Published: May 16, 2011
r 2011 American Chemical Society 11192 dx.doi.org/10.1021/jp202254u | J. Phys. Chem. C 2011, 115, 1119211203
SAMs under Water: The Impact of Ions on the Behavior of Waterat Soft Hydrophobic SurfacesAdam J. Hopkins,*,, Cathryn L. McFearin,^, and Geraldine L. Richmond*,,
Department of Chemistry and Materials Science Institute, University of Oregon, Eugene, Oregon, United States
bS Supporting Information
Since hydrophobic surfaces are by denition water hating,one might assume from a simplistic perspective that solvated ions inthe water would avoid a hydrophobic surface. While this might betrue for idealized solid hydrophobic surfaces,13 it is certainly notthe case for solvated ions near more complex biological assemblieswhere ions not only approach but also transport across theseboundary layers. Unraveling the various factors that inuence ionbehavior in these complex biological systems relative to theiridealized solid systems is not an easy task. For example, what rolesdo the polar groups of a biological assembly play, or that the moreuid nature of the hydrophobic boundary layer plays? At the air/water and solid/liquid interfaces, the polarity of adsorbates such asproteins andmacromolecules alters the structure of interfacial waterand aects ion adsorption behavior.46 Recent studies of aqueousionic solutions near simple hydrophobic liquids are shedding lighton the eect of ions on water at an extended hydrophobic uidinterface.711 From these spectroscopic7,8 and computationalstudies911 it is clear that inorganic ions do penetrate into theaqueous/hydrophobic liquid interface. Experimental support forthese conclusions comes from spectroscopically observed changesof the molecular orientation and hydrogen bonding of interfacialwater molecules when ions are present at the liquid/liquid interface.These studies show that the weak interactions between water andthe hydrophobic liquid assist in the formation of an interfacialpotential that creates an environment that can draw ions into
the interfacial region. The behavior is found to be quite distinctfrom what has been observed for interfacial water in similarexperimental1219 and computational studies conducted at theair/water interface.17,2030 Comparison of the experimental resultsobtained at the air/water and CCl4/water interface using similarions and techniques show that there is a greater tendency for theions to go to the organic liquid/water interface than the air/waterinterface. The case has recently been made that other ions, such asOH, are also attracted to hydrophobic liquid interfaces because ofthe stabilization aorded by the approach of the OH ions, whichreduces the dielectric constant at the interface and the dipoleuctuations.31,32 These ions are estimated to have a density max-imum some 67 from the Gibbs dividing surface at the liquid/liquid and air/water interfaces.31,32
Recently, there has been an increased interest in usinghydrophobic self-assembled monolayers (SAMs) chemicallyattached to a silica substrate as model systems for understandingwater and ion behavior at more complex soft hydrophobicsurfaces.33 If the behavior of water molecules at the SAM/waterjunction is largely determined by the nature of the water/hydrophobic interactions at the terminus of the monolayer, thenone can use such SAM/water systems as a model for exploring a
Received: March 9, 2011Revised: April 19, 2011
ABSTRACT: Understanding the behavior of water at hydrophobic surfaceshas been a topic of much interest for many decades. In most areas of biological,environmental, or technological relevance, the aqueous phase is not pure water,but comprises a host of ions including those associated with the acidity orbasicity of the solution. The notion that ions, including hydroxide and/orhydronium, accrue at hydrophobic interfaces is increasingly invoked as apossible explanation for the behavior of water adjacent to soft hydrophobicinterfaces such as liquids and monolayers. The focus of this study is onexploring the behavior of aqueous solutions of salts, acids and bases in contactwith hydrocarbon and uorocarbon self-assembled monolayers (SAMs) usingvibrational sum frequency spectroscopy (VSFS). The studies take a systematicapproach to understanding how each component of the SAMs interfacescontribute to the overall observed behavior of ions and water in the overallboundary region. To achieve this, the spectroscopy of the SAM/water interface in the presence and absence of aqueous phase ions,acids and bases is compared with similar measurements taken at the substrate (SiO2)/water interface and the hydrophobic liquid/water interface. The results show that the behavior of water and ions at the SAM/aqueous interface is signicantly inuenced by thesubstrate surface for both hydrocarbon and uorocarbon SAM systems. Conditions where water and ions near a SAM interfacemimic that of a liquid hydrophobic surface are identied.
11193 dx.doi.org/10.1021/jp202254u |J. Phys. Chem. C 2011, 115, 1119211203
The Journal of Physical Chemistry C ARTICLE
variety of ionic andmolecular adsorption processes atmore complexhydrophobic interfaces of biological and technological importance.One caveat to this premise is that these SAMsystems are themselvescomplex with regards to water and ion adsorption. For example,defects in the monolayers could lead to water and ions penetratingto the SiO2 substrate, potentially altering the overall interfacialelectrostatics and subsequent behavior of ions and water moleculesat the terminus of the monolayer.
The focus of this study is on exploring the behavior of aqueoussolutions of salts and acids in contact with hydrocarbon anduorocarbon SAMs using vibrational sum-frequency spectroscopy(VSFS). The studies take a systematic approach to understandinghow each component of the SAM/water interface contributes to theoverall observed behavior of ions and water in the boundary region.Spectroscopic comparison of hydrocarbon- and uorocarbon-termi-nated SAMs exploits the sensitivity of VSFS to dierences in themolecular interactions of water with these two chemical functional-ities at the interface. We are able to distinguish between the behaviorof water at the terminus of these twomonolayer systems from that ofH2O in other regions of the interface, such as those closer to thesubstrate or deeper into the bulk liquid. Contributions from thesubstrate are determined by examining and comparing the VSFSresponse from SiO2 in contact with various ion and acid solutions.Further insights into the inuence of the substrate and monolayerson interfacial water and electrolytes are obtained from both newstudies of acids and bases at the CCl4/H2O interface and previousstudies7,8,3436 of water and aqueous solutions of ions in contact withvarious hydrophobic organic liquids where a substrate is not a factor.
VSFS is uniquely suited for these studies because of itsinherent sensitivity to the orientation and dynamics of watermolecules in the interfacial region of monolayers and otherhydrophobic surfaces.3745 Detailed explanations of VSFS canbe found in the literature;4652 thus, a concise description will begiven here. The VSFS experiments within utilize a xed frequencyvisible beam (vis) that is overlapped spatially and temporally witha tunable infrared (IR) beam (IR) at the interface, whichgenerates a beam at the sum of the two incident frequencies(SF). Using the electric dipole approximation, this beam isgenerated only by molecules in a non-centrosymmetric environ-ment such as that at the solid/liquid interface. No contribution tothe VSFS beam originates from molecules in the bulk liquid orsolid environments, and thus, the signal is limited to the fewmolecular layers experiencing net orientation at the interface.
The intensity of SF is given by eq 1:
ISF j2eff j2IIRIvis n
2NR j2 1
where I(IR) is the intensity of the IR beam and I(vis) is theintensity of the visible beam. e
(2) is the eective macroscopicnonlinear susceptibility and is composed of a sum of resonantterms, i
(2), and a nonresonant term, NR(2). NR
(2) was found to besmall in the experiments within. Each resonant term can bedescribed by a discrete resonance of the form shown in eq 2.
i IR ii2
In eq 2, N is the number density of molecules, and Ai is theproduct of the Raman and IR transition moments. i is thefrequency of the IR transition,IR is the frequency of the tunable
IR, and i is the line width of the transition. These quantitiesthat are normally complex result in terms with both amplitudeand phase that can interfere with other vibrational terms. Theresonant macroscopic susceptibility, i
(2), is related to themolecular hyperpolarizability () by eq 3,
in which 0 is the permittivity of free space and the angle bracketsdenote an ensemble average over the dierent molecular orienta-tions. Choosing dierent polarizations of the visible, IR, and VSFspectra allows dierent elements of i
(2) to be investigated. Thespectra within have all been collected in the SSP polarizationscheme (S-VSF, S-vis, P-IR), which probes transition dipolemoment components normal to the interfacial plane.
The spectra have been analyzed using a nonlinear global curve-tting routine in IgorPro (Wavemetrics, Beaverton, OR) whichaccounts for spectral interferences and the deviations from idealtransition line shape by convoluting the expressions for Gaussian andLorentzian curves as initially developed by Bain and co-workers.53,54
The global t allows all spectra within a series to be analyzedsimultaneously. Each resonant mode in eq 1 is tted to eq 4:
2i Ai expiiZ 0
1i IR ii
exp i IRi
4where Ai , i , i , Li , and i are ttable parameters for theamplitude, phase, resonant frequency, Lorentzian line width, andGaussian line width of the ith mode. To reduce the number ofttable parameters, Lorentzian line widths were xed at 2 cm1
for CH modes and 5 cm1 for OH modes, and the resonantphases were only allowed values of 0 or .
Liquid/Liquid Experiments. The CCl4/H2O experimentspresented here were performed using a commercially availableVSFS system from Ekspla (Vilnius, Lithuania) and sample cellthat has been previously described.7,34 For the VSFS experi-ments, a 532 nm visible beam from the frequency doubled outputof picosecond Nd:YAG laser is overlapped spatially and tempo-rally with the output of an IR generator pumped by the samesource. The IR is tunable from 1000 to 4300 cm1. The beamenergies used were 80 J and200250 J for the visible and IRbeams, respectively. The beam angles were chosen to maintain atotal internal reflection geometry for the visible beam for allconcentrations examined: 69.5 and 75 for the visible and IRbeams, respectively. Multiple spectra of each interface were takenand averaged to achieve an acceptable signal-to-noise level. Eachspectrumwas normalized for the IR and 532 nm output as well asthe absorption of the prism.The VSFS spectra are extremely sensitive to chemical impu-
rities, so great care was taken to eliminate any eects from these.CCl4 (g99.9% , Chromosolv HPLC grade) was purchasedfrom Sigma-Aldrich and twice distilled before use. The HCl andNaOH were both purchased from Sigma-Aldrich. The HCl wasACS reagent grade, 37%, andwas used from the bottle formakingacidic solutions. NaOHpellets, 99.998%metals basis purity, wereused for the basic experiments. Although of high-grade purity,using the NaOH pellets as purchased yielded inconsistent results
11194 dx.doi.org/10.1021/jp202254u |J. Phys. Chem. C 2011, 115, 1119211203
The Journal of Physical Chemistry C ARTICLE
and often gave spectra similar to those containing small con-centrations of surfactants. To overcome this, the NaOH pelletswere baked in an oven at 220 for 12 h prior to use. Freshsolutions were prepared each day, and the pH was tested prior touse with indicator strips. The sample cell and all glassware werecleaned in NoChromix (Godax Laboratories) dissolved in con-centrated sulfuric acid, then rinsed copiously with water from aNanopure II system.Solid/Liquid Experiments. VSFS experiments at solid/liquid
interfaces were performed using a custom-built IR generationsystem. The master oscillator is a Spectra-Physics Lab 110 Nd:YAG laser (10 Hz, 600 mJ, 6.5 ns, 1064 nm) the output of whichis split to pump a potassium titanyl phospate (KTP) frequencydoubler and an optical parametric oscillator (OPA). The fre-quency doubled output of 60 mJ is used to pump a double-passKTP optical parametric oscillator (OPO). The remainder of the532 nm doubled output is then attenuated and polarizationselected before being sent toward the interface. Typically, 1.5 mJof 532 nmwas used at the sample area. OPO output is mixed with120 mJ of the 1064 nm fundamental in a two-stage (four-crystal) potassium titanyl arsenate OPA. The tunable OPAoutput ranges from 2600 cm1 to 4000 cm1 (14 mJ) with a2 cm1 bandwidth.
Samples were compression tted onto a Kel-F sample cell witha Kalrez O-ring. IR and visible beams were directed at the samplethrough a 23.1, triangular IR grade fused silica prism with a 1 in.2top surface (ISP Optics). Beam angles relative to the opticalbench surface were 17 and 23.1 for the IR and 532 nm,respectively. Samples were mated to the prism via a drop ofmicroscope immersion oil (Cargille, custom code 43421). Thespectra were normalized for IR adsorption of the prism andsample plate.The two types of coatings used in this paper are FDS and ODS
and are shown in Figure 1. FDSmonolayers were deposited using1H,1H,2H,2H-peruorodecyltriethoxysilane (FDES) (Gelest,Tullytown, PA) precursors via LangmuirBlodgett (LB) deposi-tion on a KSV Minitrough. After verifying the trough cleanlinessby measuring the surface tension of water (72.4 ( 0.2 mN/m),the trough was lled with 0.01 M hydrochloric acid to catalyzethe cleavage of the ethoxy groups. After immersing a substrate inthe dipping well, a 13 M solution of FDES precursors inCHCl3, the exact concentration determined by mass, was spreadon the surface of the liquid layer to a starting density of 1molecule per 45 2. This surface layer was then equilibrated for30 min. The barriers were then compressed to a surface pressureof 15 mN/m (27 2 mean molecular area) at which point thesubstrate was withdrawn from the trough at a rate of 5 mm/min.ODS monolayers were prepared using a solution phase
deposition technique similar to methods previously published.55
First, a freshly cleaned SiO2 substrate was humidied for 30 minover a saturated solution of Ca(NO3)2. The substrate was thensoaked in a 4:1 (vol:vol) solution of hexadecane/CCl4 with a1 mM concentration of octadecyltrichlorosilane (Gelest) for 6 h.The ODS sample was then rinsed with hexadecane, chloroform,acetone, and methanol and sonicated for 1 min in CHCl3,followed by rinsing with acetone and methanol. Lastly the SAMswere blown dry with argon. Both FDS and ODS monolayerswere cured at 110 C for 1 h. All samples were stor...