Contrasting Mechanisms of Spontaneous Adsorption at Liquid–Liquid

Article Cite This: Langmuir 2018, 34, 6170−6182

pubs.acs.org/Langmuir

Contrasting Mechanisms of Spontaneous Adsorption at Liquid− Liquid Interfaces of Nanoparticles Constituted of and Grafted with pH-Responsive Polymers Dalin Wu and Andrei Honciuc* Institute of Chemistry and Biotechnology, Zurich University of Applied Sciences, Einsiedlerstrasse 31, 8820 Waedenswil, Switzerland

Downloaded via KAOHSIUNG MEDICAL UNIV on June 15, 2018 at 13:09:13 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

S Supporting Information *

ABSTRACT: Elucidating the mechanisms responsible for spontaneous adsorption of nanoparticles (NPs) at interfaces is important for their application as emulsifiers, bubble stabilizers, or foaming agents. In order to investigate the key factors that control the spontaneous adsorption of NPs at liquid−liquid interfaces, we synthesized seven different types of NPs from pH-responsive polymers poly(2-(diethylamino)ethyl methacrylate) (PDEAEMA) and poly(2-dimethylamino)ethyl methacrylate) (PDMAEMA) via surfactant-free emulsion polymerization or via “grafting from” polystyrene (PS) NPs. The dynamic interfacial tension (IFT) measurements at the toluene−water (Tol−H2O) interface reveal that when PDEAEMA and PDMAEMA are grafted from the surface of PS NPs the solubility of the grafted pH-responsive polymers in toluene is the key factor determining the NPs’ interfacial adsorption. Under acidic conditions (pH < 6.0), PDEAEMA and PDMAEMA are protonated and show no solubility in toluene, and as a result, the grafted NPs do not adsorb at the Tol−H2O interface. Oppositely, under basic conditions (pH > 7.0), PDMAEMA dissolves in toluene and therefore the PDMAEMA-grafted NPs can adsorb at the Tol−H2O interface. Interestingly, when NPs are constituted of PDEAEMA, they can adsorb spontaneously at the Tol−H2O interface under acidic conditions (pH < 6.0) but not under basic conditions (pH > 7.0). In this case, the key factor determining the NPs’ spontaneous adsorption at the Tol−H2O interface is the degree of softness of the NPs rather than the solubility of PDEAEMA in toluene. Furthermore, we found that the adsorption of NPs constituted of PDEAEMA- (pH 2.0−6.0) and PDMAEMA-grafted PS NPs (pH 7.0−10.0) at the Tol−H2O interface is a combination of diffusion-controlled and energy-barrier-controlled. The opposite trends observed for the interfacial attachment ΔE and activation energies Ea for the “constituted of” and “grafted from” NPs with pH suggest an opposite mechanisms of adsorption at the Tol−H2O interface. Finally, the synthesized NPs prove to be effective emulsifiers, where the phase of the Pickering emulsions can be changed dynamically by pH adjustment.

1. INTRODUCTION Understanding the mechanisms of adsorption of NPs at interfaces and the governing particle−interface interactions that eventually lead to their interfacial adsorption are of significant fundamental interest and also of great practical importance across many industries. Similar to surfactants and amphiphilic macromolecules, particles with appropriate surface properties, such as solvent wettability, can also adsorb spontaneously at liquid−liquid (L−L) interfaces, thus decreasing the interfacial tension, or in the presence of convection currents with the formation of Pickering emulsions.1−6 Currently, there is increased interest in designing particles that exhibit spontaneous, meaning both thermodynamic and kinetically favored, interfacial adsorption at the L−L interface from the bulk without any energy input, like from a sonotrode or high-speed homogenizer.7−11 The groups of Richtering and To Nagi have reported that the poly(N-isopropylacrylamide) (PNIPAM) microgel particles can spontaneously adsorb at the heptane−water interface,9,12 decreasing the interfacial tension (IFT) by 33 mN/m from 43 mN/m initially.13 It was shown © 2018 American Chemical Society

that the spontaneous adsorption of PNIPAM microgels at the heptane−water interface with a measurable decrease in the IFT values was essentially caused by the degree of softness and the deformation ability of the PNIPAM microgel particles in the temperature interval from 13 to 44 °C.14,15 The softer PNIPAM microgels caused a stronger reduction of the IFT, and this explanation was backed by the theoretical calculations and simulation by others.16,17 On the other hand, particles which were subsequently surface grafted with various polymer brushes, surfactants, or amphiphilic block copolymers can also adsorb spontaneously at the L−L interface with decreasing IFT.18,19 For example, Du et al.20 reported that the (1mercaptoundec-11-yl)tetra-(ethylene glycol) (TEG)-grafted gold NPs can adsorb spontaneously at the 2,2,3,3,4,4,5,5octafluoropentyl acrylate−water interface upon decreasing the initial IFT value from 25 to 12 mN/m. Similarly, Zhang et al.8 Received: March 16, 2018 Revised: April 19, 2018 Published: May 5, 2018 6170

DOI: 10.1021/acs.langmuir.8b00877 Langmuir 2018, 34, 6170−6182

Article

Langmuir

adsorbing irreversibly at the interface. In other words, in such case, the existing particle−interface interactions play a less significant role in the interfacial adsorption of particles than the magnitude of the external energy input. The types of Pickering emulsion (w/o and o/w) are mainly determined by the surface wettability of particles at the interface by the organic solvent and water.28 Generally speaking, particles with hydrophilic surfaces, for example, unmodified SiO 2 , having better wettability in water than in organic solvent, can generate o/w Pickering emulsions,29 while particles with hydrophobic surfaces, for example, PS particles, have better surface wettability in organic solvent than in H2O, which results in w/o Pickering emulsions.30 However, if the surface of a particle is too hydrophilic or too hydrophobic, then the activation energy for their L−L interfacial attachment may be too high to overcome by mechanical stirring, thus failing to emulsify and produce Pickering emulsions. NPs in this article, due to their ability to adsorb at the L−L interfaces, proved to be efficient emulsifiers that can be used to create pH-responsive Pickering emulsions from toluene and water.

found that the methylsilyl-capped SiO2 NPs can spontaneously adsorb at the air−water interface upon decreasing the IFT value from 72 to 60 mN/m in 1800 s, while unfunctionalized SiO2 did not adsorb at the air−water interface. While both soft and grafted particles appear to spontaneously adsorb at L−L interfaces, the underlying mechanisms of interfacial adsorption and decreasing IFT may be different. Here we attempt a comparative study with respect to the effectiveness of these two classes of NPs to decrease the IFT of the Tol−H2O interface. To the best of our knowledge, such comparative studies have not yet been performed using soft and grafted NPs created from the same or very similar polymers in one report. For this study, we prepared two types of NPs. The first type has a bulk composition based on poly(2-(diethylamino)ethyl methacrylate) (PDEAEMA), which will be referred to hereinafter as PDEAEMA-constituted NPs. PDEAEMA is a popular pH-responsive polymer with pKa = 7.0−7.3,21 which is often used in drug-delivery research.22 The PDEAEMA NPs change from hard to soft gradually with the pH value changing from 10.0 to 2.0 due to the protonation in aqueous solution. In addition, through controlling the cross-linking degree (1−3%) and copolymerizing strategy with a pH-inert and hydrophobic monomer, butyl methacrylate, the softness of PDEAEMA NPs could be effectively controlled. The second type of NPs was made by grafting solid polystyrene (PS) NPs with poly(2-dimethylamino)ethyl methacrylate) (PDMAEMA), which will be referred to hereinafter as PDMAEMA-grafted PS NPs. PDMAEMA is a polymer with a pKa value of 7.4−7.8,23 but it is a water-soluble polymer in both its protonated and deprotonated states (at low temperatures). Its hydrophilicity can further increase by protonation under acidic conditions (pH < 6.0). The two types of NPs exhibit contrasting behavior with respect to their L−L interfacial adsorption function of pH. The first type, PDEAEMA-constituted NPs, can decrease the IFT of Tol−H2O only under acidic conditions (pH < 6.0). The second type, the PDMAEMA-grafted PS NPs, behaved oppositely; namely, the IFT decreased only under basic conditions (pH > 7.0) but not under acidic conditions (pH < 6.0). Moreover, the decrease in the IFT value was more dramatic upon the adsorption of the latter type of NPs, ∼30 mN/m at pH 10.0, than in the case of the former ones, ∼15 mN/m in pH 2.0. Both PDEAEMA-constituted NPs and PDMAEMA-grafted PS NPs were excellent emulsifiers for toluene and water. Even more, the emulsion phase could be dynamically changed by changing the pH to either the Tol/H2O or H2O/Tol phase. This property could be a significant advantage in encapsulations and triggered release applications.24,25 The results obtained from the dynamic IFT measurements and Pickering emulsion phase correlate well with the particle surface properties. Another important advantage of Pickering emulsions, compared to surfactant-stabilized emulsions, is their high stability due to the strong attachment of particles at the L−L interface.26 For example, the adsorption free energy of a spherical solid NP of 10 nm diameter at the hydrocarbon−water interface (IFT = 50 mN/m) having a contact angle of 90° is 3.89 × 103 kT, which is much larger than that of the corresponding molecular surfactants, only slightly larger than the thermal energy of ∼1 kT at 293 K.27 In the presence of convection currents due to mechanical agitation in emulsification experiments, the NPs acquire high kinetic energy and increased bulk diffusivities and thus easily overcome the interfacial adsorption energy barriers

2. EXPERIMENTAL SECTION 2.1. Materials. Styrene (Sty) (>99%), divinylbenzene (DVB) (80%), 2-(dimethylamino)ethyl methacrylate (DMAEMA) (>98%), 2(diethylamino)ethyl methacrylate (DEAEMA) (>99%), 2,2′-azobis(2methylpropionamidine) dihydrochloride (V-50) (>97%), 2,2′-azobis(2-methylpropionitrile) (AIBN) (>98%), (vinylbenzyl)trimethylammonium chloride (VBTMAC) (>99%), 4-vinylbenzyl chloride (4VBC) (>90%), butyl methacrylate (BMA) (>99%), copper(I) bromide (99.999%), copper(II) bromide (>99%), ethanol (EtOH) (>99%), tetrahydrofuran (THF) (>99.9%), toluene (Tol) (>99%), ethylene glycol dimethacrylate (EGDMA) (>98%), ammonium hydroxide solution (NH4OH) (28%), N,N,N′,N″,N″-pentamethyldiethylenetriamine (PMDETA) (>99%), and aluminum oxide (Al2O3 basis) (⩾98%) were purchased from Sigma-Aldrich. Sty, DMAEMA, DEAEMA, 4-VBC, BMA, and DVB were passed through Al2O3 to remove the stabilizer before use. AIBN was recrystallized from ethyl ether twice before use. Other reagents were used as received. Ultrapure water (UPW) (conductivity at 298 K: 0.055 μS/cm) was obtained from an Arium 611 VF water purification system (Startorius Stedim Biotech, Aubagne, France). 2.2. Instruments and Characterization Methods. The morphologies of synthesized NPs were characterized with the SEM (FEI Quanta FEG 250) operating at a 5−30 kV accelerating voltage in secondary electron (SE) mode under high vacuum (3 × 10−6−1.8 × 10−5 mbar). Before the measurement, the NPs were sputtered with Au in a sputter-coater (Q15OR-S sputter coater, Quorum, 20 mA for 30 s) under an Ar atmosphere (sputter vacuum: 5 × 10−2 mbar). The FTIR analyses of synthesized NPs were performed with a PerkinElmer spectrometer (Spectrum 1000). For each sample, 1000 spectra between 400 and 4000 cm−1 were acquired and averaged. The hydrodynamic diameters of NPs at a concentration of 0.5 mg/ mL at different pH values (2−10) were measured by a dynamic light scattering (DLS) instrument (Malvern Instruments, Worcestershire, U.K.) equipped with a 4 mW He−Ne laser at a wavelength of 633 nm and a set scattering angle of 173° at 25 °C. The IFT measurements were carried out via the pendant-drop and equilibrium drop-shape analysis methods. The experiments were done using a DataPhysics OCA 15Pro contact angle goniometer equipped with an automatic dosing system. The images of the drop shape were captured in real time and then digitally analyzed using the edgedetecting DataPhysics SCA 22 software module by fitting the contour of the droplet to the Young−Laplace equation. For the measurement process, we prepared an aqueous NP stock solution at concentrations of 10 and 20 mg/mL at different pH values. The liquid droplets of the PDEAEMA-constituted NP aqueous solution were generated at the apex of a straight dosing needle within the toluene phase contained in 6171

DOI: 10.1021/acs.langmuir.8b00877 Langmuir 2018, 34, 6170−6182

Article

Langmuir Scheme 1. Synthesis Routes of Different Nanoparticles Used in the Current Work

2.4. Synthesis of PDEAEMA-Constituted NPs with Different Cross-Linking Degrees. VBTMAC (500 mg) and V-50 (100 mg) were first added to 100 mL of ultrapure water (UPW) in a 250 mL round-bottomed flask. Subsequently, 10 mL of DEAEMA and EGDMA (100 or 300 μL) was added. Then the solution was deoxygenated by bubbling argon for 20 min. The reaction was further initiated by heating to 60 °C under Ar. The reaction was carried out for 24 h under stirring at 700 rpm. Finally, the PDEAEMA-constituted NPs with cross-linking degrees of 1% (PDEAEMA-1) and 3% (PDEAEMA-2) were purified by washing with EtOH four times and UPW three times continually before IFT measurements were made. 2.5. Synthesis of PDEAEMA-Constituted NPs with 200 nm Diameter and 1% Cross-Linking Degree. VBTMAC (1000 mg) and V-50 (100 mg) were first added to 100 mL of ultrapure water (UPW) in a 250 mL round-bottomed flask. Subsequently, 10 mL of DEAEMA and 100 μL of EGDMA were added. Then the solution was deoxygenated by bubbling Ar for 20 min. The reaction was further initiated by heating to 60 °C under Ar. The reaction was carried out for 24 h under stirring at 700 rpm. Finally, the PDEAEMA-constituted NPs with a cross-linking degree of 1% (PDEAEMA-3) were purified by washing with EtOH four times and UPW three times before IFT measurements were made. 2.6. Synthesis of PDEAEMA-co-PBMA Copolymerized NPs. VBTMAC (500 mg) and V-50 (100 mg) were first added to 100 mL of UPW in a 250 mL round-bottomed flask. Subsequently, 5 mL of DEAEMA, 5 mL of BMA, and 100 μL of EGDMA were added. Then the solution was deoxygenated by bubbling argon for 20 min. The reaction was further initiated by heating to 60 °C under Ar. The reaction was carried out for 24 h under stirring at 700 rpm. Finally, the PDEAEMA-co-PBMA NPs were purified by washing with EtOH four times and UPW three times continually before IFT measurements were made. 2.7. Synthesis of ATRP Initiator PS-Cl NPs. VBTMAC (50 mg) and V-50 (100 mg ) were first added to 100 mL of UPW in a 250 mL

a 20 cm cubic quartz cuvette with a corresponding dosing volume of 35 μL and a dosing rate of 35 μL/s, and for the PDMAEMA-grafted NPs, the dosing volume was 15 μL and the dosing rate was 35 μL/s. The dosing rates for generating the droplet at the apex of the tip were kept high in order to achieve good reproducibility and eliminate any lag time during the droplet expansion. If the expansion is slow, then the beginning values will be different from the IFT of the pristine interface. The same is valid if there are surface-active molecular impurities that adsorb significantly faster, on the order of milliseconds, thus pushing the starting value already significantly lower than, for example, ∼35 mN/m for the IFT for the pure Tol−H2O interface. The penetration measurement for the PDEAEMA was made directly on dried PDEAEMA bulk polymer samples, which have the same cross-linking degree (1%) with PDEAEMA-1 NPs. The procedure of synthesizing this bulk material was described in the Supporting Information. The PDEAEMA bulk polymer sample was cut into small cubes (1 × 1 × 1 cm2). The PDEAEMA cubes were then added to a 500 mL beaker containing water to ensure that the pH values remained constant for 48 h. The penetration rate was measured with a computer-controlled DCAT 21 tensiometer (DataPhysics Instruments GmbH) equipped with a solid brass penetration cone. The tensiometer measures the compensated weight (the resistance of the material to penetration) with time due to lowering of the brass cone (25 g) at a constant speed into the soft material. During the measurement, the following important parameters were used: motor speed, 0.05 mm/s; surface detection threshold, 0.05 mg; diving speed of the penetration cone, 0.05 mm/s; and maximum immersion depth, 1 mm. 2.3. Transmission Electron Microscopy (TEM). A 5 μL quantity of NP solution (0.1 mg/mL) was absorbed on 400 mesh copper grids. The grids were further stained with 2% uranyl acetate, and the negatively stained image of nanostructures was aquired on a Philips CM100 TEM at an acceleration voltage of 80 kV. 6172

DOI: 10.1021/acs.langmuir.8b00877 Langmuir 2018, 34, 6170−6182

Article

Langmuir

Figure 1. SEM images of PDEAEMA-constituted NPs at pH 7. (A) PDEAEMA-1, (B) PDEAEMA-2, and (C) PDEAEMA-co-PBMA copolymerized NPs. The scale bar is 5 μm.

Figure 2. pH-responsive behavior of PDEAEMA-constituted NPs. (A) Influence of pH values in solution (2.0−10.0) on the hydrodynamic diameters of PDEAEMA-constituted NPs. (B) Relationship of the degree of swelling of PDEAEMA-constituted NPs with the pH values in solution (2.0−10.0). round-bottomed flask. Subsequently, 9 mL of styrene, 1 mL of 4-VBC, and 100 μL of DVB were added. Then the solution was deoxygenated by bubbling argon for 20 min. The reaction was further initiated by heating to 60 °C under Ar. The reaction was carried out for 24 h under stirring at 700 rpm. Finally, the synthesized NPs were purified by washing with THF four times before carrying out ATRP of DEAEMA and DMAEMA. 2.8. Synthesis of PDEAEMA-Grafted PS NPs. PS-Cl NPs (672 mg) were dispersed in a mixture of THF (10 mL) and EtOH (10 mL) first. Then, 3.4 mg of CuBr2, 66 μL of PMDETA, and 43 mg of CuBr were added to the above solution. After the mixture was bubbled with argon for 20 min in an ice cooling bath, DEAEMA (1, 2, or 3 mL) was added to the solution. The final mixture was bubbled with argon for another 15 min in an ice cooling bath before initiating the polymerization at 60 °C under Ar for 17 h. 2.9. Synthesis of PDMAEMA-Grafted PS NPs. The procedure and quantity of chemicals used in the synthesis were the same as in the case of PDEAEMA-grafted PS NPs except that DMAEMA (3 mL) was used instead of DEAEMA. 2.10. Pickering Emulsion Preparation. A stock solution containing 4 mg/mL NPs in UPW was first prepared. The pH value of the stock solution was adjusted with HCl and NH4OH. Next, 5 mL of the stock solution containing NPs was mixed with 4 mL of Tol containing 0.01 wt % hydrophobic Hostasol yellow 3G dye. Finally, the mixture was homogenized using a Branson 450D sonifier equipped with a 1/2 in. horn for 30 s at room temperature or 70 °C.

result, the hydrophilicity of PDEAEMA and the degree of swelling and softness of the PDEAEMA-constituted NPs also increase. Four PDEAEMA-constituted NPs were synthesized with 1% (PDEAEMA-1, PDEAEMA-3) and 3% (PDEAEMA2) cross-linking degrees or were copolymerized with hydrophobic monomer BMA (PDEAEMA-co-PBMA). Four grafted PS NPs were made from starting PS-Cl NPs via ATRP grafting of PDEAEMA (PDEAEMA-g-PS-1, PDEAEMA-g-PS-2, and PDEAEMA-g-PS-3, differing in the grafted polymer lengths) and PDMAEMA (PDMAEMA-g-PS). The PS-Cl NPs had a 1% cross-linking degree to ensure that they do not dissolve in Tol during the IFT measurement and emulsification experiments. 25,30 PDMAEMA-g-PS was synthesized because PDMAEMA is hydrophilic below pKa = 7.4−7.8 and remains hydrophilic (water-soluble) above its pKa. In contrast, PDEAEMA becomes hydrophobic (water insoluble) above its pKa (7.0−7.3), which causes the PDEAEMA-grafted PS NPs to aggregate strongly. This made the IFT vs time measurement of PDMAEMA-grafted PS NPs possible under both acidic and basic conditions. The composition of synthesized NPs is given in Table S1. 3.2. Characterization and pH Responsiveness of PDEAEMA-Constituted NPs. SEM images of PDEAEMAconstituted NPs with typical soft nanogel morphologies are presented in Figure 1. Figure 2A and Table S2 show the evolution of the hydrodynamic diameters of the PDEAEMAconstituted NPs with pH; the diameters increase from 500− 800 to 1100−1350 nm upon decreasing the pH values from 10.0 to 2.0. Correspondingly, the degree of swelling is expressed as the change in the fractional diameter of the NPs at different pH values with respect to the diameters at pH 10,

3. RESULTS AND DISCUSSION 3.1. Synthesis and Characterization of NPs. The aim of this work is to investigate the main factors influencing the adsorption of NPs at the L−L interface. Scheme 1 shows the synthesis procedures of the seven types of NPs used in this work. The tertiary amine groups of PDEAEMA (pKa = 7.0− 7.3) is protonated when the pH value is below its pKa. As a

calculated with equation 6173

HDpH − HDpH10.0 HDpH10.0

× 100% (HD: hydroDOI: 10.1021/acs.langmuir.8b00877 Langmuir 2018, 34, 6170−6182

Article

Langmuir

Figure 3. SEM images of PDEAEMA- and PDMAEMA-grafted PS nanoparticles. (A) PS-Cl ARTP initiator NPs, (B) PDEAEMA-g-PS-1, (C) PDEAEMA-g-PS-2, (D) PDEAEMA-g-PS-3, and (E) PDMAEMA-g-PS. (F) Summary of the evolution of the grafted NPs’ diameters. The scale bar is 1 μm.

Figure 4. Dynamic IFT results of PDEAEMA-1 at the Tol−H2O interface at different pH values. (a) pH 10 (10 mg/mL), (b) pH 6.5 (10 mg/mL), (c) pH 5.0 (10 mg/mL), (d) pH 4.0 (10 mg/mL), (e) pH 3.0 (10 mg/mL), (f) pH 2.0 (10 mg/mL), and (g) pH 2.0 (20 mg/mL). The error bars represent the standard deviation of three measurements.

1 NPs remain positive and unchanged, well above +30 mV from pH 2.0 to 10.0 (Figure S2A). 3.3. Characterization of PDEAEMA- and PDMAEMAGrafted PS NPs. Syntheses of the PDEAEMA- and PDMAEMA-grafted PS NPs were performed via ATRP. The -Cl initiator was immobilized on the surface of the starting PS NPs. The SEM images of the obtained PS-Cl, PDEAEMA-g-PS1, PDEAEMA-g-PS-2, PDEAEMA-g-PS-3, and PDMAEMA-gPS are presented in Figure 3A−E, which show round and welldefined spherical particles. The diameters of the PDEAEMAgrafted NPs can be controlled by the grafted polymer length, increasing by 10 to 30 nm (Figure 3F) compared to those of the original PS-Cl NPs (diameter ∼150 nm). The hydrodynamic diameters of the PS-Cl and grafted NPs were also measured in native solution by DLS at pH 2, when presumably the polymer chains in the brush are fully stretched; the data is shown in Table S4. The thickness of the brush shell in the

dynamic diameter). The degree of swelling increases with decreasing pH values from 10.0 to 2.0 (Figure 2B and Table S3). Comparatively, the degree of swelling of PDEAEMA-1 increases from 0 to 162% with decreasing pH from 10.0 to 2.0, but the degree of swelling of PDEAEMA-2 and PDEAEMA-coPBMA increases only by ∼42% in the same pH interval. Evidently, the degree of cross-linking and the copolymerization of a hydrophobic monomer proved to be very effective in controlling the swelling behavior and softness of the PDEAEMA-constituted NPs. Additionally, in the 10.0 to 2.0 pH range, the DLS results in the Supporting Information (Figure S1A−C and Table S2) show that PDEAEMA-constituted NPs are well dispersed in the water and reasonably monodisperse, which allowed us to perform the dynamic IFT measurements at the Tol−H2O interface. Interestingly, the zeta potential values of PDEAEMA6174

DOI: 10.1021/acs.langmuir.8b00877 Langmuir 2018, 34, 6170−6182

Article

Langmuir

Figure 5. Dynamic IFT results of PDEAEMA-2 NPs (A) and PDEAEMA-co-PBMA NPs (B) at the Tol−H2O interface at different pH values: (a) pH 10 (10 mg/mL), (b) pH 4.0 (10 mg/mL), (c) pH 2.0 (10 mg/mL), and (d) pH 2.0 (20 mg/mL). The error bars represent the standard deviation of three measurements.

larger ΔIFT values observed at lower pH values are a consequence of the enhanced interfacial activity of PDEAEMA-1 NPs due to an increased degree of swelling or softness, in agreement with the observations of Richtering et al.15 The lowest IFT value (21.1 mN/m) was obtained at pH 2.0 (curve f, 10 mg/mL), and the dynamic IFT of PDEAEMA-1 NPs was also measured at pH 2.0 at higher concentration (curve g, 20 mg/mL). A further increase in the concentration of the PDEAEMA-1 NPs to 30 mg/mL resulted in a gel (SI, Figure S4), which may be due to the fact that all of the free water has been absorbed by the swelling PDEAEMA-1 NPs. The highest PDEAEMA-1 NP concentration for which we could perform the dynamic IFT measurement was 20 mg/mL, with the final IFT of ∼19.2 mN/m (Figure 4g). From Figure 4, it is obvious that both lower pH values and higher PDEAEMA-1 NP concentrations lead to larger ΔIFT values. In order to demonstrate that the PDEAEMA-1 NPs become softer upon decreasing the pH value of the solution, we ran penetration experiments on the PDEAEMA bulk materials with the same chemical composition and degree of cross-linking as for the PDEAEMA-1 NPs at different pH values. The preparation method and the experimental setup are given in the SI. The result in Figure S5 shows that PDEAEMA bulk material at pH 2.0 is soft, with the fastest penetration rate (g/s). The hardness of the PDEAEMA material increases progressively upon increasing the pH values of the aqueous solution. By carefully analyzing the IFT vs time curves in Figure 4b−g, it becomes evident that there are two stages describing the PDEAEMA-1 NPs adsorption at the Tol−H2O interface. The first stage happens at the very beginning, a short time of 0−1 min. After a short time, the PDEAEMA-1 NPs exhibit very fast adsorption kinetics with a rapid reduction in the IFT. At this stage in the adsorption process, after a short time of 0−1 min, the incoming PDEAEMA-1 NPs diffusing from the bulk solution meet a pristine Tol−H2O interface and will easily adsorb, causing the IFT to decrease rapidly. During the second stage, at a later time >1 min, the IFT decreases slowly and reaches the final plateau value at a very long time. This can be explained by the increased occupancy of the interface with the NPs, which impose an additional adsorption barrier due to the increasing electrostatic repulsion between PDEAEMA-1 NPs already present at the interface and those in the bulk, prohibiting the fast adsorption of PDEAEMA-1 NPs from the H2O bulk solution. As already mentioned, Richtering et al.15 demonstrated that soft particles adsorb at the L−L interface more easily than do solid particles. Similar results were obtained here by using pH-

grafted NPs increases significantly by ∼200 to ∼300 nm as compared to the starting PS-Cl NPs. The zeta potential of the PDMAEMA-g-PS NPs decreases from +64 to +8 mV with increasing pH from 2.0 to 10.0 (Figure S2B), demonstrating the protonation of PDMAEMA at pH < 7.0. The FTIR analysis of the grafted NPs there show additional vibrational bands appearing at 1728 and 1146 cm−1 originating from the vibration of −CO and −N−C−, as compared to the spectrum of the PS-Cl NPs (Figure S3), demonstrating the successful grafting of PDEAEMA and PDMAEMA on the surface of the PS NPs via ATRP. 3.4. Adsorption of NPs at the Tol−H2O Interface. The dynamic IFT measurements for slowly adsorbing particles are typically made via pendant drop tensiometry and drop-shape analysis. Here we emphasize that in such measurements two important points should be carefully considered. The first one is the good dispersibility of particles in one of the phases; if the particles are not well dispersed in the bulk solution, with the formation of aggregates, then the IFT vs time results are difficult to interpret. The second one is the need to carefully eliminate the interfacially active impurities from the solution, especially when using polymers, surfactants, and amphiphilic block copolymers as capping agents or stabilizers that are adsorbed only on the particle surface through a weak interaction, such as a double-layer interaction, hydrogen bonding, or hydrophobic interactions. The presence of such impurities will greatly interfere with the IFT measurements and very often invalidate the results. Therefore, we took great care to ensure that no impurities were present. Figure 4 shows the evolution of the IFT with time as a consequence of PDEAEMA-1 NP adsorption from the aqueous phase at the Tol−H2O interface, measured via the pendant-drop method. Under basic conditions (curve a, pH 10.0, 10 mg/mL), the IFT value remains constant at ∼34 mN/m without a significant decrease with time, while the IFT decreases dramatically under acidic condition (curves b−f at pH ⩽ 6.5, 10 mg/mL). Furthermore, the decreasing of the IFT values of PDEAEMA-1 NPs under acidic conditions strongly depends on the pH value of the aqueous solution. Lower pH values result in lower final IFT values (b−f) or larger ΔIFT values. The ΔIFT is the difference between the initial value of the IFT at t = 0 s and the plateau IFT value at t = 180 min. In the last section, we discussed that a gradual increase in the degree of swelling implies that PDEAEMA-1 NPs (Figure 2) become softer at lower pH values, understood here as compressible and deformable. Because all of the other parameters in the system remain constant, except the pH, it is likely that the increasingly 6175

DOI: 10.1021/acs.langmuir.8b00877 Langmuir 2018, 34, 6170−6182

Article

Langmuir

Figure 6. Dynamic Tol−H2O IFT results of PDEAEMA-g-PS NPs, PDMAEMA-g-PS NPs, and PDEAEMA and PDMAEMA homopolymers. (A) Dynamic Tol−H2O IFT with 20 mg/mL NPs in the water phase at pH 2.0: (a) PDEAEMA-g-PS-1, (b) PDEAEMA-g-PS-2, (c) PDEAEMA-g-PS-3, and (d) dynamic IFT of the PDEAEMA homopolymer with a concentration of 20 mg/mL at pH 2.0. (B) Dynamic Tol−H2O IFT of PDMAEMA-gPS NPs with 10 mg/mL at (a) pH 2.0, (b) pH 7.2, (c) pH 8.0, and (d) pH 10.0; (e) PDMAEMA-g-PS NPs with a concentration of 20 mg/mL at pH 10.0; (f) PDMAEMA homopolymer with a concentration of 20 mg/mL at pH 2.0; (g) PDMAEMA homopolymer with a concentration of 2 mg/mL at pH 10.0; and (h) PS-Cl NPs with a concentration of 10 mg/mL at pH 10.0. The error bars represent the standard deviation of three measurements.

with an NP concentration of 20 mg/mL are represented in Figure 6A-a−c. The IFT vs time plot remained unchanged for 5 h, which means that PDEAEMA-g-PS NPs did not adsorb at the Tol−H2O interface regardless of the PDEAEMA polymer chain length grafted onto the surface of PS NPs. We also attempted to measure the Tol−H2O IFT value of PDEAEMAg-PS NPs under basic conditions (pH 10.0), but the PDEAEMA-g-PS NPs did not disperse well in water at this pH value. In order to be able to measure the grafted NPs under basic conditions, we instead grafted PDMAEMA onto the surface of PS NPs (PDMAEMA-g-PS). The length of PDMAEMA is around 20 nm (Figure 3F-e). PDMAEMA has a pKa of 7.4−7.8 and a chemical structure similar to that of PDEAEMA but is more hydrophilic.31 Its water solubility in acidic aqueous solution (pH < 6.0) is higher than in basic aqueous solution (pH 10.0) because of the tertiary amine protonation. Figure S1-D shows that PDMAEMA-g-PS NPs are well dispersible under both acidic (pH 2.0) and basic conditions (pH 10.0), which allowed the measurement of the IFT of PDMAEMA-g-PS at the Tol−H2O interface in the pH 2.0 to 10.0 interval. The dynamic IFT measurement results obtained are presented in Figure 6B-a, and these show that PDMAEMA-g-PS NPs with a concentration of 10 mg/mL behave similarly with PDEAEMA-g-PS NPs at pH 2.0 without a noticeable decrease in the IFT value of Tol−H2O with time. On the other hand, the IFT value decreases greatly in the presence of PDMAEMA-g-PS (at 10 mg/mL), from 34 to 17.2, 15.0, and 5.5 mN/m at pH 7.2, 8.0, and 10.0, respectively, in only 60 min (Figure 6B-b−d). Moreover, the final IFT value was 4.1 mN/m (Figure 6B-e) when the PDMAEMA-g-PS concentration was 20 mg/mL at pH 10.0. Because of the rapid lowering of the IFT value at pH 10.0 when the water phase contains NPs at higher concentration, ∼20 mg/mL, the pendent drop of PDMAEMA-g-PS NPs solution quickly elongates and detaches from the tip of the needle (Figure S6). Additionally, the PS-Cl NPs do not adsorb at the Tol− H2O interface at pH 10.0 (at 10 mg/mL), and the IFT values remained constant at ∼34 mN/m (Figure 6B-h), demonstrating that the decreasing of the IFT for PDMAEMA-g-PS NPs is caused by the grafted PDMAEMA brush. In addition, the IFT vs time values of homopolymers PDEAEMA and PDMAEMA at the Tol−H2O interface at

responsive NPs. The PDEAEMA-1 NPs are softer in acidic than in basic solution due to the protonation of PDEAEMA. Additionally, the fact that the ΔIFT value is strongly pHdependent also demonstrates that the softness of PDEAEMA-1 NPs plays a key role in NP adsorption at the Tol−H2O interface. In order to further prove this hypothesis, we synthesized two other types of PDEAEMA-constituted NPs: PDEAEMA-2 with a 3% degree of cross-linking and PDEAEMA-co-PBMA NPs with a 1% degree of cross-linking in which PDEAEMA was copolymerized with a pH-inert, hydrophobic monomer. From Figure 2 and Table S3, we see that the degrees of swelling of PDEAEMA-2 NPs and PDEAEMA-co-PBMA NPs at each pH value are lower than that of PDEAEMA-1 NPs, meaning that the strategy of restricting the degree of swelling was successful. Figure 5 shows that under basic conditions (pH 10.0) neither PDEAEMA-2 nor PDEAEMA-co-PBMA NPs are not capable of lowering the IFT values (Figure 5A-a,B-a) but under acidic conditions (pH < 5.0) can decrease the IFT of the Tol−H2O interface significantly. The lowest plateau IFT values of 22.7 mN/m (Figure 5A-d) and 29.0 mN/m (Figure 5B-d) were attained by PDEAEMA-2 and PDEAEMA-co-PBMA NPs, respectively, at pH 2.0 at a concentration of 20 mg/mL. These IFT values are higher than the final IFT value of PDEAEMA-1 NPs (19.2 mN/m) under the same pH and concentration conditions (Figure 4g). The above results demonstrate that (1) increasing the degree of cross-linking and the copolymerizing inert hydrophobic monomer reduce the degree of swelling of NPs and the softness of the NPs and (2) the softer the PDEAEMA1 NPs (Figure S5), the lower the final IFT values of the Tol− H2O interface observed upon NPs adsorption. The PDEAEMA-1, PDEAEMA-2, and PDEAEMA-PBMA NPs were all capable of spontaneous interfacial adsorption at the Tol−H2O interface after the protonation of PDEAEMA polymers under acidic conditions (pH < 6.0). Next, we investigate if the other type of particles, namely, the PDEAEMA-grafted PS NPs, can also adsorb spontaneously at the Tol−H2O interface under the same pH conditions. We have synthesized three PDEAEMA-g-PS NPs with different PDEAEMA polymer chain lengths on the surface (10, 20, and 30 nm, Figure 3F-b−d). The dynamic IFT measurements of Tol−H2O under acidic conditions (pH 2.0) 6176

DOI: 10.1021/acs.langmuir.8b00877 Langmuir 2018, 34, 6170−6182

Article

Langmuir

hard NPs must be dewetted/dehydrated before being resolvated by toluene. When PDEAEMA-constituted NPs are at pH 10.0, the PDEAEMA polymers in the NPs with a relative hard structure may lose their flexibility and ability to rearrange the polymer chains, which results in no adsorption at the Tol− H2O interface at all, even though the PDEAEMA polymer has some solubility in toluene. Although plausible, this remains a conjecture at this point. However, the ability to decrease the IFT of Tol−H2O for the PDEAEMA- and PDMAEMA-grafted PS could be attributed to the better solubility/wettability of grafted polymers in toluene. In order to further test if the solubility of PDMAEMA on the surface of PDMAEMA-g-PS NPs in organic solvent is the key factor controlling the interfacial adsorption of these NPs, we have performed dynamic IFT measurements at heptane−H2O and ethyl acetate−H2O interfaces. The corresponding results are given in Figure S10 and show that at the heptane−H2O and ethyl acetate−H2O interfaces, PDMAEMA-g-PS NPs can decrease the IFT at pH 10.0 (Figure S10A-c, B-c) and lose the ability adsorb at pH 2.0 (Figure S10A-b, B-b). The solubility of the grafted PDMAEMA polymer chains (brush) in the organic phase may control both the adsorption ability of the PDMAEMA-g-PS NPs at the interfaces and the plateau IFT values that can be reached at longer times, >2 min. For example, compare the progressively lower IFTs obtained at longer times for the PDMAEMA-g-PS NPs at the heptane− H2O (∼18 mN/m, Figure S10A), Tol−H2O (∼8 mN/m, Figure 6B), and ethyl acetate−H2O interfaces (∼2 mN/m, Figure S10B), with the progressively larger solubility of the PDMAEMA polymer in heptane (4.9 mg/mL), toluene (198 mg/mL), and ethyl acetate (243 mg/mL). The solubility of PDMAEMA in the organic solvent was determined according to the procedures described in the SI. 3.5. Activation and Adsorption Energies of PDEAEMA-Constituted NPs and PDMAEMA-g-PS NPs. The interfacial attachment energy (ΔE) can be calculated if the contact angle of the three-phase line is known, which is, however, difficult to determine in situ, requiring specialized equipment to perform water freezing and vitrification.32 Instead, the following equation was applied by Dinsmore et al.20 to calculate the interfacial attachment energy (ΔE) of particles at the interface from the IFT vs time measurements

different pH values were also investigated. Figure 6A-d, 6B-f demonstrate that at pH 2.0 homopolymers PDEAEMA and PDMAEMA in their polyelectrolyte forms cannot adsorb at the Tol−H2O interface and thus decrease the corresponding IFT, behaving similarly to the PDEAEMA- and PDMAEMA-grafted PS NPs. However, PDMAEMA at pH 10.0 at a concentration of 2 mg/mL can decrease the IFT of Tol−H2O to lower than 1 mN/m in just 30 s (Figure 6B-g). After 30 s, the polymer solution pendant drop detaches from the tip of the needle (Figure S7), which indicates that nonprotonated homopolymer PDMAEMA with some solubility in Tol has a relatively high interfacial activity. Comparing the IFT results from Figures 4f and 6B-d, it is clear that PDMAEMA-g-PS NPs at pH 10.0 (10 mg/mL) have a stronger ability to decrease the IFT of Tol−H2O than does PDEAEMA-1 (10 mg/mL) at pH 2.0. However, PDEAEMA-1 NPs and PDMAEMA-g-PS NPs also have very different diameters. In order to test whether the diameter of NPs has an influence on the decrease in the IFT of Tol−H2O, we have synthesized the smaller PDEAEMA-constituted NPs, namely, PDEAEMA-3, which have the same cross-linking degree (1%) with PDEAEMA-1 NPs. The PDEAEMA-3 NPs are smaller, around 200 nm as determined from the TEM image (Figure S8), than the PDEAEMA-1 NPs, similar to the diameter of PDMAEMA-g-PS NPs (Figure 3E). Because the smaller PDEAEMA-3 NPs are very difficult to purify, due to their small sizes, we were able to measure only the dynamic IFT of the Tol−H2O interface at 1 mg/mL NP concentration in the aqueous solution. The dynamic IFT result is shown in Figure S9. It is clear that PDEAEMA-3 NPs can decrease the IFT at pH 2.0 with ΔIFT 2.5 mN/m after 3 h (Figure S9-b) and no interfacial activity at pH 10.0 (Figure S9-a). Therefore, compared to the performance PDEAEMA-1 NPs ΔIFT = 4 mN/m after 3 h for Figure S9-b, we conclude that the size of the particles in this case had little influence on the ability of the particle to lower the IFT. On the other hand, the PDMAEMAg-PS NPs with concentration 1 mg/mL at pH 10.0 can decrease the IFT from initially 34 to 9 mN/m in 50 min (Figure S9-d). Therefore, we can conclude that the stronger interfacial activity of PDMAEMA-g-PS NPs at pH 10.0 compared to that of the PDEAEMA-1 NPs at pH 2.0 is not caused by the difference in diameter. It is therefore reasonable to assume that because the PDEAEMA and PDMAEMA are hydrophilic at pH 2.0 and in their protonated form cannot be solubilized or wetted by toluene, the PDEAEMA-grafted PS NPs and PDMAEMAgrafted PS NPs cannot spontaneously adsorb at the Tol−H2O interface. However, PDEAEMA-constituted NPs can spontaneously adsorb at the interface in its protonated form (pH < 6.0), meaning that hydrophilicity is not the only parameter that controls the adsorption process of particles at the L−L interface. One possible explanation as to why PDEAEMAconstituted NPs can spontaneously adsorb at the Tol−H2O interface and decrease the IFT at pH < 6.0 is that PDEAEMA chains in the NPs with higher degrees of swelling and softness have enough flexibility that the chains can rearrange themselves to make their hydrophobic sides face toluene and their hydrophilic side face water. In this way, the surface of PDEAEMA-constituted NPs could be better wetted by toluene. Another explanation may be that the soft particles have a more efficient way of expelling the water layers separating the particle surface and Tol−H2O interface. For example, water can freely circulate through the particle. In contrast, the surface of the

ΔE = −(γ0 − γp(min))πR2/η

(1)

where γ0 is the IFT value of the clean L−L interface, γp(min) is the plateau IFT value at the maximum concentration of particles above which γp does not decrease anymore, η is the packing parameter (0.91), and R is the radius of the particles. In our calculation, we take η = 0.91 at the highest concentration of NPs at which no further decrease in γp with an increase in NP concentration is observed.20 We stated in the last section that the PDEAEMA-1 NP water solution with concentrations above 20 mg/mL at pH 2.0 produces a gel (Figure S4) and that the dynamic IFT could not be measured above 20 mg/mL. As a result, we consider that the IFT plateau values (γp), meaning interface saturation, are attained at this concentration. Table S5 in the SI shows the ΔE values of the synthesized NPs, which were calculated according to the ΔIFT measured from Figures 4, 5, and 6. The calculated values for the ΔE show a clear trend (Table S5). Among the PDEAEMA-constituted NPs, the PDEAEMA-1 NPs at pH 2.0 exhibited the strongest interfacial activity as 6177

DOI: 10.1021/acs.langmuir.8b00877 Langmuir 2018, 34, 6170−6182

Article

Langmuir

Table 1. Activation and Adsorption Energies of the PDEAEMA-1-Constituted NPs and PDMAEMA-g-PS NPs at the Tol−H2O Interface and Their Diffusivity, Effective vs Actual, at 10 mg/mL Concentration at Different pH Values NPs PDEAEMA-1

PDMAEMA-g-PS

pH 2 3 4 5 6.5 10 2 7.2 8 10

radius (nm) 678.5 632 629.5 557.5 394.5 259 84 84 84 84

D0 (m2 s−1) 3.6 3.9 3.9 4.4 4.4 9.5 2.9 2.9 2.9 2.9

× × × × × × × × × ×

Deff (m2 s−1)

−13

1.5 8.6 3.1 3.3 1.2 ∼0 ∼0 2.9 5.6 1.7

10 10−13 10−13 10−13 10−13 10−13 10−12 10−12 10−12 10−12

judged from their largest absolute value of ΔE (6.07 × 106 kT) at the Tol−H2O interface, followed by PDEAEMA-2 NPs (3.3 × 106 kT) and PDEAEMA-PBMA NPs (1.6 × 106 kT). Among the PDMAEMA-grafted PS NPs, the PDMAEMA-g-PS NPs at pH 10.0 have the largest absolute value of ΔE (1.8 × 105 kT). Many researchers use the magnitude of ΔE alone to assess the interfacial activity of particles. However, because of the different diameters of PDEAEMA-constituted NPs and PDMAEMA-gPS NPs, it is not reasonable to compare only the ΔE values for two reasons: (1) because ΔE scales with R2, predictably the largest particle, to the extent of a macroscopic object, may appear as the most interfacially active and (2) an energy barrier to interfacial adsorption may be present. It has been reported that diffusion-controlled, energy-barriercontrolled, or a combination of the two are the three main types of adsorption kinetics for particles adsorbed at interfaces.33 Pendant drop dynamic IFT measurements and Ward and Tordai theory34 are normally used to model the adsorption kinetics of NPs in the absence of an adsorption barrier. So-called Ward and Tordai theory considers that adsorption is controlled by the particle’s concentration and bulk diffusivity, followed by immediate adsorption of any particles colliding with the interface. The adsorption of NPs at the interface is much slower in the presence of an energy barrier than that described by the pure diffusive models because it is reasonable to assume that not every particle colliding with the interface also adsorbs. Liggieri et al.35 and Ravera et al.36 proposed an effective diffusion model to account for the activation energy barrier in interfacial adsorption. Therefore, an observed effective diffusion coefficient (Deff), which is different from the Stokes’ bulk diffusivity (D0), can be calculated from the dynamic IFT curves with the following equation33 D t γ = γ0 − 2NAC0ΔE eff π

× × × × ×

10−16 10−17 10−17 10−17 10−18

× 10−15 × 10−15 × 10−12

ΔE (kBT)

Ea (kBT)

slope

7.8 8.5 9.5 9.6 13.2 ∞ ∞ 6.9 6.3 0.57

−1.79 −1.05 −5.63 × 10−1 −3.44 × 10−1 −2.44 × 10−2 −1.78 × 10−2 −6 × 10−2 −1.6 × 10−1 −2.4 × 10−1 −6.2

−5.3 −4.1 −3.7 −2.2 −8.0 0 0 −1.0 −1.2 −1.7

× × × × ×

106 106 106 106 105

× 105 × 105 × 105

presence of smaller and fast-adsorbing interfacially active impurities, or particles with broad distributions. Correspondingly, the activation energy of attachment (Ea) can be calculated according to eq 3 as the logarithm of the ratio between the observed bulk-to-surface NP diffusivity Deff and bulk NP Stokes−Einstein diffusivity D0. By fitting the earlier portion of the IFT vs √t curves, t ⩽ 200 s, at all interfaces with an NP concentration of 10 mg/mL (Figures S11−S14), we were able to extract the values of Deff corresponding to all NPs. Fitting the earlier portion of the curves is justified by the fact that the incoming NPs meet a pristine interface and at a later time the electrostatic repulsion between the adsorbed and incoming NPs dominates and thus imposes an additional energy barrier.33 Here we want to learn about the NPs’ interaction with the Tol−H2O interface. The obtained D eff results for the PDEAEMA-1 NPs and PDMAEMA-g-PS NPs are given in Table 1, and for PDEAEMA-2 and PDEAEMA-PBMA NPs, they are given in Table S6. From this, it can be observed that Deff changes significantly for different NPs at the same Tol−H2O interface. The obtained effective diffusion coefficient Deff can be compared with D0 calculated from the Stokes−Einstein equation (eq 4) ⎛ E ⎞ Deff = D0 exp⎜ − a ⎟ ⎝ kBT ⎠

(3)

where Ea is the activation energy of attachment at the interface.

D0 =

kBT 6π μR

(4)

where μ is the viscosity of water and R is the radius of the NPs. The D0 values obtained from eq 4 for each NP are given in Table 1 (PDEAEMA-1 NPs and PDEMAEMA-g-PS NPs) and Table S6 (PDEAEMA-2 NPs and PDEAEMA-PBMA NPs). Basavaraj et al.33 obtained differences between Deff vs D0 of as large as 3 orders of magnitude for 10 nm silica particles at the dodecane−water interface. They claimed that the larger difference between Deff and D0 is in fact caused by the presence of a large activation energy barrier at the L−L interface. Wu and Honciuc25 obtained up to 3 orders of magnitude differences between Deff and D0 for 200 nm Janus nanoparticles. In the current situation, for PDEAEMA-1 NPs at pH from 2.0 to 6.5, the Deff value ranges from 1.49 × 10−16 to 1.24 × 10−18 m2 s−1, which is 3 to 5 orders of magnitude larger than their corresponding D0 (3.64 × 10−13−4.43 × 10−13 m2 s−1). Therefore, it is clear that the adsorption kinetics of PDEAEMA1 NPs at pH values from 2.0 to 6.5 at Tol−H2O interfaces is a

(2)

where C0 is the concentration of NPs in the bulk, γ is the measured IFT at time t, γ0 is the IFT value of the clean Tol− H2O interface, ΔE is the attachment energy (here calculated values are in Table 1), and Deff is the effective diffusion coefficient. If only a smaller number of all of the particles arriving at the interface also adsorb, then the effective diffusion coefficient Deff should always be smaller than the Stokes− Einstein bulk diffusivity, signaling that indeed an energy barrier at the interface might be present. We also emphasize that if the effective diffusivity determined from the IFT vs time curves with eq 2 is larger than the Stokes−Einstein diffusivity (D0) then this represents an unphysical situation37 and experimental conditions must be rechecked, especially with respect to the 6178

DOI: 10.1021/acs.langmuir.8b00877 Langmuir 2018, 34, 6170−6182

Article

Langmuir

Figure 7. Digital and fluorescent images of a Pickering emulsion emulsified by synthesized NPs at pH values from 2.0 to 10.0. (A) PDEAEMA-1 NPs and (B) PDMAEMA-g-PS NPs. The scale bar in the fluorescence microscope images is 200 μm.

attachment energy ΔE of PDEAEMA-1 NPs decreases from 5.3 × 106 kT to 0 and the corresponding activation energy of interfacial attachment Ea increases from 7.86 kT to very large values. On the other hand, the absolute interfacial attachment energy ΔE of PDMAEMA-g-PS NPs increases from 0 to 1.7 × 105 kT, and the corresponding activation energy of attachment Ea decrease from very large to 0.57 kT with an increase in the pH value from 2.0 to 10.0 for the same NPs concentration. Therefore, there are totally opposite decreasing and increasing trends in ΔE and Ea for PDEAEMA-1 NPs and PDMAEMA-gPS NPs with increasing pH values (Table 1 and Figure S15), which demonstrate the contrasting interfacial adsorption mechanisms at the Tol−H2O interface. As mentioned earlier, the softness of the PDEAEMA-1 NPs on one hand appears to be crucial to the particles’ spontaneous interfacial adsorption, while the solubility of the grafted polymer on the surface of the NPs in both phases, i.e., toluene and water, is responsible for PDMAEMA-g-PS spontaneous interfacial adsorption. While both materials have similar chemical natures, it is apparent that the conformational freedom of the pH-responsive polymer chains in the constituted of NPs vs surface-grafted NPs may be responsible for the markedly different interfacial activities observed in this work. 3.6. Pickering Emulsions Stabilized by pH-Responsive Nanoparticles. Controlling the pH of the solution is an efficient way to adjust the surface wetting property of pHresponsive particles and the phase of a stabilized Pickering emulsion.24,25,38−41 For example, Yang et al.40 reported that after surface modification of SiO2 NPs with pH-responsive molecules ((MeO)3SiCH2CH2CH2(NHCH2CH2)2NH2) the surface hydrophilicity of NPs can be tuned by changing the pH values and the Tol−H2O Pickering emulsion phase could be controlled. In heterogeneous catalysis, the phase inversion of Pickering emulsions can be key to the recycling of catalysts.40 Tu and Lee24 also observed pH switching behavior in Pickering emulsions stabilized by micrometer-sized particles and underlined their potential use in encapsulation and triggered release applications. Previously, we reported phase inversion in a Pickering emulsion stabilized by pH-responsive PS-PDIPAEMA Janus nanoparticles and thoroughly described the mechanisms involved.25 The main factor leading to the emulsion phase inversion is the change in polarity of the particles and consequently their affinity change with respect to the organic vs aqueous phase, a phenomenon that we have thoroughly investigated and reported on several occasions.3,25,30 In this report, we have already seen that the surface polarity of the PDEAEMA-1 and PDMAEMA-g-PS NPs can be effectively influenced by the pH value. Low pH values (pH < pKa) lead to the protonation of polymer (with NPs becoming more

combination of diffusion-controlled and energy-barrier-controlled. Comparing the Deff and D0 results in Table S6, the adsorption kinetics of PDEAEMA-2 and PEDAEMA-PBMA are the same with PDEAEMA-1 at pH values from 2.0 to 6.5. From Table S5, it should be noted that PDMAEMA-g-PS NPs at pH 10.0 with a concentration of 10 mg/mL have the lowest final IFT value (5.5 mN/m) of all of the grafted PS NPs. The calculated Deff (1.67 × 10−12 m2 s−1) for PDMAEMA-g-PS is quite similar to the D0 (2.94 × 10−12 m2 s−1), which means that there is almost no energy barrier to their adsorption at the interface of Tol−H2O at pH 10.0. However, at pH 8.0 and 7.2, the Deff values are 5.60 × 10−15 and 2.92 × 10−15 m2 s−1, which are 3 orders of magnitude smaller than their corresponding D0 (2.94 × 10−12 m2 s−1), signaling the presence of a large energy barrier experienced by the PDMAEMA-g-PS NPs while diffusing from bulk water to the Tol−H2O interface at pH 8.0 and 7.2, which vanishes at pH 10.0, in which case the adsorption is only diffusion-controlled. It can be imagined that the adsorption of NPs at the Tol−water interface, from the water phase, requires the displacement of water molecules from the NPs’ surface by toluene. Therefore, we hypothesize that the magnitude of the energy barrier is the energy cost of the NPs’ surface dehydration and solvation. Here we also note that the most interfacially active NPs are clearly the PDMAEMA-g-PS and the PDEAEMA-1, for which the interfacial attachment energy is the highest, the activation energy barrier is the lowest, and the ΔIFT is the largest. From the IFT data in Figures 4, 5, and 6 and Ea data in Table 1, it is apparent that energy barriers of 1−10 kT can be overcome by the NPs at room temperature. In only one instance, for PDEAEMA-1 at pH 6.5, the Ea is larger than 10 kT (Table 1). We also emphasize that any of the parameters above taken alone cannot entirely define the NPs’ interfacial activity. It has been proposed that the interfacial activity is the magnitude of the NPs’ attachment energy at the interface,1 but here we believe that the ability to lower the interfacial tension, ΔIFT, is also equally important to consider in making such a statement about a particle’s interfacial activity. As already stated, ΔE scales with the size of the NPs and becomes significant for microparticles and for macroscopic objects, and it may not be reflecting a true difference in the interfacial activity of objects that have different sizes; therefore, the activation energy of attachment and ΔIFT that does not scale with size should also be considered. The most interesting and surprising result can be observed when the adsorption parameters summarized in Table 1 are compared for the PDEAEMA-1 NPs and PDMAEMA-g-PS NPs. Notice that upon increasing the pH value from 2.0 to 10.0 at an NP concentration of 10 mg/mL the absolute interfacial 6179

DOI: 10.1021/acs.langmuir.8b00877 Langmuir 2018, 34, 6170−6182

Article

Langmuir

Figure 8. pH-responsive Pickering emulsion stabilized by PDEAEMA-1 NPs and dynamic phase inversion by the addition of an acid or a base. (A) Tol/H2O (pH 2.0) changes to H2O/Tol (pH 10.0) by adding NH4OH and (B) H2O/Tol (pH 10.0) changes to Tol/H2O (pH 2.0) by adding HCl. The scale bar in the fluorescence microscope images is 100 μm.

From the fluorescence microscopy images we observe that the diameter of H2O/Tol Pickering emulsion droplets stabilized by PDMAEMA-g-PS are around 200 μm at pH 10.0, which are much larger than other grafted NPs, e.g., PDMAEMA-g-PS-1 (Figure S16). We hypothesize that this observed effect could be due to the more hydrophilic nature of PDMAEMA compared to that of PDEAEMA at the same pH value, which affects the immersion depth of NPs into the Tol phase, resulting in a smaller curvature of the H2O/Tol droplets.27,42 The emulsification experiments in Figure 7 demonstrate that regardless of the NP-type, both constitutedof and grafted-with pH-responsive polymer NPs are hydrophilic at pH 2 (higher affinity for water) and more hydrophobic at pH 10 (higher affinity for toluene). In addition to the static phase inversion observed, meaning that the emulsions were prepared at the corresponding pH, the phase of the Pickering emulsion can also be changed dynamically by adding acid or base. For example, the Tol/ H2O Pickering emulsion stabilized by PDEAEMA-1 NPs at pH 2.0 can be changed to H2O/Tol by adding NH4OH and oppositely by adding HCl (Figure 8). Here we note that after adding HCl or NH4OH, gentle shaking by hand of the emulsion is necessary to induce the phase inversion in the Pickering emulsion. Clearly, the pH-induced phase inversions of Pickering emulsions is a consequence of the NPs’ surface polarity change due to protonation−deprotonation of PDEAEMA and PDMAEMA polymers, in agreement with the previous reports.25

hydrophilic) and consequently deprotonation at higher pH values (pH > pKa) (with NPs becoming more hydrophobic). In Figure 7A, the creaming up of Pickering emulsions and the formation of droplets with very strong fluorescence demonstrate the formation of aTol/H2O (i.e., o/w) Pickering emulsion at pH values below 7.0. Interestingly, at pH 8.0, a mixture of Tol/H2O and H2O/Tol Pickering emulsions can be observed in the fluorescent images, signaling that the surface of PDEAEMA-1 NPs slowly changed from hydrophilic to hydrophobic when the pH increased from 7.0 to 8.0. At pH 10.0, the Pickering emulsion settled down, and only H2O/Toltype Pickering emulsions were formed as seen from the fluorescence microscopy images. For the PDMAEMA-g-PS NPs, the Tol−H2O phase separated very fast at pH 2.0 (Figure 7B), and no stable Pickering emulsion could be obtained, which demonstrates that PDMAEMA-g-PS NPs have little or no emulsification ability. This result is likely caused by the very hydrophilic nature of the NPs’ surface. After the protonation of tertiary amine groups, the surface of PDMAEMA-g-PS NPs becomes very polar and the energy cost for dewetting the surface and resolvating it with Tol is probably very large; therefore, no interfacial adsorption was observed even with external mechanical energy input. These results are in agreement with the large Ea predicted for these PDMAEMAg-PS NPs at pH 2.0 in Table 1. At pH 4.0 and 6.0, Tol/H2Otype emulsions were obtained, while H2O/Tol-type emulsions were obtained at pH values above 7.0 (Figure 7B). Pickering emulsions stabilized by other NPs synthesized in this work, at pH 2.0 and 10.0, are presented in Figure S16. Interestingly, the PDMAEMA-g-PS NPs at pH 2.0 are the only ones that have shown a very large Ea (Table 1) and were not able to emulsify Tol and water even under strong shearing by ultrasonication (Figure 7B). In contrast, PDEAEMA-1, PDEAEMA-2, and PDEAEMA-PBMA, which have all exhibited large Ea values (Table 1 and Table S6) at pH 10, can still emulsify Tol and water under ultrasonication at room temperature (Figures 7A and S16), but no emulsion could be obtained upon gentle shaking. Upon increasing the temperature to ∼70 °C, the PDEAEMA-1 NPs are able to emulsify Tol and water at pH 10 without any ultrasonication, just gentle shaking (Figure S17), signaling that an energy barrier to adsorption at Tol/H2O is indeed present. Note that the PDEAEMA-1 NPs did not show any interfacial adsorption at room temperature in the absence of convection.

4. CONCLUSIONS We have systematically investigated the spontaneous adsorption behavior of the PDEAEMA- and PDMAEMA-grafted PS NPs and PDEAEMA-constituted NPs at the Tol−H2O interface. Although the constituted-of and grafted-with species have the same chemical properties, we concluded that these exhibit a contrasting mechanism of spontaneous interfacial adsorption at L−L interfaces. Dynamic IFT results from the pendant drop seem to indicate that the solubility of PDEAEMA and PDMAEMA in toluene is likely the key parameter controlling the spontaneous adsorption of PDEAEMA- and PDMAEMA-grafted PS NPs at the Tol−H2O interface. Namely, when PDEAEMA and PDMAEMA are in their protonated form at pH < 6.0, PDEAEMA, PDMAEMA, and their grafted PS NPs remain in bulk water without adsorbing at the Tol−H2O interface, though they can adsorb spontaneously 6180

DOI: 10.1021/acs.langmuir.8b00877 Langmuir 2018, 34, 6170−6182

Article

Langmuir ORCID

at the Tol−H2O interface in their deprotonated form at pH > 7.0 because of the increased solubility of PDEAEMA and PDMAEMA in toluene. However, PDEAEMA-constituted NPs behave oppositely. When PDEAEMA is in its protonated form at pH < 6.0, they can adsorb spontaneously at the Tol−H2O interface, and in its deprotonated form at pH > 7.0, interfacial adsorption is not observed. One possible reason for the observed behavior is that at pH < 6.0 the PDEAEMAconstituted NPs become soft due to water intake and significant swelling. We cannot explain at the moment how the softness of the NPs can be mechanistically responsible for the adsorption of particles at interfaces other than for the solid particles. However, one could hypothesize that the lower Ea of interfacial adsorption observed as the NPs become softer at lower pH has a more efficient way of expelling the water layers separating the particle surface and toluene phase as the particle approaches the interface. For example, water can freely circulate through the softer swollen particle. In contrast, for the hard NPs the surface must be dewetted/dehydrated before being resolvated by toluene. Another possibility is that polymer chains in soft particles acquire a higher flexibility to rearrange and presumably expose the hydrophobic domains preferentially to toluene. If the magnitude of the IFT drop can be used as a parameter to gauge interfacial activity, comparing the final IFT values achieved by each type of NPs leads to the conclusion that the method of grafting polymers, which have some solubility in organic solvents, is the most effective method to endow the NPs with L−L interfacial activity. In addition, the adsorption kinetics for PDEAEMA-constituted NPs at Tol−H2O interfaces in the pH range of 1.0 to 6.0 is a combination of diffusioncontrolled and energy-barrier-controlled. For PDMAEMAgrafted PS NPs, at pH from 7.0 to 8.0, the adsorption of PDEAEMA-constituted NPs at Tol−H2O interfaces is a combination of diffusion-controlled and energy-barrier-controlled, while at pH 10.0, it appears to be only diffusioncontrolled. The most important finding is that by comparing the trends in ΔE and Ea with the pH of PDEAEMA-constituted NPs with PDMAEMA-grafted PS NPs, their adsorption behavior at the Tol−H2O interface appears to follow a contrasting mechanism of interfacial attachment. Finally, the PDEAEMA-constituted NPs and PDEAEMAand PDMAEMA-grafted PS NPs can effectively emulsify Tol− H2O with the formation of Pickering emulsions. The Pickering emulsion phase can be controlled by the pH values of the solution. In acidic solution (pH 2.0), the Tol/H2O Pickering emulsion is formed, while in basic solution (pH 10.0), only H2O/Tol Pickering emulsions were observed. The main factor leading to the emulsion phase inversion is the change in polarity of the particles and their affinity with respect to the organic and aqueous phases, in agreement with the previously reported data.3,25,30

■

Andrei Honciuc: 0000-0003-2160-2484 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We are especially grateful for the financial support of the Metrohm Foundation (Herisau, Switzerland).

■

(1) Binks, B. P.; Fletcher, P. D. I. Particles Adsorbed at the Oil− Water Interface: A Theoretical Comparison between Spheres of Uniform Wettability and “Janus” Particles. Langmuir 2001, 17, 4708− 4710. (2) Binks, B. P. Particles as Surfactantssimilarities and Differences. Curr. Opin. Colloid Interface Sci. 2002, 7, 21−41. (3) Wu, D.; Chew, J. W.; Honciuc, A. Polarity Reversal in Homologous Series of Surfactant-Free Janus Nanoparticles: Toward the Next Generation of Amphiphiles. Langmuir 2016, 32, 6376−6386. (4) Zanini, M.; Marschelke, C.; Anachkov, S. E.; Marini, E.; Synytska, A.; Isa, L. Universal Emulsion Stabilization from the Arrested Adsorption of Rough Particles at Liquid-Liquid Interfaces. Nat. Commun. 2017, 8, 15701−15710. (5) Chen, Y.; Bai, Y.; Chen, S.; Ju, J.; Li, Y.; Wang, T.; Wang, Q. Stimuli-Responsive Composite Particles as Solid-Stabilizers for Effective Oil Harvesting. ACS Appl. Mater. Interfaces 2014, 6, 13334−13338. (6) Williams, M.; Warren, N. J.; Fielding, L. A.; Armes, S. P.; Verstraete, P.; Smets, J. Preparation of Double Emulsions Using Hybrid Polymer/Silica Particles: New Pickering Emulsifiers with Adjustable Surface Wettability. ACS Appl. Mater. Interfaces 2014, 6, 20919−20927. (7) Toor, A.; Helms, B. A.; Russell, T. P. Effect of Nanoparticle Surfactants on the Breakup of Free-Falling Water Jets during Continuous Processing of Reconfigurable Structured Liquid Droplets. Nano Lett. 2017, 17, 3119−3125. (8) Zhang, Y.; Wang, S.; Zhou, J.; Zhao, R.; Benz, G.; Tcheimou, S.; Meredith, J. C.; Behrens, S. H. Interfacial Activity of Nonamphiphilic Particles in Fluid−Fluid Interfaces. Langmuir 2017, 33, 4511−4519. (9) Ruhland, T. M.; Gröschel, A. H.; Ballard, N.; Skelhon, T. S.; Walther, A.; Müller, A. H. E.; Bon, S. A. F. Influence of Janus Particle Shape on Their Interfacial Behavior at Liquid−Liquid Interfaces. Langmuir 2013, 29, 1388−1394. (10) Sahiner, N.; Atta, A. M.; Yasar, A. O.; Al-Lohedan, H. A.; Ezzat, A. O. Surface Activity of Amphiphilic Cationic PH-Responsive Poly(4Vinylpyridine) Microgel at Air/Water Interface. Colloids Surf., A 2015, 482, 647−655. (11) Mourran, A.; Wu, Y.; Gumerov, R. A.; Rudov, A. A.; Potemkin, I. I.; Pich, A.; Möller, M. When Colloidal Particles Become Polymer Coils. Langmuir 2016, 32, 723−730. (12) Wu, Y.; Wiese, S.; Balaceanu, A.; Richtering, W.; Pich, A. Behavior of Temperature-Responsive Copolymer Microgels at the Oil/Water Interface. Langmuir 2014, 30, 7660−7669. (13) Li, Z.; Richtering, W.; Ngai, T. Poly(N-Isopropylacrylamide) Microgels at the Oil−water Interface: Temperature Effect. Soft Matter 2014, 10, 6182−6191. (14) Richtering, W. Responsive Emulsions Stabilized by StimuliSensitive Microgels: Emulsions with Special Non-Pickering Properties. Langmuir 2012, 28, 17218−17229. (15) Li, Z.; Harbottle, D.; Pensini, E.; Ngai, T.; Richtering, W.; Xu, Z. Fundamental Study of Emulsions Stabilized by Soft and Rigid Particles. Langmuir 2015, 31, 6282−6288. (16) Rumyantsev, A. M.; Gumerov, R. A.; Potemkin, I. I. A Polymer Microgel at a Liquid−liquid Interface: Theory vs. Computer Simulations. Soft Matter 2016, 12, 6799−6811. (17) Style, R. W.; Isa, L.; Dufresne, E. R. Adsorption of Soft Particles at Fluid Interfaces. Soft Matter 2015, 11, 7412−7419.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.langmuir.8b00877. Experimental procedure (PDF)

■

REFERENCES

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Tel: +41589345283. 6181

DOI: 10.1021/acs.langmuir.8b00877 Langmuir 2018, 34, 6170−6182

Article

Langmuir

Revealed by a Dynamic Surface Tension Probe. Langmuir 2014, 30, 710−717. (38) Kim, J.; Cote, L. J.; Kim, F.; Yuan, W.; Shull, K. R.; Huang, J. Graphene Oxide Sheets at Interfaces. J. Am. Chem. Soc. 2010, 132, 8180−8186. (39) Haase, M. F.; Grigoriev, D.; Moehwald, H.; Tiersch, B.; Shchukin, D. G. Encapsulation of Amphoteric Substances in a PHSensitive Pickering Emulsion. J. Phys. Chem. C 2010, 114, 17304− 17310. (40) Yang, H.; Zhou, T.; Zhang, W. A Strategy for Separating and Recycling Solid Catalysts Based on the PH-Triggered PickeringEmulsion Inversion. Angew. Chem., Int. Ed. 2013, 52, 7455−7459. (41) Motornov, M.; Sheparovych, R.; Lupitskyy, R.; MacWilliams, E.; Hoy, O.; Luzinov, I.; Minko, S. Stimuli-Responsive Colloidal Systems from Mixed Brush-Coated Nanoparticles. Adv. Funct. Mater. 2007, 17, 2307−2314. (42) Schrade, A.; Landfester, K.; Ziener, U. Pickering-Type Stabilized Nanoparticles by Heterophase Polymerization. Chem. Soc. Rev. 2013, 42, 6823−6839.

(18) Foster, L. M.; Worthen, A. J.; Foster, E. L.; Dong, J.; Roach, C. M.; Metaxas, A. E.; Hardy, C. D.; Larsen, E. S.; Bollinger, J. A.; Truskett, T. M.; et al. High Interfacial Activity of Polymers “Grafted through” Functionalized Iron Oxide Nanoparticle Clusters. Langmuir 2014, 30, 10188−10196. (19) Tsuji, S.; Kawaguchi, H. Temperature-Sensitive Hairy Particles Prepared by Living Radical Graft Polymerization. Langmuir 2004, 20, 2449−2455. (20) Du, K.; Glogowski, E.; Emrick, T.; Russell, T. P.; Dinsmore, A. D. Adsorption Energy of Nano- and Microparticles at Liquid−Liquid Interfaces. Langmuir 2010, 26, 12518−12522. (21) Amalvy, J. I.; Wanless, E. J.; Li, Y.; Michailidou, V.; Armes, S. P.; Duccini, Y. Synthesis and Characterization of Novel PH-Responsive Microgels Based on Tertiary Amine Methacrylates. Langmuir 2004, 20, 8992−8999. (22) Lin, W.; Nie, S.; Xiong, D.; Guo, X.; Wang, J.; Zhang, L. PHResponsive Micelles Based on (PCL)2(PDEA-b-PPEGMA)2 Miktoarm Polymer: Controlled Synthesis, Characterization, and Application as Anticancer Drug Carrier. Nanoscale Res. Lett. 2014, 9, 243−255. (23) van de Wetering, P.; Zuidam, N. J.; van Steenbergen, M. J.; van der Houwen, O. A. G. J.; Underberg, W. J. M.; Hennink, W. E. A Mechanistic Study of the Hydrolytic Stability of Poly(2(Dimethylamino)Ethyl Methacrylate). Macromolecules 1998, 31, 8063−8068. (24) Tu, F.; Lee, D. Shape-Changing and Amphiphilicity-Reversing Janus Particles with PH-Responsive Surfactant Properties. J. Am. Chem. Soc. 2014, 136, 9999−10006. (25) Wu, D.; Honciuc, A. Design of Janus Nanoparticles with PHTriggered Switchable Amphiphilicity for Interfacial Applications. ACS Appl. Nano Mater. 2018, 1, 471−482. (26) Schmitt, V.; Ravaine, V. Surface Compaction versus Stretching in Pickering Emulsions Stabilised by Microgels. Curr. Opin. Colloid Interface Sci. 2013, 18, 532−541. (27) Chevalier, Y.; Bolzinger, M.-A. Emulsions Stabilized with Solid Nanoparticles: Pickering Emulsions. Colloids Surf., A 2013, 439, 23− 34. (28) Binks, B. P.; Lumsdon, S. O. Influence of Particle Wettability on the Type and Stability of Surfactant-Free Emulsions. Langmuir 2000, 16, 8622−8631. (29) Zhang, K.; Wu, W.; Meng, H.; Guo, K.; Chen, J.-F. Pickering Emulsion Polymerization: Preparation of Polystyrene/Nano-SiO2 Composite Microspheres with Core-Shell Structure. Powder Technol. 2009, 190, 393−400. (30) Wu, D.; Binks, B. P.; Honciuc, A. Modeling the Interfacial Energy of Surfactant-Free Amphiphilic Janus Nanoparticles from Phase Inversion in Pickering Emulsions. Langmuir 2018, 34, 1225− 1233. (31) Muñoz-Bonilla, A.; Fernández-García, M.; Haddleton, D. M. Synthesis and Aqueous Solution Properties of Stimuli-Responsive Triblock Copolymers. Soft Matter 2007, 3, 725−731. (32) Isa, L.; Lucas, F.; Wepf, R.; Reimhult, E. Measuring SingleNanoparticle Wetting Properties by Freeze-Fracture Shadow-Casting Cryo-Scanning Electron Microscopy. Nat. Commun. 2011, 2, 438. (33) Dugyala, V. R.; Muthukuru, J. S.; Mani, E.; Basavaraj, M. G. Role of Electrostatic Interactions in the Adsorption Kinetics of Nanoparticles at Fluid−fluid Interfaces. Phys. Chem. Chem. Phys. 2016, 18, 5499−5508. (34) Ward, A. F. H.; Tordai, L. Time-Dependence of Boundary Tensions of Solutions I. The Role of Diffusion in Time-Effects. J. Chem. Phys. 1946, 14, 453−461. (35) Liggieri, L.; Ravera, F.; Passerone, A. A Diffusion-Based Approach to Mixed Adsorption Kinetics. Colloids Surf., A 1996, 114, 351−359. (36) Ravera, F.; Liggieri, L.; Steinchen, A. Sorption Kinetics Considered as a Renormalized Diffusion Process. J. Colloid Interface Sci. 1993, 156, 109−116. (37) Bizmark, N.; Ioannidis, M. A.; Henneke, D. E. Irreversible Adsorption-Driven Assembly of Nanoparticles at Fluid Interfaces 6182

DOI: 10.1021/acs.langmuir.8b00877 Langmuir 2018, 34, 6170−6182