8820
Langmuir 2007, 23, 8820-8826
Characterization of Bromide Ions in Charge-Stacked Zwitterionic Micellar Systems Takeshi Aoki, Makoto Harada, and Tetsuo Okada* Department of Chemistry, Tokyo Institute of Technology, Meguro-ku, Tokyo 152-8551, Japan ReceiVed April 19, 2007. In Final Form: June 2, 2007 A novel zwitterionic surfactant, N-dodecyl-N,N,N′,N′-tetramethylethylenediammoniopropanesulfonate bromide (DEPB), has been synthesized, and Br- involved in the micellar system has been characterized by potentiometry, NMR, and X-ray absorption fine structure (XAFS). Although the dissociation degree of Br- from the micelle evaluated by potentiometry almost agrees with that determined by NMR, the former is significantly smaller than the latter over the entire range of concentrations of DEPB. This is explained by assuming that the bromide ions in the micellar system have several different peripheral structures. XAFS has given significant insight into the hydration structures of Brinvolved in the system. Some of the bromide ions partitioned into the micelle are dehydrated and are directly bound by the ammonium groups in the DEP molecules. However, some of the bromide ions are still completely hydrated even when they are partitioned into the micelles. The average hydration number of the bromide ions directly bound by the ammonium groups was determined to be ∼3.3. The partial dehydration of Br- is possibly facilitated by the characteristic hydration circumstances provided by the charge-stacked structure of the surfactant and by the resulting thick palisade layer of the DEP micelle.
Molecular aggregates, such as micelles, vesicles, surface monolayers, and so forth, provide functional interfaces which enable the reactions and molecular recognitions that do not occur in bulk condensed phases.1-7 These features come from the selfassembled nature of amphiphilic molecules and also from the confined space automatically formed through their aggregation. Molecular aggregates have been utilized for the preparation of nanomaterials and biomimetic systems,8-11 the design of sensing devices,12-15 catalytic reactions,15 and molecular separation.16-19 The understanding of the phenomena occurring in these systems is not only of fundamental interest but also of essential importance for further exploitation of functional materials. Ions are often involved in the reactions and phenomena taking place on charged molecular aggregates. The interaction of ions with micelles has, for example, been well studied with various approaches, such as potentiometry,20,21 mobility measurements,4,22 * To whom correspondence should be addressed. Telephone and fax: 81-3-5734-2612. E-mail:
[email protected]. (1) Hunt, N. T.; Jaye, A. A.; Meech, S. R. Chem. Phys. Lett. 2005, 416, 89. (2) Messaoud, T.; Duplatre, G.; Michels, B.; Waton, G. J. Phys. Chem. B 2004, 108, 13137. (3) Shi, Z.; Peterson, R. W.; Wand, A. J. Langmuir 2005, 21, 10632. (4) Iso, K.; Okada, T. Langmuir 2000, 16, 9199. (5) Harada, M.; Okada, T. Langmuir 2004, 20, 30. (6) Harada, M.; Okada, T.; Watanabe, I. J. Phys. Chem. B 2003, 107, 2275. (7) Wittek, M.; Mo¨ller, G.; Johnson, M. J.; Majda, M. Anal. Chem. 2001, 73, 870. (8) Killian, J. A.; Nyholm, T. K. M. Curr. Opin. Struct. Biol. 2006, 16, 473. (9) Goto, M.; Tsukahara, T.; Sato, K.; Konno, T.; Ishihara, K.; Sato, K.; Kitamori, T. Anal. Sci. 2007, 23, 245. (10) Devillers, C. H.; Boturyn, D.; Bucher, C.; Dumy, P.; Labbe, P.; Moutet, J.-C.; Royal, G.; Saint-Aman, E. Langmuir 2006, 22, 8134. (11) Khanal, A.; Inoue, Y.; Yada, M.; Nakashima, K. J. Am. Chem. Soc. 2007, 129, 1534. (12) Kado, S.; Murakami, T.; Kimura, K. Anal. Sci. 2006, 22, 521. (13) Nakamura, T. Anal. Sci. 2007, 23, 253. (14) Kang, T.; Hong, S.; Moon, J.; Oh, S.; Yi, J. Chem. Commun. 2005, 3721. (15) Otani, W.; Kinbara, K.; Zhang, Q.; Ariga, K.; Aida, T. Chem.sEur. J. 2007, 13, 1731. (16) Welsch, T.; Michalke, D. J. Chromatogr., A 2003, 1000, 935. (17) Isoo, K.; Terabe, S. Anal. Sci. 2005, 21, 43. (18) Okada, T.; Harada, M.; Kido, T. Anal. Chem. 2005, 77, 6041. (19) Harada, M.; Kido, T.; Masudo, T.; Okada, T. Anal. Sci. 2005, 21, 491. (20) Zana, R. J. Colloid Interface Sci. 1980, 78, 330. (21) Gaillon, L.; Lelie`vre, J.; Gaboriaud, R. J. Colloid Interface Sci. 1999, 213, 287.
X-ray or neutron scattering,23-25 and chemical probing.26,27 In general, the electrostatic interaction between counterions and oppositely charged groups in amphiphilic molecules is considered as the primary driving force for the ion-involved reactions taking place on the molecular aggregates. Although this must be true, other factors also play important roles in the determination of the overall reactions. The cesium ion is, for example, more preferably partitioned into anionic surfactant micelles than other alkali ions; the affinity becomes higher when going down the periodic table.27 Similar selectivity is found in the partition of halide anions into cationic surfactant micelles.21 Such partition selectivity among identically charged ions cannot be fully interpreted only by columbic interactions. It has recently been shown that even nonionic micelles exhibit anion-governing ionic partition and that this comes from the favorable solvation of anions in the palisade layer of the nonionic micelles.22 Thus, the solvation of an ion is an important factor to be taken into account for the discussion of the ionic processes occurring on molecular aggregates. It is known that zwitterionic micelles exhibit interesting features in ionic partitions partly originating from their characteristic structures. Dodecyldimethylammoniopropanesulfonate (DDAPS, I), which has one negative and one positive group in a molecule, is the most extensively studied zwitterionic surfactant.4,6,28-31 The formal charge of the DDAPS molecule is zero, and, in turn, the DDAPS micelle is also uncharged in a pure solvent. However, (22) Ohki, T.; Harada, M.; Okada, T. J. Phys. Chem. B 2006, 110, 15486. (23) Aswal, V. K.; Goyal, P. S.; De, S.; Bhattacharya, S.; Amenitsch, H.; Bernstorff, S. Chem. Phys. Lett. 2000, 329, 336. (24) Aswal, V. K.; Kohlbrecher, J.; Goyal, P. S.; Amenitsch, H.; Bernstorff, S. J. Phys.: Condens. Matter 2006, 18, 11399. (25) Magid, L. J.; Han, Z.; Warr, G. G.; Cassidy, M. A.; Butler, P. D.; Hamilton, W. A. J. Phys. Chem. B 1997, 101, 7919. (26) Asakawa, T.; Kitano, H.; Ohta, A.; Miyagishi, S. J. Colloid Interface Sci. 2001, 242, 284. (27) He, Z. M.; O’Connor, P. J.; Romsted, L. S.; Zanette, D. J. Phys. Chem. 1989, 93, 4219. (28) Baptista, M. S.; Cuccovia, I.; Chaimovich, H.; Politi, M. J.; Reed, W. F. J. Phys. Chem. 1992, 96, 6442. (29) Chevalier, Y.; Kamenka, N.; Chorro, M.; Zana, R. Langmuir 1996, 12, 3225. (30) Kamenka, N.; Chevalier, Y.; Zana, R. Langmuir 1995, 11, 3351. (31) Okada, T. Anal. Chim. Acta 2005, 540, 139.
10.1021/la701145q CCC: $37.00 © 2007 American Chemical Society Published on Web 07/12/2007
Characteriziation of Br Ions in Micellar Systems
in most of the electrolytes, a DDAPS micelle is negatively charged.4 Although this is due partly to the polarity of the DDAPS molecule (i.e., it has an inner positive charge and an outer negative charge), it has been pointed out that the hydration nature of anions is important as well. The partition selectivity becomes lower in the order of ClO4- > I- > Br- > Cl-; a poorly hydrated anion is partitioned to a greater extent than its well hydrated counterpart.4,6 If more positive charges are introduced into the DDAPS molecule, the electrostatic interaction of the resulting micelle with anions must be enhanced. In addition, such a surfactant should still keep the characteristics of zwitterionic surfactants. According to this idea, we have designed and synthesized a novel charge-stacked surfactant, N-dodecylN,N,N′,N′-tetramethylethylenediammoniopropanesulfonate (DEP+, II). The present paper focuses on the characterization of Br- in the aqueous DEP+Br- (DEPB) micellar system with various approaches, such as potentiometry, NMR, and X-ray absorption fine structure (XAFS).
Langmuir, Vol. 23, No. 17, 2007 8821 χ(k) )
µ(k) - µb(k) - µ0(k)
k)
SjNjFj(kj)
∑ j
2
x
2m (E - E0) p2
exp(-2σj2kj2) sin[2kjrj + φj(kj)]
(2)
krj
kj )
Experimental Section
(1)
where m is the electron mass, E is the incident X-ray energy, E0 is the threshold energy, and µ(k), µ0(k), and µb(k) are the total absorption coefficient, the absorption due only to the K shell excitation of a priori isolated bromide ion, and the background absorption depending on the circumstances of the absorbing atom, such as the absorption from the other shells and long-range solvation effects, respectively. The energy giving half of the edge jump was chosen as E0. The background absorption, µb(k), was estimated with Victreen’s formula, aE -3 - bE-4 - c. The XAFS spectra in the k space, χ(k), were analyzed by curvefitting with the following equations. χ(k) )
Preparation of DEP. DEPB was synthesized by two step reactions, that is, the reaction of dodecyl bromide with N,N,N′,N′-tetramethylethylenediamine in ethanol and the following reaction of the product with propane sultone in ethanol. The evaporation of ethanol gave DEPB as a white solid. DEPB was recrystallized from acetonemethanol. A methanolic solution of recrystallized DEPB passed through a Br--form anion-exchange resin (Amberlist A26) column twice. Methanol was removed from the column effluents, and the residue was again recrystallized. Ion chromatographic measurements confirmed that the Br- content in the purified DEPB was higher than 98% of the theoretical values. The preparation of DEP chloride (DEPC) was similarly carried out with a Cl--form anion-exchange resin column. Other reagents of analytical grade were used as received. Aqueous solutions were prepared with MilliQ water. Potentiometric and 1H NMR Measurements. The concentrations of free Br- in DEPB solutions were determined at 25 °C by potentiometry with a Br--selective electrode (8005-10C, Horiba). The potential measured against a double-junction Ag/AgCl reference electrode at 25 °C was converted into the free Br- concentration according to the literature.20 A linear relation between the potential and the logarithm of the DEPB concentration was determined using DEPB solutions with concentrations below the critical micellar concentration (CMC). This linear relation was extrapolated to the working range and used as a calibration graph. 1H NMR spectra were recorded at 25 °C with an AL-300 FT-NMR spectrometer (JEOL) with sample solutions prepared in D2O. XAFS. A solution sample, NaBr or DEPB in water, was sealed in a polyethylene pouch and used for XAFS measurement at the Br K-edge. The measurements were carried out at BL9C of Photon Factory, High Energy Accelerator Research Organization in Tsukuba, Japan. Dried DEPB was mixed with BN powder very well under the dried N2 atmosphere and then sealed in a polyethylene pouch immediately before the measurements. The scattering amplitudes and phase shifts for the model systems were calculated with the FEFF8.02 program. The XAFS data were analyzed according to the literature, as briefly outlined below.6,32-34 The normalized XAFS interference function in the k-space, χ(k), is defined as
µ0(k)
x
k2 -
2m ∆E0j p2
where j is the coordination shell number, rj is the distance between the atom under study and a scattering atom, SjNj is the amplitude factor (where Sj is the amplitude reduction factor), σj is the DebyeWaller factor, E0j is the absorption edge shift, and Fj(kj) is the backscattering amplitude. The fitting parameters were the number of coordinating atoms, an energy shift, the distance between the absorbing and scattering atoms, and a Debye-Waller factor. Experimental XAFS spectra, χ(k), were Fourier-transformed with a Hanning window in the range k ) 0-6.8 Å-1. XAFS parameters were determined by curve-fitting in the range k ) 1.7-6.8 Å-1.
Results and Discussion Potentiometric and NMR Evaluation of Br- Binding to the DEP Micelle. The size of the DEPB micelle was determined by dynamic light scattering to be 4.3 ( 1.2 nm. Molecular modeling indicates that the length of the DEP+ molecule is ∼2.5 nm for the entirely extended zigzag conformation. The actual molecular length may be shorter than this length because molecular bending is expected at the ammonium nitrogen atoms. Thus, the micelle must be spherical, albeit further study should be necessary to reveal details. A number of methods have been devised to evaluate the binding of counterions to ionic micelles. Potentiometry with an ionselective electrode has been extensively used for this purpose, and it has allowed the determination of the concentration of free counterions.20,21 The free Br- concentration ([Br-]free) for the DEPB micellar system is given by
[Br-]free ) [DEP]CMC + R(CDEP - [DEP]CMC)
(3)
where CDEP and [DEP]CMC are the total concentration and CMC, respectively, of DEPB and R is the ionization degree of the DEPB micelle. The CMC of DEPB was determined to be 16 mM at 25 °C with conductivity measurements. Although a continuous decrease in the monomer dodecyl sulfate with increasing concentration above its CMC was pointed out, the constancy of the monomer concentration of DEP was assumed here because this effect is not very serious.35,36 Figure 1 shows the dependence (32) Okada, T.; Harada, M. Anal. Chem. 2004, 76, 4564. (33) Harada, M.; Okada, T.; Watanabe, I. J. Phys. Chem. B 2002, 106, 34. (34) Harada, M.; Okada, T. J. Chromatogr., A 2005, 1085, 3.
8822 Langmuir, Vol. 23, No. 17, 2007
Figure 1. Concentration dependencies of γ ) ([DEP - Br]mic)/ CDEP and the dissociation degrees of Br- from the DEP micelle (R) determined by potentiometry (RISE) and NMR (RNMR).
Figure 2. 1H NMR spectrum for DEP. The concentration of DEP was 0.1 M in D2O. The assignments were carried out on the basis of 2D NMR measurements.
of potentiometrically evaluated RISE values on CDEP. The RISE value is almost 0.2 over the entire range of CDEP studied. This value is similar to the corresponding values (R ) 0.18-0.23) reported for the dodecyltrimethylammonium bromide micelle.20,26 Although potentiometry is reliable for the determination of the concentration of free counterions in micellar solutions, it is a disadvantage that no information is provided for the counterions bound by the micelles. Small angle neutron or X-ray scattering is also an effective method to probe the structure of micelles, but it does not provide explicit information on counterion bindings.23-25 For the DEP micelles, NMR gives different perspectives from potentiometry. Figure 2 shows an NMR spectrum obtained with 0.1 M DEPB in D2O; the chemical shifts (δ) were assigned on the basis of the results of 2D NMR measurements. An increase in the concentration of DEPB resulted in the low-field shifts of the δ values of several protons, which were possibly caused either by micellization or by the binding of counterions to the micelle. Importantly, the shifts of the δ values with changing CDEP were not found for DEPC. It is known that Br- has higher (35) Sasaki, T.; Hattori, M.; Sasaki, J.; Nukina, K. Bull. Chem. Soc. Jpn. 1975, 48, 1397. (36) Moroi, Y. J. Colloid Interface Sci. 1988, 122, 308.
Aoki et al.
affinity to the positive groups on the micelle than Cl-,21 suggesting that the low-field shifts are caused by the binding of counterions. Figure 3 shows the relations between the relative δ values and CDEP. It is reasonably assumed that the terminal methyl protons (no. 1 in Figure 2) are not affected by the counterion binding. Therefore, the δ values relative to that of 1 are plotted in Figure 3. The binding of bromides should occur on the ammonium groups. Therefore, increasing CDEP causes the low-field shifts of the relative δ values for the protons bonded to the peripheral carbon atoms of the inner ammonium group (nos. 6, 8, and 10) but the opposite shifts for the protons situated near the terminal sulfonate group (nos. 4 and 5). NMR spectra of DEP were simulated with the ab initio molecular orbital program package, Gaussian03, to confirm the low-field shifts by the binding of an anion to the ammonium group. After the structural optimization of a DEP+ molecule in water, Br- was added to the system and again the structure was optimized. The optimized structure strongly depended on the initial configuration, especially on the relative position of DEP+ and Br-. Br- was finally bound on either ammonium group, and no significant energetic difference was found between the ionassociates of Br- with the inner and outer ammonium groups. The comparison of the chemical shifts simulated before and after the addition of Br- revealed that the chemical shifts of the protons present near Br- moved to the lower fields. The lowfield shifts for the 6, 8, and 10 protons thus suggest that Br- is selectively bound by the inner ammonium group. The electrostatic repulsion between adjacent DEP molecules in the micelle may be relaxed to a greater extent by the binding of Br- on the inner ammonium group. Also, the dehydration of Br- is possibly enhanced when it deeply penetrates the micelle. Preferable ionassociation thus occurs on the inner ammonium group. We focus further attention on the changes in the chemical shifts of the protons of four ammonium methyl groups (nos. 6 and 7). As shown in Figure 3, the chemical shift for the 6 protons shows a low-field shift when CDEP increases, while that of 7 is not affected by CDEP. A similar trend is seen for the 8 and 9 protons. Although an effect of CDEP on the chemical shift of the 8 protons is larger than that for the 6 protons, precise measurements are possible for the latter value because of the peak sharpness. For this reason, peaks 6 and 7 were selected. As noted above, a low-field shift was not found for 7, and thus the difference in the chemical shift between 6 and 7 (∆δ) becomes smaller as CDEP increases, as plotted in Figure 4. As far as DEPB solutions are concerned, the following simplified equilibrium for Br- binding to the DEP micelles can be assumed. Kb
DEPmic + Br- {\} DEP-Brmic Kb )
[DEP-Br]mic [DEP]mic[Br]free
(4)
where mic denotes the micelle. The following two equations also hold for the present system.
CDEP ) [DEP]mic + [DEP-Br]mic + [DEP]CMC
(5)
CBr ) [Br-]free + [DEP-Br]mic ) CDEP
(6)
where CBr is the total concentration of Br-. The NMR spectrum of the DEP monomers should not be influenced by coexistent Br-, because DEPB is completely dissociated unless it forms the micelle. Figure 4 clearly indicates that the ∆δ values are actually constant when CDEP < [DEP]CMC, strongly indicating that the
Characteriziation of Br Ions in Micellar Systems
Langmuir, Vol. 23, No. 17, 2007 8823
Figure 3. Changes in the chemical shifts (peaks 3-10) with CDEP. The δ values relative to that for proton 1 were plotted versus CDEP. The numbers refer to the protons given in Figure 2.
∆δ value is changed by Br- binding. Since the participation of DEP molecules in micellar formation does not cause the changes in the chemical shifts, DEPmic should give a ∆δ value identical to that of the monomer DEP molecules. In contrast, DEP-Brmic should give a different NMR spectrum from that for DEPmic and the DEP monomers and thereby cause a change in ∆δ. Thus, the ratio of [DEP-Br]mic to CDEP concentration (γ) should well characterize Br- binding to the DEP micelle, and the information on this ratio can be extracted from the relation between ∆δ and CDEP.
γ)
[DEP-Br]mic CDEP
(7)
This parameter can be related to R by the following equation.
R)
CDEP(1 - γ) - [DEP]CMC CDEP - [DEP]CMC
(8)
Substituting eqs 4-6 into eq 7 gives
Kb )
γ (1 - γ){(1 - γ)CDEP - [DEP]CMC}
(9)
A difference in the chemical shift between 6 and 7 should be constant (∆δ ) ∆δ1) when CDEP < [DEP]CMC. The minimum ∆δ (∆δ2) must be reached when the Br- binding to the micelle is saturated. The experimentally measured ∆δ values should be within the range between these two extremes, ∆δ1 and ∆δ2, and they are represented by
∆δ ) (1 - γ)∆δ1 + γ∆δ2
(10)
As CDEP increases to above [DEP]CMC, ∆δ becomes smaller; this change can be explained by the above model with the parameters Kb ) 68 (σ ) 19) M-1 and ∆δ2 ) 0.036 (σ ) 0.001) ppm, as depicted by the solid curve in Figure 4. In Figure 1, γ and the calculated R values (denoted by RNMR to distinguish from RISE) are plotted versus CDEP. The RNMR value is slightly larger than the RISE value over the entire CDEP range. Marked differences between these two dissociation degrees at CDEP < 0.04 M are due mainly to the uncertainty of the CMC. In contrast, the difference over the intermediate concentration range (0.04-0.1 M) deserves further discussion. Figure 5 schematically illustrates the local structures of bromides partitioned into the DEP micelle: direct ion-pair, Br-
Figure 4. Change in difference in the chemical shift between peaks 6 and 7 (∆δ) with CDEP. The solid curve shows the result of curvefitting based on eqs 4-7 with the parameters Kb ) 68 M-1 and ∆δ2 ) 0.036 ppm. The dashed line represents the corresponding change in ∆δ for DEPC. Details are given in the text.
associated with the ammonium group via water molecules (watershared ion-associate), and hydrated Br- partitioned into the micelle. The direct ion-pair formation should be accompanied by the partial dehydration of Br-, and the resulting strong electrostatic interaction should induce the largest shifts of the δ values. Even though the water-shared ion-associate may also cause the low-field shifts of δ, its effect must be much smaller than that of the direct ion-pair. Although we cannot explicitly discriminate the effect of the direct ion-pair from that of the water-shared ion-associate, the contributions from them are involved in the RNMR values. In this case, the difference between RNMR and RISE can be interpreted by the presence of hydrated bromide ions, which are partitioned into the micelle but do not form ion-associates with the ammonium groups of the DEP molecules. On the contrary, if the water-shared ion-associate does not cause the low-field shift of the δ values, the difference between RNMR and RISE implies the existence of partitioned hydrated Br- and/or water-shared ion-associate. Importantly, both species are completely hydrated. Therefore, in any case, there are at least two types of bromides, which have different hydration numbers. The hydration of Br- in the DEP micelle is discussed below in more detail. Hydration Structure of Br- in the DEP Micelle. XAFS is an effective method to probe the short-distance structure around a targeted atom and has, for example, elucidated the solvation
8824 Langmuir, Vol. 23, No. 17, 2007
Aoki et al.
Figure 6. XAFS spectra at the Br K-edge obtained with solid DEPB (dashed curve), Br- (NaBr) in water, and the 0.03-0.15 M DEPB solutions. The arrows indicate increasing CDEP (0.03, 0.05, 0.1, and 0.15 M). Obvious changes in the phase shifts as well as decreasing intensities with increasing CDEP are seen.
Figure 5. Possible local structures of Br- in the DEPB micellar system.
structures of ions under various circumstances.6,32-34,37 We have focused our special attention on the solvation structures of Brand have found that its hydration structures are partly perturbed when it is bound by positive groups, such as ammonium groups in ion-exchange resins and surface monolayers.6,32-34 From this perspective, the local structure of Br- in the DEP micelle is of fundamental interest because of the charge-stacked structure of DEP and the thick hydrophilic layer of the micelle. As stated above, we have inferred that there are several hydration states for Br- in the DEP micelle on the basis of the determination of the R values. XAFS is expected to evidence the solvation structures of these species. Figure 6 compares the XAFS χk3 spectra at the Br K-edge obtained with the dried DEPB solid, an aqueous DEPB solution, and an aqueous NaBr solution. A number of studies have indicated that the hydration number of Br- is six, and therefore, the spectrum for aqueous Br- should represent the structure of Br-(H2O)6.38-45 The spectrum for the DEPB solid is obviously different from that of hydrated Br- in the phase shift and oscillation intensity. In a dried DEP solid, Br- should be directly bound by the ammonium groups in the DEP molecules, and thus, this spectrum reflects the direct ion-pair of Br- and DEP. The curve-fitting with Br-C as a model gave the fitting parameters, rBr-C ) 3.65 Å, N ) 9.2, and σ ) 0.20 Å. The coordination distance agrees with those determined for Br- bound on the ammonium groups in an ion-exchange resin32-34 and for Br- dissolved in aprotic (37) Ohki, T.; Harada, M.; Okada, T. J. Phys. Chem. B 2007, 111, 7245. (38) Ohtaki, H.; Radnai, T. Chem. ReV. 1993, 93, 1157. (39) Ohtomo, N.; Arakawa, K.; Takeuchi, M.; Yamaguchi, T.; Ohtaki, H. Bull. Chem. Soc. Jpn. 1981, 54, 1314. (40) Licheri, G.; Piccaluga, G.; Pinna, G. Chem. Phys. Lett. 1975, 35, 119. (41) Licheri, G.; Piccaluga, G.; Pinna, G. J. Chem. Phys. 1975, 63, 4412. (42) Wakita, H.; Ichihashi, M.; Mibuchi, T.; Masuda, I. Bull. Chem. Soc. Jpn. 1982, 55, 817. (43) Caminiti, R.; Cucca, P. Chem. Phys. Lett. 1982, 89, 110. (44) Raugei, S.; Klein, M. L. J. Chem. Phys. 2002, 116, 196. (45) de Barros Marques, M. I.; Cabaco, M. I.; Sousa Oliveira, M. A.; Alves Marques, M. Chem. Phys. Lett. 1982, 91, 22.
solvents,46 in which methyl carbon atoms are the primary scattering groups. If Br- is placed on the tripod composed of two methyl groups and one methylene group of the ammonium groups of DEP, three carbon atoms and one nitrogen atom are located at similar distances apart from Br- and act as scattering groups. Taking the similar backscattering ability of a nitrogen atom to a carbon atom into account, we can say that N ) 4 is reasonable for Br- directly bound on an ammonium group. Although there still remains some ambiguity for the solid-phase structure of DEPB, the above coordination number (N ) 9.2) suggests that Br- is possibly sandwiched between two ammonium groups in the dried DEPB. The spectrum obtained with the aqueous DEPB solution has a phase shift similar to that for the aqueous NaBr solution, suggesting that water is a predominant scattering group for Br- in the DEPB solution. However, the oscillation intensity of the spectrum for a DEPB solution is lower than that for Brin water, and disagreements in the phase shift are also seen (indicated by arrows in Figure 6). These spectral features become more obvious as the concentration of DEPB increases. A usual XAFS analysis procedure was first applied to the spectra obtained with aqueous DEPB solutions by assuming oxygen atoms as scattering groups. A result of the curve-fitting for 0.15 M DEPB is depicted in Figure 7A; the parameters were determined to be rBr-O ) 3.08 Å, N ) 3.6, and σ ) 0.15 Å. The distance between Br- and O is obviously shorter than that estimated for Br- (r ) ∼3.2 Å) in water,32,33,46 and the agreement between the experimental spectrum and the calculated one is not very good. These come from the assumption of single scattering groups (Br-O). We studied the local structure of Br- in anion-exchange resins soaked in water and found that some of the bromide ions are directly bound by the trimethylammonium ion groups in resins and are simultaneously hydrated by one to three water molecules.32 In the local structure of Br-, two different scattering groups, the oxygen atoms in water and the carbon (and possibly nitrogen) atoms in the ammonium group, contribute to the XAFS spectrum obtained at the Br K-edge. Our analyses indicated that the methyl carbon atoms in the ion-exchange group are located at r ) ∼3.6 Å.33,34 A similar situation is realized in the present case as well. If Br- directly bound by the ammonium group of DEP simultaneously interacts with water molecules, two different (46) Tanida, H.; Sakane, H.; Watanabe, I. J. Chem. Soc., Dalton Trans. 1994, 2321.
Characteriziation of Br Ions in Micellar Systems
Langmuir, Vol. 23, No. 17, 2007 8825
Figure 7. Results of curve-fittings. (A) Usual XAFS analysis with Br-H-O as a model. (B) Simplified fitting based on eq 11 involving the spectra obtained for solid DEPB and hydrated Br- as components. The solid curves are the experimental spectrum for 0.15 M DEPB. The dashed curves are the results of curve-fittings. Table 1. Parameters Determined on the Basis of Eq 11 and Hydration Numbers of Br- in the DEP Micellar Systems DEP/M
a1a
a2a
Nentire () 6a1)
Nmic
Non-mic
0.03 0.05 0.1 0.15
0.82 (0.01) 0.79 (0.01) 0.73 (0.01) 0.71 (0.01)
0.12 (0.21) 0.28 (0.21) 0.29 (0.21) 0.31 (0.21)
4.92 4.74 4.38 4.26
3.84 4.20 4.09 4.07
3.38 3.75 3.62 3.58
a
Standard deviations are in parentheses.
scattering groups should be present at different interaction distances. In such a case, a two-shell model is often employed for XAFS spectral analyses. However, the intrinsically long coordination distances of Br- have made difficult the application of a two-shell model. A simplified technique was proposed to overcome this problem and to estimate the contributions from two scattering groups.32 Equation 2 indicates that a XAFS χ spectrum is basically proportional to N. When there are two scattering groups around a targeted atom, the resulting spectrum should be the superimposition of two individual spectra obtained for single scattering groups (χ1 and χ2).
χ ) a1χ1 + a2χ2
(11)
where a1 and a2 are coefficients representing the contributions from χ1 and χ2, respectively. The appropriate selection of χ1 and χ2 is important for the successful utilization of eq 11. As noted above, there are two scattering groups, that is, the oxygen atoms in water molecules and the carbon atoms in the methyl ammonium groups, in close vicinity to Br- in aqueous DEPB solutions. Therefore, the XAFS spectrum obtained for hydrated Br- was selected as χ1, and that obtained with the dried DEP solid was selected as χ2. Figure 7B illustrates the results of the least-square fitting to the spectrum for an aqueous 0.15 M DEP solution with eq 11. The calculated spectrum agrees well with the experimental one, and the parameters were determined to be a1 ) 0.71 (σ ) 0.10) and a2 ) 0.31 (σ ) 0.21). The results obtained by the application of eq 11 are summarized in Table 1. Two parameters are changed in opposite ways; that is, a1 decreases but a2 increases with increasing DEP concentration. When a1 ) 1 and a2 ) 0, the spectrum corresponds to that of Br-(H2O)6; that is, the average hydration number of Br-(Nentire)
Figure 8. Summarized characterization of bromide ions in the DEPB systems. The y-axis represents the ratio of a particular species to the total concentration of DEPB. The solid thick curve represents the boundary between the micellar bound bromides and free bromides, whereas the black dashed curve distinguishes free bromides into two categories: bromides as countercations of the monomer DEP+ () CMC) and bromides dissociated from the DEPB micelle. The red dashed curve indicates the boundary between completely hydrated bromides and partly desolvated bromides.
should be equal to 6a1. As shown in Table 1, Nentire decreases with increasing DEP concentration due to the decrease in the relative contribution from free Br- as a counterion of the DEP monomer. This contribution can be corrected by the following equation:
Nmic )
(Nentire - 6[DEP]CMC/CDEP) (CDEP - [DEP]CMC)/CDEP
(12)
where Nmic represents the average hydration number of the counterions of the DEP micelles. The Nmic value () 4.1-4.2) is almost constant irrespective of CDEP. The low value can be seen for CDEP ) 0.03 M; this comes from the uncertainty of the CMC. The constant Nmic value supports the constant dissociation degree of the counterions (R) from the micelle. The hydration number of Br- dissociated from the micelle should also be six. Thus, the average hydration number of Br- bound by the micelle (Non-mic) is given by
Non-mic )
Nmic - 6R 1-R
(13)
Substituting RISE into eq 10 allows us to estimate the Non-mic values, which range from 3.6 to 3.8 for CDEP ) 0.05-0.15 M as listed in Table 1. The Non-mic values for the DEP micelle indicate that if all of the bromide ions on the micelle are directly bound on the ammonium groups, they should simultaneously be hydrated by 3.6-3.8 water molecules. As discussed above, the difference between RISE and RNMR can be interpreted by the existence of two different hydration structures for the bromide ions in the DEP micelles, that is, completely hydrated Br- and partially dehydrated Br- forming a direct ion-pair with the ammonium group. Although it is not known whether the former is simply partitioned into the micelle or forms water-shared ion-associates, both should be Br(H2O)6-. Thus, the difference between RISE and RNMR originates from the existence of Br(H2O)6- in the micelle. The correction of this contribution to Non-mic gives the hydration number of the bromide ions affecting the NMR spectra, which is equal to ∼3.3. This hydration number almost agrees with the maximum hydration number () 3.0) of Br- directly bound by the active groups in the anion-exchange resin soaked in water,32 possibly suggesting that the low-field shifts of δ are
8826 Langmuir, Vol. 23, No. 17, 2007
caused only by the direct ion-pairs. If the water-shared ionassociates also influence the NMR spectra of DEP, the hydration number of Br- directly bound to the ammonium groups in the micelle should be smaller than 3.3. Figure 8 summarizes the results of the characterization of Br- in aqueous DEP systems, where it was assumed that the water-shared ion-associates do not cause the low-field shifts of the NMR signals. The ratios of the individual concentrations of Br- species to CBr are taken as the y-axis. Hydrated bromides partitioned into the micelle occupy ∼10% of CBr involved in the entire system.
Conclusion The bromide ions in the DEP micellar solution were characterized by potentiometry, NMR, and XAFS. The chargestacked structure of the DEP molecule has allowed the efficient utilization of NMR to probe the states of Br- in the system, and it has implied that Br- partitioned into the micelle can adopt at least two different solvation configurations. In the present study, the hydration number of Br- directly bound to the DEP molecule
Aoki et al.
has been determined to be ∼3.3. This number suggests that approximately three water molecules are removed when Br- is directly bound by the ammonium group. Similar results were reported for Br- in an anion-exchange resin32 and zwitterionic surface monolayer.6 Thus, this may be the general nature of Brunder water-shortage circumstances, in which the water structure is perturbed by interaction with the hydrophilic moieties of surfactant molecules and the polymer domains of ion-exchange resins. Further compilation of data is expected to allow the understanding of not only the hydration nature of ions in such media and but also ion-recognition selectivity provided by functional materials. Acknowledgment. This work was in part supported by a Grant-in-Aid for Scientific Research from the Japan Society for the Promotion of Science and by the Salt Science Research Foundation. LA701145Q
