Article pubs.acs.org/Langmuir
Interactions of Na+, K+, Mg2+, and Ca2+ with Benzene Self-Assembled Monolayers M. Rimmen,* J. Matthiesen, N. Bovet, T. Hassenkam, C. S. Pedersen, and S. L. S. Stipp Nano-Science Center, Department of Chemistry, University of Copenhagen, Universitetsparken 5, 2100 Copenhagen, Denmark S Supporting Information *
ABSTRACT: Interactions between cations and organic molecules are found throughout nature, from the functionality and structure of proteins in humans and animals to the exchange of ions in minerals in soil and oil reservoirs with the fluid phases. We have explored the behavior of the s-block elements that are most common in the natural world, namely, Na+, K+, Mg2+, and Ca2+. Specifically, we investigated how these ions affect the interactions between surfaces covered by self-assembled monolayers (SAMs) terminated with benzene molecules. We used a flat oxidized silicon substrate and an atomic force microscopy (AFM) tip that were both functionalized with 11-phenoxyundecane-1-thiol and measured the adhesion force between them in solutions of each of the four chloride salts. We observed that the adhesion increased in the order of the Hofmeister series: K+ < Na+ ≈ Mg2+ < Ca2+. Supplementary evidence from X-ray photoelectron spectroscopy (XPS) allowed us to conclude that K+ binds in the benzene layers, creating a positive surface charge on the benzene-covered surfaces, thus leading to lower adhesion in KCl solutions than in pure water. Evidence suggested that Ca2+ does not bind to the surfaces but forms bridges between the layers, leading to higher adhesion than in pure water. In Na+ and Mg2+ solutions, adhesion is quite similar to that in pure water, indicating a lack of interaction between these two ions and the surfaces, or at least that the interaction is too weak to be detected by our measurements. The results of our studies clearly show that even a nonpolar, hydrophobic molecule, such as benzene, has a role to play in the behavior of aqueous solutions and that it interacts differently depending on which ions are present. Even ions from the same column in the periodic table behave differently.
■
INTRODUCTION The behavior of organic molecules in aqueous solution is often affected by the cations that are also present. Organic molecule− cation interactions control reactions throughout nature. In animals, alkali and alkaline earth metal ions affect the solubilities of various proteins1,2 and modify the conformation and stability of ribonucleic acid (RNA).3 The balance between alkali and alkaline earth metal ions in soils, which is controlled by the cation-exchange capacity of minerals,4 particularly clay minerals,5 contributes to the fertility of soils and their ability to sequester toxic compounds from groundwater. The effect of cations on the adsorption capacity of nonpolar compounds in natural porous media is as much a question in groundwater remediation as it is in developing more effective strategies for enhancing oil recovery from reservoirs.6 Differences in the character of atoms, namely, the electron configuration, ion size, charge, and polarizability, determine the extent of interaction. This is known as the specific ion effect. In aqueous solution, electron configuration determines an ion’s abilities to hydrate (bind water around it) and to hydrolyze (release protons or hydroxyl ions to become a hydroxylated species). The hydrated radius of an ion is the radius of the ion with its water shell. Small or strongly charged ions, that is, those with a higher ionic potential, attract more water molecules, which increases the hydrated radius. This plays a role in its interaction with organic species. The sizes of the hydrated radii of the most common alkali and alkaline earth ions are listed in Table 1, along with their ionic potentials (i.e., the ratio between © 2014 American Chemical Society
Table 1. Ionic and Hydrated Radii and Ionic Potentials of Alkali and Alkaline Earth Metals8 ion
ionic radius (Å)
ionic potential (nC/m)
hydrated radius (Å)
ionic potential (nC/m)
Li+ Na+ K+ Mg2+ Ca2+
0.68 0.99 1.33 0.65 0.99
2.36 1.62 1.20 4.93 3.24
3.82 3.58 3.31 4.28 4.12
0.42 0.45 0.48 0.75 0.78
charge and radius). They follow the order Mg2+ > Ca2+ > Li+ > Na+ > K+. Hofmeister1,7 demonstrated that the electron configuration of an adsorbing cation influences protein solubility. Proteins “salt out”, decreasing their dissolved concentration in the order Ca2+ < Mg2+ < Li+ < Na+ < K+, which is almost identical to the increase in hydrated radii. Hofmeister’s work demonstrated that a thicker hydration sphere decreases binding efficiency. His results can be explained by the energy required to release bound water before a tight bond can form between the cation and the organic molecule. In the human body, specific ion effects are responsible for the function of Na+/K+-ATPase,9,10 which is also known as the Na+/K+ pump. ATPase selectively transports Na+ and K+ across cell membranes so that three Na+ ions are transported out of Received: May 14, 2014 Revised: July 3, 2014 Published: July 8, 2014 9115
dx.doi.org/10.1021/la5018664 | Langmuir 2014, 30, 9115−9122
Langmuir
Article
the cell for every two K+ ions that are transported in. This effectively removes positive charge from the cell and creates a resting potential across the cell membrane. At some point, this resting potential becomes strong enough that no more ions can be moved out of the cell. Such a potential is used by neurons to send electrical impulses through the body. Computational studies have shown that, in the gas phase, a Na+ ion binds to benzene rings more strongly than a K+ ion,11 but in aqueous solution, the order is reversed. This is because, in aqueous solution, an ion must first break bonds with some of the water molecules in its hydration shell before it can associate with benzene. Because K+ binds fewer water molecules, it needs to break fewer bonds. K+ can also form a stronger bridge than Na+ between two benzene molecules in aqueous solution.12 K+ is favored because the interaction distance between K+ and benzene is longer than that between Na+ and benzene. For Na+ to bind to two benzene rings, the entire hydration shell of Na+ must be ejected. The longer interaction distance of K+ allows some water to remain, decreasing the overall energy required for K+ to form bonds.12 In aqueous solution, K+ can even bind three benzene rings,13 which agrees well with gas-phase studies. In the gas phase, the K+ bond with a benzene ring is slightly stronger than that with a water molecule.14 Interactions between alkaline earth metal ions and benzene have also been investigated. Rodriguez-Cruz and Williams15 used electrospray mass spectrometry in the gas phase and showed that Mg2+ and Ca2+ can exchange water to bind with benzene. The rate is highest when there are four or fewer water molecules around the ion. The presence of more than four water molecules shields the ions well enough to prevent interaction with benzene. In aqueous solution, the cations are fully hydrated, so interaction with benzene is weaker. Cheng et al.16 performed computational studies on benzene−Me2+X− complexes in a vacuum and demonstrated that the identity of the specific halide counterion, X−, present was important. They extended their study to include all halide ions and showed that the interactions for these complexes were similar to those for benzene−Me+. They confirmed that Be2+, with its small size and divalent charge, binds more strongly than Mg2+, which, in turn, binds more strongly than Ca2+, again consistent with the Hofmeister series. In our research group, we have studied adhesion behavior between various organic molecules and surfaces, in solutions containing various ions,5,17−19 by using an atomic force microscopy (AFM) tip that is functionalized with a selfassembled monolayer20 (SAM) of specific organic compounds so that it represents a cluster or droplet of organic molecules. This form of AFM is called chemical force microscopy (CFM) and has been known since the mid-1990s.21 The investigation of ions interacting with these SAMs was started shortly thereafter.22 This method has been developed into singlemolecule AFM, in which only one molecule is attached to the tip and it is therefore possible to investigate the rupture force of molecular bonds.23 Thiols are used to attach molecules to tips because they make very strong bonds with a gold layer that can be evaporated onto both an AFM tip and a flat substrate. This ensures that the end group of the functionalized molecule protrudes from the surface. Here, we aim to understand the interaction between benzene molecules in aqueous solutions containing Na+, K+, Mg2+, and Ca2+. To do this, we functionalized both a tip and a substrate surface with 11-phenoxyundecane-1-thiol, yielding a benzenecovered surface on both the tip and the substrate, and we
compared the adhesion between the two surfaces in pure water and in chloride solutions containing each of the four cations (Figure 1). We chose to attach the benzene ring through an
Figure 1. Sketch of the tip and substrate functionalized with SAMs of benzene-terminated molecules.
oxygen atom because the interaction of this compound with ions is closer to that of pure benzene than a compound without oxygen. Phenol, which has an oxygen atom in the same type of structure as our compound, binds K+ by 2.1 kJ/mol more energy than pure benzene,11 which has no oxygen. Toluene (methylbenzene), which could be considered a model for an alkyl chain bound to benzene and which does not have an oxygen in its structure, binds K+ by 6.3 kJ/mol more than benzene,11 so toluene is less similar to benzene than phenol is.
■
METHODS
To make the solutions and the SAMs, we used only compounds of reagent grade or better, supplied by Sigma-Aldrich Chemicals. We used ultrapure, deionized water, drawn from a Milli-Q column; its resistivity, R, was 18 MΩ·cm or higher. Synthesis of Thiol. The molecules for functionalizing the tip and substrate were prepared with a thiol group on one end of an alkane chain of 11 carbon atoms and a benzene group attached on the other end through an oxygen atom. The benzene alkanethiol was prepared using a three-step synthesis (see Scheme 1). The final step was performed with the gold-coated tip or substrate present in the solution because the free thiol is unstable. Over time, the thiol oxidizes to disulfide, but the thioacetate that forms in the second step is stable, so this precursor can be stored without risk of oxidation. 11-Bromo-1-phenoxyundecane. Potassium carbonate (K2CO3) was dried in an oven at 120 °C for 3 days to remove the water of crystallization. Acetone (25 mL) was dried by adding anhydrous magnesium sulfate (MgSO4) and stirring for 3 h. The MgSO4 was filtered off, and the dry acetone was used as a solvent. To the acetone were added 1,11-dibromoundecane (3.00 g, 2.25 mL, 9.56 mmol), phenol (C6H5OH, 894 mg, 9.50 mmol), and anhydrous potassium carbonate (1.34 g, 9.70 mmol). The mixture was stirred, heated to boiling, and allowed to reflux for 20 h. After the mixture had cooled to room temperature, water (50 mL) and ether (50 mL) were added, and 9116
dx.doi.org/10.1021/la5018664 | Langmuir 2014, 30, 9115−9122
Langmuir
Article
Scheme 1. Synthesis of a Benzene-Functionalized Thiola
a The first step is the reaction between phenol (C6H5OH), 1,11-dibromoundecane, and dry potassium carbonate. The product of the first step is then reacted with potassium thioacetate and finally with aqueous ammonia.
The spring constant was then determined by fitting a Lorentzian function to the thermal spectrum of the cantilever.24 The tip was approached to the substrate with a velocity of 7.8 μm/s until a repulsive force of 500 pN between the tip and substrate was reached. The tip was then retracted at the same velocity. The adhesion force (Figure 2), extracted from each of the 400 force curves taken
the mixture was shaken and separated. The ether phase was washed with water (50 mL), and the water phase was combined with the first water phase and extracted with ether (50 mL). The ether phases were combined and dried by adding anhydrous MgSO4 and stirring for 30 min. The MgSO4 was filtered off, and the ether phase was evaporated away. The crude product was a yellow oil that was separated by column chromatography on silica gel with an eluent of 3:7 CH2Cl2/ C7H16. This yielded the product as an almost colorless oil (1.026 g, 33% of the theoretical yield). 11-Phenoxyundecyl Thioacetate. We dissolved 11-bromo-1phenoxyundecane (923 mg, 2.82 mmol) in methanol (25 mL) and added potassium thioacetate (KSCOCH3, 656 mg, 5.74 mmol). The mixture was stirred, heated to boiling, and refluxed for 20 h. Concentrated ammonium chloride (NH4Cl, 13 M, 25 mL) was added, and the mixture was extracted with ether (3 × 50 mL). The ether phases were combined and dried by adding anhydrous MgSO4 and stirring for 30 min. The MgSO4 was filtered off, and the ether phase was evaporated away. The crude product was a colorless oil that was separated by column chromatography on silica gel with an eluent of 1:1 CH2Cl2/C7H16. This yielded the product as a grayish white solid (518 mg, 90% of the theoretical yield). 11-Phenoxyundecane-1-thiol. 11-Phenoxyundecyl thioacetate (65 mg, 0.2 mmol) was dissolved in ethanol (50 mL) to make a 4 mM solution. The vessel containing the solution was wrapped with aluminum foil and stored in a refrigerator to minimize light exposure. Five milliliters of this solution was mixed with concentrated aqueous ammonia (NH3, 13 M, 5 μL) in a Petri dish to form the product along with acetamide [CH 3 C(O)NH 2 ]. This solution was used immediately hereafter to produce the SAM. Preparation of the SAM. To prepare the substrates, we used polished silicon wafers that had developed an amorphous silica (SiO2) layer during exposure to air. We cleaved 1 cm2 squares from the wafer and cleaned them by UV ozone treatment to remove adventitious carbon, and then we used an electron beam evaporator to deposit a layer of titanium that was about 2 nm thick, followed by a layer of gold that was about 100 nm thick. Substrates for the XPS experiments were prepared in the same way. To deposit the SAMs, AFM tips and gold-coated silica wafers were rinsed with ultrapure water and then ethanol, dried with a jet of N2, and treated with UV/ozone for 30 min. Next, they were put into the solution containing 11-phenoxy-1-undecanethiol and acetamide. The closed Petri dish was wrapped with Parafilm and covered with aluminum foil, and the solution was allowed to react with the tips and the substrates for 20 h at room temperature (21 °C). The finished tips and substrates were transferred to a Petri dish of pure ethanol for 30 min to release excess acetamide and residual 11-phenoxyundecane-1thiol. Each tip and substrate was used immediately for AFM experiments. Chemical Force Microscopy (CFM). All CFM experiments were performed using an Asylum Research MFP-3D atomic force microscope. All tips were Olympus Biolever AFM probes with two types of cantilevers having nominal spring constants of about 6 and 30 pN/nm. These tips are gold-coated and are therefore ideal for attachment of the thiol end of the SAM molecules. For these experiments, we used only cantilevers with a spring constant of ∼6 pN/nm. Before each experiment, the deflection sensitivity of the cantilever was measured by fitting the contact part of a force curve.
Figure 2. Example of a force curve with the approach part in red and the retraction part in blue. The adhesion force is the difference between the minimum on the retraction curve (D) and the flat part of the retraction curve (E), which, in this case, is about 500 pN. over a 2 × 2 μm2 area (20 × 20 points, 100 nm between each point) was then plotted to produce a force map, with color used to indicate the extent of adhesion. All experiments were performed in a fluid cell with both tip and substrate submerged in the solution. The volume of the fluid cell was roughly 3 mL. We performed three sets of experiments. In Set 1, four solutions were tested: ultrapure deionized water, called Milli-Q, and solutions of 100 mM NaCl, KCl, and CaCl2. In Set 2, five solutions were used: Milli-Q and solutions of 100 mM NaCl, KCl, MgCl2, and CaCl2. In Set 3, five solutions were used: Milli-Q and solutions of 100 mM NaCl and KCl and 33 mM MgCl2 and CaCl2. The concentrations of MgCl2 and CaCl2 were lower to make the ionic strengths identical in all four salt solutions. All solutions in all sets had a pH of 5.5 and a temperature, T, of 25 °C. To enable absolute comparisons of the results for various solutions, the same tip and substrate were used for all of the experiments of each set. It is impossible to prepare two tips that are exactly the same size and shape and with identical SAM coverage, so in each set, the adhesion measured in the chloride solutions were compared to that in ultrapure water. Our procedure for comparing the adhesions in the various solutions was the same in all three sets. First, two force maps were obtained in Milli-Q water. Then the solution was changed to a salt solution, and two new force maps were obtained. Each time, the solution was exchanged by extracting about three-quarters of the contents of the fluid cell with a syringe and then injecting the same amount of the new solution. This was repeated four times, effectively replacing more than 99% of the original solution. We did not remove and replace all of the solution at once because we would not have been able to keep the tip at the same location on the substrate and to avoid evaporation and possible precipitation of salts on the tip or surface. After each measurement in a salt solution, the solution was replaced by Milli-Q water again. This procedure was repeated, exchanging solutions of 9117
dx.doi.org/10.1021/la5018664 | Langmuir 2014, 30, 9115−9122
Langmuir
Article
Figure 3. Typical force maps and average adhesion obtained from the experiments of Set 1.
Figure 4. Average adhesion in ultrapure water and (A) 100 mM NaCl from Set 1, (B) 100 mM KCl from Set 1, (C) 100 mM CaCl2 from Set 1, and (D) 100 mM MgCl2 from Set 2. Each pair of results, for example, the first two bars labeled water, were obtained in the same solution, simply repeating the data collection, to test for the time dependence of adhesion. nitrogen to minimize salt precipitation on the substrate. We used the lack of the presence of Cl− as an indication that salts had not precipitated, as has been shown previously.26
various compositions, to demonstrate reproducibility throughout each set of experiments. X-ray Photoelectron Spectroscopy (XPS). This is a technique that is especially sensitive to the composition of the near-surface region, that is, the top 10 nm. We used a Kratos Axis UltraDLD vacuum chamber and a monochromatic AlKα X-ray source (energy = 1486.6 eV, power = 150 W). The data were fitted using Shirley background subtraction and the commercial software CasaXPS. Binding energies were calibrated to the Au 4f7/2 peak at 84 eV. Using XPS, we confirmed the formation of a SAM by examining the S 2p peak. The binding energy for S bound to Au is 162−163 eV,25 which we could identify. We could see only a small peak for S at 164 eV, which indicates a few stray thiol benzene molecules trapped in layers rather than bonded to the gold. Three additional substrates functionalized with SAMs terminated with benzene groups and three substrates with SAMs terminated with CH3 groups were examined using XPS. One benzene-terminated sample and one CH3-terminated sample were immersed in Milli-Q water for 30 min as the controls. Another sample of each was exposed to 100 mM KCl for 30 min, and a third sample of each was exposed to 100 mM CaCl2 for 30 min. The samples were removed from solution, and the remaining fluid was immediately removed with a jet of
■
RESULTS Figure 3 shows four force maps from Set 1, taken with the same tip and substrate, in solutions of 100 mM NaCl, KCl, and CaCl2. Adhesion in 100 mM KCl is lower than that in water, but adhesion in 100 mM CaCl2 is higher than that in water. Adhesion is slightly higher in 100 mM NaCl than in water. To quantitatively compare the results obtained in the various solutions, we plotted average adhesion as a function of experiment (see Figure 4). Figure 4A shows a plot of the average adhesion from sequential force maps obtained in water and NaCl. Each bar represents the average adhesion of one force map. Throughout the experiments, we kept the tip and substrate the same and varied only the solution in contact. Adhesion in NaCl was ∼8% higher than that in water. Figure 4B shows the average adhesion in KCl solutions, compared with that in ultrapure water. On average, the 9118
dx.doi.org/10.1021/la5018664 | Langmuir 2014, 30, 9115−9122
Langmuir
Article
that Na+ and Mg2+ do not interact with the benzene layers or, at least, do not interact strongly enough for us to see it. The lower adhesion in K+ solutions and the higher adhesion in Ca2+ solutions are consistent in all three experiments, but in Set 3, the adhesion difference between water and CaCl2 solution is only slightly larger than the variation in adhesion. The change in adhesion for Set 3 is smaller because, in these experiments, to test the effect of ionic strength, CaCl2 was only one-third as concentrated as in Sets 1 and 2. Table 2 presents the average adhesion values, FAD, shown in Figure 5. It also lists the differences between the adhesions measured in the salt solutions and water
adhesion was about 20% lower in KCl than in water. These results also demonstrate that the solution can be exchanged without changing the properties of the tip or substrate and that the adhesion in this case does not depend on the composition of the previous solution. Figure 4C shows that the average adhesion in CaCl2 is consistently higher than that in pure water, by 33% on average. In contrast with the behavior observed for the monovalent cations, adhesion does not return to the initial value when the solution is changed back to pure water. This indicates that Ca2+ ions cannot be effectively removed, even after replacing the solution with two additional complete exchanges of ultrapure water (Figure 4C, labeled as re-water). When the solution was replaced by 100 mM KCl, adhesion decreased to that initially observed in KCl (Figure 4B), suggesting that K+ replaces Ca2+. Replacing the KCl with water reproduces the original water adhesion before CaCl2 was injected. The change in adhesion with CaCl2 indicates that Ca2+ interacts with the benzene but in a different way than K+ does. In summary, for Set 1, the adhesion between the benzene surfaces in solutions of NaCl did not change significantly from the adhesion in water alone. There was less adhesion in the KCl solution than in water, and the adhesions in water before and after KCl were the same. CaCl2 caused a considerably higher adhesion, and the effect did not decrease during subsequent solution replacement by water. However, adhesion did decrease after replacement by KCl solution and was restored to the initial value after a new injection of water (Figure 4C). In Set 2, we measured all of the solutions again using a new tip and new substrate. The absolute adhesion was not the same as in Set 1, because the tip and substrate were different, but the trends in the change in adhesion were similar (the measured adhesion values are presented in the Supporting Information). In Set 2, we also measured the effect of 100 mM MgCl2 solution, which was not measured in Set 1. Figure 4D shows a comparison of data from pure water and 100 mM MgCl2. In the solutions of MgCl2, adhesion was not much different than that in pure water. Set 3 was performed with water and solutions all having the same ionic strength, namely, 100 mM, so the MgCl2 and CaCl2 solutions had a concentration of 33 mM. The average adhesion values for all three sets of experiments are shown in Figure 5. Overall, the results from the experiments in NaCl and MgCl2 are not significantly different from those in water, indicating
ΔFAD = FAD,ion − FAD,water
and the percentage relative adhesion difference, called the adhesion response ⎤ ⎡1 Resp = ΔFAD ⎢ (FAD,ion + FAD,water)⎥ × 100 ⎦ ⎣2
The adhesion response in KCl solutions was negative in all cases. The response for NaCl was small and changed sign between experiments, indicating no or little effect from NaCl. In the experiments with MgCl2, in solutions of both 100 and 33 mM, the responses were also small and of opposite sign, which also indicates a limited effect from Mg2+. The response in CaCl2 was significant; the lower response in Set 3 indicates a concentration-dependent response. XPS was used to further investigate the adhesion changes caused by KCl and CaCl2; the data are presented in the Supporting Information. Spectra from the benzene SAM that had been immersed in water only to serve as the control had no peaks at binding energies representing potassium or calcium. Spectra from the KCl-exposed substrate had peaks at 293.1 and 295.8 eV, corresponding to K 2p3/2 and K 2p1/2, respectively, but there were no peaks in the region for photoelectrons from chlorine. This indicates that potassium bonded to the benzene substrate but it was not present as KCl, thus showing that surface K did not result from precipitation of KCl. No counterions could be identified with XPS, but at the nearneutral pH of these experiments, we assume that OH− balances the charge. Oxygen from this very small amount of OH− cannot be distinguished from O in the SAM layer itself. XPS peak intensity ratios also show that, for each S, there was about 0.12 K, so ∼10% of the benzene molecules on the SAM had K+ associated with them. On the sample exposed to CaCl2, there was no sign of a peak in the Ca binding energy region, indicating that the substrate exposed to CaCl2 solutions had not adsorbed calcium. Spectra from the three substrates that were prepared with the CH3 SAM and exposed to potassium and calcium solutions showed no sign of adsorbed K or Ca. This provides further evidence for benzene rings being able to bind to potassium.
■
DISCUSSION Benzene is hydrophobic and only very slightly soluble in water (1.8 g/L26). It is planar because of its six delocalized πelectrons, which are located equally on each side of the hexagonal carbon ring, so benzene has no dipole moment but is very polarizable. Benzene rings stack so the π-electrons of one ring interact with the π-electrons of the next. This arrangement, called π-stacking, is the preferred orientation for benzene rings.
Figure 5. Average adhesions in water and all four solutions from all three sets of experiments. Each point is an average of all of the experiments performed in that solution, so the uncertainty is within the size of the symbol in most cases. 9119
dx.doi.org/10.1021/la5018664 | Langmuir 2014, 30, 9115−9122
Langmuir
Article
Table 2. Average Adhesion (FAD), Adhesion Change (ΔFAD) between Water and Salt Solutions, and Response (Resp) of the Adhesion to Added Salt Set 1: 100 mM salt solutions
Set 3: 100 mM NaCl and KCl solutions, 33 mM MgCl2 and CaCl2 solutions
Set 2: 100 mM salt solutions
solution
FAD (pN)
ΔFAD (pN)
Resp (%)
water NaCl KCl MgCl2 CaCl2
727 ± 10 805 ± 9 575 ± 8
0 77 −153
0 10 −24
952 ± 31
225
27
FAD (pN) 1130 1120 892 1170 1340
± ± ± ± ±
ΔFAD (pN)
Resp (%)
0 −10 −238 40 205
0 −9 −24 4 17
20 20 14 20 20
FAD (pN) 1050 981 891 976 1100
± ± ± ± ±
20 13 11 8 40
ΔFAD (pN)
Resp (%)
0 −67 −157 −72 45
0 −7 −16 −7 4
Scheme 2. Resonance Structures of Our Benzene Compound, Showing the Donating Effect of Oxygen Lone Pairs into the Ring, Which Almost Cancels the Effect of Oxygen Being More Electronegative
In these experiments, it is likely that π-stacking helps to stabilize the benzene thiol layers on the tip and the substrate. The compound used in our experiments, 11-phenoxyundecane-1-thiol, has an oxygen atom linking benzene to the chain. The benzene ring on this compound has been found to behave very similarly to free benzene in ion adsorption experiments.11 This is the result of two opposing effects: one that pulls electrons toward the oxygen and one that pushes away electrons from the oxygen lone pairs. The first effect comes from oxygen being more electronegative than carbon, so the oxygen pulls electrons from the single bonds in the benzene ring. The second effect comes from the overlap between the lone-pair electrons on the oxygen atom and the π-electron cloud of the benzene ring, which donates these π-electrons into the ring (Scheme 2). Because these two effects almost cancel each other, the interaction energy between our compound and a potassium ion derived from computational modeling, −68.3 kJ/mol, is very close to that between potassium and benzene, −66.2 kJ/mol.11 One might wonder whether the ions could bind directly to the oxygen, but if this were the case, we would have seen decreased adhesion for all ions because they all bond readily with oxygen. Therefore, we conclude that this oxygen does not play an active role in binding the ions. The studies on the interactions between benzene and ions that we have cited all investigated the interaction energies.11−16 One of these was a gas-phase study14 and revealed little about the interactions in aqueous solution. Other studies were strictly computational.11−13,16 Instead of measuring the interaction energies, our experiments measured the adhesion force, FAD. The tip and substrate were different in the three sets of experiments, and because no two tips or substrates are identical, the absolute adhesions cannot be compared between the three sets. However, the changes in adhesion with changes in solution can be compared. We can determine the theoretical adhesion force between the tip and the substrate from the sum of three contributing forces: van der Waals, FvdW; electric double layer, FEDL; and hydrophobic interaction, FHI FAD = FvdW + FEDL + FHI
double layer interactions. This theory has been shown to give good estimates of the forces measured by AFM in a wide range (1−100 mM) of salt concentrations.18,29 If one assumes that the tip and substrate remain undeformed when they come into contact, the van der Waals force between the tip and substrate can be estimated using FvdW = −
AR 6D2
(3)
where A is the Hamaker constant; R represents the radius of the tip; and D represents the distance between the tip and the substrate, which we assumed to be about 0.2 nm, roughly the diameter of a naked atom, the closest that two objects can ever get to each other. Ederth et al.30 measured the Hamaker constant for the interaction between two surfaces coated with 16-mercaptohexadecanol and found it to be 3.4 × 10−20 J for distances between 7 and 15 nm, where the effect of gold on A is significant. Here, we are calculating the force at contact, where D = 0.2 nm, and we assume that the effect of gold is negligible because it is far from the interacting surfaces. Using Lifshitz theory,31,32 we determined the Hamaker constant for benzene interacting with benzene across water to be 3.8 × 10−21 J. Using this value for A, along with D = 0.2 nm and R = 30 nm, we estimated the van der Waals force to be FvdW = 475 pN, which means that the van der Waals interaction is responsible for approximately half of the adhesion force in our experiments. For benzene surfaces in ultrapure water, FEDL is small because the benzene molecules are not charged. Whereas several studies show that OH− binds at air/water or oil/water interfaces,33 benzene rings only form very weak bonds to anions because such ions would be repelled by the π-ring. Therefore, we have assumed that the surface charge density of the benzene SAM layer is negligible, so that FEDL ≈ 0. This means that the rest of the adhesion arises entirely from the hydrophobic interaction, which cannot be determined theoretically. With XPS, we demonstrated that K+ is associated with the SAM, so in KCl solutions, the benzene-covered surface is charged. Therefore, the electric double layer force is no longer zero, and we can set FEDL equal to the difference in the adhesion force between water and KCl. The average value of FEDL from the three sets of experiments is 183 pN, a repulsive force, which explains the decrease in adhesion when water is replaced by KCl solution. An approximate expression for FEDL is18
(1)
In this equation, we have assumed a hard-wall Pauli repulsion, represented by a minimum distance between the surface and tip of 0.2 nm. We have used the theory presented by Derjaguin, Landau, Verwey, and Overbeek (DLVO)27,28 to estimate two of the contributing forces, namely, the van der Waals and electric 9120
dx.doi.org/10.1021/la5018664 | Langmuir 2014, 30, 9115−9122
Langmuir FEDL =
Article
2πλDR [(σT 2 + σS2)e−2D / λD + 2σTσSe−D / λD] ε0ε
Although the strong presence of Ca2+ could be explained by binding to silica,34 XPS showed no evidence of exposed silica, and this hypothesis cannot explain the removal of Ca2+ by K+.
(4)
■
where R represents the tip radius; ε0 represents the dielectric permittivity of vacuum; ε is the relative permittivity of the solution; D is the tip−substrate distance; σT and σS are the surface charge densities of the tip and substrate, respectively; and λD represents the Debye length ⎛ 2F 2I × 103 ⎞−1/2 λD = ⎜ ⎟ ⎝ ε0εRT ⎠
CONCLUSIONS We have shown a new way to investigate the effect of ionic solutions on benzene-covered surfaces. We have measured the adhesion between benzene-functionalized tips and benzenefunctionalized substrates in water and chloride solutions of Na+, K+, Mg2+, and Ca2+. In solutions of 100 mM NaCl and MgCl2, there was minimal effect on adhesion, compared with pure water. Adhesion decreased considerably in 100 mM KCl, and XPS showed that potassium was present in the near-surface region, which indicates that K+ binds inside the benzene layers. This makes the surfaces positively charged and decreases the adhesion between them. In 100 mM CaCl2 solution, adhesion was considerably higher than in the other solutions examined. However, when the substrate was dried and put under a vacuum for XPS experiments, no Ca2+ was observed in the near-surface region. Thus, we propose that hydrated Ca2+ ions form bridges between the tip and the substrate, thereby increasing adhesion. If Ca2+ left when the sample was dried for analysis in a vacuum, this could explain why these measurements showed no Ca2+ in the near-surface region. Overall, adhesion forces were found to decrease in the order Ca2+ > Mg2+ ≥ Na+ > K+, which nicely follows the Hofmeister series, Ca2+ > Mg2+ > Na+ > K+, suggesting that interactions between benzene surfaces in ionic solutions are controlled by ion hydration. Our results probed aspects of the interaction between hydrophobic benzene molecules, such as those found in biological systems, that are influenced by specific ions in aqueous solution. Ion specificity for certain molecules or surfaces, where these molecules are present, can be exploited by organisms, where carbon-containing molecules are responsible for ion transport and dissolved metals, such as Ca, Mg, K, and Na, help to control the behavior of proteins.
(5)
In this relationship, F represents the Faraday constant, and I is the ionic strength in molarity (M), which is multiplied by 103 to convert it to units of mol/m3. R is the gas constant, and T is the temperature. In the 100 mM KCl solution, the Debye length is 0.97 nm. The tip and substrate are identical in composition, so their surface charges can be assumed to be the same, and eq 4 can be simplified to FEDL =
4πλDRσ 2 −2D / λD + e−D / λD) (e ε0ε
(6)
From eq 6, we estimated σ = 15 mC/m2. We can rewrite the surface charge in terms of the number of K+ ions bound to each benzene molecule on average by σ K+ per benzene = eρ (7) where e represents the elementary charge and ρ is the number of benzene molecules per square meter. The surface area per thiol over a perfect CH3-terminated SAM would be 0.215 nm2,20 which leads to ρ = 1/0.215 nm2 = 4.65 × 1018 m−2. Because the benzene group is bulkier than CH3, the surface area for our SAM might be larger, resulting in a lower value for ρ. The number of K+ ions per benzene is then about 0.02 = 2%. If we assume that each K+ binds between two benzene rings to form a sandwich, then 4% of the benzene rings would share a K+. The XPS experiments showed about one K+ for every 10 benzene molecules, which is about five times more than the DLVO calculations suggest. However, the expression for FEDL (eq 6) is known to be uncertain for D < λD, and the XPS analysis requires a vacuum, meaning that the sample must be dehydrated, whereas the CFM measurements were made in solution. Considering the sources of difference, the agreement is reasonable. XPS showed no Ca2+ on the benzene surfaces, which indicates that Ca2+ does not bind strongly to the benzene layer, as K+ does. Ca2+ has a larger hydrated radius than K+, that is, its water bonding is stronger. For Ca2+ to link benzene rings, this water would have to be expelled, but this is unlikely because of the strong hydration of Ca2+. One hypothesis is that hydrated Ca2+ acts as bridges between the tip and substrate, thereby increasing adhesion. If the hydration energy of Ca2+ is higher than the energy gained from binding to the benzene layer, the Ca2+ would be lost when the sample was dried and set under a vacuum. That could explain why we observed no Ca2+ using XPS. The observation that adhesion force did not return to its original value when the CaCl2 solution was replaced with water indicated that Ca2+ binds so strongly between the two benzene layers that it cannot be removed by rinsing with distilled water. When K+ is added, it binds in the layers, which makes them charged so the Ca2+ bond weakens and Ca2+ is removed.
■
ASSOCIATED CONTENT
S Supporting Information *
Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org/.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS We thank the NanoGeoScience group members for encouragement and discussion, Klaus Bechgaard for advice on the synthesis of the benzenethiol compound, and Keld West and Maria Bruun Pedersen for general support in the laboratory. This work was part of the Nano-Sand Project, funded by the BP Exploration Operating Company Limited, under the Pushing Reservoir Limits ExploRe Program. A small contribution came from the Engineering and Physical Sciences Research Council (EPSRC Program Grant EP/I001514/1) for the Materials Interface with Biology (MIB) consortium. 9121
dx.doi.org/10.1021/la5018664 | Langmuir 2014, 30, 9115−9122
Langmuir
■
Article
Distributions in Molecular Assemblies. J. Am. Chem. Soc. 1995, 117, 7943−7951. (23) Friddle, R. W.; Noy, A.; De Yoreo, J. J. Interpreting the Widespread Nonlinear Force Spectra of Intermolecular Bonds. Proc. Natl. Acad. Sci. U.S.A. 2012, 109, 13573−13578. (24) Hutter, J. L.; Bechhoefer, J. Calibration of Atomic-Force Microscope Tips. Rev. Sci. Instrum. 1993, 64, 1868−1873. (25) Bourg, M. C.; Badia, A.; Lennox, R. B. Gold-Sulfur Bonding in 2D and 3D Self-Assembled Monolayers: XPS Characterization. J. Phys. Chem. B 2000, 104, 6562−6567. (26) Stipp, S. L. S.; Hochella, M. F., Jr.; Parks, G. A.; Leckie, J. O. Cd2+ uptake by calcite, solid-state diffusion, and the formation of solid solution: Interface processes observed with near-surface sensitive techniques (XPS, AES and LEED). Geochim. Cosmochim. Acta 1992, 56, 1941−1951. (27) Arnold, D.; Plank, C.; Erickson, E.; Pike, F. Solubility of Benzene in Water. Ind. Eng. Chem. Chem. Eng. Data Ser. 1958, 2, 253− 255. (28) Dejarguin, B. V.; Landau, L. Theory of the Stability of Strongly Charged Lyophobic Sols and of the Adhesion of Strongly Charged Particles in Solution of Electrolytes. Acta Physicochim. URSS 1941, 14, 633−662. (29) Verwey, E. J. W.; Overbeek, J. T. G. Theory of the Stability of Lyophobic Colloids; Elsevier: Amsterdam, 1948. (30) Ederth, T.; Claesson, P.; Lieberg, B. Self-Assembled Monolayers of Alkanethiolates on Thin Gold Films as Substrates for Surface Force Measurements. Long-Range Hydrophobic Interactions and Electrostatic Double-Layer Interactions. Langmuir 1998, 14, 4782−4789. (31) Lifshitz, E. M. The Theory of Molecular Attractive Forces between Solids. Sov. Phys. JETP 1956, 2, 73−83. (32) Dzyaloshinskii, I. E.; Lifshitz, E. M.; Pitaevskii, L. P. The General Theory of van der Waals Forces. Adv. Phys. 1961, 10, 165− 209. (33) Beattie, J. K.; Djerdev, A. M.; Warr, G. G. The Surface of Neat Water is Basic. Faraday Discuss. 2009, 141, 31−39. (34) Wang, X.; Lee, S. Y.; Miller, K.; Welbourn, R.; Stocker, I.; Clarke, S.; Casford, M.; Gutfreund, P.; Skoda, M. W. A. Cation Bridging Studied by Specular Neutron Reflection. Langmuir 2013, 29, 5520−5527.
REFERENCES
(1) Hofmeister, F. Zur Lehre von der Wirkung der Salze. Arch. Exp. Pathol. Pharmakol. (Leipzig) 1888, 24, 247−260. (2) Pedersen, K. O. On Charge and Specific Ion Effects on Sedimentation in the Ultracentrifuge. J. Phys. Chem. 1958, 62, 1282− 1290. (3) Ginsburg, A.; Carroll, W. R. Some Specific Ion Effects on the Conformation and Thermal Stability of Ribonuclease. Biochemistry 1965, 4, 2159−2174. (4) Mohsen, J. Effect of Sodium and Magnesium on Kinetics of Potassium Release in Some Calcareous Soils of Western Iran. Geoderma 2008, 145, 207−215. (5) Skovbjerg, L. L.; Hassenkam, T.; Makovicky, E.; Hem, C. P.; Yang, M.; Bovet, N.; Stipp, S. L. S. Nano Sized Clay Detected on Chalk Particle Surfaces. Geochim. Cosmochim. Acta 2012, 99, 57−70. (6) Zou, B.; McCool, C. S.; Green, D. W.; Willhite, G. P. A Study of the Chemical Interactions between Brine Solutions and Dolomite. SPE Reservoir Eval. Eng. 2000, 3, 209−215. (7) Lo Nostro, P.; Ninham, B. W. Hofmeister Phenomena: an Update on Ion Specificity in Biology. Chem. Rev. 2012, 112, 2286− 2322. (8) Conway, B. E. Ionic Hydration in Chemistry and Biophysics; Studies in Physical & Theoretical Chemistry; Elsevier Science: Amsterdam, 1981. (9) Skou, J. C. The Influence of Some Cations on an Adenosine Triphosphatase from Peripheral Nerves. Biochim. Biophys. Acta 1957, 23, 394−401. (10) Hall, J. E.; Guyton, A. C. Textbook of Medical Physiology, 11th ed.; Elsevier Saunders: Philadelphia, PA, 2006. (11) Sayyed, F. B.; Suresh, C. H. Quantitative Assessment of Substituent Effects on Cation−π Interactions Using Molecular Electrostatic Potential Topography. J. Phys. Chem. A 2011, 115, 9300−9307. (12) Kumpf, R. A.; Dougherty, D. A. A Mechanism for Ion Selectivity in Potassium Channels: Computational Studies of Cation−π Interactions. Science 1993, 261, 1708−1710. (13) Duan, M.; Song, B.; Shi, G.; Li, H.; Ji, G.; Hu, J.; Chen, X.; Fang, H. Cation⊗3π: Cooperative Interaction of a Cation and Three Benzenes with an Anomalous Order in Binding Energy. J. Am. Chem. Soc. 2012, 134, 12104−12109. (14) Sunner, J.; Nishizawa, K.; Kebarie, P. Ion−Solvent Molecule Interactions in the Gas Phase. The Potassium Ion and Benzene. J. Phys. Chem. 1981, 85, 1814−1820. (15) Rodriguez-Cruz, S. E.; Williams, E. R. Gas-Phase Reactions of Hydrated Alkaline Earth Metal Ions, M2+(H2O)n (M = Mg, Ca, Sr, Ba and n = 4−7), with Benzene. J. Am. Soc. Mass Spectrom. 2001, 12, 250−257. (16) Cheng, Y.-H.; Liu, L.; Fu, Y.; Chen, R.; Li, X.-S.; Guo, Q.-X. Counterion Effects on the Cation−π Interaction between Alkaline Earth Cations and Benzene. J. Phys. Chem. A 2002, 106, 11215−11220. (17) Hassenkam, T.; Pedersen, C. S.; Dalby, K.; Austad, T.; Stipp, S. L. S. Pore Scale Observation of Low Salinity Effect on Outcrop and Oil Reservoir Sandstone. Colloids Surf. A 2011, 390, 179−188. (18) Hilner, E.; Andersson, M. P.; Hassenkam, T.; Matthiesen, J.; Stipp, S. L. S. The Influence of Salinity on the Adhesion of a Nonpolar Model Oil Component to Reservoir Sandstone. Energy Fuels, to be published. (19) Hassenkam, T.; Mitchell, A. C.; Pedersen, C. S.; Skovbjerg, L. L.; Bovet, N.; Stipp, S. L. S. The Low Salinity Effect Observed on Sandstone Model Surfaces. Colloids Surf. A 2012, 403, 79−86. (20) Love, J. C.; Estroff, L. A.; Kriebel, J. K.; Nuzzo, R. G.; Whitesides, G. M. Self-Assembled Monolayers of Thiolates on Metals as a Form of Nanotechnology. Chem. Rev. 2005, 105, 1103−1169. (21) Frisbie, C. D.; Rozsnyai, L. F.; Noy, A.; Wrighton, M. S.; Lieber, C. M. Functional Group Imaging by Chemical Force Microscopy. Science 1994, 265, 2071−2074. (22) Noy, A.; Frisbie, C. D.; Rozsnyai, L. F.; Wrighton, M. S.; Lieber, C. M. Chemical Force Microscopy: Exploiting Chemically-Modified Tips to Quantify Adhesion, Friction, and Functional Group 9122
dx.doi.org/10.1021/la5018664 | Langmuir 2014, 30, 9115−9122
