Article pubs.acs.org/Langmuir
Interactions between Colloidal Particles in the Presence of an Ultrahighly Charged Amphiphilic Polyelectrolyte Danfeng Yu,† Hui Yang,† Hui Wang,† Yingxian Cui,‡ Guang Yang,‡ Jian Zhang,‡ and Jinben Wang*,† †
Key Laboratory of Colloid, Interface and Chemical Thermodynamics, Institute of Chemistry, Chinese Academy of Sciences, No. 2, 1st North Street, Zhongguancun, Beijing 100190, P. R. China ‡ State Key Laboratory of Offshore Oil Exploitation, CNOOC Research Institute, No. 6, Taiyanggong South Street, Chaoyang District, Beijing 100027, P. R. China ABSTRACT: A novel amphiphilic polyelectrolyte denoted as PAGC8 and a traditional amphiphilic polyelectrolyte denoted as PASC8 were prepared. PAGC8 consisted of gemini-type surfactant segment based on 1,3-bis (N,N-dimethyl-N-octylammonium)-2-propyl acrylate dibromide, while PASC8 incorporated acryloyloxyethyl-N,N-dimethyl-N-dodecylammonium bromide as single chain surfactant units within its repeat unit structure. Turbidity, stability, and zeta potential measurements were performed in the presence of PAGC8 and PASC8, respectively, to evaluate their effectiveness in inducing solid/liquid separations. It was found that the maximum transmittance was observed before the zeta potential values reached the isoelectric point, implying that not only charge neutralization but also charge-patch mechanism contributed to the separation process. Colloid probe atomic force microscopy technique was introduced to directly determine the interactions between surfaces in the presence of ultrahighly charged amphiphilic polyelectrolyte. On the basis of the AFM results, we have successfully interpreted the influence of the charge density of the polyelectrolytes on the phase stability. Electrostatic interaction played the dominant role in the flocculation processes, although both electrostatic interaction and hydrophobic effect provided contributions to the colloidal dispersions. The attractions upon surfaces approach in the case of PAGC8 were significantly larger than that of PASC8 due to the higher charge density. The strong peeling events upon retraction in the presence of PAGC8 implied that the hydrophobic effect was stronger than that of PASC8, which displayed the loose pulling events. A strong attraction was identified at shorter separation distances for both systems. However, these interactions cannot be successfully described by the Derjaguin−Landau−Verwey−Overbeek (DLVO) theory of colloid stability due to the participation of charge-patch and strong hydrophobic effect. To account for the additional interactions, we proposed an extended DLVO empirical model to explain the non-DLVO forces in the systems. A reasonable physical model was also proposed to further describe the interactions between surfaces in the two amphiphilic polyelectrolyte systems.
1. INTRODUCTION Polyelectrolytes are an interesting class of macromolecules that can exhibit attractive electrostatic interactions with oppositely charged surfaces via the charge distributed along the backbone or in their side chains.1,2 They are widely used to improve solid/liquid separation processes or to control the stability of colloidal suspensions3,4 and thus are employed in the fields of wastewater treatment, sludge removal, pulp and paper production, as well as in the pharmaceutical and cosmetics industries.5,6 Many of these applications rely on the ability of polyelectrolytes to attach to charged particles and modify the interparticle forces.7 The stability or flocculation of colloidal systems in the presence of oppositely charged polyelectrolyte molecules is regulated by the adsorption of the polyelectrolytes onto the particle surfaces.8,9 Polymers with very high charge densities can bring out a great number of charges at the adsorption sites, resulting in enhanced attractive electrostatic interactions between positive patches on one surface and negative patches on another surface.10 Therefore, increasing the charge density is © 2014 American Chemical Society
an effective method to improve the efficiency of solid/liquid separations. The significant parameter affecting the solution properties of a polyelectrolyte is the linear charge density, whose definition has been promoted by Manning in his counterion condensation theory. In this model, the polyelectrolyte is assumed to be a linear charged cylinder whose charge density is evaluate by the value of ξ, which was further calculated and explained in detail in the following section.11 The homopolymer PASC8 and PAGC8, with higher values of ξ than that of the conventional polycations, were regarded as high charge density and ultrahigh charge density polyelectrolytes in this study.12,13 Polyelectrolyte with ultrahigh charge density can drastically modify the properties of dispersed systems. However, research on the interactions between surfaces in the presence of polyelectrolytes with ultrahigh charge density is still less developed and far from being completely understood.14 Received: July 30, 2014 Revised: October 10, 2014 Published: November 14, 2014 14512
dx.doi.org/10.1021/la503033k | Langmuir 2014, 30, 14512−14521
Langmuir
Article
investigation. In addition, a reasonable physical model with respect to the force interactions was proposed to further interpret the discrepancy of forces in the presence of the two polymers. The experimental force−distance curves were fitted according to the DLVO theory and the extended DLVO model incorporating patch−charge interactions and the hydrophobic effect.28
Amphiphilic polyelectrolytes can exhibit different types of intra/intermolecular associations between hydrophobic moieties besides electrostatic interactions.15 Moreover, cationic polyelectrolytes incorporating hydrophobic chains have a greater ability to enhance the performance of solid/liquid separation systems involving large amounts of surfactants, oils, or dyes.7,16 We have developed a new kind of polyelectrolyte (PAGC8) with the merits of both ultrahigh charge density and amphiphilic properties, which we have called “geminized amphiphilic polyelectrolytes”.17 This polyelectrolyte was used to enhance the efficiency of solid/liquid separation and showed excellent performance in regulating the phase behavior in the present work. This kind of geminized amphiphilic polyelectrolyte can thus provide remarkable performance in enhancing separations and is not only of academic interest but also has potential applications in various industrial fields. Colloid probe atomic force microscopy (CP-AFM) was used to obtain a direct and comprehensive picture of the interactions acting between surfaces in the presence of the two amphiphilic polyelectrolytes.18,19 A spherical particle was glued to a cantilever according to the cantilever-moving technique, which was reformed in this study. The entire attachment process can be performed via AFM without the need for other microscopes. So far, no systematic studies using CP-AFM to characterize the interaction forces encountered in the presence of ultrahighly charged amphiphilic polyelectrolyte have been reported, and the mechanism for the hydrophobic effect is still under debate.20,21 Amphiphilic polyelectrolytes adsorb onto a surface as a result of short-range and long-range forces incorporating electrostatic forces, van der Waals forces, and specific interactions.22−24 The interactions between surfaces adsorbed polyelectrolyte molecules are strongly influenced by the properties of the polyelectrolyte, such as its charge density and whether it possesses hydrophobic chains. The polyelectrolyte molecules adsorb onto a surface to form laterally heterogeneous sites, thus creating localized domains bearing positive or negative charges, which leads to additional electrostatic patch-charge attractions.25,26 The hydrophobic chains of amphiphilic polyelectrolytes protruding from the surfaces will display hydrophobic effect in addition to the electrostatic interaction. The classical DLVO (Derjaguin, Landau, Verwey, and Overbeek) theory, which considers only two types of forces (the electrical doublelayer repulsion and the London−van der Waals attraction), predicts that a suspension becomes unstable near the isoelectric point (IEP) where the interactions between the particles are dominated by attractive van der Waals forces.27 Further away from the IEP, electrostatic repulsions become dominant and the suspension becomes stable. We have attempted to extend the DLVO theory to include the additional interactions encountered in the novel amphiphilic polyelectrolyte systems to further understand how these interactions influence the solid/liquid separation process. We present an investigation on the interactions between silica surfaces in the presence of PAGC8 and PASC8. Turbidity, zeta potential, stability, and AFM measurements were performed to investigate the influence of these polyelectrolytes on the solid/liquid separation process exhibited by the colloidal dispersions. CP-AFM force measurements were performed in the presence of the two highly charged amphiphilic polyelectrolytes. Silica spheres and silicon substrates were used as examples of negatively charged surfaces during this
2. EXPERIMENTAL SECTION 2.1. Materials. Two amphiphilic polyelectrolytes were prepared in our laboratory.17 One was a cationic homopolymer of 1,3-bis (N,Ndimethyl-N-octylammonium)-2-propyl acrylate dibromide with ultrahigh charge density and double hydrophobic chains in each repeat unit (geminized amphiphilic polyelectrolyte, PAGC8) (Figure 1a). The
Figure 1. General chemical structure of the polyelectrolytes PAGC8 (a), PASC8 (b), and PASC1 (c). other polyelectrolyte was a cationic homopolymer of acryloyloxyethylN,N-dimethyl-N-octyl ammonium bromide bearing high charge density and single hydrophobic chains in each repeat unit (PASC8, Figure 1b). The highly charged hydrophilic polyelectrolyte (PASC1, Figure 1c) was purchased from Beijing Chemical without further purification (1.3 × 104 g/mol) and was used in the zeta (ζ) potential and flocculent size distribution experiments. Millipore Milli-Q grade water was used in all experiments. Silica particles of 50 nm in diameter (Aladdin, China) were used for the turbidity, stability, and ζ potential measurements. Silica spheres of 10 μm (Aladdin, China) were used as the colloidal probe. The AFM cantilever (Asylum Research, Santa Barbara, CA) with a nominal spring constant of 0.35 N/m was used throughout this study. Silicon wafers were cut (0.5 cm × 0.5 cm) and successively washed with acetone (10 min), ethanol (10 min), and ultrapure water (10 min) before they were dried with highly pure nitrogen and then kept in an O2 plasma cleaner for 30 min (Harrick Plasma cleaner, PDC32G, USA) to yield a highly hydrophilic surface. The surface roughness of these substrates was measured via AFM in the tapping 14513
dx.doi.org/10.1021/la503033k | Langmuir 2014, 30, 14512−14521
Langmuir
Article
mode. In a typical area (1 μm × 1 μm) the standard deviation from the mean height was 1 nm. 2.2. Methods. 2.2.1. Turbidity Measurements. The aqueous polyelectrolyte solutions and silica dispersions were prepared using Millipore Milli-Q grade water. The concentration of the model suspension used here was 1.5 g/L. The suspension was sonicated for 30 min using an ultrasonicator (KQ3200E, Kunshan instruments, China) and subsequently stirred for 15 min at 300 rpm to fully disperse the powder. The measurements were performed in a 10 mL glass container with water circulation, and the temperature was kept at 25.0 °C. A 5 mL suspension was added to the glass container under stirring before the polyelectrolyte solutions were gradually added and the mixture was stirred for an additional 3 min. After the dispersion was allowed to settle for 3 min, the transmittance (T %) of the supernatant was measured at 450 nm using a probe colorimeter (PC920, Brinkmann, Germany). Before the measurements were performed, the probe was immersed into Millipore Milli-Q grade water and the transmittance (T %) was calibrated to 100%. The original concentration of the polyelectrolytes used in this experiment was 1 g/L. 2.2.2. Stability Measurements. The stability of the dispersions was measured using a dispersion analyzer (LUMiSizer, LUM, Germany). This system was equipped with an accelerated centrifuge and, at the same time, could collect spectral changes of the sample based on the Stokes Law and the Beer−Lambert Law.28 The original concentration of the suspension used for this experiment was 10 g/L. The polymers with different concentrations were added to the fully dispersed silica suspension. The suspensions were subsequently injected into the polycarbonate tube, which was collocated to the dispersion analyzer after it had been stirred for 3 min. The tubes were placed into the dispersion analyzer, and the centrifuge speed was set to 300 rpm after the temperature was equilibrated at 25.0 °C. The transmittance of the suspensions according to the position of the tube were measured and used to analyze the stability of the suspension. 2.2.3. Zeta Potential Measurements. The same suspension used for the turbidity measurements was also used for the ζ potential measurements. The ζ potentials of the supernatant in the presence and absence of polymers (PAGC8, PASC8, and PASC1) were measured using a Zetasizer (Nano-ZS, Malvern, England) instrument after the suspension had been settled for 3 min and calculated according to the Smoluchowski theory. The measured electrophoretic mobilities (U) were converted into ζ potentials according to the equation U = εζ/η.30 Each ζ potential value was derived from the average of three measurements. 2.2.4. Flocculent Size Distribution Measurements. The same suspension used for the turbidity measurements was also used for the flocculent size measurement. The flocculent size distributions of the suspension in the presence of polymers (PAGC8, PASC8, and PASC1) were measured using a Zetasizer (Nano-ZS, Malvern, England). A 2 mL suspension was prepared and added to the cell; then, the polyelectrolyte solution was added directly to the cell and mixed in a controlled manner, for example, 5 s on a vortex mixer or turning 10 times overhead. Afterward the cell was put into the instrument and the measurement was started. 2.2.5. AFM Force-Measuring Technique. AFM force measurements were conducted between the colloidal probe and silicon wafer surface in aqueous solutions.31 A Multimode VIII atomic force microscope (Digital Instruments, Bruker, USA) with an O-ring liquid cell was used in contact mode. A silica sphere of 10 μm in diameter was glued to the cantilever with an epoxy resin according to the reformative cantilevermoving technique. An example of such a probe is shown in Figure 2. The actual spring constants of the cantilevers bearing the colloidal particle were in the range of 0.3 to 0.4 N/m, which were determined using the “thermal tune” method introduced by Hutter and Bechhoefer.32 Force curves were obtained by converting the cantilever deflections (mV) and piezotube displacements in accordance with Hooke’s law. The colloidal probes were washed with ethanol before they were dried with nitrogen and used to determine the forces with a freshly cleaved silicon surface. The AFM fluid cell was rinsed with ethanol and dried with nitrogen prior to the experiments. The forces
Figure 2. Representative SEM image of the colloidal probe. between the flat surface and the colloidal probe were measured as a function of the separation distance. The ramp rate for an approach/ retraction cycle was adjusted to 1 Hz. At least 50 force curves were recorded at different locations on the surface for each experimental condition. Force curves were analyzed using the NanoScope Analysis software (Bruker, Germany). 2.2.6. Theory and Calculation. 1. Charge Density. To clarify the impact of the charge density of the two amphiphilic polyelectrolyte molecules on the oppositely charged colloidal particles, we calculated the linear charge density according to Manning’s theory33 ξ=
e2 εkTb
(1)
where e represents the elemental charge, ε denotes the dielectric constant of the solvent, k denotes the Boltzmann constant, and T represents the temperature in kelvin. Meanwhile, b corresponds to the average spacing between two vicinal ionized groups on the polyion based on a repeating unit length of 0.25 nm of a polyethylene unit.34 The charge densities of the two polyelectrolytes calculated using the above formula was found to be 5.68 and 11.34 for PASC8 and PAGC8, respectively. These values were much larger than those of conventional polycations that are used as flocculants.29,35,36 2. DLVO Theory. Interfacial interactions between particles and a substrate in aqueous media are typically described in terms of the DLVO theory, in which the van der Waals forces37 and electrostatic double-layer interactions38 can be expressed as shown in eq 2
Ft = Fvdw + Fedl
(2)
where Ft represents the total interaction force, FvdW corresponds to the van der Waals force, and Fedl denotes the electrical double-layer force.39 The key point in calculating the van der Waals forces is to determine the Hamaker constant. For a spherical tip interacting with a flat substrate, the theoretical van der Waals force can be approximated by eq 3
Fvdw =
− A131R 6d 2
(3)
where A131 is the Hamaker constant, R, denotes the radius of the spherical particle, and d is the separation distance. An approximation of the Hamaker constant is obtained by using a simplification of the Lifshitz theory40 of two solid phases interacting across a medium as expressed by eq 4
A131 =
2 3hνe (n12 − n32)2 3 ⎛ ε1 − ε3 ⎞ kT ⎜ ⎟ + 4 ⎝ ε1 + ε3 ⎠ 16 2 (n12 + n32)3/2
(4)
−23
where k denotes Boltzmann’s constant (1.381 × 10 J/K), T is the absolute temperature (in K), h is Planck’s constant (6.626 × 10‑34 J·s), and ve denotes the absorption frequency (3 × 1015 Hz). Meanwhile, ε1 14514
dx.doi.org/10.1021/la503033k | Langmuir 2014, 30, 14512−14521
Langmuir
Article
denotes the dielectric constant of silica (3.8), ε3 corresponds to the dielectric constant of the medium (80), n1 is the refractive index of silica (1.448), and n3 is the refractive index of the medium (1.333). The double-layer force (Fedl), modeled with the boundary conditions at constant surface potential is described by Fψ in eq 5
Fψ =
2 2 2πεε0 2ψ1ψ2 exp(d /λD) − (ψ1) − (ψ2) RλD exp(2d /λD) − 1
(5)
where ε0 (8.854 × 10‑12 C2/J·m) is the permittivity of vacuum, and ε (80) denotes the relative permittivity (the dielectric constant) of the solution. Meanwhile, ψ1 and ψ2 correspond to the surface potentials of the substrate and probe surfaces, respectively. The surface potentials were approximated by the ζ potentials (mV) of the silicon surface and the silica particle attached to the AFM tip. λD denotes the Debye length as described by eq 6 λD =
Figure 3. Transmittance curves of silica dispersions that are mixed with PAGC8 and PASC8 at different concentrations. The inset shows the macroscopic flocculation process of silica suspension (1.5 g/L) in the presence of PAGC8 (1 g/L) at different concentrations.
εε0kBT 2ce 2
(6)
where e is the electron charge (1.602 × 10‑19 C) and c corresponds to the concentration.41 3. Extended DLVO Model. In the present paper, the DLVO theory was extended to provide a more complete interpretation of the force− distance experimental data. Specifically, a term describing the structural forces including the hydrophobic and charge-patch interactions was introduced, as expressed in eq 7 Ft = Fvdw + Fedl + Fs
The addition of PAGC8, which possessed a higher charge density and pronounced hydrophobicity, caused the solid/ liquid separation to occur in a more efficient manner than was achieved with PASC8. Consequently, the solution became transparent at a lower polyelectrolyte concentration than was observed in the PASC8 system. The very highly charged PAGC8 tended to adsorb onto the spherical particles with more effective charge neutralization than was encountered with PASC8, which had a lower charge density. The greater hydrophobicity (due to the gemini double alkyl tails) endowed the PAGC8 molecules with an excellent ability to facilitate the solid/liquid separation. In the case of PASC8, however, electrostatic interactions represented the main force driving the flocculation. PASC8 exhibited a relatively weak hydrophobic effect, and thus the solid/liquid separation was less efficient. Polyelectrolyte molecules adsorbed at the solid surface possessed some uncompensated charge sites, and the aggregation of dispersed particles through the charge-patch mechanism is also reasonable besides the charge neutralization. The charge-patch mechanism is typical for highly charged polyelectrolytes, which usually adsorb with local surface overcharging.9 Both PAGC8 and PASC8 have the potential and capacity to exhibit charge-patch interactions, in addition to the charge neutralization. 3.2. Effects of the Amphiphilic Polyelectrolytes on the Stability of Colloidal Dispersion. The stability of the silica dispersions was analyzed as a function of the polymer concentration. The transmittance of the silica dispersions increased with increases in the concentrations of PAGC8 and PASC8, during which the negatively charged particles were gradually neutralized by the positively charged polymers. However, the transmittance began to decrease and the system tended to become stable as the polyelectrolyte concentration was increased further, during which repulsive electrostatic double-layer forces as expected from classical DLVO theory became mainly responsible for the stabilization. Figure 4 shows the stability curves of a silica dispersion that was mixed with PAGC8 (a) and PASC8 (b) at polyelectrolyte concentrations of 3 (a) and 6 mg/L (b), respectively. The results demonstrated that the PAGC8−silica system exhibited a higher transmittance (80%) than was provided by the PASC8− silica system (60%). A more compact sediment was also observed in the presence of PAGC8 (4 mm) than was observed in the PASC8−silica system (8 mm), indicating that the
(7)
where Fs represented the structural force including both the hydrophobic and charge-patch interactions. This structural force is often expressed as a double-exponential function as described by eq 8
Fs = R[C1e−d / D1 + C2e−d / D2]
(8)
where C1 and D1 correspond to the non-DLVO forces observed at separation distances (d) of
