, 14 (21), 5980−5983.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing ﬁnancial interest.

■

ACKNOWLEDGMENTS XR experiments were conducted at Sector 12-IDD and Sector 33-IDD of the Advanced Photon Source at Argonne National Laboratory. This work and the use of the Advanced Photon Source are supported by the U.S. Department of Energy, Oﬃce of Science, Oﬃce of Basic Energy Sciences, Division of Chemical Sciences, Biosciences and Geosciences, under Contract DE-AC02-06CH11357. We thank Lynda Soderholm for her comments on the manuscript. We also thank Paul Fenter and Sang Soo Lee for fruitful discussions and access to their atomic force microscope.

■

REFERENCES

(1) Tasker, P. A.; Plieger, P. G.; West, L. C. Metal complexes for hydrometallurgy and extraction. Compr. Coord. Chem. II 2003, 759. (2) Jal, P. K.; Patel, S.; Mishra, B. K. Chemical modification of silica surface by immobilization of functional groups for extractive concentration of metal ions. Talanta 2004, 62 (5), 1005−1028. (3) Demir, S.; Brune, N. K.; Van Humbeck, J. F.; Mason, J. A.; Plakhova, T. V.; Wang, S.; Tian, G.; Minasian, S. G.; Tyliszczak, T.; Yaita, T.; Kobayashi, T.; Kalmykov, S. N.; Shiwaku, H.; Shuh, D. K.; Long, J. R. Extraction of Lanthanide and Actinide Ions from Aqueous Mixtures Using a Carboxylic Acid-Functionalized Porous Aromatic Framework. ACS Cent. Sci. 2016, 2 (4), 253. (4) Tavlarides, L. L.; Bae, J. H.; Lee, C. K. Solvent Extraction, Membranes, and Ion Exchange in Hydrometallurgical Dilute Metals Separation. Sep. Sci. Technol. 1987, 22 (2−3), 581. (5) Ellis, R. J.; Audras, M.; Antonio, M. R. Mesoscopic Aspects of Phase Transitions in a Solvent Extraction System. Langmuir 2012, 28 (44), 15498−15504. (6) Bu, W.; Mihaylov, M.; Amoanu, D.; Lin, B.; Meron, M.; Kuzmenko, I.; Soderholm, L.; Schlossman, M. L. X-ray Studies of Interfacial Strontium−Extractant Complexes in a Model Solvent Extraction System. J. Phys. Chem. B 2014, 118 (43), 12486−12500. (7) Bu, W.; Yu, H.; Luo, G.; Bera, M. K.; Hou, B.; Schuman, A. W.; Lin, B.; Meron, M.; Kuzmenko, I.; Antonio, M. R.; Soderholm, L.; Schlossman, M. L. Observation of a Rare Earth Ion−Extractant Complex Arrested at the Oil−Water Interface During Solvent Extraction. J. Phys. Chem. B 2014, 118 (36), 10662−10674. (8) Scoppola, E.; Watkins, E.; Li Destri, G.; Porcar, L.; Campbell, R.; Konovalov, O.; Fragneto, G.; Diat, O. Structure of a liquid/liquid interface during solvent extraction combining X-ray and neutron reflectivity measurements. Phys. Chem. Chem. Phys. 2015, 17 (23), 15093−15097. (9) Bernardis, F. L.; Grant, R. A.; Sherrington, D. C. A review of methods of separation of the platinum-group metals through their chloro-complexes. React. Funct. Polym. 2005, 65 (3), 205−217. (10) Baybarz, R.; Weaver, B.; Ivinser, H. Isolation of transplutonium elements by tertiary amine extraction. Nucl. Sci. Eng. (La Grange Park, IL, U.S.) 1963, 17 (3), 457−462. (11) Kumar, V.; Jha, M. K.; Jeong, J.; Gupta, D.; Lee, J.-c. Solvent extraction of platinum using amine based extractants in different solutions: A review. Hydrometallurgy 2014, 142, 60−69. (12) Moyer, B. A.; Delmau, L. H.; Fowler, C. J.; Ruas, A.; Bostick, D. A.; Sessler, J. L.; Katayev, E.; Dan Pantos, G.; Llinares, J. M.; Hossain, 11538

DOI: 10.1021/acs.langmuir.6b03479 Langmuir 2016, 32, 11532−11539

Article

Langmuir (34) Argekar, S. U.; Kirley, T. L.; Schaefer, D. W. Determination of structure-property relationships for 3-aminopropyltriethoxysilane films using x-ray reflectivity. J. Mater. Res. 2013, 28 (08), 1118−1128. (35) Bu, W.; Vaknin, D.; Travesset, A. How accurate is PoissonBoltzmann theory for monovalent ions near highly charged interfaces? Langmuir 2006, 22 (13), 5673−5681. (36) Walter, S. R.; Youn, J.; Emery, J. D.; Kewalramani, S.; Hennek, J. W.; Bedzyk, M. J.; Facchetti, A.; Marks, T. J.; Geiger, F. M. In-Situ Probe of Gate Dielectric-Semiconductor Interfacial Order in Organic Transistors: Origin and Control of Large Performance Sensitivities. J. Am. Chem. Soc. 2012, 134 (28), 11726−11733. (37) Zhu, M.; Lerum, M. Z.; Chen, W. How To Prepare Reproducible, Homogeneous, and Hydrolytically Stable AminosilaneDerived Layers on Silica. Langmuir 2012, 28 (1), 416−423. (38) Park, C.; Nagy, K. L.; Fenter, P. A.; Sturchio, N. C. Hydration and Distribution of Ions at the Mica-Water Interface. Phys. Rev. Lett. 2006, 97 (1), 016101. (39) Naidoo, K. J.; Klatt, G.; Koch, K. R.; Robinson, D. J. Geometric Hydration Shells for Anionic Platinum Group Metal Chloro Complexes. Inorg. Chem. 2002, 41 (7), 1845−1849. (40) Truflandier, L. A.; Autschbach, J. Probing the Solvent Shell with 195Pt Chemical Shifts: Density Functional Theory Molecular Dynamics Study of PtII and PtIV Anionic Complexes in Aqueous Solution. J. Am. Chem. Soc. 2010, 132 (10), 3472−3483. (41) Clark, A. E.; Samuels, A.; Wisuri, K.; Landstrom, S.; Saul, T. Sensitivity of Solvation Environment to Oxidation State and Position in the Early Actinide Period. Inorg. Chem. 2015, 54 (13), 6216−6225. (42) Dinescu, A.; Clark, A. E. Thermodynamic and Structural Features of Aqueous Ce(III). J. Phys. Chem. A 2008, 112 (44), 11198− 11206. (43) D’Angelo, P.; Spezia, R. Hydration of Lanthanoids(III) and Actinoids(III): An Experimental/Theoretical Saga. Chem. - Eur. J. 2012, 18 (36), 11162−11178. (44) Nichols, K. P.; Pompano, R. R.; Li, L.; Gelis, A. V.; Ismagilov, R. F. Toward Mechanistic Understanding of Nuclear Reprocessing Chemistries by Quantifying Lanthanide Solvent Extraction Kinetics via Microfluidics with Constant Interfacial Area and Rapid Mixing. J. Am. Chem. Soc. 2011, 133 (39), 15721−15729.

11539

DOI: 10.1021/acs.langmuir.6b03479 Langmuir 2016, 32, 11532−11539

Atomic-Scale Evidence for Potential Barriers and ... - ACS Publications