ARTICLE pubs.acs.org/Langmuir
Adsorption Behavior of Hydrophobin and Hydrophobin/Surfactant Mixtures at the SolidSolution Interface Xiaoli L. Zhang,† Jeffrey Penfold,*,†,‡ Robert K. Thomas,† Ian M. Tucker,§ Jordan T. Petkov,§ Julian Bent,|| and Andrew Cox|| †
Physical and Theoretical Chemistry Laboratory, Oxford University, South Parks Road, Oxford, U.K. STFC, Rutherford Appleton Laboratory, Chilton, Didcot, Oxon, U.K. § Unilever Research and Development Laboratory, Port Sunlight, Quarry Road East, Bebington, Wirral, U.K. Unilever Research Laboratories, Sharnbrook, MK44 1LQ Beds, U.K.
)
‡
bS Supporting Information ABSTRACT: The adsorption of surface-active protein hydrophobin, HFBII, and HFBII/surfactant mixtures at the solidsolution interface has been studied by neutron reflectivity, NR. At the hydrophilic silicon surface, HFBII adsorbs reversibly in the form of a bilayer at the interface. HFBII adsorption dominates the coadsorption of HFBII with cationic and anionic surfactants hexadecyltrimethyl ammonium bromide, CTAB, and sodium dodecyl sulfate, SDS, at concentrations below the critical micellar concentration, cmc, of conventional cosurfactants. For surfactant concentrations above the cmc, HFBII/surfactant solution complex formation dominates and there is little HFBII adsorption. Above the cmc, CTAB replaces HFBII at the interface, but for SDS, there is no affinity for the anionic silicon surface hence there is no resultant adsorption. HFBII adsorbs onto a hydrophobic surface (established by an octadecyl trimethyl silane, OTS, layer on silicon) irreversibly as a monolayer, similar to what is observed at the airwater interface but with a different orientation at the interface. Below the cmc, SDS and CTAB have little impact upon the adsorbed layer of HFBII. For concentrations above the cmc, conventional surfactants (CTAB and SDS) displace most of the HFBII at the interface. For nonionic surfactant C12E6, the pattern of adsorption is slightly different, and although some coadsorption at the interface takes place, C12E6 has little impact on the HFBII adsorption.
’ INTRODUCTION Biosurfactants are of much current interest1,2 because of their potential applications in areas involving bioremediation,3 biosustainability, and biodegradation.4 Glycolipids1 such as the rhamnolipids, sophorolipids, and mannostylerythitol lipids and lipopeptides and proteins such as surfactin and hydrophobin are those most commonly considered and studied. Much of their attraction and potential lies in their novel surface adsorption properties, and yet many of these aspects are poorly characterized or understood. Hydrophobin is a small (710 kDa), highly surface-active globular protein that is produced by filamentous fungi. Hydrophobin exists as two major types, HFBI and HFBII, where HFBI is the more hydrophobic species. This study focuses on the more water-soluble HFBII, which has been produced here via the fermentation of a modified yeast. The crystal structure of HFBII has been established.5,6 It is characterized by a conserved pattern of eight cysteine residues that make four intramolecular disulfide bridges, and this makes the protein very compact and robust. The structure is nearly globular, with a central β-barrel structure and a small segment of an R-helix. The surface activity and functionality of HFBII arise primarily from a hydrophobic patch r 2011 American Chemical Society
consisting of side-chain residues of leucine, valine, and analine, which occupy some 20% of the surface area of the protein. The biological function of hydrophobin is diverse, and hydrophobins are both secreted at and retained within fungal structures. The functionality of the hydrophobins is related to the pronounced surface activity and hence usually involves surfaces and interfaces.2 Hydrophobins act as a coating or protective (hydrophobic) surface, aid adhesion, and promote surface modification. The ability of hydrophobins to lower the surface tension, aid the growth of aerial hyphae, coat spores and aerial hyphae, and promote the attachment of fungi to solid surfaces shows their pronounced surface properties.2 The unusually pronounced surface properties have generated much interest in potential industrial applications as surface coatings and protective agents and in adhesion. The other major attraction is the ability to stabilize dispersions, such as foams and emulsions, and this has considerable potential in aerated foods such as ice cream and other food-based emulsions and foams.2 Received: May 9, 2011 Revised: July 27, 2011 Published: July 28, 2011 10464
dx.doi.org/10.1021/la201711p | Langmuir 2011, 27, 10464–10474
Langmuir In spite of the overriding importance of the surface properties of hydrophobin, there have been relatively few detailed studies. Kisko et al.7 and Kallio et al.6 reported the formation of monolayer films at the airwater interface from Brewster angle microscope, grazing incidence diffraction, and X-ray reflectivity measurements. In a more recent study,8 we used NR to study the adsorption of HFBII and the competitive adsorption of HFBII with surfactants SDS, CTAB, and C12E6 at the airwater interface. The NR measurements showed that HFBII adsorbed to form a dense monolayer that was ∼30 Å thick at the airwater interface and that it totally dominates the adsorption when mixed with conventional surfactants SDS, CTAB, and C12E6 at concentrations below the surfactant cmc, whereas above the cmc HFBII is displaced by the surfactant. Others911 have reported more highly organized surface assemblies, forming rodlet-type structures, at a variety of different interfaces. Otherwise, there is very little detailed information about the adsorption of HFBII and HFBII/surfactant mixtures at different interfaces. Kisko et al.12 used small-angle X-ray scattering, SAXS, to study the solution self-assembly of HFBI and HFBII. The predominant structure obtained was in the form of hydrophobin tetramers, where the self-assembly was largely driven by the hydrophobic interactions. Zhang et al.13 have recently confirmed the formation of hydrophobin tetramers in solution using SANS and have identified and characterized the structure of HFBII/surfactant mixed aggregates. To maximize the potential of hydrophobin for applications, a more detailed understanding of the surface and self-assembly properties of hydrophobin, and especially of the interaction with conventional surfactants and more flexible proteins such as β-casein, is essential. In this article, neutron reflectivity, NR, has been used to probe the adsorption behavior of HFBII and HFBII in combination with the anionic, cationic, and nonionic surfactants of sodium dodecyl sulfate, SDS, hexadecyltrimethyl ammonium bromide, CTAB, and hexaethylene monododecyl ether, C12E6, at different solidsolution interfaces, both hydrophilic and hydrophobic. This is one of a series of articles that will explore some of the fundamental physicochemical properties of HFBII/surfactant and HFBII/protein mixtures at different surfaces and in solution.
’ EXPERIMENTAL DETAILS Specular neutron reflectivity measurements were made on the D17 reflectometer14 at the Institute Laue Langevin, Grenoble, France, using the white beam time-of-flight method. Measurements were made in the Q range (where Q is the wave vector transfer normal to the surface of interface, z is direction, Q = 4π sin θ/λ, λ is the neutron wavelength, and θ the grazing angle of incidence) of 0.009 to 0.26 Å1 using the neutron wavelength range of 2.5 to 20 Å and two different grazing angles of incidence, 0.8 and 3.0. The reflectivity measurements for the set of two angles typically took ∼60 min. The resolution of Q, ΔQ/Q, was ∼4%. The neutron beam is incident at grazing incidence at the solidsolution interface by transmission through the upper crystalline silicon phase. The surface of the silicon, Æ111æ, supplied by Crystran, was polished to a rms, and the illuminated area was ∼30 30 surface roughness of e5 Å mm2. The upper (unused) surface of this crystal was protected using a shaped Teflon spacer that contacted only the outer 2 mm of the crystal. The cell uses a solvent volume of ∼5 mL, and the exchange of ∼2030 mL of solution using a manual syringe (at a rate of ∼10 to 20 mL/min) provides efficient sample changing and rinsing. When exchanging the solution in the cell, the existing solution was initially removed, and the next solution was the flushed through approximately
ARTICLE
four to six times. The data were normalized for the incident beam spectral distribution, measurement time, and detector efficiency and established on an absolute scale by reference to the direct beam intensity and a standard water scatterer. Specular neutron reflectivity provides information about the composition and concentration profile in the direction perpendicular to the surface or interface on a molecular scale, probing distances of ∼10 to , and is described in detail elsewhere.15,16 The specular ∼3000 Å reflectivity, R(Q), can be described using the kinematic approximation16 or in terms of the optical description of reflectivity in thin films.17 The latter approach is used in the modeling of the reflectivity data reported in this article. The simplest model, with the least number of layers, consistent with the data and assessed by least-squares, is adopted. A key feature of the technique that is exploited here is that for neutrons the refractive index can be manipulated using H/D isotopic substitution, where H and D have vastly different scattering powers for neutrons. This gives rise to a difference in the scattering-length density F(z) or refractive index, n(z), where n(z) is defined as nðzÞ ¼ 1
λ2 FðzÞ 2π
ð1Þ
H/D isotopic substitution can be applied to the adsorbate and solvent, and modeling the reflectivity data from the different combinations simultaneously can minimize uncertainties in the model.16 It relies on there being no significant isotopic dependence on the adsorption and the structure of the adsorbed layer, and this is well established for these types of systems.16 The neutron scattering lengths, Σb, and molecular volumes associated with the different components used in this study are summarized in Table 1 in the Supporting Information. The deuterium-labeled surfactants (alkyl chain deuterated) of SDS and CTAB (abbreviated d-SDS and d-CTAB, respectively) and deuterated OTS (d-OTS) were synthesized using established synthesis procedures.18,19 Hydrogenous SDS and CTAB (h-SDS and h-CTAB, respectively) were obtained from Sigma. SDS and CTAB were recrystallized from ethanol/acetone mixtures. Hydrogeneous C12E6 (h-C12E6) was obtained from Nikkol and used as supplied. Surface tension measurements with the absence of a minimum at the cmc were used as a criterion for acceptable purity. The cmc values for the different surfactants are 8 103 M for SDS, 9 104 M for CTAB, and 9 105 M for C12E6,20 and their corresponding molecular weights, MWs, are 288, 364, and 406 Da. Class II hydrophobin HFBII (MW ∼7 kDa) was produced at an external fermentation company (BAC) using a yeast fermentation route and was purified by a two-phase extraction route at Unilever Research, Vlaardingen, as described in detail elsewhere.21,22 The extracted and purified material was freeze dried before use. All of the NR measurements were made in D2O, which was obtained from Fluorochem. All glassware and the Teflon bases of the liquidsolid cells were cleaned using alkali detergent (Decon 90) and rinsed thoroughly in high-purity water (Elga Ultrapure). All of the measurements were made at a fixed temperature of 30 C. The silicon surfaces were prepared using a mild piranha treatment, as described in detail elsewhere,23 to ensure a hydrophilic oxide surface of well-defined thickness and density. The hydrophobic surfaces were prepared by the deposition of d-OTS following established procedures.24 NR measurements were made for HFBII and HFBII/surfactant (SDS, CTAB, and C12E6) mixtures at the hydrophilic and hydrophobic surfaces (established by a d-OTS layer). The bare hydrophilic and hydrophobic surfaces were characterized in D2O. d-OTS was chosen to provide a well-defined hydrophobic surface with an optimal contrast for the subsequent HFBII and HFBII/surfactant adsorption measurements. NR measurements were made in D2O to characterize the bare hydrophilic and hydrophobic OTS surfaces used in the adsorption measurements. The mixed HFBII/surfactant solutions were prepared 10465
dx.doi.org/10.1021/la201711p |Langmuir 2011, 27, 10464–10474
Langmuir in D2O and stored for several hours at a temperature above the Krafft point of the surfactant. At the hydrophilic surface, the HFBII adsorption (at an HFBII concentration of 0.2 mg/mL) in D2O was measured, followed by rinsing in D2O and a subsequent measurement after rinsing in D2O. The adsorption of the SDS/HFBII and the CTAB/HFBII mixtures in D2O was then measured on different fresh hydrophilic surfaces at surfactant concentrations that were less than and greater than the cmc and for h and d surfactants. Because the initial measurements of the HFBII adsorption on the hydrophilic surface showed (see later description for details) that the HFBII adsorption was reversible, the subsequent HFBII/surfactant mixed adsorption measurements were made for HFBII/surfactant mixed solutions. For the HFBII/SDS adsorption onto hydrophilic silica measurements, the following detailed sequence of measurements was made: (i) 0.2 mg/mL HFBII/D2O (ii) rinsing and measurement in D2O (iii) 0.1 mg/mL HFBII: 2.5 104 M h-SDS mixture in D2O (iv) rinsing and measurement in D2O (v) 0.1 mg/mL HFBII: 2.5 104 M d-SDS mixture in D2O (vi) rinsing and measurement in D2O (vii) 0.1 mg/mL HFBII: 102 M h-SDS mixture in D2O. A broadly similar sequence of measurements was made for the hbf2/ CTAB mixed adsorption on hydrophilic silica and is given as follows: (i) 0.2 mg/mL HFBII/D2O (ii) rinsing and measurement in D2O (iii) 0.1 mg/mL HFBII: 5 105 M h-CTAB mixture in D2O (iv) rinsing and measurement in D2O (v) 0.1 mg/mL HFBII: 5 105 M d-CTAB mixture in D2O (vi) rinsing and measurement in D2O (vii) 0.1 mg/mL HFBII: 103 M h-CTAB mixture in D2O (viii) rinsing and measurement in D2O (ix) 0.1 mg/mL HFBII: 103 M d-CTAB mixture in D2O (x) rinsing and measurement in D2O. A broadly similar sequence of measurements was made for HFBII/SDS, HFBII/CTAB, and HFBII/C12E6 mixed adsorption onto the hydrophobic (established using a d-OTS layer) silica surface. Because HFBII now adsorbs irreversibly (with respect to rinsing in solvent, D2O) on the OTS surface (see later discussion for details), the measurement sequence is slightly modified compared to the measurements on the hydrophilic surfaces. Here, following the initial HFBII adsorption and rinsing in D2O, the subsequent exposure of the HFBII-coated OTS surfaces is to h- and d-surfactant solutions in D2O, not to HFBII/surfactant mixtures. As such, the following sequence of measurements was made for SDS: (i) 0.2 mg/mL HFBII/D2O (ii) rinsing and measurement in D2O (iii) 5 104 M h-SDS/D2O (iv) 2 102 M h-SDS/D2O (v) rinsing and measurement in D2O (vi) 2 102 M d-SDS/D2O (vii) rinsing and measurement in D2O. An almost identical sequence of measurements was made for CTAB, where the CTAB concentration below the cmc was 104 M and the CTAB concentration above the cmc was 2 103 M. For C12E6, similar measurements were also made, but using only h-C12E6 such that the sequence of measurements was the following: (i) 0.2 mg/mL HFBII/D2O (ii) rinsing and measurement in D2O (iii) 2 105 M h-C12E6/D2O (iv) rinsing and measurement in D2O (v) 2 104 M h-C12E6/D2O. Some repeated sample exchanges and additional NR measurements in D2O were made to confirm that the solvent exchange and surface rinsing were efficient and complete.
ARTICLE
Neutron reflectivity measurements of the bare hydrophilic and OTScoated hydrophobic surfaces were made in D2O in order to provide an initial characterization of the bare solid surfaces. Figure 1a in the Supporting Information shows the reflectivity data for one of the hydrophilic surfaces used in this study in D2O. The second surface that was used produced, within error, identical data. The NR data for the hydrophilic surface are consistent with a well-defined thin oxide layer at the silicon/D2O interface with a thickness of 10 ( 1 Å and a scattering-length density of (2.7 ( 0.2) 106 Å2. This corresponds to an oxide layer that has a density that is ∼80% of that of an idealized oxide layer and is broadly similar and consistent with that reported elsewhere for the same mild piranha treatment.23 Figure 1b in the Supporting Information shows the NR data for one of the hydrophobic d-OTS-coated silicon surfaces in D2O used in this study. Three different OTS surfaces were used, and within experimental error, the data are identical for each surface. The pronounced interference fringe observed in the reflectivity data is consistent with the formation of a well-defined OTS layer with a thickness of ∼30 Å. The solid line in Figure 1b in the Supporting Information is a model calculation for the simplest model that is consistent with the data. Accounting for the oxide layer on the silicon, as previously described, the d-OTS layer can be described by two layers. The two-layer model consists of a inner layer (adjacent to the oxide layer) with a thickness of 18 ( 1 Å and a scattering-length density of (6.5 ( 0.2) 106 Å2 and an outer layer (adjacent to the solvent) with a thickness of 17 ( 2 Å and a scattering-length density of (4.3 ( 0.3) 106 Å2. The structure of the OTS layer is broadly similar to that reported elsewhere24 using a similar method of deposition. The inner layer corresponds to a dense close-packed layer with a packing close to ideal (95% of the density of a close-packed layer), and the outer layer is a more disordered, less dense layer with a volume fraction of ∼70%.
’ RESULTS AND DISCUSSION HFBII Adsorption into Hydrophilic and Hydrophobic Solid Surfaces. Neutron reflectivity measurements were made for
HFBII adsorption onto hydrophilic and OTS-coated hydrophobic silicon surfaces in D2O for an HFBII concentration of 0.2 mg/ mL, and the NR data for the adsorption onto both types of surface are shown in Figure 1. The NR data in Figure 1a shows that HFBII adsorbs onto the hydrophilic silica surface to form a thin, well-defined layer with a mean thickness of 42 ( 1 Å and a scattering-length density of (3.5 ( 0.1) 106 Å2. This corresponds to a dense closepacked layer with a volume fraction, j, of HFBII of ∼ 80%. The thickness of the adsorbed layer is systematically thicker than that observed at the airwater interface, where a mean thickness of ∼31 ( 2 Å was observed.8 This would imply a different orientation of the molecule at the two different interfaces. However, because the approximate dimensions of this globular protein are 24 27 30 Å3,6 the differences are more consistent with bilayer formation at the solid hydrophilic interface. The bilayer is probably in the same basic dimer formation observed as one of the crystal structures, with the hydrophobic patches at the center of the dimer. It is well established that conventional surfactants adsorb at hydrophilic surfaces with structures related to their aggregate state, and it is not surprising that hydrophobin should be similar in this respect. Although the NR data show the adsorption of HFBII at the hydrophilic interface for an HFBII concentration of 0.2 mg/mL, the data after rinsing in D2O are similar/identical to that of the bare interface (before HFBII adsorption). Hence, rinsing the surface in D2O results in the complete removal of the adsorbed HFBII from the interface. 10466
dx.doi.org/10.1021/la201711p |Langmuir 2011, 27, 10464–10474
Langmuir
ARTICLE
Figure 1. (a) NR data for HFBII adsorption onto a hydrophilic surface in D2O, (black) bare hydrophilic surface preadsorption, (blue) a bare interface, post HFBII adsorption and rinsing in D2O, and (red) + 0.2 mg/mL HFBII. The solid line is a calculated curve for a single layer with a thickness of 42 ( 1 Å and a scattering-length density of (3.5 ( 0.1) 106 Å2. (b) NR data for HFBII adsorption onto an OTS hydrophobic surface in D2O, (black) bare OTS surface preadsorption, (blue) OTS surface after rinsing in D2O, and (red) + 0.2 mg/mL HFBII. The solid line is a calculated curve for a single layer of HFBII on top of the OTS with a thickness of 20 ( 1 Å and a scattering-length density of (3.5 ( 0.2) 106 Å2.
Although HFBII adsorbs as a dense layer at the hydrophilic interface, the adsorption is reversible and hence HFBII is only loosely bound to the surface. The NR data in Figure 1b for the adsorption of HFBII onto the OTS-coated hydrophobic surface shows a different pattern of behavior. The differences in the NR data in D2O and for 0.2 mg/ mL HFBII in D2O show a pronounced shift to lower scattering vectors, Q, in the interference fringe arising from the OTS layer and an increase in its visibility. This is associated with the adsorption of HFBII onto the OTS hydrophobic surface. The solid line in Figure 1b is for a model calculation where the HFBII adsorbed layer is described by a single uniform layer on top of the OTS with a thickness of 20 ( 1 Å and a scattering-length density of 3.5 ( 0.1 106 Å2. The parameters describing the
underlying OTS layer are unaffected by the HFBII adsorption. The adsorbed HFBII layer corresponds to a close-packed layer with a volume fraction, j, of HFBII of ∼0.8. This density is broadly similar to that reported at the airwater and the hydrophilic solidsolution interfaces (as discussed earlier). However, the layer thickness is systematically smaller than was observed at the airwater interface (31 Å) and at the hydrophilic solidsolution interface (42 Å). Compared to the overall dimensions of the globular protein (24 27 30 Å3) this would imply a different orientation at the surface compared to that at the airwater interface and possibly some overlap with the outer part of the underlying OTS layer. The data modeling is not consistent with this latter implication, but this may be associated with the insensitivity to such detailed variations. 10467
dx.doi.org/10.1021/la201711p |Langmuir 2011, 27, 10464–10474
Langmuir
ARTICLE
However, a significant difference in the adsorption behavior of HFBII onto the OTS hydrophobic surface compared to the hydrophilic surface is that rinsing in D2O changes the reflectivity pattern very slightly. This means that HFBII is not desorbed on rinsing in D2O and that its adsorption is irreversible to solvent exposure. The slight changes upon rinsing are consistent with a modest reorganization of the adsorbed layer, but overall the adsorption is now much more robust compared to that on the hydrophilic surface. The complete set of model parameters used to describe the HFBII adsorption onto the three different OTS surfaces that have been used in this study is summarized in Table 2 in the Supporting Information. Hydrophobin forms a dense layer at both the hydrophilic and hydrophobic solid surfaces, with a volume fraction of ∼0.8. This compares with the slightly lower volume fraction at the air water interface of ∼0.7.8 In this respect, hydrophobin is different from other proteins that partially denature on adsorption,25 whereas hydrophobin is more robust.2,8,12,13 However, relatively high volume fractions are observed in the adsorption of other compact surface-active molecules such as the lipoprotein surfactin, where adsorbed volume fractions of >0.9 are reported.26 HFBII/Surfactant Adsorption onto the Hydrophilic Silica Surface. The neutron reflectivity data in Figure 2 show the effect of the adsorption from a 0.1 mg/mL HFBII and 2.5 104 M h-SDS mixture in D2O. Compared to the reflectivity for the bare hydrophilic surface (also shown in the figure), the reflectivity shows a pronounced interference fringe consistent with strong
adsorption from the mixture at the interface. The reflectivity for the same HFBII/SDS mixture, but with d-SDS, is almost identical. In this case, d-SDS is effectively contrast matched to the D2O solvent, and hence this implies that the adsorbed layer is predominantly HFBII. The solid line in the figure shows the model calculation for a single layer with a thickness of 43 ( 2 Å and a scattering-length density of (3.5 ( 0.2) 106 Å2. The reflectivity and the associated model parameters are very close to those observed for the HFBII adsorption alone (Figure 1a). The key model parameters for the HFBII/h-SDS and HFBII/d-SDS adsorption are summarized in Table 1. This confirms the supposition that the adsorption is predominantly HFBII and that there is little or no SDS adsorption. This is not entirely unexpected because SDS would not normally have any affinity for the anionic surface of silicon. There are, however, circumstances where SDS will adsorb to such surfaces with a cosurfactant, such as was observed with a nonionic cosurfactant onto a hydrophilic silicon surface.27 As illustrated in Figure 2, rinsing in D2O after HFBII/h-SDS and HFBII/d-SDS exposure results in the removal of the adsorbed HFBII layer, consistent with the observations for HFBII alone. The data for the adsorption for a 0.1 mg/mL HFBII and 2.5 104 M SDS mixture (for a concentration that is less than the cmc) contrast markedly with the equivalent data for SDS concentrations greater than the cmc (2 102 M). The data are not shown in Figure 2 but are identical to the reflectivity for the bare hydrophilic surface. This indicates quite clearly that there is no longer any HFBII adsorption when the SDS
Figure 2. Neutron reflectivity data for HFBII/SDS adsorption onto a hydrophilic surface, (cyan) Si/D2O, preadsorption. (Black) 0.1 mg/mL HFBII: 2.5 104 M h-SDS. (Blue) 0.1 mg/mL HFBII: 2.5 104 M d-SDS. (Red) Si/D2O post HFBII/h-SDS adsorption and rinsing in D2O. (Green) Si/ D2O post HFBII/d-SDS adsorption and rinsing in D2O. The solid line is a calculated curve for an adsorbed HFBII layer (for the parameters in Table 1).
Table 1. Key Model Parameters for HFBII/SDS Adsorption onto a Hydrophilic Silica Surface sample
d1 ((1 Å)
F1 ((0.2 106 Å2)
d2 ((2 Å)
F2 ((0.2 106 Å2)
0.1 mg/mL HFBII/2.5 104 M h-SDS/D2O
10
2.7
43
3.5
rinsed in D2O 0.1 mg/mL HFBII/2.5 104 M d-SDS/D2O
10 10
2.7 2.7
43
3.3
rinsed in D2O
10
2.7 10468
dx.doi.org/10.1021/la201711p |Langmuir 2011, 27, 10464–10474
Langmuir
ARTICLE
concentration is greater than the cmc. This is consistent with the previous observations of mixed surfactant/HFBII adsorption at the airwater interface.8 It implies, as was inferred for the
Figure 3. Neutron reflectivity for HFBII/CTAB adsorption onto a hydrophilic surface: (a) for 5 105 M CTAB (c < cmc), (cyan) Si/ D2O, preadsorption, (black) 0.1 mg/mL HFBII/5 105 M h-CTAB/ D2O, (blue) 0.1 mg/mL HFBII/5 105 M d-CTAB/D2O, (red) Si/ D2O, post HFBII/h-CTAB adsorption and rinsing in D2O, and (green) Si/D2O, post HFBII/d-CTAB adsorption and rinsing in D2O and (b) for 103 M CTAB (c > cmc), (cyan) Si/D2O, preadsorption, (black) 0.1 mg/mL HFBII/103 M h-CTAB/D2O, (blue) 0.1 mg/mL HFBII/103 M d-CTAB/D2O, (red) Si/D2O, post HFBII/h-CTAB adsorption and rinsing in D2O, and (green) Si/D2O, post HFBII/d-CTAB adsorption and rinsing in D2O. The solid lines are model calculations for the parameters summarized in Table 2.
airwater interface studies, that in this case SDS/HFBII bulk complex formation is more energetically favorable than the surface adsorption, and this effect is further reinforced here by the negligible affinity of SDS for the anionic silicon surface. A similar sequence of measurements was made for the HFBII/ CTAB mixture, but a more complex pattern of adsorption behavior is observed. This is illustrated in Figure 3a for a CTAB concentration of 104 M (at a concentration that is less than the cmc) and in Figure 3b for a concentration of 2 103 M CTAB (at a concentration that is greater than the cmc). For a CTAB concentration of 104 M (at a concentration that is less than the cmc), the adsorption from a 0.1 mg/mL HFBII/ 5 105 M h-CTAB/D2O mixture (Figure 3a) shows a broad interference fringe that indicates some adsorption but is different from what is observed for HFBII/SDS mixtures. The data can no longer be described by a single layer at the interface. The simplest model used to fit the data is a two-layer model (for parameters listed in Table 3) with an inner layer of ∼30 Å and an outer layer of ∼10 Å. The scattering-length densities of the two layers are consistent with the coadsorption of HFBII and h-CTAB, with a greater amount of CTAB in the outer layer compared to that in the inner layer. On rinsing in D2O, the interference fringe becomes more pronounced and can now be described by a single uniform layer with a thickness of ∼40 Å. The thickness and scattering-length density are now similar to what is observed for the HFBII adsorption alone. Similar measurements for d-CTAB resulted in the formation of a more pronounced interference fringe (Figure 3a) and gave rise to a relatively small change on rinsing. The model required to reproduce that data is broadly similar to that for h-CTAB (Table 2), but the scattering-length density values are higher, reflecting the contribution from d-CTAB rather than h-CTAB. A notable feature, compared to HFBII/SDS adsorption, is that for the HFBII/CTAB mixture rinsing in D2O does not remove the adsorbed layer and indeed a substantial amount remains. It is not possible from this limited range of measurements to quantify the amount of each component (HFBII, CTAB, and D2O) at the interface adsorbed in each layer. However, the variation with rinsing and with CTAB contrast strongly suggests that a mixed HFBII/CTAB layer is formed. Furthermore, the model parameters associated with the adsorbed layer that remains after rinsing are not consistent with a layer of only CTAB adsorbed onto the hydrophilic silica surface,28 where a thinner layer, ∼ 30 Å, with a lower scattering-length density was reported. A similar sequence of measurements was made for HFBII/ CTAB mixtures, but at a CTAB concentration that was greater than the cmc (103 M). The resulting reflectivity profiles are broadly similar to those observed at the lower CTAB concentration and are shown in Figure 3b. However the general trends in
Table 2. Key Model Parameters for HFBII/CTAB Adsorption onto Hydrophilic Silica sample
d1 ((1 Å) F1 ((0.2 106 Å2) d2 ((2 Å) F2 ((0.2 106 Å2) d3 ((2 Å) F3 ((0.2 106 Å2)
0.1 mg/mL HFBII/5 105 M h-CTAB/D2O
10
2.7
28
2.9
rinsed in D2O
10
2.7
40
3.2
0.1 mg/mL HFBII/5 105 M d-CTAB/D2O
10
2.7
33
4.0
rinsed in D2O
10
2.7
46
3.8
0.1 mg/mL HFBII/103 M h-CTAB/D2O
10
2.7
22
rinsed in D2O 0.1 mg/mL HFBII/103 M d-CTAB/D2O
10 10
2.7 2.7
30 42
rinsed in D2O
10
2.7
34 10469
10
1.5
14
3.7
3.2
21
0.5
3.3 4.2
12
2.2
3.9
12
3.3
dx.doi.org/10.1021/la201711p |Langmuir 2011, 27, 10464–10474
Langmuir
ARTICLE
Table 3. Key Model Parameters for HFBII/SDS Adsorption onto an OTS-Coated Hydrophobic Surface d1 ((1 Å)
F1 ((0.2 106 Å2)
d2 ((2 Å)
F2 ((0.2 106 Å2)
d-OTS/D2O
16
6.7
13
5.0
0.2 mg/mL HFBII/D2O
16
6.7
13
rinsed in D2O
16
6.7
13
5 104 M h-SDS/D2O
16
6.7
2 102 M h-SDS/D2O
16
rinsed in D2O
16
2 102 M d-SDS/D2O rinsed in D2O
sample
d3 ((2 Å)
F3 ((0.2 106 Å2)
5.0
19
3.4
5.0
18
3.7
13
5.0
18
3.7
6.7
13
4.3
13
3.9
6.7
16
4.0
16
7.1
15
5.0
17
6.9
16
4.8
Figure 4. Neutron reflectivity for HFBII and SDS adsorption onto an OTS hydrophobic surface, (cyan) d-OTS/D2O, (black) 0.2 mg/mL HFBII/D2O, (red) 5 104 M h-SDS, (green) d-OTS/D2O post HFBII, SDS adsorption and rinsing in D2O, (blue) 2 102 M h-SDS/D2O. The solid lines are model calculations for the parameters summarized in Table 3.
Figure 3b and the parameters in Table 2 imply a greater CTAB adsorption and a greater difference on rinsing in D2O. In detail, the model parameters for the HFBII/h-CTAB adsorption are consistent with two-layer model, where the inner layer of mixed HFBII/CTAB is similar to that observed at the lower CTAB concentration and the outer layer is thicker and corresponds to a greater amount of CTAB adsorption. Furthermore, the impact of rinsing in D2O is that more HFBII and CTAB remain adsorbed at the surface. The measurements with d-CTAB reinforce the results with h-CTAB for CTAB concentrations above the cmc. For the HFBII/d-CTAB mixture, the data are described as a single layer. For the isotopic combination HFBII/h-CTAB, a two-layer structure was required because of the greater contrast provided by h-CTAB. At both CTAB concentrations (above and below the cmc), the specific affinity of CTAB for the hydrophilic silica surface results in a different pattern of adsorption with HFBII and in its response to rinsing in D2O compared to that of SDS. CTAB acts as a “glue” that makes the adsorption now partially irreversible, whereas with SDS it was entirely reversible. HFBII/Surfactant Adsorption onto the OTS Hydrophobic Surface. In Figure 4, the impact of SDS on a preadsorbed layer of HFBII on an OTS hydrophobic surface is illustrated.
The initial d-OTS surface (as described earlier in detail) gives rise to a pronounced interference fringe and is described by two layers at the interface and an inner dense and an outer less dense layer. HFBII (at a concentration of 0.2 mg/mL) adsorbs strongly (as described earlier), resulting in a more pronounced interference fringe that is shifted to lower Q values, consistent with a dense layer of HFBII, ∼20 Å, adsorbed onto the OTS surface. The addition of 5 104 M h-SDS (for concentrations that are less than the cmc) results in no measurable change in the reflectivity. This indicates that SDS has no impact on OTS or on the HFBII adsorbed layer and at this concentration it does not adsorb onto HFBII or OTS. The measurement at a higher SDS concentration of 2 102 M (for concentrations greater than the cmc) results in a different reflectivity profile, where the interference fringe is shifted to higher Q values. This corresponds to a thinner layer on top of the OTS, as illustrated in Table 3. The outermost layer is now only 13 Å and is too thin to correspond to an HFBII layer. Hence at this higher SDS concentration, SDS has displaced HFBII at the OTS surface. This is directly analogous to what is observed at the airwater interface, as recently reported by Zhang et al.8 Rinsing in D2O after HFBII and SDS (for SDS concentrations above and below the cmc) ultimately results in 10470
dx.doi.org/10.1021/la201711p |Langmuir 2011, 27, 10464–10474
Langmuir
ARTICLE
Figure 5. Neutron reflectivity for HFBII and CTAB on an OTS hydrophobic surface, (cyan) d-OTS/D2O, (yellow) 0.2 mg/mL HFBII/D2O, (red) 104 M h-CTAB/D2O, (green) d-OTS/D2O, post CTAB and HFBII adsorption and rinsing in D2O, (blue) 2 103 M h-CTAB/D2O. The solid lines are model calculations for the parameters summarized in Table 4.
Table 4. Key Model Parameters for HFBII/CTAB Adsorption onto an OTS-Coated Hydrophobic Surface sample
d1 ((1 Å)
F1 ((0.2 106 Å2)
d2 ((2 Å)
F2 ((0.2 106 Å2)
d-OTS/D2O
18
6.3
16
4.4
0.2 mg/mL HFBII/D2O rinsed in D2O
18 18
6.3 6.3
16 16
4.4 4.4
d3 ((2 Å)
F3 ((0.2 106 Å2)
19 20
3.7 4.3
104 M h-CTAB/D2O
18
6.3
15
3.6
19
4.5
2 103 M h-CTAB/D2O
18
6.3
16
4.4
14
2.7
rinsed in D2O
18
5.6
22
3.8
2 103 M d-CTAB/D2O
18
6.3
15
5.6
18
5.1
rinsed in D2O
18
6.3
16
5.3
17
5.1
the desorption of both HFBII and SDS. However, as illustrated in Figure 4 and in Table 3, the final OTS layer is slightly different from its initially deposited state. The measurements in h-SDS, d-SDS (data not shown in Figure 4), and after rinsing (parameters in Table 3) indicate that some surfactant is incorporated into the outer region of the OTS layer but that there is no corresponding interpenetration for HFBII alone. A broadly similar pattern of behavior for HFBII/CTAB mixed adsorption onto an OTS-coated hydrophobic surface was observed, and this is summarized in Figure 5 and Table 4. As shown in Figure 5 and Table 4, the initial OTS layer results in a well-defined interference fringe, which is further enhanced and shifted to lower Q values by the adsorption of HFBII. The thickness and density of the OTS and HFBII layers are similar to those discussed earlier in the context of HFBII/SDS adsorption onto an OTS surface. Rinsing in D2O (not shown in the figure) makes an imperceptible difference in the reflectivity and hence the adsorption (which is also reflected in the model parameters in Table 4). Exposure to 104 M h-CTAB/D2O (for concentrations that are less than the cmc) makes little or no difference in the reflectivity (Figure 5), and hence there is no CTAB adsorption or disruption of the HFBII layer. At the higher surfactant concentration, the reflectivity is different and the layer adsorbed onto the OTS surface is now substantially thinner. This is broadly
similar to what was observed for SDS at a concentration greater than the cmc. Hence, HFBII is now replaced by CTAB at the interface. This is as a result of the HFBII/CTAB bulk complex formation being more energetically favorable than adsorption, as discussed earlier in the context of SDS/HFBII adsorption at the hydrophobic surface. Rinsing the surface with D2O leads to the formation of a slightly different OTS surface. Subsequent measurements for d-CTAB (for concentrations greater than the cmc, Table 4) confirm that some CTAB is incorporated into the outer region of the OTS layer, as was observed for SDS. A similar sequence of measurements was made for HFBII and C12E6 adsorption onto the OTS hydrophobic surface, as illustrated in Figure 6. The initial OTS layer is similar to the previous ones described earlier (Table 1 in the Supporting Information), and the HFBII adsorption is very similar. What is noticeable, however, is that the addition of C12E6 (at concentrations above and below the cmc) makes very little difference to the measured reflectivity. Furthermore, rinsing in D2O after exposure to C12E6 restores the reflectivity to that obtained for the initial HFBII adsorption onto the OTS surface. Hence, it is evident that C12E6, at concentrations above and below the cmc, does not remove HFBII from the surface. The key model parameters are summarized in Table 5. 10471
dx.doi.org/10.1021/la201711p |Langmuir 2011, 27, 10464–10474
Langmuir
ARTICLE
Figure 6. Neutron reflectivity for HFBII and C12E6 adsorption onto an OTS hydrophobic surface, (cyan) d-OTS/D2O, (yellow) 0.2 mg/mL HFBII, (blue) 2 105 M h-C12E6/D2O, (red) 2 104 M h-C12E6/D2O, and (green) rinsing in D2O after adsorption. The solid lines are model fits for the parameters summarized in Table 5.
Table 5. Key Model Parameters for HFBII/C12E6 Adsorption onto an OTS-Coated Hydrophobic Surface sample
d1 ((1 Å)
F1 ((0.2 106 Å2)
d2 ((2 Å)
F2 ((0.2 106 Å2)
d-OTS/D2O 0.2 mg/mL HFBII/D2O
17 16
6.9 6.7
13 13
5.0 5.0
21
3.4
2 105 M h-C12E6
16
6.7
10
6.4
19
2.8
2 104 M h-C12E6
16
6.7
10
6.4
19
2.4
The minimal changes in the reflectivity profiles and the associated model parameters also imply that there is very little C12E6 adsorption at concentrations both below and above the cmc. The parameters do indicate a slight compression of the outer part of the OTS layer and some minor incorporation of C12E6 into the HFBII layer. This is in contrast to what was observed at the other model hydrophobic surface studied, the airwater interface.8 In that case, the HFBII/SDS, HFBII/ CTAB, and HFBII/C12E6 mixtures all behaved in a similar way. That is, below the surfactant cmc the surface is dominated by HFBII, and above the surfactant cmc HFBII is displaced from the surface and replaced by surfactant. However, at the airwater interface some HFBII remained at the interface for C12E6 concentrations above the cmc. It should be noted that the experimental details for the measurements at the hydrophobic solidsolution interface and for the airwater interface are slightly different. At the hydrophobic solid surface, a preadsorbed layer of HFBII is exposed to surfactant solutions at concentrations below and above the cmc whereas the equivalent measurements at the airwater interface were made for mixed HFBII/surfactant solutions, with surfactant concentrations greater than or less than the cmc. General Discussion. The dimensions of the adsorbed HFBII layer at the different interfaces show some interesting differences compared to the overall dimensions, 24 27 30 Å3, of the protein.6 At the airwater interface, the mean thickness is ∼31 Å,8 and this is consistent with other reported measurements.7 It was inferred that this corresponds to an orientation at
d3 ((2 Å)
F3 ((0.2 106 Å2)
the interface with the hydrophobic patch on the surface of the protein being adjacent to the air phase. At the hydrophobic OTS surface, the results presented here are consistent with a mean layer thickness that is significantly smaller, ∼20 Å. Hence, it is assumed that this corresponds to a different orientation of the OTS surface. This would imply an orientation where the hydrophobic patches are orthogonal to the surface and presumably adjacent to each other within the layer. This has some similarities with the hydrophilic dimers that are postulated to form as a precursor to tetramer formation in solution.6 The slight differences between the adsorbed layer thickness and the protein dimensions would imply a slight flattening or distortion of the protein at the interface or a small overlap/penetration of the protein into the outer layer of the OTS surface. From the data presented, it would be difficult to be too confident about a distinction between either mechanism. At the hydrophilic interface, the adsorbed layer thickness is ∼43 Å, which is substantially larger than that formed at the airwater or hydrophobic liquidsolid interface and significantly larger than any of the protein dimensions. This is consistent with a bilayer formed at the interface but from the dimensions indicates a bilayer where the hydrophobic patch is orthogonal to the interface. It is then likely that the surface structure originates from the dimer and tetramer structures that have been demonstrated to form in solution.6,7 At the hydrophilic surface, HFBII adsorbs to form a dense layer at the interface, but its adhesion is relatively weak. That is, it is completely removed by rinsing in aqueous solvent whereas at 10472
dx.doi.org/10.1021/la201711p |Langmuir 2011, 27, 10464–10474
Langmuir the hydrophobic surface the adsorbed layer that forms is more robust with respect to rinsing in aqueous solvent. The weak electrostatic interaction between HFBII and the hydrophilic silica surface is not surprising because HFBII is only weakly anionic, with a surface potential of only a few millivolts. However, given the structural considerations discussed above, it is not entirely clear why the adsorption should be more resilient to rinsing in aqueous solvent at the hydrophobic surface. In the coadsorption of HFBII with the conventional cosurfactants, the general pattern that emerges is that at concentrations below the surfactant cmc the HFBII adsorption is unaffected by the cosurfactant. Whether there is any surfactant coadsorption depends entirely upon the nature of the underlying surface and the cosurfactant. Above the surfactant cmc, the common pattern of adsorption behavior is that the surfactant removes HFBII from the surface in favor of HFBII/surfactant solution complex formation, as was observed at the airwater interface,8 under slightly different circumstances, as discussed earlier. Whether the HFBII is replaced by surfactant at the interface again depends upon the nature of the underlying surface and the cosurfactant. An obvious manifestation of this selectivity occurs for HFBII/ SDS adsorption at the hydrophilic surface, where the lack of affinity of SDS for the anionic silica surface results in no SDS adsorption above the cmc. An exception to the general trend above the cmc exists for the HFBII/CTAB mixture where some HFBII and CTAB remain at the surface after rinsing in aqueous solvent. Here, the interaction between HFBII and CTAB profoundly changes the nature of the surface interaction. At the hydrophobic surface, for surfactant concentrations greater than the cmc, surfactant adsorption occurs for both SDS and CTAB, where the surfactant adsorption is now entirely driven by the hydrophobic interaction.
’ SUMMARY The adsorption of HFBII and its coadsorption with conventional surfactants SDS, CTAB, and C12E6 at different solid solution interfaces have been studied by NR. At the hydrophilic solid surface, HFBII adsorbs to form a layer that is ∼43 Å thick, which, given the approximate dimensions of the protein, is thicker than that expected for a monolayer, and is assumed to form a bilayer at the interface. Surprisingly, the adsorption is reversible, and rinsing in solvent removes the adsorbed layer. When HFBII is coasdsorbed at the hydrophilic surface with SDS at a concentration below the SDS cmc, the SDS has no impact on HFBII adsorption. For SDS concentrations greater than the cmc, HFBII is removed from the surface in favor of HFBII/SDS solution complex formation, and the resulting surface has neither adsorbed SDS nor adsorbed HFBII. In the coadsorption of HFBII and CTAB at the hydrophilic surface, a slightly different pattern of adsorption is found. For CTAB concentrations that are less than the cmc, a mixed HFBII/CTAB layer is formed in which the distribution of CTAB is not uniform. A similar pattern of adsorption occurs for CTAB concentrations above the cmc but with greater CTAB adsorption. Hence for CTAB concentrations above and below the cmc, a mixed HFBII/ CTAB layer is formed. In contrast to the observations for SDS/ HFBII, on rinsing in solvent and for surfactant concentrations greater than the cmc only partial desorption occurs, and substantial amounts of HFBII and CTAB remain at the surface. This implies that CTAB induces an increased irreversibility of the adsorption of HFBII at the interface.
ARTICLE
HFBII adsorbs irreversibly onto a hydrophobic surface, created by an OTS layer. A thinner adsorbed layer, ∼20 Å, is observed and is hence consistent with a different molecular orientation at this hydrophobic surface compared to that at the airwater interface where a layer of ∼31 Å was observed.8 Notably, the adsorption is now irreversible to rinsing in solvent. Exposure to an SDS solution has no impact on the HFBII adsorbed layer for SDS concentrations below the cmc. For concentrations greater than the cmc, HFBII is entirely removed from the surface. This is in part analogous to what is observed at the airwater interface8 for the coadsorption of HFBII/ surfactant mixtures. A similar pattern of behavior is observed for CTAB, except that on rinsing in solvent some CTAB remains at the surface. In contrast to the observations for both SDS and CTAB, nonionic surfactant C12E6 has little impact on HFBII adsorption at the hydrophobic surface. These results illustrate that the nature of HFBII adsorption is different at the different solid surfaces and different from the adsorption at the airwater interface. Apart from adsorbing in different orientations and structures, it can also exhibit extremes of reversible and irreversible adsorption. Furthermore, the nature of the conventional cosurfactant can strongly influence the detailed nature of competitive adsorption at different interfaces. Finally, these results provide an important insight into how HFBII adsorption can be manipulated, and this will have important implications for foam and emulsion stability.
’ ASSOCIATED CONTENT
bS
Supporting Information. Model parameters associated with the analysis of the NR data from the OTS surfaces, NR data for the initial characterization of the hydrophilic and hydrophobic surfaces, and neutron scattering lengths and molecular volumes for the different materials. This material is available free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION Corresponding Author
*E-mail: jeff
[email protected].
’ ACKNOWLEDGMENT The contribution from John Chapman, Chris Sidebottom, and Tim Litchfield regarding the production and purification of HFBII from modified yeast is gratefully acknowledged. The provision of beam time on the D17 reflectometer at the ILL, Grenoble, France, and the invaluable assistance of Bob Cubitt, Andrew Wildes, and Michel Bonnaud are acknowledged. ’ REFERENCES (1) Kitamoto, D.; Morita, T.; Fukuoka, T.; Konishi, M.; Inura, J. Curr. Opin. Colloid Interface Sci. 2009, 14, 315. (2) Linder, M. B. Curr. Opin. Colloid Interface Sci. 2009, 14, 356. (3) Milligan, C. N. Environ. Pollut. 2005, 133, 183. (4) Muthusamy, K.; Gopalakrishnan, S.; Kochupappy, T.; Sivachidambaram, R.; Sivachidambaram, P. S. Curr. Sci. 2008, 94, 736. (5) Hahanpaa, J.; Szilvray, G. R.; Kaljunen, H.; Maksimainen, M.; Linder, M. B.; Rouvinen, J. J. Prot. Sci. 2006, 15, 2120. (6) Kallio, J. M.; Linder, M. B.; Rouvinen, J. J. Biol. Chem. 2007, 282, 28733. (7) Kisko, K.; Szilvray, G. R.; Vuorimaa, E.; Lemmetyinen, H.; Linder, M. B.; Torkkeli, M.; Serimaa, R. Langmuir 2009, 25, 1612. 10473
dx.doi.org/10.1021/la201711p |Langmuir 2011, 27, 10464–10474
Langmuir
ARTICLE
(8) Zhang, X. L.; Penfold, J.; Thomas, R. K.; Tucker, I. M.; Petkov, J. T.; Bent, J.; Cox, A.; Campbell, R. A. Langmuir 2011accepted for publication. (9) Paanenen, A.; Vuorimaa, E.; Torkkeli, M.; Penttila, M.; Kauranen, M.; Ikkala, O.; Lemmetyinen, H.; Serimaa, R.; Linder, M. B. Biochemistry 2003, 42, 5253. (10) Wosten, H. A. B.; Schuren, F. H. J.; Wessels, J. G. L. EMBL J. 1994, 13, 5845. (11) Wosten, H. A. B.; deVries, O. M. H.; Wessels, J. G. H. Plant Cell 1993, 5, 1567. (12) Kisko, K.; Szilvay, G. R.; Vainio, U.; Linder, M. B.; Serimaa, R. Biophys. J. 2008, 94, 198. (13) Zhang, X. L.; Penfold, J.; Thomas, R. K.; Tucker, I. M.; Petkov, J. T.; Bent, J.; Cox, A.; Grillo, I. Langmuir 2011; DOI: 10.1021/La 2020226. (14) http://www.ill.eu/D17/ (15) Penfold, J.; Thomas, R. K. J. Phys.: Condens. Mater. 1990, 2, 1369. (16) Lu, J. R.; Thomas, R. K.; Penfold, J. Adv. Colloid Interface Sci. 2000, 84, 143. (17) Heavens, O. S. Optical Properties of Thin Solid Films; Dover Publications: New York, 1991. (18) Lu, J. R.; Morrocco, A.; Su, T. J.; Thomas, R. K.; Penfold, J. J. Colloid Interface Sci. 1993, 158, 303. (19) Lu, J. R.; Hromadova, M.; Simister, E. A.; Thomas, R. K.; Penfold, J. J. Phys. Chem. 1994, 98, 1159. (20) Van Os, N. M.; Haak, J. R.; Rupert, L. A. M. Physico-Chemical Properties of Selected Anionic, Cationic, and Nonionic Surfactants; Elsevier: Amsterdam, 1993. (21) Basheva, E. S.; Kralchevsky, P. A.; Danov, K. D.; Stoyanov, S. D.; Blijdenstein, T. B. J.; Pelan, E. G.; Lips, A. Langmuir 2011, 27, 4481. (22) Bailey, M.; Askolin, S.; Horhammer, N.; Tenkanen, M.; Linder, M.; Penttila, M.; Nakari-Setala, T. Appl. Microbiol. Biotechnol. 2002, 58, 721. (23) Penfold, J.; Staples, E.; Tucker, I. Langmuir 2002, 18, 2967. (24) Fragneto, G.; Thomas, R. K.; Rennie, A. R.; Penfold, J. Science 1995, 267, 657. (25) Clarkson, J. R.; Cui, Z. F.; Darton, R. C. J. Colloid Interface Sci. 1999, 215, 323. (26) Shen, H. H.; Thomas, R. K.; Chen, C. Y.; Darton, R. C.; Baker, S. C.; Penfold, J. Langmuir 2009, 25, 4211. (27) Penfold, J.; Staples, E.; Tucker, I.; Thomas, R. K. Langmuir 2000, 16, 8879. (28) Penfold, J.; Tucker, I.; Petkov, J. T.; Thomas, R. K. Langmuir 2007, 23, 8357.
10474
dx.doi.org/10.1021/la201711p |Langmuir 2011, 27, 10464–10474
