pubs.acs.org/Langmuir © 2010 American Chemical Society
Visualization of Ion Distribution at the Mica-Electrolyte Interface Siu-Hong Loh* and Suzanne P. Jarvis Conway Institute of Biomolecular and Biomedical Research, University College Dublin, Belfield, Dublin 4, Ireland Received March 22, 2010. Revised Manuscript Received May 17, 2010 Local ionic environments within nanometer proximity of a surface play a major role in the interactions which occur there and can be of critical importance in, for example, colloid suspensions, as well as biological function. Such environments often vary significantly from bulk properties, as we show here by the direct imaging of a range of monovalent (Liþ, Naþ) and divalent (Ca2þ, Mg2þ) cations distributed at the liquid-solid interface of mica. We image local charge distributions relative to the atomic lattice of mica and adjacent structured water and explain how their location is influenced by the electrostatic characteristics of the underlying lattice.
Introduction Interfacial interactions involving electrolytes control a broad range of natural and technological phenomena including biological function and technologies based on colloid suspensions. Understanding such interactions at the atomic scale is fundamental to our ability to manipulate and utilize the physical and chemical properties of biological and nonbiological materials in aqueous solutions and humid environments. The basal plane of muscovite mica has been used extensively both as the interacting surface in solid-liquid interactions as studied by surface forces apparatus and as a substrate for the immobilization of biomolecules in atomic force microscopy (AFM). It has become a popular choice for interfacial studies, due to its perfect cleavage along the (001) plane to achieve an atomically flat surface, such as the distribution and mobility of adsorbed ions.1,2 The crystal structure consists of complex aluminosilicate layers (Figure 1a and b) and has a nominal composition of KAl2(Si3Al)O10(OH)2. Replacement of Si4þ ions by Al3þ ions yields net negative charges, and these are compensated for by interlayer Kþ ions, which bind two aluminosilicate layers together through electrostatic forces. In electrolyte solution, the negative charges of a cleaved mica surface are balanced by absorbed cations.3 Previous studies using X-ray reflectivity have provided some useful information on the interaction of cations with aqueous-mineral interfaces.4,5 Here, we investigate highly localized cation charge distributions relative to the atomic lattice of mica using frequency modulation atomic force microscopy (FM-AFM). Experimental Section Sample Preparation. Four electrolyte solutions with the same physiological salt concentration of 150 mM (LiCl, NaCl, MgCl2, and CaCl2) were prepared by dissolving high-purity grade salts (Sigma-Aldrich, Dublin, Ireland) in pure water (Milli-Q water, 18.3 MΩ-cm). High-grade V-1 muscovite mica discs (SPI supplies, *Corresponding author. Tel: þ353-1-716-6780; Fax: þ353-1-716-6777; E-mail:
[email protected].
(1) Xu, L.; Salmeron, M. Langmuir 1998, 14, 2187. (2) Xu, L.; Salmeron, M. Langmuir 1998, 14, 5841. (3) Skipper, N.; Chang, F.-R.; Sposito, G. Clays Clay Miner. 1995, 43, 285. (4) Park, C.; Fenter, P. A.; Nagy, K. L.; Sturchio, N. C. Phys. Rev. Lett. 2006, 97, 016101. (5) Schlegel, M. L.; Nagy, K. L.; Fenter, P.; Cheng, L.; Sturchio, N. C.; Jacobsen, S. D. Geochim. Cosmochim. Acta 2006, 70, 3549.
9176 DOI: 10.1021/la1011378
West Chester, PA) were glued onto a Teflon substrate holder and cleaved with adhesive tape. 100 μL of NaCl (150 mM) was immediately deposited onto the freshly cleaved mica surface and the substrate holder was placed in the AFM for imaging. These procedures were repeated for the other three electrolyte solutions (LiCl, MgCl2, and CaCl2). AFM Measurements. A bespoke low-noise AFM operating in frequency modulation (FM) mode was used for all measurements.6 The cantilever oscillation amplitude was kept constant and typically was in the range 1.5-2.5 A˚. Instead of using a conventional analogue feedback controller for the implementation of FM-AFM, a new digital feedback controller integrated as part of an Asylum Research (Santa Barbara, USA) bipolar controller combined with gain calculation algorithms was used for robust feedback tuning.7 Force curves were obtained as frequency shift versus distance curves at the cantilever fundamental resonance frequency and converted to force using the method of Sader and Jarvis.8 All FM-AFM images presented in this paper were obtained at a constant frequency shift of the second flexural mode of the cantilever (deflection noise density at the second resonance equal to 8.5 fm/(Hz)1/2). Gold-coated (detector side) silicon cantilevers with typical tip radius of 2 nm (Nanosensors: SSS-NCHAuD) and force constant of around 13 N/m were used in all measurements. Cantilever stiffness was calibrated using the method reported by Hutter and Bechhoefer.9 In water, the first and second resonance frequencies for this type of cantilever are approximately 100 kHz and 630 kHz, respectively.
Results and Discussion On approaching the mica surface, the AFM tip was subject to an oscillatory force interaction as shown in Figure 2b. This is due to the discrete layering of water into solvation shells adjacent to the mica surface.10 Points of equivalent frequency shift (force) are indistinguishable for the feedback electronics, and as a consequence, an image can be obtained at a number of quasi-stable locations at quantized distances from the mica surface. These distances correspond to the size of a water molecule (∼2.8 A˚) and the number of discrete locations depends on the number of solvation shells typically observed for a particular tip-solution (6) Fukuma, T.; Jarvis, S. P. Rev. Sci. Instrum. 2006, 77, 043701. (7) Kilpatrick, J. I.; Gannepalli, A.; Cleveland, J. P.; Jarvis, S. P. Rev. Sci. Instrum. 2009, 80, 023701. (8) Sader, J. E.; Jarvis, S. P. App. Phys Lett. 2004, 84, 1801. (9) Hutter, J. L.; Bechhoefer, J. Rev. Sci. Instrum. 1993, 64, 1868. (10) Israelachvili, J. N.; Pashley, R. M. Nature 1983, 306, 249.
Published on Web 05/20/2010
Langmuir 2010, 26(12), 9176–9178
Loh and Jarvis
Figure 1. Crystal structure of muscovite mica. (a) Cleaved surface (001) plane (Kþ ions not shown). (b) [110] projection.
Figure 2. (a) FM-AFM 3D height image in phosphate buffer solution with discrete layering of water adjacent to the mica surface. (b) Oscillatory force curve showing points of equivalent frequency shift (force) that corresponds to quasi-stable locations at quantized distances from the mica surface. These distances correspond to the size of a water molecule (∼2.8 A˚).
sample combination.11 By careful comparison of a typical force profile adjacent to the surface, with the features observed during imaging, it is possible to attribute specific features in the image either to the direct tip-mica interaction (as shown in purple in Figure 2) or to a tip-mica interaction mediated through a single layer of water (as shown in blue in Figure 2). In this way, the images shown in Figure 3 can be attributed to surface features directly on the mica lattice, which are not mediated via any water molecules. Figure 3 shows FM-AFM images of the cleaved surface of muscovite mica taken in the four electrolyte solutions. A honeycomblike pattern with hexagonal array structures is seen in all the images in Figure 3. Two arrangements can be associated with the (11) Fukuma, T.; Higgins, M. J.; Jarvis, S. P. Biophys. J. 2007, 92, 3603.
Langmuir 2010, 26(12), 9176–9178
Letter
Figure 3. FM-AFM images of the cleaved surface of muscovite mica taken in electrolyte solutions. All images have similar size: 3.5 nm 2.5 nm. Features which are greater than 50 pm above the honeycomb lattice are highlighted in red. (a) Image taken in 150 mM LiCl solution, Δf=þ40 Hz, A=0.20 nm, scanning speed: 60 nm/s. (b) Image taken in 150 mM NaCl solution, Δf=þ10 Hz, A=0.20 nm, scanning speed: 60 nm/s. (c) Image taken in 150 mM MgCl2 solution, Δf = þ100 Hz, A = 0.16 nm, scanning speed: 60 nm/s. (d) Image taken in 150 mM CaCl2 solution, Δf = þ200 Hz, A = 0.18 nm, scanning speed: 78 nm/s. Gold-coated (detector side) silicon cantilevers (Nanosensors: SSS-NCHAuD) were used. The cantilevers were actuated at the second flexural mode (in the range 600-650 kHz depending on the cantilever).
mica lattice: an arrangement of oxygen ions with ditrigonal symmetry as shown in Figure 4a and a hexagonal array of Si/Al ions as shown in Figure 4b. The building block that forms the tetrahedral layer is the SiO4 tetrahedron, which consists of one silicon ion surrounded by four oxygen ions. From the [110] projection view of the crystal structure (Figure 1b), the tetrahedral sheet, which consists of an oxygen-silicon-oxygen (or partially oxygenaluminum-oxygen) layer, indicates that the Si/Al ions reside below the outer oxygen basal plane. If AFM imaging is performed in the absence of cations, the AFM tip will first interact with the oxygen ions in the basal plane. In this case, we would expect to observe an arrangement with ditrigonal symmetry in the images. (Note that the arrangement with ditrigonal symmetry is different from the hexagonal array of Si/Al ions, as small equilateral triangles, which are formed by groups of oxygen triads, are seen in the first case but not in the second). However, it is apparent that all AFM images in Figure 3 show a hexagonal arrangement matching that shown in Figure 4b. The AFM tip cannot interact directly with the Si/Al ions because these ions are beneath the oxygen ions in the basal plane. Therefore, in order to produce the hexagonal array that mirrors the hexagonal array of Si/Al ions (seen in all images in Figure 3), we suggest that the AFM tip must be interacting with cations that absorb above the oxygen triads to generate the protrusions observed in Figure 3. Note that the location of each cation absorbed above an oxygen triad is vertically on top of the corresponding tetrahedron Si/Al ion. This is illustrated in the structural model shown in Figure 4c. It could be observed qualitatively that protruding features were randomly distributed on top of the oxygen triads throughout the scanned images (the z-scale color table in Figure 3 has been selected to highlight in red those features which are greater than 50 pm above the honeycomb lattice). We propose that the protruding features above oxygen triads can be attributed to the absorption of cations, particularly for oxygen triads that contain Al3þ ions, due to excess negative charges from the surrounding oxygen ions. The protrusions observed in Figure 3 should be considered to indicate the preferred locations of highly mobile cations averaged DOI: 10.1021/la1011378
9177
Letter
Figure 4. (a) Arrangement of oxygen ions with ditrigonal symmetry. (b) Hexagonal array of Si/Al ions. (c) Structural model showing protruding features randomly distributed on top of the oxygen triads.
over the data acquisition time (∼60 s for a single image). Due to this slow data acquisition time, the mobile cations would be better considered as a diffuse cloud of charge rather than discrete objects detected at specific locations. Thus, although taking charge considerations into account we would expect preferential absorption of cations above Al3þ, this would not be to the total exclusion of cations above Si4þ over the time taken to acquire a single image. This could manifest itself in the images as a somewhat larger protrusion above Al3þ. As can be seen from the images, not all protrusions are the same size, indicating a variation in the interaction above different oxygen triads on the surface. In other words, the variation in size of protrusions in the AFM images might be indicative of variations in the underlying ions (e.g., Al3þ or Si4þ) in the tetrahedral layer closest to the cleaved surface. At the atomic scale, the protruding structures are seen to be irregular and distributed randomly in all the images taken in electrolyte solutions. Moreover, from the images shown in Figure 3, there are hexagons that exhibit clear distortion, as indicated by arrows. The ability to reveal such protrusions and lattice defects at the atomic scale strongly suggests that the AFM tips used in all measurements have an atomically sharp asperity. True atomic resolution was achieved in all the images taken in a liquid environment. The imaging results, particularly in the case of MgCl2
9178 DOI: 10.1021/la1011378
Loh and Jarvis
(Figure 3c) and CaCl2 (Figure 3d), agree well with Monte Carlo simulation data published recently.12,13 In these studies, a comparison of the potential of mean force (PMF) profiles revealed that cation absorption above an oxygen triad of a tetrahedral substitution is energetically much more favorable than that in a ditrigonal cavity, in the case both of Mg2þ and Ca2þ cations. In the case of LiCl and NaCl, there are relatively more protrusions in the images shown (Figure 3a and b), as compared to MgCl2 and CaCl2 (Figure 3c and d). As charged entities, the number of monovalent cations required to balance the negatively charged surface of mica would be higher than divalent cations, thus providing a plausible explanation for the higher number of protrusions. Previous studies using X-ray reflectivity have derived electron density profiles and the adsorption height of Ca2þ from which it was suggested that divalent cations such as Ca2þ are partially hydrated near the surface and could reside anywhere on the surface.5 In other words, in this case the oxygen triads were not the only sorption sites, but instead, the cations could also be found equally distributed over all the ditrigonal cavities of the muscovite surface. This would appear to contradict our findings. However, our AFM measurements differ from X-ray reflection measurements in that they measure the interaction of an AFM tip at the mica-electrolyte interface, whereas X-ray reflection measures the nonperturbed mica-electrolyte interface. As the AFM tip approaches within 2 A˚ of the mica lattice, water molecules are excluded from the intervening space, while it may still be possible for the smaller unhydrated cations to occupy this region (bare ionic radii, Liþ (60 pm), Naþ (95 pm), Mg2þ (65 pm), and Ca2þ (99 pm)). Our results indicate that unhydrated cations do indeed occupy this near-surface region to balance the negative charges of the cleaved mica, as water is excluded by the close proximity of the incoming solid surface.
Conclusions In summary, in the presence of electrolytes we have observed the distribution of 50 pm protrusions above the location of an apparently random number of the oxygen triads in the mica lattice, suggesting that mobile cations preferentially reside above these oxygen triads. The lateral imaging resolution demonstrated here indicates that FM-AFM could serve as a powerful means to provide new insights into the influence of ions on interactions at aqueous-mineral interfaces. Further, the direct interaction of the AFM tip with the solid-fluid interface highlights the nature of the distributed charge encountered by an incoming object interacting with the interface at separations of less than the dimensions of a single water molecule. Such features are not accessible by other experimental techniques that do not mechanically interact directly with the interface despite the obvious importance of this near-surface region in the interactions of colloids and biological molecules in ionic solutions. Acknowledgment. This work was supported by Science Foundation Ireland (grant no. 07/IN1/B931). The authors thank Jason Kilpatrick for assisting with the preparation of figures and Khizar Sheikh for advice on AFM imaging parameters. (12) Meleshyn, A. J. Phys. Chem. C 2009, 113, 12946. (13) Meleshyn, A. J. Phys. Chem. C 2009, 113, 17604.
Langmuir 2010, 26(12), 9176–9178
