J . Phys. Chem. 1985,89, 4868-4873
4868
beam intensity when a voltage is applied to the rods.
Results and Discussion Findings of the previously reported experiments' indicate that the monomer ( H 2 0 ) and dimer ((H20)2)are polar and that the higher polymer [(H20),, N = 3-61 are nonpolar. In the present study, the quadrupole electric deflection of water clusters was extended to cluster sizes containing up to 17 water molecules. As expected, both H 2 0 and ( H 2 0 ) 2were found to focus. The observed strong focusing of (H20)2, detected as the protonated monomer [H+(H,O)], at low voltage is indicative of a first-order Stark effect; this species was also found to focus in the hexapole fieldI3 as expected. The quadrupole focusing of H 2 0 is much weaker, indicative of a second-order Stark effect; it did not display focusing in the hexapole field.I3 The strong focusing of ( H 2 0 ) 2 is consistent with the calculated s t r u c t ~ r e ' ~which J ~ has been verified by molecular beam electric resonance spectroscopy." Structural calculation^^^^^* performed on higher polymers predict cyclic structures for clusters containing five or more waters. There is no agreement as to whether the trimer and tetramer have open chain or closed ring structures. The results of the present quadrupole electric deflection experiments showed that the species (H20), with N = 3 to 17 do not focus; rather they show defocusing ranging from about 5 to 10%when the beam obstacle is removed and the field is on. Such defocusing indicates that these clusters are nonpolar and is strong evidence favoring the predicted cyclic structures. [It is worthy of note that the odd members of the acetic acid clusters were found to display focusing.' The results enabled definite assignments for N ranging to 4 and indicated the presence of higher-order clusters having polar structures.] N o evidence for the predicted9 formation of ion pairs was obtained, perhaps because polarizability effects dominate the focusing behavior. Simple primary alcohols interact through hydrogen bonds similar to those between water molecules. Recent ab initio mo(15) Del Bene, J.; Pople, J. J. Chem. Phys. 1970, 52, 4858; Owicki, J.; Shipman, L.; Scheraga, H. J . Phys. Chem. 1975, 79, 1794. (16) Kistenmacher, H.; Popkie, H.; Clementi, E.; Watts, R.J. Chem. Phys. 1974,60,4455. Kistenmacher, H.; Lie, G.; Popkie, H.; Clementi, E. J. Chem. Phys. 1974, 61, 546. (17) Dyke, T.; Mack, K.; Muenter, J. J. Chem. Phys. 1977, 66, 498. (18) Hankins, D.; Moskowitz, J.; Stillinger, F. J . Chem. Phys. 1970,53, 4544.
lecular orbital calculations on the menthanol dimer19 suggest that the dimer should be bound by a near linear hydrogen bond and, thus, be a polar species. Similar calculations on higher methanol polymersZosuggest cyclic nonpolar structures similar to those for water. 5 ~ 1 8 Both the monomer and dimer of methanol were found to exhibit refocusing characteristic of polar structures. The results of the quadrupole electric deflection experiment on the species ( C H , O H ) , N = 3 to 17, show that these species rather strongly defocus (about 7-14%), consistent with the theoretically predictedI9 nonpolar cyclic structures and the earlier finding that the trimer and tetramer are nonpolar.2 While no ab initio calculations on ethanol clusters presently exist, on heuristic grounds these species are expected to have structures similar to those for methanol. The detected ion clusters had the stoichiometry H+(C2HsOH),; the distribution also displays an intensity drop between H+(C2H50H)4and H+(C2HSOH)s. In accord with the methanol results, the quadrupole electric deflection of ethanol monomer and dimer also showed these species to exhibit refocusing characteristics of polar structures. Again, all clusters (C2HSOH), ( N = 3-13) display strong defocusing (1O-17%), an indication of nonpolar cyclic structures. Study of the electric deflection of water clusters has been extended in the present work to a degree of aggregation of 17 molecules. In agreement with earlier structures' for clusters up to the hexamer, and in accord with p r e d i c t i o n ~ , ~the ~ Jpresent ~ findings are consistent with the clusters having a cyclic structure. The results of the molecular beam electric deflection experiments performed on clusters of methanol, and ethanol, indicate that, in agreement with the situation for the smaller clusters found by Odutola et a1.,2the higher polymers are nonpolar supporting the cyclic closed ring structures predicted the~retically.'~
Acknowledgment. The authors gratefully acknowledge the US. Army Research Office, Grant No. DAAG-29-79-0133, which supported the experimental phases of the work, and Grant No. DAAG29-82-K-0160, which supported the later phases of writing enabling publication. Registry NO. HZO, 7732-18-5; CH3OH, 67-56-1; CZHSOH, 64-17-5. (19) Del Bene, J. J . Chem. Phys. 1971,55,4633. (20) Curtiss, L. J . Chem. Phys. 1977, 67, 1144.
A Comparative Study of Organic Counterion Binding to Micelles with the Fourier Transform NMR Self-Diffusion Technique Mikael Jansson* and Peter Stilbs Institute of Physical Chemistry, Uppsala University, S - 751 21 Uppsala, Sweden (Received: February 25, 1985; In Final Form: June 13, 1985) Multicomponent NMR self-diffusion measurements on D 2 0 solutions of decylammonium acetate, chloroacetate, and dichloroacetate are reported. The association behavior of ionic surfactant systems, as evaluated for a comparison of individual component diffusion data (quantified in terms of the degree of counterion binding, the cmc, and the nonmicellar amphiphile concentration above the cmc), was found to be distinctly sensitive to the counterion character. The cmc decreases in the series CH,COO-, CH2C1COO; and CHC12C00- while the degree of counterion binding increases rather strongly. In competitive ion-binding experiments in mixed surfactant systems the differences in ion binding are dramatically amplified.
Introduction N M R based pulsed-gradient spin-echo (PGSE) self-diffusion measurements have recently emerged as a powerful tool for the investigation of counterion binding in polyelectrolyte systems.'-" (1) Lindman, B.; Puyal, M.-C.; Kamenka, N.; RymdBn, R.; Stilbs, P. J . Phys. Chem. 1984, 88, 5048. (2) Lindman, B.; Kamenka, N.; Puyal, M.-C.; Brun, B.; Jonsson, B. J . Phys. Chem. 1984, 88, 5 3 .
0022-3654/85/2089-4868$01.50/0
With the Fourier transform modification of the technique (FTPGSE)one can simultaneously monitor the self-diffusion coefficients of several components and, through a comparative procedure, conveniently quantify aggregation and substrate- and ion-binding processes in solution. Recent investigations in this (3) Stilbs, P.; Lindman, B. J. Phys. Chem. 1982, 85, 2587. (4) Stilbs, P.; Lindman, B. J . Magn. Reson. 1982, 48, 132.
0 1985 American Chemical Society
Organic Counterion Binding to Micelles
The Journal of Physical Chemistry, Vol. 89, No. 22, 1985 4869 2
4
0
25
30
35
45
A
h
A
50
h
A
-
t
.
A
55
h
60
6(ms)
Figure 1. An example of a competitive multicomponent self-diffusion FT-PSGEexperiment on a mixture of decylammonium acetate, chloroacetate, and dichloroacetate. Above is the normal 'HNMR spectra, and below a series of spin-echo spectra, where the rate of the decrease of the peak magnitudes with increasing duration of the gradient pulse directly reflects the respective molecular self-diffusion rates. For experimental details, see ref 1 1 and
.*L. 1
field concern, e.g., ion binding to polyelectrolytes," micellar agg r e g a t i ~ nand , ~ substrate binding to vesicles6 and cy~lodextrins.~ The vast majority of studies concerning counterion binding to micelles have been performed on surfactants with inorganic counterions, but there is a growing interest in the solution properties of purely "organic" amphiphiles.8-10 The aggregation behavior of the generally larger and more hydrophobic organic counterions can, by different arguments, be anticipated to be different from that of small and hydrophilic inorganic counterions and is therefore of considerable interest to investigate experimentally. In a previous paper3 Stilbs and Lindman briefly reported the aggregation of decylammonium dichloroacetate in water. The present communication reports a more extensive and comparative FT-PGSE 'H N M R study on D20solutions of decylammonium acetate, chloroacetate, and dichloroacetate. The aim of this communication is to compare the association behavior of the three surfactants in terms of the counterion binding, the cmc, and the free amphiphile concentration above cmc. Measurements were also performed on mixtures of the different counterions in order to directly compare the ion binding in a competitive situation.
Experimental Section Materials. Sample Preparation. The decylammonium acetate, chloroacetate, and dichloracetate amphiphiles were synthesized (5) Lindman, B.; Nilsson, P.-G.;Stilbs, P.; WennerstrBm H. Pure Appl. Chem. 1984, 281. ( 6 ) Stilbs, P.; Arvidson, G.; Lindblom, G. Chem. Phys. Lipids 1984, 35, l O A.. _-
(7) RymdCn, R.; Carlfors, J.; Stilbs, P. J. Inclusion Phenom. 1983, I , 159. (8) Underwood, A. L.; Anacker, E. W. J. Colloid Interface Sei. 1984,100,
by neutralizing aqueous solutions of acetic acid (EGA), chloroacetic acid (Grave), and dichloroacetic acid (EGA) with decylamine (Merck, all chemicals of puriss quality), added dropwise with stirring. The water was removed by freeze drying, and the raw product was then recrystallized in a mixture of water and ethanol, and finally stored in a desiccator under vacuum. Samples were prepared directly in precision 5-mm N M R tubes mostly through incremental dilution with D20.Trace amounts of the highly hydrophobic and essentially completely micellarly solubilized hexamethyldisiloxane were added in order to quantify the actual self-diffusion coefficient of the micelle at each studied concentration (see next subsection). Methods. The measurements were performed on protons at 99.6 MHz at a temperature of 38 f 0.5 "C (the reason being that the Krafft point is about 25 OC for decylammonium dichloroacetate) on a standard JEOL FX-100 Fourier transform N M R spectrometer equipped with a 5-mm proton probe, operating at 2.3 T using procedures described earlier.11,12 The magnetic field gradients were generated through the standard modulation coils at the probe (connected antiphase) and were calibrated against literature values of self-diffusion coefficients of trace amounts of protons in heavy water. An example of a self-diffusion measurement on a mixture of decylammonium acetate, chloroacetate, and dichloroacetate from the competitive ion binding experiments described below is shown in Figure 1. The monitoring of the association process is based on the difference in intrinsic selfdiffusion coefficients for the amphiphile monomers, counterions, and micelles. The translational mobility of an amphiphile or a counterion is considerably reduced (1-2 orders of magnitude) when diffusing with the micelle. The
128.
(9) Underwood, A. L.; Anacker, E. W. J. Phys. Chem. 1984,88, 2390. (10) Broxton, T. J. Ausr. J . Chem. 1981, 34, 2313.
(11) Stilbs, P.; Moseley, M. E. Chem. Scr. 1980, 15, 176. (12) Stilbs, P.; Moseley, M. E. Chem. Scr. 1980, 15, 215.
4870
Jansson and Stilbs
The Journal of Physical Chemistry, Vol. 89, No. 22, 1985
TABLE I: Measured Self-Diffusion Coefficients and Deduced Quantities of Decylammonium Acetate in D20 concn, M 0.029 0.040 0.057 0.070 0.095 0.111 0.120 0.125 0.142 0.180 0.200 0.220 0.250
D+
D-
D.
Cmict,M
CmiL,M
crre+, M
P
0.82 f 0.03" 0.82 f 0.03 0.82 f 0.03 0.70 f 0.02 0.54 f 0.01 0.50 f 0.02 0.492 f 0.01 0.474 f 0.01 0.414 f 0.01 0.321 f 0.006 0.309 f 0.004 0.267 f 0.005 0.222 f 0.005
1.29 f 0.03 1.30 f 0.03 1.29 f 0.03 1.15 f 0.02 0.98 f 0.01 0.95 f 0.02 0.954 f 0.01 0.944 f 0.01 0.867 f 0.01 0.800 f 0.006 0.803 f 0.004 0.755 f 0.006 0.693 f 0.006
0.17 f 0.01 0.15 f 0.01 0.147 f 0.01 0.130 f 0.01 0.130 f 0.01 0.122 f 0.007 0.124 f 0.006 0.111 f 0.006 0.1 13 f 0.005
0.040 0.052 0.058 0.062 0.083 0.128 0.146 0.171 0.21 1
0.027 0.034 0.036 0.038 0.05 1 0.076 0.084 0.101 0.128
0.055 0.058 0.062 0.063 0.059 0.052 0.054 0.049 0.039
0.66 0.65 0.63 0.62 0.62 0.60 0.58 0.59 0.61
Error limits correspond to 80% stastical confidence intervals, regarding random errors only. bAll self-diffusion coefficients are given in units of m2 s-l. TABLE 11: Measured Self-Diffusion Coefficients and Deduced Owntities of Decylammonium Chloroacetate in D20 concn, M 0.022 0.029 0.040 0.044 0.057 0.067 0.080 0.086 0.111 0.120 0.139 0.143 0.179 0.200 0.250
D+
D-
D,
Cm,c+, M
CmiL, M
CfI,+, M
P
0.78 f 0.03' 0.80 f 0.03 0.72 f 0.02 0.69 f 0.03 0.59 f 0.03 0.52 f 0.02 0.46 f 0.01 0.37 k 0.01 0.343 f 0.007 0.325 f 0.01 0.285 f 0.008 0.253 f 0.003 0.208 f 0.002 0.200 f 0.002 0.162 f 0.001
1.18 f 0.03 1.16 f 0.03 1.12 f 0.03 1.08 f 0.02 0.94 f 0.02 0.90 f 0.01 0.80 f 0.01 0.74 f 0.01 0.715 f 0.01 0.681 f 0.01 0.644 f 0.008 0.604 f 0.005 0.567 f 0.005 0.561 f 0.004 0.536 f 0.003
0.12 f 0.02 0.12 f 0.03 0.12 f 0.01 0.10 f 0.01 0.114 f 0.008 0.101 f 0.007 0.100 f 0.007 0.102 f 0.005 0.091 f 0.003 0.093 f 0.004 0.082 f 0.002
0.018 0.028 0.040 0.054 0.074 0.082 0.103 0.1 12 0.150 0.170 0.222
0.01 3 0.017 0.029 0.035 0.048 0.055 0.069 0.076 0.101 0.1 14 0.147
0.039 0.039 0.039 0.032 0.036 0.038 0.036 0.031 0.029 0.030 0.027
0.71 0.63 0.71 0.65 0.65 0.67 0.68 0.67 0.67 0.67 0.66
Error limits correspond to 80% statistical confidence intervals, regarding random errors only. bAll self-diffusion coefficients are given in units of m2 s-I
(partial) binding process is therefore reflected in a sensitive manner in the time-averaged self-diffusion coefficient of the bound species. Since there is a fast exchange between amphiphiles or counterians associated with the micelles and the aqueous phase on the N M R time scale the following two-site model applies
By rearranging eq 1 one finds for the amount of micellized amphiphile, C,,,', the amount of associated counterions, C,,;, and the amount free amphiphiles, Cfree+that Cmicf = Ctot(D+O- D+)/(D+O - DmiJ
(2) (3)
where DOMrepresents the time-averaged diffusion coefficient of the species (amphiphile, solubilizate, or counterion), p s the fraction of micellarly bound species, D,,, the diffusion coefficient of the micelle, and Dfreethe diffusion coefficient of the "free" species in the aqueous phase. As mentioned in the preceding subsection, the actual monitoring of D,,, was made by measuring the self-diffusion coefficient of added hexamethyldisiloxane (HMDS), assuming complete micellar solubilization. (It was noted that the spin-echo signal of the solubilizate vanishes completely below cmc, as a result of desolubilization and transport to the solution/air interface.) One might argue that, due to incorporation of the solubilizate HMDS, the micelle size is slightly large for the HMDS-containing micelles as compared to the micelles without HMDS, and therefore the measured diffusion coefficient could be somewhat lower than the actual D,,,.This will not cause serious calculation errors since the evaluated parameters are rather insensitive to changes in D,,, however. The two-site model certainly has limitations. When applied to counterions it only differentiates between associated and nonassociated counterions (within that model framework), which is a crude simplification if we consider the continuous distribution of counterions in the vicinity of a charged micellar surface or other polyelectrolyte-like systems. Nevertheless, the model successfully describes various similar association phenomena.'-2.4.'3 (13) Lindman, B.; Puyal, M.-C.; Kamenka; N.; Brun, B.; Gunnarsson, G. J . Phys. Chem. 1982, 86, 1702.
(4) rspectively, where D+O and D? are the self-diffusion coefficients of the free amphiphile and the free counterion, measured above cmc, and D+, D-, and D,the observed self-diffusion coefficients of the decylammonium ion, the counterion, and the hexamethyldisiloxane molecule. The values of C,,,' and C,,,- are determined by the degree of association to the micelle of the amphiphile and the counterion, respectively. In order to quantify the degree of association of the counterion to the micelle, we introduce the degree of counterion binding, /3, defined as Cmic-/Cmic+.
Results The self-diffusion coefficients of the dwylammonium ion (D+), the counterion (D-), and the solubilizate (D,) were determined as a function of surfactant concentration (in D20). C,,,', C,,;, Cfree+,and p were calculated as described in the Experimental Section. Tables 1-111 summarize the experimental results of the self-diffusion measurements on decylammonium acetate, monochloroacetate, and dichloroacetate. Competitive experiments were also performed on decylammonium micelles with two or three counterions present. Measurements were made on all possible binary and ternary combinations of acetate, chloroacetate, and dichloroacetate. The pertinent parameters are summarized in Table IV. An error propagation analysis on the evaluated values of 0 and Cfre+indicates that D+O and D? are the most critical parameters when evaluating @ and Cf1,,+via the present measurement ap-
Organic Counterion Binding to Micelles
The Journal of Physical Chemistry, Vol. 89, No. 22, 1985 4871
TABLE III: Measured Self-Diffusion Coefficients and Deduced Quantities of Decylammonium Dichloroscetate in D 2 0 concn, M
Dt
D-
0.019 0.024 0.034 0.045 0.057 0.060 0.078 0.080 0.108 0.109 0.141
0.78 f 0.03' 0.79 f 0.03 0.64 f 0.01 0.50 f 0.01 0.42 f 0.008 0.36 f 0.01 0.31 f 0.01 0.31 f 0.01 0.225 f 0.005 0.202 f 0.004 0.170 f 0.005 0.138 f 0.005 0.136 f 0.003 0.106 f 0.002
1.15 f 0.03 1.16 f 0.03 0.96 k 0.01 0.78 f 0.01 0.71 f 0.01 0.62f 0.01 0.58 f 0.01 0.56 f 0.01 0.483 f 0.008 0.438 f 0.005 0.422 f 0.005 0.398 f 0.005 0.409 f 0.002 0.371 f 0.002
0.171
0.197 0.239 a
D,
Cm,s+,
0.14 f 0.02 0.11 f 0.01 0.10 f 0.008
M
Clnkc-,
0.009 0.019 0.031 0.036 0.053 0.054 0.083 0.086 0.117 0.146 0.174 0.218
0.067 f 0.008 0.073 f 0.007 0.072 f 0.008 0.048 f 0.005 0.035 f 0.002 0.036 f 0.002 0.025 f 0.002 0.045 f 0.005 0.039 k 0.005
M
0.006 0.016 0.024 0.029 0.041 0.043 0.065 0.070 0.092 0.114 0.132 0.168
8
M
Cfree',
0.025 0.025 0.026 0.024 0.025 0.025 0.025 0.024 0.024 0.024 0.023 0.021
0.73 0.80 0.77 0.80 0.78 0.80 0.79 0.81 0.79 0.78 0.76 0.77
Error limits correspond to 80% statistical confidence intervals, regarding random errors only. All self-diffusion coefficients are given in units of mz s-l.
TABLE IV: Measured Self-Diffusion Coefficients and Deduced Quantities of Mixtures of Decylammonium Acetate, Chloroacetate, and Dichloroacetate counterions acetate chloroacetate acetate dichloroacetate chloroacetate dichloroacetate acetate chloroacetate dichloroacetate
(1)
(2) (3)
(4)
a
concn, M
D+ 0.203 f 0.01" 0.203 f 0.01 0.144f 0.003 0.144f 0.003 0.110 f 0.001 0.110 f 0.001 0.149 f 0.008 0.149 f 0.008 0.149 f 0.008
0.120 0.120 0.120 0.125 0.120 0.122 0.085 0.085 0.086
D-
D.
Cmr+,M
0.716 f 0.01 0.454f 0.01 0.743 f 0.004 0.261 f 0.001 0.508 f 0.002 0.276 f 0.001 0.790 f 0.01 0.483 f 0.01 0.268 & 0.01
0.096 f 0.02 0.096 f 0.02 0.054 f 0.002 0.054 f 0.002 0.038f 0.001 0.038 f 0.001 0.071 f 0.008 0.071 i 0.008 0.071 f 0.008
0.102 0.102 0.105 0.110 0.109 0.111 0.076 0.076 0.077
CA;.
M
0.058 0.080 0.047 0.101 0.071 0.096 0.030 0.053 0.070
8 0.57 0.79 0.44 0.92 0.65 0.87 0.39 0.70 0.91
Error limits correspond to 80% statistical confidence intervals, regarding random errors only. All self-diffusion coefficients are given in units of
m2s-l. 1.0
I I
I
0.9
d
I Lo
N E
?
5
0.7
0.8
1
i
e 0.7
0.6
0.01 0.0
"
"
0.1
"
"
I
0.2
[C10H21NH3CH2C1C00-]
'
"
'
I
0.3
i n D20
Figure 2. Self-diffusion coefficients of chloroacetate ( O ) , the decylammonium ion (A),and hexamethyldisiloxane (0), as a function of total surfactant concentration at 38 "C.
proach.14 As seen from eq 2 and 3, C,+ and Cm; (and therefore also /3 and Cfre+)are particularly sensitive toward small changes in D+O and D-O a t low surfactant concentrations. This is due to numerical uncertainty in (D+O- D+) and (D-O - D-), respectively, become increasingly in the low concentration range. 3/ and Cfree+ insensitive to variations in D+O and D? in the high concentration range. This is manifested in Figures 3 and 5 as an increased experimental scatter of @ with respect to Cfrcc+toward lower concentrations.
Discussion Figure 2 displays the measured quantities of decylammonium chloroacetate, illustrating the established component self-diffusion (14)The values of D+" and D-O are taken from the low concentration data. D-O for acetate, chloroacetate, and dichloroacetae were determined to be 1.30 X and 1.16X lo4 m2 s-', respectively, white D+O for the 1.18 X m* s-'. decylammonium ion was 0.81 X
0.5
0
I
I
0.1
0.2
[SURFACTANT] M
in
3
D20
Figure 3. Evaluated 8-values for dichloroacetate (A),chloroacetate (o), and acetate ( 0 ) counterions, as a function of surfactant concentration. The lines are only intended as an aid to the eye. The &values at the lowest surfactant concentrations are more uncertain than those at higher concentrations, as discussed in the text.
behavior of an ionic surfactant. Self-diffusion coefficients of the decylammonium ion and the counterion are relatively constant below cmc and decrease above cmc with increasing concentration, in line with previous observations (see, e.g., ref 1 and 2). Within the two-site model framework this is explained in terms that the translational mobility of an ion or a molecule diffusing with the micelle is considerably reduced and therefore the time-averaged value (the observed quantity) decreases, in proportion to the fraction bound (cf. eq 1). A slight decrease of D, with increasing concentration is evident in the data of Figure 2, in line with previous observations on micellar systems.'-3 This is usually explained with electrostatic micelle-micelle repulsion^.'^ (15) Mazo, R.M.J Chem. Phys. 1965,43, 2873.
4872 The Journal of Physical Chemistry, Vol. 89, No. 22, 1985
d
m
N E
m
I 0 d
0.5
1
1
Jansson and Stilbs
0.06
i
1
\ O+
0.0' 0.0
"
"
I
20.0
'
'
"
I
40.0
[SURFACTANT]-'/
M-'
'
'
'
'
J
0.00
i n 020
Figure 4. Self-diffusion coefficient of the decylammonium ion with acetate (A), chloroacetate ( 0 ) , or dichloroacetate ( O ) , as a function of the inverse surfactant concentration. The two straight lines intersect at
cmc, according to the phase separation model.
The Degree of Counterion Binding. Figure 3 illustrates the evaluated @-valuesfor the three counterions at different surfactant concentration. The invariance of counterion association over a large concentration range, established for several systems,I6l8 is also observed here. The @-valuesare distinctly different, however. The average value of @ for acetate is close to 0.60, for chloroacetate 0.67, and for dichloroacetate 0.78. For a full qualitative interpretation of the p-values one must map out all possible sources of ion-micelle attraction. In the present case there are differences in both counterion polarizability and hydrophobicity and further investigation is therefore required. A systematic study of micellar binding of organic counterions is now underway in our laboratory. It is interesting to note that the largest counterion, dichloroacetate, is the most strongly bound. For inorganic counterions it is established that a smaller hydrated radius is correlated with a higher degree of a s s o c i a t i ~ n . ~In ~ *a~simple ~ model this is explained by the ability of smaller ions to create more compact electric double layers. The results indicate that the organic counterion binding is not affected by counterion size in this way. This is in line with previous findings of Mukerjee et al.z' when studying the association behavior of tetraalkylammonium ions with dodecyl sulfate micelles. A recent study by Underwood and Anackers on decyltrimethylammonium micelles with inter alia the same counterions as in the present communication indicates a rather different counterion association behavior. The acetate ion was found to be the most associated, in contrast to our findings on decylammonium micelles in the present paper. The investigated systems, as well as the underlying method (based on light scattering), are different, and the subject needs further investigation to clarify this point. Experiments were also performed in order to study competitive effects on the association behavior of the counterions. As a general net result of the various ion-binding processes in a competitive situation the preferential binding of the most associated counterion in the single-ion case is amplified even further. The results are summarized in Table IV. When comparing the degree of association in Tables 1-111 with those in Table IV we find in all four mixed systems effects of this kind. the most pronounced ionsegregation effect is found in the acetate/dichloroacetate system; for acetate is reduced from 0.60 (in the single ion situation) to (16) Zana, R. J. J . Colloid Interface Sci. 1980, 78, 330. (17) Khan, A.; Sderman, 0.;Lindblom, G . J. Colloid Interface Sci. 1980, 78, 217. (18) Vikingstad, E. J Colloid Inferface Sci. 1980, 73, 260. (19) Wennerstrom, H.; Lindman, B. Phys. Rep. 1979, 52, 1. (20) Lindman, B.; Wennerstrom, H. Top. Curr. Chem. 1980, 87, 1 (21) Mukerjee, P ; Mysels, K. J.; Kapaun, P. J . Phys Chem. 1967, 71, 4166
I
0.0
60.0
I
0.1
,
0.2 [SURFACTANT] M i n 020
0.3
Figure 5. Concentration of the free decylammonium ion with acetate ( O ) , chloroacetate (O), or dichloroacetate(A) as surfactant counterion, as a function of surfactant concentration. The lines are only intended as an aid to the eye. The C,' at the lowest surfactant concentrations are more uncertain than those at higher concentrations,as discussed in
the text. 0.44, and for dichloracetate @ is raised from 0.78 to 0.92. The p-value for acetate is even further reduced when both dichloroacetate and chloroacetate are present. Cmc Effects. The phase separation model is an useful first approximation for micelles but deviations from it are well documented." According to this model, a plot of D+ vs. C;l, will give two straight lines intersecting at the cmc. Figure 4 illustrates that the data are in good qualitative and quantitative aggreement with this model, giving a cmc of 0.061 M for decylammonium acetate, 0.043 M for decylammonium chloroacetate and 0.027 M for decylammonium dichloroacetate. There is a strong correlation between the degree of counterion binding and the cmc for the counterions. The self-consistent qualitative explanation of this phenomena is that a higher degree of counterion binding to the ionic micelle reduces to a higher extent the repulsive forces between the hydrophilic parts of the molecules. The micelle configuration is therefore stabilized relative to a solution of free surfactants and micelle formation occurs at lower concentrations. Effects on Cf1,+. As predicted theoretically already by Hartley in 193522and in line with several recent experimental investigations on ionic s ~ r f a c t a n t s , ' ~the ~ *free ~ ~ ~surfactant ~' concentration shows a maximum around cmc and decreases with increasing surfactant concentrations. A qualitative phenomenological description is based on an entropic r e a ~ n i n g . ' ~ ~There . ' ~ is an uneven counterion distribution in the solution as the counterions are associated to a high degree with the micelle, resulting in a negative entropic contribution to the micelle formation. As the surfactant concentration is increased above the cmc the concentration close to the micelles is nearly constant ( p is constant) while the counterion concentration in the solution is steadily increased due to counterion dissociation. The negative entropy contribution then becomes reduced and the micelles are stabilized relative to the free surfactant ions, giving lower Cfm+at higher surfactant concentrations. These arguments are nicely demonstrated with the organic counterions studied in the present paper. The diverse p-values, caused by dissimilar counterion polarizability or hydrophobicity, lead to different amounts of dissociated counterions at a specific surfactant concentration (if inorganic counterions were used, the effect would not be so large since the p-values are rather independent of the counterionZ0). Figure 5 illustrates how CfI,,+ decreases more rapidly for acetate than for chloroacetate and dichloroacetate, and also more rapidly for chloroacetate relative (22) Hartley, G. S. 'Aqueous Solutions of Paraffin Chain Salts"; Hermann: Paris, 1936. (23) Gunnarsson, G.; Jonsson, B.; Wennerstrom, H. J . Phys. Chem. 1980. 84, 3114.
J. Phys. Chem. 1985,89,4873-4875 to dichloroacetate. The degree of counterion dissociation is largest for decylammonium acetate and therefore the negative entropy contribution is reduced to a higher degree when increasing the surfactant concentration. Conclusions It is demonstrated that the Fourier transform NMR self-diffusion technique gives direct and easy access to information on organic counterion binding to micelles even in the presence of
4873
several counterions. The degree of counterion binding, the cmc, and the free amphiphile concentration above cmc were demonstrated to be distinctly different for decylammonium acetate, chloroacetate, and dichloroacetate. In a competitive situation, an amplification of the differences in ion binding was demonstrated. Registry No. Decylammonium acetate, 2016-38-8; decylammonium chloroacetate, 78961-19-0;decylammonium dichloroacetate, 98087-68-4.
Vapor-Phase Dipole Moment Values from Solution Measurements Miiximo B a h t Departamento de Fisica, Facultad de Ciencias Exactas y Naturales, Universidad de Buenos Aires, Pabellh 1 , C. Universitaria, 1428 Buenos Aires. Argentina (Received: March 4, 1985; In Final Form: June 27, 1985)
An equation proposed by H. Mechetti to describe the permittivity of very dilute polar/nonpolar solutions as a function of concentration was used to calculate the molecular dipole moment of the solutes. The procedure is very simple, and the results are in good agreement with those obtained through more elaborate known methods. Furthermore, an adequate choice of solvents leads to results that are in very good agreement with the corresponding vapor-phase values.
Introduction The molecular dipole moment, as a very useful parameter for structural studies, should ideally be calculated from vapor-phase measurements, to avoid any kind of intermolecular interactions. This is only possible with compounds that have a vapor pressure high enough to allow them to be handled as vapors in the measuring devices. Therefore, the number of compounds that can be studied is limited. To overcome this problem, calculations have been proposed based on measurements made on very dilute solutions of the candidate substance in nonpolar solvents. It was assumed that under these conditions there would be no solute/solute or solute/solvent interactions. The first equation proposed by Peter J. W. Debyel in 1921 was followed by others that have been extensively described and discussed in the literatures2 The initial success of such equations was attributed to the absence of interactions. This appeared to be confirmed by the linear behavior, a t very high dilutions, of permittivity (e), density (d) (or specific volume Q, and the refractive index (n) as a function of concentration. However, comparing solution and vapor-phase measurements on low-temperature boiling compounds, we found discrepancies in some cases as high as 30% (chloroform3),although in general they are between 8% and 15% (water, nitromethane, methanol, trichloroethane, and acetone4. This was originally attributed to a so-called solvent effect that has been carefully discussed over the years since 1934; under the more adequate denomination of solute/solvent interactions or solvation.6 The latter appears to be the cause for the differences observed between vapor-phase and solution values. It is a known fact that for dilute binary liquid systems the permittivity (e) is a linear function of concentration. However, due to the above-mentioned interactions, this behavior cannot be described by a simple additive law. This was discussed in a previous paper,’ and as a result, eq I was proposed by H. Mechetti based on a model that proved adequate for the permittivity of dilute solutions of water and lower alcohols in nonpolar solvents: €12
= €1
+ 4*
+ l ) e 1 h 2 + 2)ro2 (nZ2 + 3(2c, + n22)2(el n22)3kT
+
+
Member of the Carrcra del Investigador Ciendfico, CONICET.
0022-3654/85/2089-4873$01.50/0
In this equation, el is the permittivity of the pure solvent, e12the permittivity of the solution, n22the refractive index of the pure solute (squared), a2 the radius of the molecule calculated from the experimental specific volume (q,N2 the solute concentration expressed as the number of molecules per milliliter of solution, po the vapor-phase dipole moment of the solute molecule, k Boltzmann’s constant, and T the temperature. If e l , n2, V2,and e12 are measured, eq I can be rearranged to calculate po:
3(2q
+ n2)(2e1+ nzZ)
J
Equation I1 indicates that vapor-phase values of the molecular dipole moment could be calculated from single solution measurements. The procedure is very simple since no additional calculations are necessary, like permittivity and specific volume slopes or total polarization values, as is the case with other well known methods (i.e,, that of Halverstadt and Kumler*).
(1) Peter W. J. Debye, “Polar Molecules”, Dover Publications Inc., New York, 1921. (2) H. Bradford Thompson, J. Chem. Educ., 43,66 (1966). (3) A. L. McClellan, “Tables of Experimental Dipole Moments”, Vol. 2, Rahara Enterprises, El Cerrito, CA, 1974, p 45. (4) A. L. McClellan, ref 3, pp 25, 42, 42, 64, 82. ( 5 ) F. Horst Mueller, Trans. Faraday SOC.,30, 729 (1934). (6) S.D. Christian, A. A. Taha, and D. W. Gash, Rev. Chem. SOC.,24, 20 (1970). (7) M. Bar6n and H. Mechetti, J . Phys. Chem., 86,3464 (1982). (8) J. F. Halverstadt and W. D. Kumler, J . Am. Chem. SOC.64, 2988 (1942). (9) A. B. Lindenberg, C. R . Seances Acad. Sci., Ser. C, 262,1504 (1966). (IO) M. Barbn, An. Asoc. Quim. Argent., 67,203 (1979). (11) H. Lumbroso and C. Andrieu, Bull. SOC.Chim. Fr., 3201 (1966). (12) A. A. Abdurakhmanov, R. A. Regimova, and L. M. Imanov, Phys. Len.A , 32A, 123 (1970). (13) J. P. de Jongh and H. A. Dijkerman, J. Mol. Specrrosc., 25, 129 (1968). (14) H. Loozenga, Mol. Phys., 9, 501 (1965). (1 5 ) S. Weiss, J . Phys. Chem., 70,3 146 (1 966). (16) R. Holm, M. Mitzlaff, and H. Hartmann, 2.Narurforsch., A: Asrrophys., Phys. Phys. Chem., 23A,307 (1968).
0 1985 American Chemical Society
