2008
J. Phys. Chem. B 2007, 111, 2008-2014
NMR Studies of Aggregation and Hydration of Surfactants Containing Amide Bonds Maria Stjerndahl,†,§ Dan Lundberg,*,†,‡ Hailing Zhang,‡ and Fredric M. Menger‡ Applied Surface Chemistry, Department of Chemical and Biological Engineering, Chalmers UniVersity of Technology, 412 96 Go¨teborg, Sweden, and Department of Chemistry, Emory UniVersity, Atlanta, Georgia 30322 ReceiVed: September 7, 2006; In Final Form: NoVember 29, 2006
The consequences of including amide bonds into the structure of short-chain nonionic surfactants have been studied. Of particular interest were the possible effects of the hydrogen bonding ability of the amide group on the micellar shape. The aggregate structure and hydration of two different amide-containing surfactants, C7H15CO-NH-(CH2CH2O)4H and C7H15CO-(NH-C3H6-CO)2N(CH3)2, were investigated using NMR diffusometry (pulsed gradient spin echo NMR) as the main technique. Data from experiments on the surfactants, the hydrophobic probe molecule hexamethyldisilane (HMDS), and water were interpreted to gain information about the solution structures, and the results were compared to those on a previously studied alcohol ethoxylate surfactant of similar size, C8E4. Both of the amide-containing surfactants form small micelles within the whole investigated concentration range. At the critical micelle concentration, the aggregates are most probably spherical, and with increasing surfactant concentration there are indications of either a minor aggregate growth or agglomeration of the micelles. In addition, it was found that the presence of amide groups in the surfactant inhibits the intermicellar transport of HMDS, which occurs in the C8E4 system. From measurements on water diffusion in the three surfactant systems, it could be concluded that the surfactant hydration is higher when amide bonds are present.
Introduction The amide group, plentiful in nature and an important actor in numerous processes of life, has over the past decades found its way into the molecular structures of various synthetic surfactants. The incentives for including an amide bond in a surfactant molecule can vary. Today, an important general driving force for developing new surfactants is a pursuit of substances that are more benign to the environment than traditional ones.1,2 This is often accomplished by inserting one or several labile bonds within the molecular structure (most commonly introduced between the hydrophilic and the hydrophobic moieties of the surfactant),3 a modification that obviously also may have the negative side effect of impairing the applicability of the surfactant. It has been shown, however, that certain surfactants having an amide group as the cleavable bond exhibit very good biodegradation profiles in combination with an excellent chemical stability,4 a combination of properties that makes these surfactants attractive. The inclusion of an amide bond into a surfactant also renders the compound altered physicochemical properties, which sometimes can give the substance an improved performance.5 These differences can often be traced back to the ability of the amide bond to form intermolecular hydrogen bonds. The two major classes of amide-containing surfactants are poly(ethylene glycol) alkyl amides (exemplified by tetra(ethylene glycol) mono-n-octyl amide in Figure 1a) and carbohydrate amides of fatty acids.5 An amide bond is more hydrophilic than an ethylene oxide group, and exchange of the * Corresponding author. E-mail:
[email protected]. † Chalmers University of Technology. § Present address: Department of Chemistry, University of Florida, Gainesville, Florida 32611. ‡ Emory University.
Figure 1. Molecular structures of (a) tetra(ethylene glycol) mono-noctyl amide, that is, the Amide; (b) the Peptoad; and (c) tetra(ethylene glycol) mono-n-octyl ether, that is, the Ether.
latter for the former in a poly(ethylene glycol) ether surfactant makes the molecule more hydrophilic, which, in turn, is manifested in increased critical micelle concentrations and cloud points.6,7 An investigation of the effect of the same modification on the surface pressure vs area per molecule isotherms revealed attractive forces between the amide groups in an adsorbed monolayer.7 It turns out that addition of an amide bond gives a larger increase in the intermolecular attraction than what is caused by an additional methylene group. The presence of attractive intralayer forces between the amide groups was further supported by mean-field lattice model calculations, which also indicated that the boundary between the hydrophobic and hydrophilic regions of an adsorbed surfactant layer is more welldefined on inclusion of this bond.7 In the case of carbohydrate amides of fatty acids, there is commonly a very strong headgroup-headgroup attraction due to the collective interactions exerted by the amide and hydroxyl groups. Long-tail (C12-C20) compounds of this class form very stable, solid-like monolayers, whereas substances carrying
10.1021/jp0658492 CCC: $37.00 © 2007 American Chemical Society Published on Web 02/03/2007
Aggregation and Hydration of Amide Surfactants shorter tails (C6-C10) can give gels consisting of loose networks of precipitated, often regularly helical, surfactant fibers.5 In a recent publication, Menger et al. present a study of a new class of surfactants, with the headgroups made up of multiple amide groups separated by short hydrocarbon spacers (an example is shown in Figure 1b).8 A series of these compounds, dubbed peptoads, with variation in the length of the spacers as well as the hydrophobic tails, were synthesized and characterized with respect to their basic physicochemical properties. The solubility of the peptoads is strongly dependent on the length of the hydrophobic tail and the nature of the headgroup terminus. Peptoads having hydrophobic chains exceeding seven carbons in length, as well as those carrying terminal NH groups, exhibit very low water solubility. The low solubility is explained to be due to the combination of hydrophobic interactions between the hydrocarbon chains and the hydrogen bonds between the amide groups. Furthermore, it was found that some of the short-chain peptoads show concentration-dependent clouding and miscibility gaps. The differences in molecular structure between clouding and nonclouding surfactants are small, and it appears to be a delicate balance between surfactant-surfactant and surfactant-water interactions that strongly influence the behavior of the peptoads in aqueous solutions. The spontaneous curvature of a surfactant aggregate is closely related to the relative sizes of the hydrophilic and hydrophobic parts of the surfactant. An increase in the attractive forces between the headgroups can cause a decrease in the effective headgroup area and, hence, a decrease in the curvature of the surfactant film. In turn, this is a driving force for aggregate growth. Is it possible that the average attraction between amide groups of adjoining amide-containing surfactants is strong enough to affect the shape of the micelles? If so, can these interactions give rise to a significant micellar growth at low concentrations, even for a surfactant with a short hydrophobic part? Or, on the other hand, will the hydration of the amide group play the leading role, with the expected outcome of an increased effective headgroup area? To our knowledge, these questions have not been addressed in any previous study. The aggregation and hydration of two amide-containing nonionic surfactants with hydrophobic parts of similar size were investigated in this work; namely, the poly(ethylene glycol) alkyl amide and the peptoad presented in Figure 1a and b. The peptoad shown in Figure 1b gives isotropic solutions within the whole studied concentration range. Reference experiments were performed with the previously studied alcohol ethoxylate C8E4,9-11 which was chosen because of its structural similarity to the studied oxyethylene amide (see Figure 1a and c). An additional benefit of using this compound is that it is “borderline” with regard to aggregate size, in that it shows a tendency for micellar growth at slightly elevated temperatures.10 Hence, one can expect that quite subtle changes in headgroup-headgroup interactions caused by the presence of amide groups can have significant effect on the aggregate size. For the sake of simplicity, the three studied surfactants will be referred to as just the Amide, the Peptoad and the Ether in the following text. The investigations were performed using pulsed gradient spin echo (PGSE) NMR as the main technique. Diffusion data from experiments on the surfactants, the hydrophobic probe molecule hexamethyldisilane, and water were interpreted to gain information about the solution structures. Materials and Methods Materials. The tetra(ethylene glycol) mono-n-octyl amide and the Peptoad were synthesized using the procedures presented
J. Phys. Chem. B, Vol. 111, No. 8, 2007 2009 in previous publications.4,8 The purities of the surfactants are estimated to >99% and >97%, respectively. Tetra(ethylene glycol) mono-n-octyl ether (>98%) was obtained from Fluka; hexamethyldisilane (HMDS) (98%), from Aldrich; and deuterium oxide (99.8 atom % D), from Dr Glaser AG. The critical micelle concentrations (CMC) of the Ether, Peptoad, and Amide surfactants are 10, 40, and 60 mM, respectively.4,8,12,13 The differences in the CMC’s in D2O and H2O were assumed to be negligible.14,15 Sample Preparation. Samples were prepared by diluting 800 mM stock solutions of of the Amide, the Peptoad, or the Ether and ∼1 mM HMDS in D2O to the appropriate surfactant concentrations. The self-diffusion coefficients of the monomeric surfactants were determined in samples containing 2 mM of the respective compounds. These dilute samples did not contain HMDS. NMR Diffusion Experiments. All NMR diffusion experiments were performed at 20 °C on a Varian Inova 500 MHz spectrometer equipped with a Doty Scientific 16-38 Diffusion Probe using a Hahn-echo sequence16,17 and sine-bell-shaped pulsed field gradients. The delay between the gradient pulses, ∆, and the width of the gradient pulses, δ, were kept constant at 70 and 4 ms, respectively, while the strength of the pulsed gradient, G, was linearly incremented from 0.01 up to 2.5 T/m (maximum varied among experiments and samples) in 16 or 32 steps. The self-diffusion coefficients, D, of the different components were achieved from the attenuation of the relevant echo peaks by linear least-square fits to a modified version of the StejskalTanner equation,16,18
ln(I/I0) ) -(γGδ)2D(4∆ - δ)/π2
(1)
where I is the measured signal intensity; I0, the signal intensity in the absence of gradient pulses; and γ, the magnetogyric ratio of protons. For all studied surfactants and concentrations, the observed echo decays were single-exponential and gave very good fits to eq 1. Samples were inserted into the probe at least 30 min prior to the experiments to allow for thermal equilibrium to be attained. 1H NMR for Investigation of Peak Widths and Shapes. Spectra for investigation of the concentration dependence of the width and shape of the NMR peaks were run on a Varian INOVA 600 MHz spectrometer. Results and Discussion A very simple way to find an indication of significant micellar growth in an aqueous surfactant solution is to examine how the viscosity of the samples depends on the surfactant concentration. If elongated aggregates with a large axial ratio are formed in a solution, this is usually manifested by a significant increase in viscosity. We did not perform any quantitative rheology experiments on our samples, but from visual inspection of the solutions, it could be concluded that no significant change in viscosity occurs with an increased concentration for any of the studied compounds. In the investigated composition ranges, all samples were low-viscosity solutions. Another approach in the pursuit of qualitative evidence for large aggregates in a surfactant solution is to study the concentration dependence of width and band shape for the peaks in proton NMR spectra. If large aggregates are formed, the resonances can be expected to broaden and show a characteristic shape with a wide base. Such peaks can be described as superpositions of Lorentzian lines of different widths19,20 and
2010 J. Phys. Chem. B, Vol. 111, No. 8, 2007
Stjerndahl et al.
Figure 2. Partial 1H NMR spectra of the Peptoad at three different concentrations. The peaks represent the methylene and methyl groups of the hydrophobic chain. 0.1 ppm ) 60 Hz.
can have half-height bandwidths of hundreds of hertz. For all surfactants studied in this work, there were very small changes in both half-height width and band shape of the NMR signals with increasing concentration. Figure 2 shows the 1H NMR spectra of three different Peptoad solutions, in which the concentration is ranging from below the CMC up to the maximum studied concentration. The other two surfactants gave similar results. The fact that no significant change of the NMR peaks is observed with increasing concentration, together with the lack of an appreciable increase in viscosity, showed at an early stage of this study that there is not much reason to believe that there is a dramatic aggregate growth with increasing concentration in any of the investigated systems. The translational mobilities of the different species present in a solution are dependent on the formation of self-assemblies and other interactions between the components. Hence, by investigating the concentration dependence of the self-diffusion coefficients of the solutes and the solvent in a surfactant solution, one can get information on the size and shape of possible micelles, aggregate-aggregate interactions, and hydration. This is also a well-established approach for such purposes, which have given invaluable contributions to the understanding of surfactant systems.21,22 Thus, to achieve some detailed information on the aggregation and hydration in the presently investigated systems, the PGSE-NMR technique was used to determine the self-diffusion coefficients of the surfactants, the water, and the highly hydrophobic probe molecule HMDS in samples with different surfactant concentrations. In Figure 3, the results for the different surfactants and the HMDS are shown. The diffusion coefficients of the monomers were 3.7 × 10-10, 3.5 × 10-10, and 3.8 × 10-10 m2 s-1 for the Amide, the Peptoad, and the Ether, respectively. In a micellar solution for which only one self-diffusion coefficient is obtained for the surfactant, as is the case in the systems of interest here, the exchange between surfactant monomers in the bulk and surfactant molecules residing in micelles occurs at a rate faster than the time-scale of the NMR diffusion experiment. Under the assumption that the micelles can be regarded as discrete entities, the observed self-diffusion coefficient, Dsobs, is then a weighted average of the diffusion coefficients at the two sites according to
Dsobs ) psmonoDsmono + psmicDsmic ) (1 - psmic)Dsmono + psmicDsmic (2) where psmono and psmic are the fractions of free and aggregated surfactant, respectively, and Dsmono and Dsmic are the diffusion coefficients for the surfactant monomers and the micelles.9,23
Figure 3. The observed diffusion coefficients, Dsobs (filled symbols) of the Amide, the Peptoad, and the Ether surfactants, together with the diffusion coefficient of the hydrophobic probe molecule HMDS, Dhmds, (open symbols) in the respective systems, as a function of surfactant concentration.
HMDS is a very hydrophobic substance that when added to a micellar solution is heavily partitioned to the hydrophobic regions of the aggregates. Consequently, its self-diffusion coefficient can often be taken to represent the diffusion coefficient of the aggregates. In Figure 3, one can see that for the Amide and the Peptoad, Dsobs is higher than the observed diffusion coefficients of the HMDS, Dhmds, for all studied concentrations and that the former is approaching the latter with increasing surfactant concentration. Evaluation of Dsobs using eq 2, Dsmono, and the relevant CMC’s (CMC ) psmic × cstotal) shows that Dhmds to a good approximation represent Dsmic for
Aggregation and Hydration of Amide Surfactants
J. Phys. Chem. B, Vol. 111, No. 8, 2007 2011
these two surfactants. In turn, this supports that both the Amide and the Peptoad form micelles with discrete hydrophobic cores and that interaggregate transport of HMDS is slow. The Ether system, on the other hand, shows a different behavior. Rather than decreasing monotonically as in the other systems, Dhmds levels out with increasing surfactant concentration, and for the highest concentration, it is even higher than Dsobs (Figure 3c). At least for the two highest concentrations, Dhmds is higher than the values of Dsmic predicted using eq 2. This phenomenon is previously observed for polyoxyethylene surfactants and has been explained to be due to the presence of significant intermicellar interactions, which in turn increases the probability for a hydrophobic probe molecule to be transferred between the aggregates.9,23 Hence, Dhmds does not reflect the aggregate diffusion in this system. The interactions between polyoxyethylene surfactant micelles are shown to be due to attractive forces between the headgroups.10,24,25 Thus, considering the fact that the Amide carries an oxyethylene chain of the same size as the Ether, it is interesting to find that there is no considerable contribution from interaggregate transport to Dhmds in the Amide system. Apparently, the presence of the amide bond prevents intermicellar exchange of HMDS. One possible explanation to this observation is that the amide bond impairs the attractive forces between the aggregates, an idea that is also supported by the fact that the cloud points for amide surfactants of this type are significantly higher than those for related ether surfactants with hydrophobic parts of the same size.6,25-27 It is well-known that the degree of attractive intermicellar interactions increases as the cloud point for an oxyethylene surfactant is approached.10,24 An alternative way to account for the difference between the amide and ether systems with regard to the transport mechanism for HMDS is that the layer of amide bonds forms a “mantle” that is too polar to allow for transport of the solubilized hydrophobic species. This latter idea can be related to the former findings that indicate a sharper, more well-defined boundary between the hydrocarbon and headgroup regions in aggregates of poly(ethylene glycol) alkyl amide surfactants, as compared to that in aggregates formed by polyether surfactants lacking the amide functionality.7 As the volume fraction of micelles in a solution increases with surfactant concentration, the micelle diffusion will be increasingly obstructed. In the case of spherical aggregates at low concentrations, the decrease in the diffusion coefficient, Dsmic, as a function of the volume fraction of micelles, Φ, can often be described by a first-order expression according to
Dsmic ) Dmic 0 (1 - kΦ)
(3)
where Dmic 0 is the diffusion coefficient of the micelles extrapolated to infinite dilution and k is a constant that usually varies between 1.7 and 2.5, depending on the surfactant type and the degree of interactions between the aggregates.28-30 For the systems where Dsmic are known (as Dhmds), that is, those containing the Amide or the Peptoad, eq 3 can be applied to evaluate if these surfactants form spherical aggregates. In Figure 3, one can see that for the Amide, the decrease in Dsmic is roughly linear in the whole investigated composition range, whereas the slope decreases at higher concentrations for the Peptoad (c ) 400 mM S Φ ≈ 0.12). If one subtracts the fraction of monomeric surfactant and uses a density of unity for the surfactants when estimating the volume fraction of micelles, a fit of eq 3 to the initial linear part of the plots in Figure 3 gives Dmic values of 7.8 × 10-11 and 6.6 × 10-11 m2 s-1 for the 0
Amide and the Peptoad, respectively. The corresponding values of k are 2.9 and 3.9, that is, significantly higher than expected for aggregates of spherical geometry. A deviation in k of this order can be explained by either a slight micellar growth or by interactions between the aggregates. In the following, these possibilities (with an emphasis on the former) will be discussed in some more detail. can be related to the hydroFor spherical aggregates, Dmic 0 dynamic radius of the aggregates, RH, via the Stokes-Einstein equation,
Dmic 0 )
kBT 6πηRH
(4)
where kB is the Boltzmann constant and η is the viscosity of the solvent at the experimental temperature, T. Equation 4, together with the Dmic 0 values presented above and a viscosity of D2O of 1.251 mPas, gives hydrodynamic radii of 2.3 and 2.6 nm for the Amide and Peptoad, respectively. These values should be compared to the extended lengths of the Amide and the Peptoad surfactants, which can be estimated to 2.9 and 2.8 nm, respectively, using known bond lengths and taking the bond angles into consideration.24,31 The headgroups of polyoxyethylene surfactants are flexible, and in a micelle of such surfactants, a significant average portion of the ethoxylene chains bend back toward the micellar core.23,32 As a result, the radius of the micelles can be expected to be somewhat smaller than the extended length of the surfactant molecule, something that is true also if hydration is taken into account.32 For instance, in micelles of the alcohol ethoxylate C12E8, the thickness of the headgroup region is only about 70% of the length of an extended octa(ethylene glycol) chain.32 Since the difference between the length of the extended Amide surfactant and the corresponding experimental micellar radius can be accounted for by a similar degree of chain bending, it seems fair to conclude that at least at concentrations near the CMC, the micelles of this surfactant are very close to spherical. Although the length of an extended Peptoad is slightly smaller than that of the Amide, it turns out that the apparent radius of the Peptoad micelles is larger. One possible explanation to this observation is that the shape of these micelles deviates slightly from spherical symmetry (see below), but it should also be kept in mind that the headgroup of the Peptoad can be expected to be less flexible than the tetra(ethylene glycol) chain of the Amide surfactant. This is due to both the larger volume per unit length of the “oligo-amide” and the resonance stabilization of the amide groups, which restrict rotation around these. Hence, the experimental hydrodynamic radius can still be consistent with aggregates of spherical symmetry. In a case of spheroidal micelles, Dmic 0 can be related to the aggregate dimensions by a modified version of eq 4, where a correction factor F(r) has been attached,
Dmic 0 )
kBT F(r) 6πηb
(5)
where b is the short semiaxis of the spheroid, and r is the axial ratio r ) a/b, where in turn, a is the long semiaxis of the spheroid. For a prolate aggregate,
F(r) )
ln(r + (r2 - 1)1/2) (r2 - 1)1/2
(6)
2012 J. Phys. Chem. B, Vol. 111, No. 8, 2007
Stjerndahl et al.
and for an oblate aggregate,
F(r) )
arctan(r2 - 1)1/2 (r2 - 1)1/2
(7)
When micelles grow and attain a nonspherical shape, their smallest dimension is always, by geometrical constraints, roughly the length of two extended surfactant molecules. Equations 6 and 7 infer that if this smallest dimension is set to /Dsphere decreases steeply with b, and b is kept constant, Dspheroid 0 0 an increasing r for both prolate and oblate aggregates. This also means that quite small deviations from spherical symmetry will give unreasonable values of the radius for an assumed spherical micelle if it is estimated by means of eq 4. Taken together, the above facts further justify that the first formed micelles (with increasing concentration) are spherical for both the Amide and the Peptoad. If the high values of k achieved from the fits of the diffusion data to eq 3 can be explained by micellar growth into spheroidal aggregates with an increasing surfactant concentration, how large are these aggregates? By using an approach presented by Jonstro¨mer et al.,32,33 one can approximate the concentration dependence of Dsmic by applying an expression of the same form as eq 3, with the difference that the hydrodynamic volume fraction of the spheroid, Φspheroid , is used rather than the H can be estimated by physical volume fraction (i.e., Φ). Φspheroid H
) Φspheroid H
Φ r F(r)3 n
(8)
where F(r) is as defined above, whereas n is 1 for prolates and 2 is for oblates. An estimate of r for an aggregate with a certain self-diffusion coefficient, Dsmic, can be achieved from eqs 3 and 5-8 in the following manner: (1) b and k are set to certain reasonable values, (2) a value of r is assumed, (3) this r is used in eq 5 and eq 6 or 7 to calculate the corresponding Dmic 0 , (4) is calculated using eq 8, and (5) a the corresponding Φspheroid H and Dsmic is calculated using eq 3 and the values of Φspheroid H that where achieved in steps 3 and 4. Steps 2-5 are Dmic 0 repeated until the calculated Dsmic agrees with the observed self-diffusion coefficient. This approach to evaluate diffusion data from NMR measurements has been shown to give values of r that are in good agreement with results from other techniques.33 If k is set to 2 and b to 2.3 and 2.6 nm for the Amide and the Peptoad, respectively, the protocol described above gives axial ratios of tentative prolate micelles at the highest studied concentration (800 mM) of 1.9 and 2.5 for the two surfactants. If the micelles instead are assumed to be oblate, the corresponding values of r are 1.6 and 2.0. In this context, it is worth mentioning that for the Peptoad, the prolate aggregate shape is supported by molecular dynamics simulations performed on this compound.8,34 These indicate a slight one-dimensional growth of the micelles with increasing surfactant concentration. It can also be noted that it is far more common that micelles grow into prolate or wormlike micelles than into oblate ones. To conclude the results of the above calculations, it is obvious that if any micellar growth is occurring in the studied systems, this is very limited. An alternative, or complementary, explanation to the values of k is that the micelles that are formed at the CMC are agglomerating when the surfactant concentration is increased, which gives an increase in the effective size of the diffusing entities. Indications on this phenomenon have previously been
Figure 4. The normalized self-diffusion coefficients of water vs the volume fraction of hydrophobic tails residing in micelles for the Amide (9), the Peptoad (4), and the Ether (O). The solid line represents an estimate of the contribution from obstruction to the decrease in the water diffusion.
observed; for instance, for C8E4 (i.e., the Ether surfactant used as a reference substance in this study).10,11,35 However, on the basis of the present data, it is, unfortunately, not possible to conclude if the concentration dependence of Dsmic for the studied systems is due to aggregate growth, agglomeration of the micelles, or a combination of the two. The translational mobility of water in a micellar solution is strongly dependent on the surfactant concentration. The selfdiffusion is reduced due to the combination of two effects. First, the hydrophobic cores of the micelles exclude a fraction of the total volume for the diffusing water and thereby cause an average lengthening of its diffusion pathway. This is the obstruction effect. Second, water molecules are hydrating the surfactant headgroups and, thus, have a reduced mobility when residing in the polar regions of the micelles. Figure 4 shows the results from the measurements on water in the three studied systems. The self-diffusion coefficients are presented as normalized to the self-diffusion coefficient of water in bulk, Dw0 , and are plotted versus the volume fraction of the surfactant tails that are residing in micelles, Φhc, that is, the part of the surfactant fraction that can be assumed to completely obstruct water diffusion. The total volumes of the hydrophobic cores were calculated on the basis of the assumption that these consist of C7H15 chains in the Amide and Peptoad solutions and C8H17 chains in the Ether case. The density of the hydrocarbon parts was assumed to be 0.7 g/cm3. In general, the effects from obstruction and hydration on the decrease of water diffusion in a surfactant solution are inseparable. However, with some knowledge about the shape of the aggregates, it is possible to predict the contribution from obstruction and, hence, extract information about the degree of hydration. If the hydrophobic cores are approximated as hard spheres, the decrease in diffusion coefficient of the solvent due to obstruction can be described by the following expression,36
Dwobs )
Dw0 1 + Φhc/2
(9)
where all the parameters are defined above. The difference in obstruction of solvent diffusion exerted by spherical or prolate aggregates is strikingly small, even when the axial ratio of the prolates is large.36 With the volume fractions in question, the maximum difference in obstruction is on the order of a few percent. Hence, since it is shown above that the micellar growth
Aggregation and Hydration of Amide Surfactants
J. Phys. Chem. B, Vol. 111, No. 8, 2007 2013 To compensate for the obstruction effect, eq 10 (without the pwmonoDwmono term) can be combined with eq 9 to give
Dw0 Dwobs ) (1 - pwmic) + pwmicDwmic 1 + Φhc/2
(11)
Equation 11 can then be rewritten to yield an expression for an apparent fraction of water that diffuses with the micelles.
pwmic )
Figure 5. Hydration numbers as a function of concentration for the Amide (9), the Peptoad (4), and the Ether (O).
is very limited in the studied systems, the error imposed by assuming spherical aggregates in the following discussion is expected to be negligible. By means of eq 9, the contribution from obstruction to the decrease in the water diffusion can be estimated, and a plot of such a prediction is included in Figure 4. As is evident from this figure, and which is known from previous work on nonionic surfactants,32,37 hydration of the surfactant gives a major contribution to the decrease in water diffusion. It is also interesting to note that the surfactants containing amide groups affect the water diffusion to a larger extent than the Ether surfactant. In the same way as for the surfactants, the observed selfdiffusion coefficient for the water, Dwobs, is the weighted average of the diffusion coefficients in the different environments where it resides. Under the assumption that the solvent has distinct diffusion coefficients when present as free water and when hydrating either monomeric or aggregated surfactant, Dwobs can be described by an expression of the same form as eq 2,
Dwobs ) pwfreeDwfree + pwmonoDwmono + pwmicDwmic
(10)
where Dwfree, Dwmono, and Dwmic are the diffusion coefficients for free water, water hydrating surfactant monomers, and water hydrating surfactant molecules residing in micelles, respectively, and pwfree, pwmono, and pwmic are the fractions of water at each site. It has been shown that since the hydration water cannot be considered as bound to surfactants, but rather has a decreased self-diffusion coefficient when present in the headgroup region, the view described by eq 10 is not a perfect description of the real situation.32,36 It is assumed, however, to be a good enough model for the present discussion. It has been shown that the effect on water diffusion per surfactant molecule is significantly higher for monomers than for molecules residing in micelles.38 Since the surfactants included in this study have relatively high CMC’s, the effect of hydration of the monomeric surfactant, that is, the pwmono Dwmono term in eq 10, cannot be disregarded without consideration. As can be seen in Figure 4, Dwobs is somewhat lower than Dw0 at the CMC, that is, at Φhc ) 0, for both the Amide and the Peptoad. However, since pwmono decreases gradually with increasing surfactant concentration above the CMC, the errors in the estimated micelle hydration caused by disregarding monomer hydration should be small if it is evaluated at concentrations well above the CMC.
Dw0 - Dwobs 1 + Φhc/2 Dw0 - Dwmic 1 + Φhc/2
(12)
In turn, pwmic can be used to calculate a hydration number, H, by applying
H ) pwmic
nwater nsurf
(13)
where nwater is the number of water molecules, and nsurf is the number of surfactant molecules present in the system. Figure 5 presents the hydration numbers of the studied compounds, as calculated using eqs 12 and 13, as a function of the surfactant concentration. When calculating pwmic, the experimental values of Dhmds were used for Dmic for the Amide and the Peptoad, whereas for the Ether, Dmic values were calculated using eq 2 and the experimental values of the CMC and Dsmono. As can be seen in Figure 5, and in line with what could be suggested from Figure 4, the surfactants containing amides have significantly higher hydration numbers than the Ether. When interpreting Figure 5, it is important to remember the meaning of H as calculated from eqs 12 and 13, that is, an apparent number of water molecules that has the same diffusion coefficient as the micelles. As mentioned above, the hydration water cannot be considered to be actually bound to the aggregates, but rather, to have a decreased diffusion coefficient. Hence, in addition to the obvious reason for an increase in H, that is, that the actual number of water molecules residing in the headgroup region of the micelles increases, the change in H can alternatively be explained by an increase in the strength of the surfactant-water interactions (which corresponds to a lower actual diffusion coefficient in the headgroup region). Within the frames of the presently applied model, it is not possible to separate the contributions from the two explanations. It should be noted that the concentration dependence of H is mainly an apparent effect of the limitations in the applied twosite model,36 possibly with a contribution from the more strongly hydrated monomers at concentrations close to the CMC. Conclusions We have studied some of the effects of including amide bonds in nonionic surfactants with short hydrophobic tails. In particular, we wanted to find out what effects the hydrogen bonding ability of the amide group may have on the shape of the aggregates that are formed by such surfactants. The aggregation and hydration of two amide-containing surfactants were compared to that of an oxyethylene-based surfactant of similar size. It appears that both the amide-containing surfactants form small micelles within the whole investigated concentration range. At the CMC, these are most probably spherical. However, the decrease in the self-diffusion coefficient of the micelles with
2014 J. Phys. Chem. B, Vol. 111, No. 8, 2007 increasing surfactant concentration is stronger than what is generally observed for aggregates of spherical symmetry. This observation can be explained by either a minor aggregate growth (into prolates or oblates with axial ratios below 2.5), agglomeration of the micelles, or a combination of the two. Although both the Amide and the Ether surfactants carry oxyethylene chains of the same size, there is a large difference in the intermicellar transport of a hydrophobic solubilizate. For the Ether case, there is, as is often observed for surfactants of this class, a significant transport, but in the Amide system, such transport seems to be absent. This difference in behavior can be a result of either decreased intermicellar headgroupheadgroup interactions or that the amide bonds form a hydrophilic shield with low permeability for the solubilizate. From measurements on water diffusion in the three surfactant systems, it could be concluded that the surfactant hydration is higher when amide bonds are present. This observation is in line with the higher CMC values of the Amide and Peptoad as compared to the Ether. Acknowledgment. This work was financially supported by the Swedish Research Council (MS) and by an NIH grant to F.M.M. (D.L. and H.Z.). The authors thank Prof. Krister Holmberg for suggesting these experiments and for valuable discussions. We are also grateful to the Swedish NMR Center for granting spectrometer time. Mr. Patrik Jarvoll and Dr. Magnus Nyde´n are acknowledged for input on various NMRrelated issues. References and Notes (1) Balson, T.; Felix, M. S. B. Biodegradability of non-ionic surfactants. In Biodegradability of Surfactants; Karsa, D. R., Ed.; Blackie: Glasgow, 1995. (2) Swisher, R. D. Surfactant Biodegradation, 2nd ed.; Marcel Dekker: New York, 1987. (3) Stjerndahl, M.; Lundberg, D.; Holmberg, K. Cleavable Surfactants. In NoVel surfactants, 2nd ed.; Holmberg, K., Ed.; Marcel Dekker: New York, 2003; Vol. 114; p 317. (4) Stjerndahl, M.; Holmberg, K. J. Surfact. Deterg. 2005, 8, 331. (5) Lif, A.; Hellsten, M. Nonionic Surfactants Containing an Amide Group. In Nonionic Surfactants, Organic Chemistry; van Os, N. M., Ed.; Marcel Dekker: New York, 1998; Vol. 72; p 177. (6) Kjellin, M. Structure-property relationships of surfactants at interfaces and polyelectrolyte-surfactant aggregates. Ph.D. Thesis, Royal Institute of Technology, 2002. (7) Kjellin, U. R. M.; Claesson, P. M.; Linse, P. Langmuir 2002, 18, 6745.
Stjerndahl et al. (8) Menger, F. M.; Zhang, H. Langmuir 2005, 21, 10428. (9) Faucompre´, B.; Lindman, B. J. Phys. Chem. 1987, 91, 383. (10) Glatter, O.; Fritz, G.; Lindner, H.; Brunner-Popela, J.; Mittelbach, R.; Strey, R.; Egelhaaf, S. U. Langmuir 2000, 16, 8692. (11) Stubenrauch, C.; Nyde´n, M.; Findenegg, G. H.; Lindman, B. J. Phys. Chem. 1996, 100, 17028. (12) The previously reported value for the Peptoad was 20 mM; however, reevaluation of surface tension data resulted in a CMC of 40 mM. (13) Ohta, A.; Takiue, T.; Ikeda, N.; Aratono, M. J. Solution Chem. 2001, 30, 335. (14) Chang, N. J.; Kaler, E. W. J. Phys. Chem. 1985, 89, 2996. (15) Mukerjee, P.; Kapauan, P.; Meyer, H. G. J. Phys. Chem. 1966, 70, 783. (16) Stejskal, E. O.; Tanner, J. E. J. Chem. Phys. 1965, 42, 288. (17) Stilbs, P. Prog. Nucl. Magn. Reson. Spectrosc. 1987, 19, 1. (18) Price, W. S.; Hayamizu, K.; Ide, H.; Arata, Y. J. Magn. Reson. 1999, 139, 205. (19) Olsson, U.; So¨derman, O.; Guering, P. J. Phys. Chem. 1986, 90, 5223. (20) Ulmius, J.; Wennerstro¨m, H. J. Magn. Reson. 1977, 28, 309. (21) Furo, I. J. Mol. Liq. 2005, 117, 117. (22) So¨derman, O.; Stilbs, P.; Price, W. S. Concepts Magn. Reson., Part A 2004, 23A, 121. (23) Nilsson, P. G.; Wennerstro¨m, H.; Lindman, B. J. Phys. Chem. 1983, 87, 1377. (24) Evans, D. F.; Wennerstro¨m, H. The Colloidal Domain: Where Physics, Chemistry, Biology and Technology Meet, 2nd ed.; Wiley-VCH: New York, 1999. (25) Mitchell, D. J.; Tiddy, G. J. T.; Waring, L.; Bostock, T.; McDonald, M. P. J. Chem. Soc. Faraday Trans. 1 1983, 79, 975. (26) The cloud points (CP) for the Ether and the Amide surfactants are 40 and >100 °C (i.e., not detectable), respectively. The CP’s of the corresponding C12-amide and ether surfactants are 4 and 52 °C, respectively. (27) Stjerndahl, M.; Holmberg, K. J. Colloid Interface Sci. 2005, 291, 570. (28) Evans, G. T.; James, C. P. J. Chem. Phys. 1983, 79, 5553. (29) Nyde´n, M. Measuring micelle size and shape. In Handbook of Applied Surface and Colloid Chemistry; Holmberg, K., Ed.; John Wiley & Sons: New York, 2001; Vol. 2; p 281. (30) Ohtsuki, T.; Okano, K. J. Chem. Phys. 1982, 77, 1443. (31) Ayward, G.; Findlay, T. SI Chemical Data, 5th ed.; John Wiley & Sons: New York, 2002. (32) Jonstro¨mer, M.; Jo¨nsson, B.; Lindman, B. J. Phys. Chem. 1991, 95, 3293. (33) So¨derman, O.; Jonstro¨mer, M.; van Stam, J. J. Chem. Soc., Faraday Trans. 1993, 89, 1759. (34) Menger, F. M.; Zhang, H.; de Joannis, J.; Kindt, J. T. Solubilization of Paclitaxel (Taxol) by Peptoad Self-Assemblies. To be published in Langmuir. (35) Strunk, H.; Lang, P.; Findenegg, G. H. J. Phys. Chem. 1994, 98, 11557. (36) Jo¨nsson, B.; Wennerstro¨m, H.; Nilsson, P. G.; Linse, P. Colloid Polym. Sci. 1986, 264, 77. (37) Nilsson, P. G.; Lindman, B. J. Phys. Chem. 1983, 87, 4756. (38) Lindman, B.; Wennerstro¨m, H.; Gustavsson, H.; Kamenka, N.; Brun, B. Pure Appl. Chem. 1980, 52, 1307.
