pH and Surface Charge Switchability on Bifunctional Charge

Jan 2, 2018 - Phone: 785-532-6371 (D.A.H.)., *E-mail: [email protected]. Phone: 804-828-1298 (M.M.C.). Cite this:Langmuir 34, 2, 663-672. Abstract. ...
0 downloads 12 Views 2MB Size
Download PDF
Article Cite This: Langmuir XXXX, XXX, XXX−XXX

pubs.acs.org/Langmuir

pH and Surface Charge Switchability on Bifunctional Charge Gradients Kayesh M. Ashraf,† Md Rezaul K. Khan,† Daniel A. Higgins,*,‡ and Maryanne M. Collinson*,† †

Department of Chemistry, Virginia Commonwealth University, Richmond, Virginia 23284-2006, United States Department of Chemistry, Kansas State University, Manhattan, Kansas 66506-0401, United States

‡

S Supporting Information *

ABSTRACT: Multifunctionalized pH-sensitive silica gradients containing acidic and basic functional groups have been prepared to evaluate how the spatial arrangement of active sites on a surface influences the surface charge and pH switchability. The gradient surfaces were prepared using controlled rate infusion in such a manner that the individual gradients in the strong acid (sulfonic acid) and in the weak base (propylamine) align, whereas a gradient in the weakly acidic silanol groups opposes them. The relative amounts of the three species were varied by controlling the composition of the deposition solution, whereas the hydrophobicity of the underlying surface was set by using base layer-coated substrates prepared from either tetramethoxysilane or tetramethoxysilane/octyltrimethoxysilane mixtures. Results from X-ray photoelectron spectroscopy confirm that aligned gradients are formed in both amine and sulfonic acid groups, and the relative amounts bound to the surface follow that expected from the solution composition. Water contact angle measurements show a 40°−50° change across the length of the gradient, the exact values being dependent on the hydrophobicity of the base layer. Zeta potential measurements on gradient mimics reveal that there is a pH where the net charge on the gradient surface is predicted to have a constant but nonzero value. Static contact angle measurements and modeling confirm this prediction. At a pH acidic of this value, the gradient in charge runs in one direction, whereas at a pH basic of this value, the gradient in charge runs in the other direction. This point can be strategically moved from acidic values to basic values by changing the relative amounts of acidic and basic functionalities on the surface. The origin of this unique pH switchability can be found in acid−base chemistry. By modeling the charge along the gradient surface using a simple equilibrium model, a distribution of pKa values were noted in these materials.

■

INTRODUCTION

ing oppositely charged weak electrolytes, such as poly(acrylic acid) and poly(2-vinylpyridine) or plasma polymerized amine and carboxylic acid polymers.12−15 Multifunctionalized acid−base silica materials have also served as heterogeneous catalysts to enhance coupling reactions.7,16−18 The acidic and basic catalytic components separately activate two different substrates, ultimately yielding a lower energy reaction pathway. Reactions such as hydrolysis of an acetal and subsequent Henry reaction,17,19 aldol condensation,18,20 and Knoevenagel reactions,21 for example, can be accelerated by using a surface-bound catalyst that has both the components (acid and base) acting cooperatively. The acid component will activate one component for

Multicomponent charge gradients provide the means to create a continuously varying surface charge density along a single substrate. Such charge gradients can be used as a smart tool to evaluate how subtle changes in surface chemistry and charge can influence the adsorption of proteins, platelets, DNA, and other biological species on a surface in one single experiment.1−6 Surface charge gradients can also provide a high throughput means to evaluate acid−base chemistry on a surface as well as obtain much deeper insights into cooperative effects between different functional groups on surfaces, which are critically important in separation science and catalysis.7−9 Examples of charge gradients that have been prepared include −N(CH3)3+ gradients formed by attaching amine-terminated reagents (−NH2) followed by quaternarization,10 or NH3+ gradients via protonation.11 Two-component charge gradients have been prepared from polyelectrolyte brushes, incorporat© XXXX American Chemical Society

Received: July 5, 2017 Revised: December 11, 2017

A

DOI: 10.1021/acs.langmuir.7b02334 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

ethanol, TMOS, OTMOS, HCl, and water in a 400:200:70:70:70 volume ratio. Each sol was then spin-coated on the clean Si wafer at 5000 rpm for 30 s. The base layer-coated substrates were dried via desiccation overnight. Sols used to form the gradients were freshly prepared. The aminosilane sol was prepared by mixing x mL (x = 0.5, 1, or 2) of APTEOS with 20 mL of ethanol and y mL (y = 0.125, 0.25, or 0.5 mL, respectively) of water. The mercaptosilane sol was prepared by stirring x mL (x = 0.125, 0.25, or 0.5) of MPTMOS with 20 mL of EtOH and y mL (y = 0.125, 0.25, or 0.5 mL, respectively) of 0.1 M HCl for 5 min using a magnetic stirrer. Then, z mL (z = 0.12,5, 0.25 or 0.5 mL) of 0.3 M NH4OH was added to the mixture and stirred for additional 25 min. NH4OH increases the pH of the sol and helps catalyze the condensation reaction of the mercaptosilane to the surface silanol groups. The sol was used 90 min after addition of the base. A total of six different samples were prepared for the experiment, as shown in Table 1. Two different base layers were used for making

electrophilic addition, while the base component will increase the nucleophilicity of the other substrate.7,16−18,22 Bifunctional surface imprinting of silica with thiol and amine has also been reported as a means to spatially organize functional groups.23 In a previous work, we have strategically modified a silica surface with both sulfonic acid and amine functional groups in such a fashion so as to create aligned and opposed gradients.24 The gradients were prepared via the controlled rate infusion (CRI) method,25,26 where a gradient of amine on a tetramethoxysilane (TMOS)-derived base layer-coated substrate was prepared first, followed by a gradient of thiol functional groups over the amine gradient. The second gradient was prepared either in an aligned or opposed direction relative to the amine gradient. Depending on whether the individual gradients were aligned or opposed to each other, the gradients exhibited entirely different properties in terms of their wettability and net surface charge.24 The most interesting results were observed on the aligned multicomponent gradient mimics, which showed a distinct point of constant charge (PCC), where the zeta potential (ζ) along the gradient at pH ≈ 6.6 was constant.24 In the present work, we have studied in detail the origin of this unique PCC both in terms of its charge and pH by varying the ratio of the acidic and basic functionalities. To further modulate the surface properties, the aligned gradients were prepared on two different base layer-coated silicon wafers. One was chosen to be hydrophilic and prepared from a sol containing hydrolyzed TMOS, and the other was chosen to be hydrophobic and prepared from a sol containing octyltrimethoxysilane (OTMOS). The gradients were characterized by X-ray photoelectron spectroscopy (XPS) to see how the components vary across the surface. Wettability was evaluated from static water contact angle (WCA) measurements, which also further confirmed the gradient formation. The variation in the local charge density along the length of the samples was studied by an electrokinetic analyzer, which measured the zeta potential (ζ) at different pH values. The distribution of charge along the length of the bifunctionalized samples strongly depended on the type and arrangement of the charged groups at the interface and can be modeled using simple acid−base equilibrium equations. The model not only predicts the presence of a distinct PCC for aligned gradients but also indicates that there is a distribution of pKa values on the surface. The stimulus-responsive materials explored in these studies have numerous applications, including being used as tools to regulate droplet motion, to study and control protein adsorption and biofouling, and to investigate cooperative interactions between neighboring groups, which are indispensable to the field of cooperative catalysis and separation science.

■

Table 1. Volumes of the Two Organosilanes Used To Make the Different Gradient Samples sample

APTEOS (mL) in 20 mL EtOH

MPTMOS (mL) in 20 mL EtOH

volume ratio R = (NH2/SH)

T1:0.5 T2:0.25 T1:0.125 T0.5:0.125 C81:0.5 C82:0.25

1 2 1 0.5 1 2

0.5 0.25 0.125 0.125 0.5 0.25

2 8 8 4 2 8

these gradients; one was chosen to be hydrophilic, for example, the TMOS-derived base layer, and the other was chosen to be hydrophobic, for example, the OTMOS-derived base layer. For clarity, each sample is designated as T or C8, where T stands for the TMOS-derived base layer-coated substrate and C8 stands for the OTMOS-derived base layer. The subscripts indicate the volumes (mL) of silanes used to make the gradients and the uniformly modified substrates. The first number indicates the volume of aminosilane, and the second number indicates the mercaptosilane volume in each of the deposition solutions. So T1:0.25 indicates that 1 mL of amine was mixed with 20 mL of ethanol to prepare the amine sol, and separately 0.25 mL of MPTMOS was mixed with 20 mL of ethanol to the prepare the sulfhydryl (SH) sol. Gradient Preparation. The multicomponent gradients on the base layer-coated substrates were prepared by CRI in the manner depicted by Scheme 1. CRI involves infusing a silica sol into a vial containing a vertically aligned substrate.25,26 In this work, the aminosilane is deposited first followed by mercaptosilane, as the surface-bound amine helps catalyze the condensation of the hydrolyzed mercaptosilane to the substrate, as shown in other work.27 A syringe pump (Harvard Apparatus, PHD 2000 infusion) controls the rate of infusion of the modifying sol. During infusion, the base layer-coated substrate is exposed to the aminosilane solution for 30 min along its length, thus producing a gradual spatial variation in amine modification from the bottom to the top.25,26 The amine gradient was then modified with the thiol in a similar fashion. In this case, the glass vial contains the vertically aligned amine gradient in such a fashion so that the two gradients (amine and thiol) align in the same direction. These gradients are called aligned multicomponent gradients. The time gap between the two infusions was 12 h. To convert these gradients into multicomponent charge gradients, the substrates were immersed in 30% H2O2 at 65 °C for 20 min to convert SH to SO3− and simultaneously NH2 to NH3+, as previously described.24 These gradients were then characterized using XPS, WCA, and zeta potential measurements. Characterization. XPS was performed with a Thermo Fisher ESCALAB 250 imaging X-ray photoelectron spectrometer [Al Kα (1486.68 eV), 500 μm spot size, 50 eV pass energy, and 0.1 eV step

EXPERIMENTAL SECTION

TMOS (98%), OTMOS (99%; C8) 3-aminopropyltriethoxysilane (APTEOS, 99%), and 3-mercaptopropyltrimethoxysilane (MPTMOS, 98%) were purchased from Acros Organics and used as received. Silicon substrates (University Wafer, B-doped, ⟨111⟩) were sonicated with ethanol and Millipore water for 5 min and then cleaned with piranha solution (Caution: piranha solutions are extremely dangerous and react violently with organic materials) via sonication for 10 min followed by thorough rinsing with water. The substrate was then spincoated with either a TMOS-based sol or a hybrid sol prepared from OTMOS and TMOS in an ∼1 to 3 volume ratio. The TMOS-based sol was prepared by mixing ethanol, TMOS, 0.1 M HCl, and water in a 400:200:70:70 volume ratio. The hybrid sol was prepared from B

DOI: 10.1021/acs.langmuir.7b02334 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

were obtained from the Helmholtz−Smoluchowski equation and reflect the charge on the siloxane sample relative to the PP reference. The ζ of the PP reference was measured as a function of pH and then subtracted from the measured ζ for the full channel at each pH to obtain the accurate ζ for the substrate alone. The ζ and isoelectric point (IEP) of the PP reference sheet was consistent with the Anton Paar electrokinetic analyzer specifications (IEP = 4 ± 0.2).

Scheme 1. Fabrication of Bifunctional Charge Gradients on Base Layer-Coated Substrates of Different Polarities by Programmed CRI

■

RESULTS AND DISCUSSION Gradient Fabrication. Gradient fabrication begins by spin-coating a silica sol prepared from either TMOS or a C8/ TMOS mixture onto a silicon or glass substrate to form a homogeneous base layer for covalently attaching the organoalkoxysilanes.25 Use of the TMOS or C8/TMOS sol affords control over surface hydrophobicity. Using CRI, the base layercoated substrate was exposed first to a solution containing APTEOS and then after 12 h to a solution of MPTMOS, both in a time-dependent fashion. As a result of the time-dependent condensation reaction between the surface silanol groups and the hydrolyzed APTEOS or MPTMOS precursors, the organic groups were covalently bonded to the substrate in a gradient fashion such that the top of the substrate has mostly surface silanol groups and the bottom has amine, thiol, and unreacted silanol groups. Because the surface-immobilized amine groups catalyze the condensation of the mercaptoalkoxysilanes,24 it is likely that the mercapto and amine functionalities are located near each other. By changing the concentration of the APTEOS and MPTMOS in the original silane solution, four multifunctional chemical gradients on TMOS-derived base layers and, for proof-of-principle, two on C8/TMOS-derived base layers were prepared. The simultaneous oxidation of the SH to sulfonate groups and protonation of the amine groups to form NH3+ groups was achieved via immersion in 30% H2O2 for 20 min. In these materials, it is also possible that an opposing gradient of surface silanol groups is present. Gradient Characterization. XPS. The presence and extent of modification of the surfaces with NH3+ and SO3− as well as the degree of protonation/oxidation were evaluated using XPS. Figure 1 shows an overlay of six N 1s and S 2p XPS spectra for the T1:0.5 and T2:0.25 charge gradients acquired at ∼3 mm

size]. The samples were placed on top of the conducting tape on a 5 cm × 2 cm sample holder. XPS spectra were acquired at constant intervals (typically every 3 mm) across the wafer, starting ∼3 mm from the edge. The spectra were calibrated by taking the C 1s peak as 284.6 eV. Quantification of the N 1s and S 2p multiplex peak areas was done by commercially available software (Avantage version 4.4). For the calculation of the normalized peak area, Wagner sensitivity factors were used. Contact angle measurements were made with a Rame-Hart contact angle goniometer by the sessile drop method. The volume of the drop of water (or buffer) was 1 μL. Zeta potentials (ζ) were determined using an Anton-Paar SurPASS instrument and a clamping cell, including a standard electrokinetic analyzer, as previously described.24 Samples used for ζ measurements were 5 cm × 2.5 cm. Flat polypropylene (PP) sheets (0.0125 in., 5 × 2 cm2, McMaster-Carr) were used as a reference. Streaming potentials were measured in the pH range of 3−10, as monitored by two Ag/AgCl electrodes at 20 ± 2 °C and at a maximum pressure difference of 500 mbar. The average value of four measurements was recorded. The electrolyte was 0.001 M KBr (99%) (Sigma-Aldrich). The ζ values

Figure 1. XPS N 1s and S 2p spectra in the T1:0.5 (A,B) and T2:0.25 (C,D) charge gradients. Both the samples exhibit a gradual increase in the intensity of S and N from the top of the substrate (red) to the bottom (purple), as depicted by the arrows. C

DOI: 10.1021/acs.langmuir.7b02334 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir intervals along the film length. The N 1s peak in the XPS spectra shown in Figure 1A appears as a doublet when fitted, giving rise to binding energies of ∼400 and ∼402 eV, which correspond to free amine and protonated/H-bonded amines, respectively.28 As expected, and as is consistent with the previous work, the peak intensity is higher for the protonated form of the amine group.24 Upon oxidation with hydrogen peroxide, the thiol becomes SO3−, and nearby amine groups become protonated. Neighboring SO3− groups act as counterions for nearby NH3+ groups.24 The disappearance of the SH peak at ∼162.5 eV indicates that SH groups have been oxidized, but some free amine remains, as evidenced by the N 1s peak at 400 eV (Figure 1A,C). This result indicates that the surface concentration of thiol is lower than that of the amine and is consistent with the (2×, 4×, 8×) lower concentration of thiol relative to amine in the respective deposition solutions. For S 2p in the charged gradients, the spectra consist of one peak at ∼168 eV (see Figure 1B,D), in agreement with the reported values for SO3−.29−31 For both gradients, the N 1s and S 2p signals increase from the top to the bottom of the substrate, indicative of the presence of multicomponent gradients aligned in the same direction. Note that for sample T1:0.5, some SH remains unoxidized, as shown by a small shoulder near 163 eV. It is likely that the 20 min immersion time in H2O2 was not sufficient to oxidize all SH groups, given the relatively high concentration of SH present in the T1:0.5 sample. Thus, this particular sample may also contain SH groups. Figure S1 shows the N 1s and S 2p XPS spectra along the length of the substrate for C81:0.5 and C82:0.25. Similar to T samples, C8 samples also exhibit a N 1s doublet. The free amine is at ∼400 eV, and the protonated amine is at ∼402 eV. As noted in the T samples, the majority of the amine is in the protonated form, but unprotonated amine groups are still present on the surface. The S 2p peak is found to be at 168.5 eV, representing the S in the SO3− group. These peaks were found to increase from the top of the gradients to the bottom, indicating the formation of aligned gradients of N and S on the more hydrophobic base layer. The peak intensity for N 1s is much higher on the TMOS-derived base layer compared to the C8-derived base layer. The lower degree of amine modification in the C8 samples is attributed to the steric effect of the bulky C8 organic group as well as a reduced amount of surface silanol groups.32 Likewise, the peak intensity of the S 2p peak is also higher on the TMOS-derived base layer-coated substrate. Because it is possible that amine groups can be oxidized by H2O2, an additional experiment was carried out. In this experiment, two control samples were prepared by soaking the TMOS-derived base layer-coated substrate in amine (APTEOS/EtOH; 1:20 mL) for 30 min. One of the control samples was then soaked in H2O2 for 20 min. The other was not. The N 1s XPS spectra acquired for the H2O2-treated and untreated samples were overlaid, as shown in Figure S2. As can be seen, no change in the position and shape of the N 1s peak (∼400 eV) in the H2O2-treated and untreated samples was observed. Furthermore, no additional peaks can be seen for the oxidized amine groups. For example, N 1s occurs in nitrate at 406 eV33 and in nitrite at 405 eV.33 Nitroso also occurs at 400 eV.34 However, it is deemed unlikely because there is no change in the peak intensity at 400 eV before and after treatment. These results confirm that H2O2 does not oxidize the amine functional group but only the sulfur.

To better understand the gradient profile, we plotted the integrated areas under the N 1s and S 2p peaks for all T and C8 gradient samples, and these results are shown in Figure 2A,B. This figure also includes the result from the T1:0.25

Figure 2. Variation of the XPS peak area for nitrogen (N, black solid lines) and sulfur (S, red solid lines) as a function of distance along the gradient samples. (A) T samples; (B) C8 samples. (A) Yellow square (T2:0.25), purple circle (T1:0.25), red diamond (T1:0.5), blue triangle (T1:0.125), and dark green circle (T0.5:0.125) and (B) as mentioned in the plot.

sample24 for comparison. For all gradient profiles, the amount of both N and S increases from the top (25 mm) to the bottom (0 mm). This is expected, as the bottom was exposed to the sols for a longer period of time than the top. Samples prepared with the same amount of APTEOS in the sols T1:0.5 (red diamond line) and T1:0.25 (purple circle line; from ref 24) show similar integrated peak areas that are lower than that found for the sample prepared with double the amount of APTEOS [T2:0.25 (yellow square)], as expected. It can also be noted in Figure 2 that the N and S surface concentrations are higher in the T samples compared to the C8 samples, as mentioned earlier. Also noteworthy is that the amount of immobilized SH on T2:0.25 is greater than that observed on T1:0.25. The surfaceimmobilized amine groups catalyze the condensation of thiol, resulting in a greater amount of SH on the surface prepared with the higher amine (T2:0.25).27 The normalized peak area of nitrogen to sulfur using Wagner sensitivity factors along with instrument factors were obtained, and the values ranged from 2 for the T1:0.5 sample to 8 for the T1:0.125 sample. The values for the T2:0.25 and T0.5:0.125 samples were 4.5 and 6.7, respectively. No significant variation was observed with distance. WCA Analysis (Wettability). WCA measurements were performed to evaluate the wettability of the multicomponent charge gradients as a function of distance. To see the spatial variation across the gradient, the WCAs were acquired at 3 mm intervals lengthwise on the four representative samples: T2:0.25 (A), C82:0.25 (B), T1:0.5 (C), and C81:0.5 (D), as shown in Figure S3. As expected, the gradients on the TMOS-derived base layer were more hydrophilic than those prepared on the D

DOI: 10.1021/acs.langmuir.7b02334 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir C8-derived base layer-coated substrate. Likewise, the contact angle was low on the highly modified end (e.g., bottom) because of the presence of the charged species, whereas the least modified end (e.g., top) shows a character more like the siloxane base layer. The contact angles gradually increase from 13° and 12° at the bottom to 43° and 41° at the top of the gradient for T2:0.25 and T1:0.5, respectively. The C8-derived base layer was significantly more hydrophobic with a WCA of 92°− 96° at the top compared to 46°−47° at the bottom (more modified end). Collectively, both XPS and WCA analyses confirm the presence of a gradient along the length of both substrates. Electrokinetic Measurements. Electrokinetic measurements are a particularly valuable means to obtain detailed information regarding the charge at various locations along the gradient surface. The net charge depends on the functional groups present on the surface, their location relative to each other, and the pH of the solution. In this work, the measurement of streaming potentials resulting from the pressure-driven flow of an electrolyte solution across a substrate was used to evaluate the surface charge through measurement of the zeta potential (ζ) via the Helmholtz− Smoluchowski equation ζ=

ΔE κη ΔP εε0

(1)

where ΔE is the streaming potential in volts, ζ is the zeta potential in volts, η is the viscosity of the solution in pascal seconds, κ is the solution conductivity in siemens per meter, ε is the dielectric constant of the solution, ε0 is the dielectric permittivity of vacuum, and ΔP is the pressure in Pa.35 From the measurement of ΔE and ΔP, the zeta potential can be obtained because all except these two variables are constants. The ζ calculated in this manner has been used to characterize charged surfaces including glass modified with aminosilanes,36 silica nanoparticles,37 and aminosilane gradients.11,24 For flat substrates, which were the subject of analysis in the current work, a commercial clamping cell was used. With a gradient substrate, however, only the average ζ can be measured.24 To obtain information about the surface charge at discrete positions along the gradient, six uniformly modified samples (size: 2 × 5 cm2) were prepared, each of which represents a certain surface chemical composition along the length of the multicomponent gradient. Table 2 summarizes

Figure 3. Change of ζ with pH for the gradient mimics (samples 1 → 6) prepared on the TMOS-derived base layers. (A−E) correspond to T1:0.125, T1:0.25, T1:0.5 to T2:0.25, and T0.5:0.125, respectively. The gradient bar to the left and right of the profiles represents the direction of modification. B was reported in previous work (ref 24).

SiOH groups exhibit a pKa of ≈ 8.39 On a surface, however, the pKa values of acidic and basic functionalities are different from their solution values and can also depend of the local chemical environment.40−43 Using chemical force microscopy, the effective pKa for surface-immobilized 2-aminoethanethiol was estimated to be 7, ∼3 orders of magnitude lower than the solution values.41 For acids, the reverse is true: an increase in effective pKa is observed. For propylsulfonic acid immobilized on mesoporous silica, in particular, the pKa was found to be 2.8 compared to 1.6 in solution.44 Because all three groups are ionizable, all three will contribute to the overall surface charge and pH switchability. As can be seen in Figures 3 and S4, the zeta potential shifts negative as the pH of the solution increases and the surface bound species becomes deprotonated, leading to the formation of SO3−, SiO−, and NH2. The larger the amount of SO3− on the surface, as in the case of T1:0.5 or C81:0.5, the more negative the surface is at an acidic pH; the zeta potential−pH profile approaches that obtained for the base layer-coated substrate, as shown by the black line. Also noteworthy in each of the profiles is the appearance of a crossover at a point, which we term as PCC, where the ζ of all six samples is the same. This

Table 2. Exposure Time in the Silane Solutions sample id time in APTEOS sol, min time in MPTMOS sol, min

1 5 5

2 10 10

3 15 15

4 20 20

5 25 25

6 30 30

the length of time the substrate was exposed to each of the silanes for the six samples. Sample 6 mimics the bottom or the highly modified end of the gradient, whereas sample 1 mimics the top or the least modified end. As before, the samples were immersed in H2O2 for 20 min to oxidize/protonate the surface. The ζ as a function of pH for the mimics of T and C8 gradient samples are presented in Figures 3 and S4, respectively. As described in the previous work, ζ largely depends on the concentration of the charged species on the surface, their pKa values, their relative position, and the solution pH.24 In solution, the pKa values of SO3H and NH3+ are ∼−2.0 and ∼10.0, respectively,20,38 whereas the majority of E

DOI: 10.1021/acs.langmuir.7b02334 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir PCC shifts its position depending on the composition of the surface and the siloxane base layers, as discussed in more detail below. At pH values acidic of the PCC, the surface charge becomes more positive with increasing modification, whereas at values basic of the PCC, the surface charge becomes more negative with increasing modification. This result is described in more detail below. Figure 3A−C shows the ζ−pH profile from ∼3 to ∼10 when the volume of APTEOS in the original sol was kept constant and that of MPTMOS decreased (R = 2 → 8). As the pH of the solution increases, the zeta potential becomes more negative because of the deprotonation of the surface-bound functional groups, leading to the formation of SO3−, SiO−, and NH2. When the surface concentration of protonated amine groups is high as in Figure 3A,B, the zeta potential is fairly positive at acidic pH. As R decreases to 2, the ζ becomes significantly more negative, particularly at low pH values, and the ζ−pH profile approaches that of the base layer (black curve). These surfaces are richer in negatively charged SO3− groups because the volume of MPTMOS in the deposition sol was increased by 2−4-fold relative to APTEOS. Also evident is a shift in the pH at the IEP to more acidic values as R decreases. Figure 3D,E shows the ζ−pH profiles for samples T2:0.25, and T0.5:0.125, where the volume of aminosilane is decreased by a factor of 4 and the volume of the mercaptosilane is decreased by a factor of 2. As the ratio R decreases from 8 to 2 and the amount of surface-bound amine drops, the curves shift toward that acquired for the base layer-coated substrate, particularly at pH < 8. The pH at the IEP for the different surfaces also shifts to more acidic values as the ratio R decreases. Most noteworthy is that the pH at the PCC significantly shifts from a basic value to an acidic value as the amount of amine on the surface decreases (Figure 3D vs 3E). The ζ−pH profiles for the two samples prepared on the more hydrophobic base layers, for example, C81:0.5 and C82:0.25, are presented in Figure S4A,B, respectively. Results similar to that described for the T samples are observed. The shape of the ζ versus pH curves depends on the ratio R. The sample with the greater amount of amine on the surface has a more positive charge relative to the one with a lower fraction. Likewise, sample C81:0.5 having a smaller R exhibits ζ behavior similar to the base layer-coated substrate, and the pH at the IEPs shifts to more acidic values. Again, the pH at the PCC also noticeably shifts to more acidic values with a decrease in R. To better evaluate how the ζ and pH at the PCC change with the composition of charged groups on the surface, these values were obtained from the plots shown in Figures 3 and S4. Each line was fitted using linear regression and the intersection points X (pH) and Y (zeta potential) for all unique pairs of lines determined. The individual averages and standard deviations for X and Y intersection points were determined from those values, and the results are shown in Figure 4. They are very informative. The most straightforward comparison can be made of the samples prepared with the same amount of amine but decreasing volumes of mercaptosilane (0.5 mL → 0.125 mL). As can be seen in Figure 4 (depicted by the long arrow), as the surface-bound amine stays approximately constant while SO3− decreases (R increases), the pH at the PCC shifts slightly from 6.4 → 7.3, and the zeta potential becomes increasingly positive. For the R = 4 samples (red bars in Figure 4), it can be

Figure 4. Zeta potential and pH at the PCC as a function of the ratio of NH3+/SO3− (R).

seen that ζ at the PCC is more positive for the sample with the lower amount of SO3− on the surface (T0.5:0.125 vs T1:0.25) even though the amount of amine on the surface is also smaller. Also, the pH at the PCC is significantly more basic for the sample with the higher amount of amine (T1:0.25) even though the amount of SO3− on the surface is slightly higher. Likewise, samples with R = 8 (green bars in Figure 4) can be compared. The sample with the higher amount of amine (T2:0.25) on the surface results in a more basic pH at the PCC. The ζ, however, is more negative, undoubtedly because of the greater amount of SO3− on the surface. What these data indicate is that the pH at the PCC is predominately determined by the amine groups because under these conditions (pH at PCC ≈ 4.5−9), sulfonic acid will be deprotonated (SO3−) on all seven surfaces. The extent of protonation/deprotonation of amine and silanol groups, however, is very sensitive to the pH in this region. As evident in Figure 4, the high surface density of amine groups as in T2:0.25 leads to a more basic PCC, whereas the low density of amine groups as in T0.5:0.125 leads to a much more acidic PCC. The net charge at the PCC for each of these seven surfaces, however, is very sensitive to the amount of SO3− on the surface. When the amount of SO3− groups on the surface is the lowest, the zeta potential is the most positive (e.g., T1:0.125 and T0.5: 0.125). The amine groups and silanol groups also contribute to the net charge as evident upon comparison of T1:0.25 and T2:0.25, but not as strongly as the SO3−. Theoretical modeling is consistent with these statements (see below). In the case of C8 samples, for example, C81:0.5 and C82:0.25, the pH at that PCC becomes more basic from 4.5 to 6.9 as the amount of amine on the surface increases, whereas the ζ changes slightly to more negative values. The higher amine modification leads to higher thiol condensation on the surface, thereby increasing the coverage of SO3− groups on the surface. Thus, the ζ for the C82:0.25 sample is slightly more negative at the PCC than it is for the C81:0.5 sample. The variation of ζ with surface modification (R) at two different pH values (3 and 9) was analyzed, and the results are shown in Figure 5. For simplicity, we focused on the T and C8 samples (T2:0.25, T0.5:0.125 and C82:0.25, C81:0.5). At acidic pH, the ζ values are positive because of the relatively high concentration of amine groups (and hence NH3+) on the surface. Most noteworthy, ζ becomes more positive as the extent of modification increases (sample 1 → sample 6). This is in direct contrast to that observed at pH 9, which depicts a negative slope. Under these conditions, ζ becomes increasingly negative as the extent of modification increases. At pH = 9, the surface charge is dominated by the SO3− groups as well as deprotonated silanol groups. For the T2:0.25 sample, which has a PCC of ∼9.1, the slope is near 0. That is, ζ is largely invariant F

DOI: 10.1021/acs.langmuir.7b02334 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

Figure 5. Changes in ζ with the degree of modification for the gradient mimics (samples 1 → 6) prepared on the TMOS and C8-derived base layers at a low pH (pH = 3) and a high pH (pH = 9). As can be seen, the charge gradients present at pH 3 and 9 oppose each other.

with the degree of modification of the surface. In other words, electrokinetic measurements obtained on gradient mimics indicate that when the pH of the solution is at the PCC, no charge gradient will exist. Furthermore, the direction of the charge gradient could be switched by simply altering the pH of a contacting aqueous solution. To further prove that at the PCC, no charge gradient exists and the six gradient mimics represent an actual gradient, static contact angles were measured for two gradient films (T2:0.25 and C82:0.25) at the pH corresponding to the PCC for each (9.1 and 6.9, respectively) as well as at pH 3 and ∼5−6 (Milli-Q water) for comparison. On the basis of the data obtained on the gradient mimics, at the pH corresponding to the PCC, no gradient in wettability should be observed because the charge is constant along the length of the substrate. However, at pH values more acidic or basic than this value, a gradient in charge and thus wettability should be observed. The results shown in Figures S5 and 6 confirm this hypothesis. At pH 3 (malonic acid buffer) and 5.5 (Milli-Q water), a gradient in the contact angle for both T and C8 samples is observed. At pH 5.5 (MilliQ water), the WCA changes from 104° and 44° to 50° and 13° for C8 and T samples, respectively. However, at pH ∼PCC (6.9 and 9.1), WCAs are constant along the length of the substrate: ∼92° and ∼32° for C8 and T, respectively. Acid−Base Model. The origins of the PCC were further explored by modeling of the charge along the gradient surface. A simple equilibrium model was employed, assuming the surface is exposed to an aqueous solution. Surface charge was assumed to arise from covalently bound −NH3+, −SO3−, and −Si−O− groups. Each of these was taken to be in equilibrium with their conjugate weak base or acid, that is, −NH2, −SO3H, and −Si−OH. For simplicity, it was assumed that all are active. The surface density of each charged species was then calculated as follows Q NH + = FN 3

[H+] [H+] + Ka NH3+

Figure 6. WCA profile for T2:0.25 and C82:0.25 samples at various pH values. For pH 3, a malonic acid buffer was used, whereas a N-(2hydroxyethyl)piperazine-N′-ethanesulfonic acid buffer was used for pH 6.9. For pH 9.1, a boric acid buffer was used.

Q SO − = FS 3

Q SiO− = FSi

KaSO3H +

[H ] + KaSO3H

(3)

KaSiOH [H ] + KaSiOH

(4)

+

Here, Fi represents the formal surface density of the nitrogen-, sulfur-, and silanol-based species. The other parameters have their usual meaning. The three expressions were combined to determine the surface charge density Q = Q NH + − Q SO − − Q SiO−

(2)

3

G

3

(5) DOI: 10.1021/acs.langmuir.7b02334 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

Figure 7. Surface charge as a function of solution pH for the gradient compositions shown in the insets (FN blue, FS red, and FSi black) for three different volume ratios of precursor silanes. For the data shown in A−C, each group in the model is assumed to exhibit a single, homogeneous pKa value. The pKa exhibited by the −NH3+ groups was assumed to be 7.3, that for −SO3H, 2.8 and that for the −Si−OH groups, 6.4. In the model shown in D−F, the pKa values were allowed to vary over a defined range, as described in the text. The colored lines indicate the charge observed at different positions along the gradient. The gradient bars to the left and right of the profiles depict the direction of each gradient, as shown in Figures 3 and S4.

values exhibited by the −NH3+ groups were assumed to be uniformly distributed from 5.3 to 9.3, those for −SO3H from 1.8 to 3.8, and those for the −Si−OH groups from 4.4 to 8.4. The slope of the curves near the PCC are clearly less steep and more consistent with the experimental data when the pKa values are distributed over a small range. Again, a distinct PCC is present in the aligned gradients, the position of which depends on the gradient compositions. This PCC is not present in the opposed gradients obtained experimentally24 or modeled herein using the acid−base theory (Figure S6). Origin and Significance of the PCC. Cooperative interactions between the acidic (silanol groups, sulfonic acid) and basic functionalities (amine) lead to the presence of a distinct pH where the zeta potential (e.g., charge) was constant with position or composition. In many respects, the PCC is similar to an IEP, except that the average charge is not zero, but rather a finite value that can range from ∼−28 to ∼−110 mV, depending on the surface composition. This finite negative charge arises in part from the strongly acidic functional groups that remain deprotonated over the pH titration range. This unique form of cooperativity is only found in aligned gradients because the individual groups that define a particular region are positioned in a manner that promotes favorable interactions with each other on a given substrate. On an opposed gradient, the strongly acidic and basic sites are at opposing ends of the surface and thus may only have the ability to cooperatively interact in a small segment located in or near the middle of the substrate, if their respective concentrations are high enough. Importantly, the PCC can be moved from acidic (∼4) to basic pH values (∼9) by changing the relative amounts of strong acid, weak acid, and weak base on the surface. At pH values acidic or basic of the PCC, the surface charge increases or decreases with the extent of modification,

Figure 7(A−C) shows the modeled surface charge density in charge/nm2 as a function of solution pH for the gradient compositions shown in the insets (FN blue, FS red, FSi black) for three different volume ratios of precursor silanes (A−C). Point 0 corresponds to the position where the surface density of the added functional groups is low, and point 10 corresponds to the position where it is high. In this model, each group is assumed to exhibit a single, homogeneous pKa value. The pKa exhibited by the −NH3+ groups was assumed to be 7.3, that for −SO3H, 2.8 and that for the −Si−OH groups, 6.4. What is immediately evident is that all curves intersect at a single value (e.g., the PCC), confirming that its appearance is expected for aligned gradients of acidic and basic moieties. Also consistent with what is observed experimentally is that the PCC shifts to more basic values as the ratio (R) becomes larger. Upon close comparison of Figure 7(A−C)with Figures 3 and S4, it is clear that the overall shape of the modeled versus experimental plots are different. Near the PCC, the change in the charge density versus pH is significantly steeper for the modeled curves compared to the experimental curves. Previous work has established that the extent of protonation/ deprotonation on a surface can be much more complicated than that in solution.40−43 It thus may not be appropriate to represent each ionizing species with a single pKa value, given the subtle differences in solvation and local chemical environments that may result from lateral interactions between charged groups along the length of the gradient films.45 Figure 7(D−F) shows the plot of the modeled surface charge densities in charge/nm2 as a function of position, for model gradients in which the weak acid/based groups have a range of pKa values. These were obtained for the gradient profiles (FN, FS, and FSi) shown in the insets. Here, the pKa H

DOI: 10.1021/acs.langmuir.7b02334 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir respectively. The ability to switch not only the magnitude of charge but also the direction of the charge gradient is of particular importance to the development of pH-responsive smart materials.14 Unique to the present study is the presence of a PCC where cooperativity between the charged groups on the surface (SO3−, NH3+, and SiO−) keeps the net charge constant. Although a chemical gradient is still present, no charge gradient exists.

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. Phone: 785-532-6371 (D.A.H.). *E-mail: [email protected]. Phone: 804-828-1298 (M.M.C.).

■

CONCLUSIONS Multicomponent pH-sensitive gradients have been prepared by incorporating sulfonic acid groups and amine groups on hydrophilic or hydrophobic silica-modified surfaces. Such gradients are chemically complex, incorporating an intricate distribution of NH3+, SO3−, and SiO− species from one end of the substrate to the other. The addition of hydrophobicity into the base layer can further serve as a tool to modulate the surface properties. In conjunction with pH, the relative concentrations of the different surface-immobilized functional groups determine the net charge on the surface as well as the steepness (profile) of the charge gradient. These concentrations can be easily changed by changing the volume of the organosilane in the depositing solution and by changing the chemical composition of the base layer. As expected, increasing the SO3− fraction on the surface leads to more negative zeta potentials, whereas increasing the NH3+ fraction on the surface leads to more positive zeta potentials. Interestingly, however, each of the gradient films depicts a unique pH, where the net charge along the length of the substrate is approximately constant. Modeling using simple acid−base equilibrium equations establishes that this point can only be observed on aligned bifunctional gradients because of the cooperative interactions between the acidic and basic functional groups on the surface. This unique pH switchability of surface charge demonstrated herein is particularly valuable for control over the movement of charged species along a surface as well as to regulate the droplet motion and control protein adsorption. For example, a strongly anionic surface obtained when the pH is above the PCC should exhibit a reduction in adsorption for those proteins that have a net negative charge. Similarly, a zwitterionic-like surface, which is present at the PCC, may also resist protein adsorption. In addition, the ability to spatially arrange acidic and basic sites in close proximity to each other is essential to their application as cooperative catalysts to enhance the selectivity of organic reactions. Charge gradients provide a time-effective approach to study not only cooperativity but also electrostatic surface interactions, which are fundamentally important in ion chromatography. Future directions will proceed in this manner.

■

charged density versus solution pH for opposed gradients (PDF)

ORCID

Daniel A. Higgins: 0000-0002-8011-2648 Maryanne M. Collinson: 0000-0001-6839-5334 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS M.M.C. and D.A.H. gratefully acknowledge support by the U.S. National Science Foundation (DMR-1404805 and DMR1404898). The authors acknowledge NSF CHE-0820945 MRI Program for acquisition of an X-ray photoelectron spectrometer for research and education at VCU. We also acknowledge the support of the VCU Nanomaterials Core Characterization (NCC) facility and Dr. Dmitry Pestov for his help with the XPS data acquisition. We also thank Professor Kenneth Wynne for use of the electrokinetic analyzer and Kallol Roy for his assistance with control samples.

■

REFERENCES

(1) Genzer, J. Surface-Bound Gradients for Studies of Soft Materials Behavior. Annu. Rev. Mater. Res. 2012, 42, 435−468. (2) Genzer, J.; Bhat, R. R. Surface-bound soft matter gradients. Langmuir 2008, 24, 2294−2317. (3) Lin, X.; He, Q.; Li, J. Complex polymer brush gradients based on nanolithography and surface-initiated polymerization. Chem. Soc. Rev. 2012, 41, 3584. (4) Ruardy, T. G.; Schakenraad, J. M.; van der Mei, H. C.; Busscher, H. J. Preparation and characterization of chemical gradient surfaces and their application for the study of cellular interaction phenomena. Surf. Sci. Rep. 1997, 29, 3−30. (5) Li, Z.; Ashraf, K. M.; Collinson, M. M.; Higgins, D. A. Single Molecule Catch and Release: Potential-Dependent Plasmid DNA Adsorption along Chemically Graded Electrode Surfaces. Langmuir 2017, 33, 8651−8662. (6) Hulander, M.; Lundgren, A.; Faxälv, L.; Lindahl, T. L.; Palmquist, A.; Berglin, M.; Elwing, H. Gradients in surface nanotopography used to study platelet adhesion and activation. Colloids Surf., B 2013, 110, 261−269. (7) Margelefsky, E. L.; Zeidan, R. K.; Davis, M. E. Cooperative catalysis by silica-supported organic functional groups. Chem. Soc. Rev. 2008, 37, 1118−1126. (8) Brunelli, N. A.; Venkatasubbaiah, K.; Jones, C. W. Cooperative catalysis with acid−base bifunctional mesoporous silica: impact of grafting and co-condensation synthesis methods on material structure and catalytic properties. Chem. Mater. 2012, 24, 2433−2442. (9) Huh, S.; Chen, H.-T.; Wiench, J. W.; Pruski, M.; Lin, V. S.-Y. Cooperative catalysis by general acid and base bifunctionalized mesoporous silica nanospheres. Angew. Chem., Int. Ed. 2005, 44, 1826−1830. (10) Lin, Y. S.; Hlady, V. The desorption of ribonuclease A from charge density gradient surfaces studied by spatially-resolved total internal reflection fluorescence. Colloids Surf., B 1995, 4, 65−75. (11) Kannan, B.; Nokura, K.; Alvarez, J. C. Fabrication of Surface Charge Gradients in Open-Tubular Capillaries and Their Characterization by Spatially Resolved Pulsed Streaming Potential Measurements. Langmuir 2013, 29, 15260−15265.

ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.langmuir.7b02334. N 1s high resolution XPS spectra acquired on the C8 gradient, uniformly modified amine-modified substrates with and without treatment with hydrogen peroxide, zeta potential titration curves on the C8 gradient mimics, photographs of water droplets and their corresponding contact angle profiles on the TMOS and C8-derived base-coated substrates, and modeled I

DOI: 10.1021/acs.langmuir.7b02334 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir (12) Ionov, L.; Houbenov, N.; Sidorenko, A.; Stamm, M.; Luzinov, I.; Minko, S. Inverse and reversible switching gradient surfaces from mixed polyelectrolyte brushes. Langmuir 2004, 20, 9916−9919. (13) Ionov, L.; Houbenov, N.; Sidorenko, A.; Stamm, M.; Minko, S. Smart microfluidic channels. Adv. Funct. Mater. 2006, 16, 1153−1160. (14) Mierczynska, A.; Michelmore, A.; Tripathi, A.; Goreham, R. V.; Sedev, R.; Vasilev, K. pH-tunable gradients of wettability and surface potential. Soft Matter 2012, 8, 8399. (15) Tai, F.-I.; Sterner, O.; Andersson, O.; Ekblad, T.; Ederth, T. pH-control of the protein resistance of thin hydrogel gradient films. Soft Matter 2014, 10, 5955−5964. (16) Li, P.; Cao, C.-Y.; Liu, H.; Yu, Y.; Song, W.-G. Synthesis of a core−shell−shell structured acid−base bifunctional mesoporous silica nanoreactor (MS-SO3H@MS@MS-NH2) and its application in tandem catalysis. J. Mater. Chem. A 2013, 1, 12804−12810. (17) Huang, Y.; Xu, S.; Lin, V. S.-Y. Bifunctionalized mesoporous materials with site-separated Bronsted acids and bases: Catalyst for a two-step reaction sequence. Angew. Chem., Int. Ed. 2011, 50, 661− 664. (18) Zeidan, R. K.; Hwang, S.-J.; Davis, M. E. Multifunctional heterogeneous catalysts: SBA-15-containing primary amines and sulfonic acids. Angew. Chem., Int. Ed. 2006, 45, 6332−6335. (19) Bass, J. D.; Solovyov, A.; Pascall, A. J.; Katz, A. Acid−base bifunctional and dielectric outer-sphere effects in heterogeneous catalysis: A comparative investigation of model primary amine catalysts. J. Am. Chem. Soc. 2006, 128, 3737−3747. (20) Zeidan, R.; Davis, M. The effect of acid−base pairing on catalysis: An efficient acid−base functionalized catalyst for aldol condensation. J. Catal. 2007, 247, 379−382. (21) Hruby, S. L.; Shanks, B. H. Acid−base cooperativity in condensation reactions with functionalized mesoporous silica catalysts. J. Catal. 2009, 263, 181−188. (22) Yu, C.; He, J. Synergic catalytic effects in confined spaces. Chem. Commun. 2012, 48, 4933−4940. (23) Bass, J. D.; Katz, A. Bifunctional surface imprinting of silica: Thermolytic synthesis and characterization of discrete thiol−amine functional group pairs. Chem. Mater. 2006, 18, 1611−1620. (24) Ashraf, K. M.; Giri, D.; Wynne, K. J.; Higgins, D. A.; Collinson, M. M. Cooperative Effects in Aligned and Opposed Multicomponent Charge Gradients Containing Strongly Acidic, Weakly Acidic, and Basic Functional Groups. Langmuir 2016, 32, 3836−3847. (25) Kannan, B.; Dong, D.; Higgins, D. A.; Collinson, M. M. Profile Control in Surface Amine Gradients Prepared by Controlled-Rate Infusion. Langmuir 2011, 27, 1867−1873. (26) Kannan, B.; Higgins, D. A.; Collinson, M. M. Aminoalkoxysilane reactivity in surface amine gradients prepared by controlled-rate infusion. Langmuir 2012, 28, 16091−16098. (27) Dewoolkar, V. C.; Kannan, B.; Ashraf, K. M.; Higgins, D. A.; Collinson, M. M. Amine-phenyl multi-component gradient stationary phases. J. Chromatogr. A 2015, 1410, 190−199. (28) Moses, P. R.; Wier, L. M.; Lennox, J. C.; Finklea, H. O.; Lenhard, J. R.; Murray, R. W. X-ray Photoelectron Spectroscopy Of AlkylamineSilanes Bound To Metal Oxide Electrodes. Anal. Chem. 1978, 50, 576−585. (29) Corum, L. E.; Hlady, V. Screening platelet−surface interactions using negative surface charge gradients. Biomaterials 2010, 31, 3148− 3155. (30) Ding, Y.-X.; Streitmatter, S.; Wright, B. E.; Hlady, V. Spatial variation of the charge and sulfur oxidation state in a surface gradient affects plasma protein adsorption. Langmuir 2010, 26, 12140−12146. (31) Shen, C.-H.; Lin, J.-C. Surface characterization and platelet compatibility evaluation of binary mixed self-assembled monolayers containing novel sulfonic acid terminated alkanethiol. Colloids Surf., B 2010, 79, 156−163. (32) Ashraf, K. M.; Wang, C.; Nair, S. S.; Wynne, K. J.; Higgins, D. A.; Collinson, M. M. Base Layer Influence on Protonated Aminosilane Gradient Wettability. Langmuir 2017, 33, 4207−4215. (33) Ivashenko, O.; van Herpt, J. T.; Feringa, B. L.; Browne, W. R.; Rudolf, P. Rapid reduction of self-assembled monolayers of a disulfide

terminated para-nitrophenyl alkyl ester on roughened Au surfaces during XPS measurements. Chem. Phys. Lett. 2013, 559, 76−81. (34) Batich, C. D.; Donald, D. S. X-Ray Photoelectron Spectroscopy Of Nitroso Compounds: Relative Ionicity Of The Closed And Open Forms. J. Am. Chem. Soc. 1984, 106, 2758−2761. (35) Bard, A. J.; Faulkner, L. R. Electrochemical Methods: Fundamentals and Applications; John Wiley & Sons, Inc.: New York, 2001. (36) Metwalli, E.; Haines, D.; Becker, O.; Conzone, S.; Pantano, C. G. Surface characterizations of mono-, di-, and tri-aminosilane treated glass substrates. J. Colloid Interface Sci. 2006, 298, 825−831. (37) Hayes, R. A.; Böhmer, M. R.; Fokkink, L. G. J. A Study of silica nanoparticle adsorption using optical reflectometry and streaming potential techniques. Langmuir 1999, 15, 2865−2870. (38) Hall, H. K. Correlation of the Base Strengths of Amines. J. Am. Chem. Soc. 1957, 79, 5441−5444. (39) Ong, S.; Zhao, X.; Eisenthal, K. B. Polarization of water molecules at a charged interface: second harmonic studies of the silica/water interface. Chem. Phys. Lett. 1992, 191, 327−335. (40) Creager, S. E.; Clarke, J. Contact-Angle Titrations of Mixed ωMercaptoalkanoic Acid/Alkanethiol Monolayers on Gold. Reactive Vs Nonreactive Spreading, and Chain Length Effects on Surface pKa Values. Langmuir 1994, 10, 3675−3683. (41) van der Vegte, E. W.; Hadziioannou, G. Acid−base properties and the chemical imaging of surface-bound functional groups studied with scanning force microscopy. J. Phys. Chem. B 1997, 101, 9563− 9569. (42) Vezenov, D. V.; Noy, A.; Rozsnyai, L. F.; Lieber, C. M. Force titrations and ionization state sensitive imaging of functional groups in aqueous solutions by chemical force microscopy. J. Am. Chem. Soc. 1997, 119, 2006−2015. (43) Kim, K.; Kwak, J. Faradaic impedance titration of pure 3mercaptopropionic acid and ethanethiol mixed monolayers on gold. J. Electroanal. Chem. 2001, 512, 83−91. (44) Cinlar, B.; Shanks, B. H. Characterization of the acidic sites in organic acid functionalized mesoporous silica in an aqueous media. Appl. Catal., A 2011, 396, 76−84. (45) Kholin, Y.; Zaitsev, V. Quantitative physicochemical analysis of equilibria on chemically modified silica surfaces. Pure Appl. Chem. 2008, 80, 1561−1592.

J

DOI: 10.1021/acs.langmuir.7b02334 Langmuir XXXX, XXX, XXX−XXX