Langmuir 1994,10, 1140-1145
1140
Alkanediyl-a,o-bis(dimethylalkylammonium bromide) Surfactants. 4. Ultrasonic Absorption Studies of Amphiphile Exchange between Micelles and Bulk Phase in Aqueous Micellar Solutions M. Frindi and B. Michels Laboratoire d'Ultrasom et de Dynamique des Fluides Complexes (LUDFC), URA 851 du CNRS, Universitk Louis Pasteur, 4 rue Blaise Pascal, F-67070 Strasbourg Cedex, France
H. Levy and R. Zana* Institut Charles Sadron (CRM-EAHP), CNRS- ULP, 6 rue Boussingault, 67083 Strasbourg Cedex, France Received December 6,1993. I n Final Form: February 14,1994' The critical micelle concentrations (cmc's) of two gemini surfactants, the propane- and hexanediyla,@-bis(octyldimethy1ammoniumbromide), referred to as 8-3-8and 8-64,have been determined by spectrofluorometry, ultrasonic absorption, and electrical conductivity measurements. The isothermal volume change upon micellization of 8-6-8 h a been obtained from density measurements. The kinetics of the exchange of 8-3-8 and 8-6-8 between micelles and bulk phase has been investigated by the ultrasonic absorption relaxation method in the concentration range from slightly below cmc to about 2 cmc. All the solutionsinvestigatedshowed that in the frequencyrange 0.5-85 MHz a singlerelaxationprocess is associated with the exchange equilibrium as in the case of conventional surfactants. It therefore appears that the two alkyl chains of gemini surfactante exit from and enter into the micelles simultaneously. The rate constants for these processes are close to those found for conventional cationic surfactants of similar nature and having nearly the same cmc as the gemini surfactants.
Introduction Micelles present in aqueous solutions of amphiphiles at concentrations above the critical micelle concentration (cmc) are not static assemblies. They constantly exchange amphiphiles with the bulk (intermicellar) phase. The kinetics of these exchanges has been extensively investigated' and appears to be well understood on the basis of the theory of micellar kinetics of Aniansson and The analysis of the experimental data on the basis of this theory has consistently yielded for the rate constant k+ of incorporation of an amphiphile in a micelle (Figure 1A) values close to those calculated for a diffusion-controlled process (rate of reaction close to the rate of collision between reactants, here micelle and amphiphile), in the range 3 X 108 to 5 X 109 M-' s-l.l The rate constant k- for the exit of a surfactant from a micelle (Figure 1A) is usually found to depend very much on the hydrophobicity of the amphiphile and to decrease by a factor close to 3 per additional methylene group in the hydrophobic moiety of conventional surfactants: ionic,' nonionic? and zwitterionic.4 In addition to providing information on the micelle kinetics, the data also yield values of the polydispersity in the micelle aggregation number.14 The same kinetic behavior (nearly diffusion-controlled incorporation) was reported for several two-chain surfactants, the dialkylmethylammonium bromides,6 and for a
* To whom correspondence should be addressed.
Abstract published in Advance ACS Abstracter, April 1, 1994. J.; Zana, R. Surfactant Solutiom: New Methods of Investigation; Zana, R., Ed.; Dekker: New York, 1987; Chapter 8. (2) Aniansson, E. A. G.; Wall,S. J. Phys. Chem. 1974,78,1024; 1975, (1) Lang,
79, 857.
(3)Frindi, M.; Michels, B.; Zana, R. J. Phys. Chem. 1991, 95, 4832;
1992,96,6095 and 8137.
(4)Frindi, M.; Michels, B.; Zana, R. Submitted for publication.
(5) Lang, J. J. Phys. Chem. 1982, 86, 892.
A : One step exchange
B : Two-step exchange
Figure 1. Schematic representation of the exchange of a Surfactantbetween micelles and bulk. (A) conventional surfactant;(B)dimeric surfactant. The alkyl chains in the micelle core are not represented for the sake of clarity.
bolaform surfactant, the docosanediyl-a,w-bis(trimethy1ammonium bromide).B The only amphiphiles for which the incorporation was found to be much slower (by 4 orders of magnitude) than for a diffusion-controlled process appear to be the gangliosides, surfactants with two long chains and a very large and chemically complex polysaccharide head group.' The kinetic study was performed by light scattering and requires additional confirmation by an independent method. (6) Zana, R.; Yiv, S.; Kale, K. J. Colloid Interface Sci. 1980, 77,486. ( 7 ) Cantu, L.; Corti, M.; Salin, P. J. Phys. Chem. 1991,95,5981.
0743-7463/94/2410-1140$04.50/00 1994 American Chemical Society
Ultrasonic Absorption Studies of Amphiphile Exchange
Recently we have become interested in a new class of amphiphiles, the alkanediyl-a,@-bis(alky1dimethylammonium bromidela of general chemical formula These com-
CmHzm+ 1
C n h I +1
pounds can be referred to as gemini, or bipolar amphiphiles. They may be called “siamese” surfactants since, like siamese twins, they are made of two identical amphiphilic moieties connected at the level of the head group, here an adjustablepolymethylene chain containings carbon atoms. These surfactants have been represented by the abbre. ~ can formally be considered as viation m - s - m , 2 B ~ They the dimers of the two-unequalchain amphiphiles CmH2,,+l( C , / ~ H , + I ) N + ( C H ~ ) ~The B ~ . “dimeric” surfactants are reminiscent of both conventional (one head groupone alkyl group) and bolaform (two head groups in a! and o positions-one alkanediylgroup) surfactants. At this time we have well characterized the effect of m and s on the cmc and micelle ionization degree,& the behavior at the air-water interface,sbthe microstructure” and aggregation behaviofl in aqueous solution, the formation of lyotropic liquid crystallinephases,&andthe rheology of the aqueous solutionsafof the m-s-m,2Br surfactants. Our program of research includes also the synthesis of the “trimeric” surfactants: m-s-m-s-m,3Br. Several groups have been or are presently involved in the study of various aspecta of similar or different dimeric surfactants.@-19 The dimeric surfactants raise interestingquestionswhen considering the manner in which they exit fromlincorporate into their micelles. Two possible mechanisms are represented in Figure 1B. The one-step mechanism is very similar to that for conventional Surfactantsbut would involve the simultaneousexit of the two alkyl chains, while the two-step mechanism (exit of one chain, then the other) may result in a different kinetic behavior with two characteristic rate constants and relaxation times. This possibility, together with our interest in checking whether the Aniansson and Wall2 model also describes the dynamic behavior of complex surfactants such as the dimeric ones, led us to perform an ultrasonic absorption relaxation study of the exchange process in aqueous micellar solutions of dimeric surfactants. Since the
Langmuir, Vol. 10, No.4, 1994 1141
ultrasonic relaxation method only permits studies of processeswith characteristictimes in the 1ps to 1ne range, we have selected two dimeric surfactants witha short alkyl chain (asin our studies of nonionic surfactants3),the 8-38,2Br and 8-6-8,2Br (the counterions are omitted below), in order to be able to also check the effect of the spacer carbon number on the exchange kinetics. For a complete analysis of the relaxation data, we have determined the cmc of these surfactants by conductivity, pyrene fluorescence probing, and ultrasonic absorption, and the volume change upon micelle formation from densitymeasurements as in similar studies of other surfactants.3 It is shown below that the exchange process is characterized by a single relaxation time and that the results are well explained on the basis of Aniansson and Wall2 theory for micellar kinetics.
Materials and Methods Materials. The surfactanta were synthesized by reacting octyl or hexanediylbromide with propanediyl-l,3-bis(dimethylamine) 1,6-bis(dimethylamine),using a slight excew (about 10%)of the bromide, under reflux in absolute ethanol for 48h. After rotatory evaporation of the ethanol and of the unreacted compounds (under 1 mm vacuum for the latter), the resulting solids were recrystallized 3 times from ethyl acetate-ethanol mixtures. The elemental analysis (C, H, N, Br)and Karl-Fisher determination of the water content confirmed the surfactants to be at least 98% pure. Pyrene was the same as in previous studies.g Methods. The electrical conductivity method was used to determine the cmc of the surfactants and the degree of micelle ionization.2o The conductivities were measured using an autobalanced conductivity bridge (Wayne-Kerr B 905)operated at 1 kHz in conjunction with a Tacussel (France) conductivity cell (type TE 100). The plota of the ultrasonic absorption a/f2 (a = absorption coefficient, f = ultrasonic frequency) versus surfactant concentration, C, at a constant and sufficiently low frequency cf 0.7 MHz), were also used for cmc determination^.^ Finally,the cmc’s were determined by spectrofluorometryfrom the plots of the ratio IJI3 of the intensities of the f i t and the third vibronic peake in the fluorescence emission spectrum of pyrene solubilized at very low concentration (5X 10-7 to 10.8 M) in the surfactant solutions,8 versus C. The fluorescence spectra were recorded using a Hitachi F 4010 spectrofluorometer, at an excitation wavelength of 335 nm and a bandwidth of 1.5 nm both at the excitation and emission. The isothermal volume change upon micellization AV# was obtained from the plot of the surfactant apparent molal volume $V against C? The densities required for obtaining the values of $V were measured with a densimeter Anton-Parr DMA 60. The ultrasonic absorption alp was measured using the same experimental apparatus as in recent studies? The ultrasonic relaxation spectra were analyzed as previously described.3 Ail measurements were performed at 25 “C except for some ultrasonic relaxation measurements which were also performed at 40 O C . The surfactant concentration C is expressed in moles of surfactant per liter. The concentration in alkyl chains is evidently twice larger.
(8) (a) Zana, R.; Benrraou, M.; Rueff, R. Lungmuir 1991,7,1072.(b) Alami, E.;Marie, P.; Beinert, G.; Zana, R. Langmuir 1993,9,1465. (c) Zana, R.;Talmon, Y.Nature 1993,362,228.(d) Zana, R.;Benrraou, M. Unpublished reaultq (e) Alami, E.; Levy, H.; h a , R.; Skoulioe, A. Langmuir 1993,9,940. (0Kern,F.;Lequeux, F.; Zana, R.; Candau, S. Langmuir, in press. (9) Mass, R. A.; Li,J. M. J . Am. Chem. SOC.1992,114,9228. (10)Fuhrhop, J.-H.; Hungerbuhler, H.; Siggel, U. Langmuir 1990,6, 1295,and referencea therein. (11)Kim,J.-M.; Thompson,D.H. Langmuir 1992,8,637. (12)Pinazo, A.; Du,M.; Solane, C.; Pee, M. A.; Erra, P.; Infante, M. Results and Discussion R. JAOCS, J. Am. OiL Chem. SOC.1993,70, 37. (13)Devinsky, F.;Lacko, I.; Imam, T.J . Colloid Interface Sci. 1991, 1. Cmc Determinations. Figure 2 shows the plot of 143,336. Devinsky, F.;Maearora, L.; Lacko, I. J . Colloid Interface Sci. 1986,106,235. the Id13 ratio against C for the two surfactants. These (14)Abid, S.K.;Hamid, S. M.;Sherrington,D. C. J. Colloid Interface plots have the usual sigmoidalshape, with a rapid decrease Sci. 1987,120,245. of 11/13at C slightly below cmc and a leveling off a t high (15)Perreira, H.C.; Lukenbach, E. R.; Lindemann, M. K. JAOCS, J. Am. Oil. Chem. SOC.1979,56,1015. C, corresponding to rather high 11/13 values, about 1.61 (16)Deinega,Y.;Ul’berg,Z.R.;Maro&ko,L.G.;Rudi,V.P.;Deniaenko, and 1.67 for the 8-3-8 and 8-6-8 surfactants, respectively. V. P. Kolloidn. Zh. 1974,36,649. For the sake of comparison we have measured 11/13 in (17)Zhu, Y.P.; Masuyama, A.; Nakatauji, Y.; Okahara, M. J . Jpn. Oil. Chem. SOC.1993,42,86,and referencea therein. aqueous solutions of octyltrimethylammonium bromide (18)Zhu, Y. P.;.huyama, A.; Kirito, Y.; Okahara, M.; h e n , M. P. (OTAB), which can be considered to represent, approxJAOCS, J. Am. Ihl. Chem. SOC.1992,69,626. (19)Rozycka-Roszak,B.;Witek, S.;Pnestalski, S. J . Colloidlnterface (20)Zana, R.J. Colloid Interface Sci. 1980, 78, 330. sci. 1989,i31,181.
Frindi et al.
1142 Langmuir, Vol. 10, No. 4, 1994
c* 7 1
1.9
lot
‘V
t
l.’
1.6
0
2
0 20
40
60
80
100
C(mW
3000
2500
E
v
2000
1
t
60
80
100
120
Figure 4. Variation of the electrical conductivity with the surfactant concentration for 8-3-8(+) and 8-6-8(0). The plot for 8-6-8has been shifted by 2 mS for the sake of clarity.
surfactant
0, 1500
500
)I,1 --
loooI
_,i
20
40
60
80
,
100 120 140
C(mW
Figure 3. Variation of alp with the surfactant concentration for 8-3-8(+) and 8-6-8(0).
imately, the monomer of 8-34, at C = 400 mM, i.e. a concentration well above the cmc of OTAB,20and found a value of 1.56, close to that for 8-3-8. All 11/13 values for OTAB, 8-3-8and 8-6-8are high, indicating that the pyrene is not well shielded from water in the micelles formed by these three surfactants, probably because of their small size. The cmc of each surfactant has been obtained by the usual extrapolation procedure as the concentrations correspondingto the intercept of the rapidly decreasing part of the plot and that nearly horizontal part at high Ce3The cmc values are 57 and 73 mM for 8-3-8 and 8-6-8, respectively. Figure 3 shows that the variation of the ultrasonic absorption a/f2with C for the two surfactants has a sharp break, followed by a steep and nearly linear increase of alp. The extrapolation of this part of the plot to the value of a/f2 for the solvent yields cmc values of 55 and 70 mM for 8-3-8and 8-6-8,respectively, in good agreement with the fluorescence results. The plots in Figure 4 which represent the variation of the electrical conductivity with the surfactant concentration show two breaks for each surfactant. The high concentration breaks yield cmc values of 55 and 73 mM, in good agreement with the values from fluorescence probing and ultrasonic absorption. The first break occurs at a lower concentration of about 20 mM for both 8-3-8 and 8-6-8. In a study of the cmc of dimeric surfactants,” the concentration at the first break was taken as the cmc of the 8-6-8, as the conductivity plots did not extend to sufficiently high concentration. This error is now corrected. The micelle ionization degrees, ai, for the two surfactants obtained as the ratio of the slopes at C < 20
cmcO
cmcb
cmd 53
ai
0.27 73 0.33 0 From theZl/Zaw C plots. b From the a/f2v8 C plots at a frequency close to 0.7 MHz. From the conductivity w C plots. 83-8 8-64
a
0
40
Table 1. Values of the Cmc (mM)of the Investigated Surfactants at 25 O C
2
0
20
C(mW
Figure 2. Variation of IllIa with the surfactant concentration for 8-3-8(+) and 8-6-8(0).
h
I
0
120
57 73
65 70
mM and C > cmc are 0.27 and 0.33 for 8-3-8 and 8-6-8, respectively. An increase of ai with the spacer chain length was also found for the 12+12,2Br surfactants.” At this stage, the interpretation of the first break in the K vs C plots in Figure 4 is speculative. This break is not seen inI1/I3 vs C plots and no excess ultrasonic absorption is observed at C > 20 mM up to a concentration close to the cmc. Therefore it does not correspond to an association between alkyl chains and is more likely due to ion-pairing between one B r a n d the dimethylammoniumhead groups. Such ion-pairingbetween B r and ammonium head groups has been previously assumed, on the basis of conductivity measurements, in aqueous solutions of bolaform surfactants14 and in the analysis of surface tension data for dimeric ~urfactants.13~~~ The cmc values from fluorescence,ultrasonic absorption, and electrical conductivity are listed in Table 1,together with the degree of micelle ionization. 2. Isothermal Volume Change upon Micellization. Figure 5shows that the plot of the apparent molal volumes dv, of 8-6-8, calculated from the density data, at various surfactant concentrations, has the behavior expected for micellar systems. The apparent molal volume at the cmc is 492 f 1cm3/mol. This value is very close to that which can be calculated from the partial molal volume of nonyltrimethylammonium bromide 4vo(NTAB) = 240.7 and the partial molal volume increments 4v0(H) = 10.7 cm3/mol and 4vo(CH2) = 15.9 ~m~/mo1.~3 Indeed one can write 4Vo(8-6-8) = 24,O(NTAB) + 24$(CH,) - 24,’(H) = 491.8 cm3/mol (1) The usual equation3 +v(8-6-8) = 4v,mc+ AVTo(C- cmc)/C
(2)
has been fitted to the data at C > cmc in order to extract (21)Devinsky, F. Privata communication.
(22) De Lisi, R.; Ostiguy, C.; Perron,G.; Deenoyers, J. E. J. Colloid Znterface Sci. 1979, 71, 147. (23)Perron, G.;Desnoyers, J. E. Fluid Phase Equilib. 1979, 2, 239.
Langmuir, Vol. 10, No. 4,1994 1143 Table 2. Values of A, B, and fa for the Solutions
502
Investigated
-
500
-
498
-
C (mM)
496
-
50 55
494
-
492
-
8-34
h
I
z E % 2
lo1'% (m-l sz) 24 24 25 25 26 25 26
60
B
70 80 90 100 I
490
1 O W (m-182) a b
C(mM) 60 70
23 25 26 26 27 27 28
80
f 1500 E
90 100 110 120
I '2
a
1016A (m-1st)
1 V f (Hz) ~
78 990 2590 2890 3495 3190 3450
1.10 0.45 0.67 0.92 1.12 1.41 1.61
8-6-8 101sA (m-182)
b
a 73 390 880 1315 1625 1620 1650
19.7 19.4 18.4 19.3
92 258 345 510
1 V f (Hz) ~ a b 5.10 1.53 1.70 2.11 2.48 2.81 3.10
4.37 4.67 4.83 5.08
*
Results at 25 O C . Resulta at 40 O C . 6
Y
5 I
4
I
h
0
L
0.1
10
1
100
; = 3
v
c
2
f(MHz)
Figure 6. Ultrasonicabsorption spectra for a 100 mM solution
of 8-6-8 at 26 O C (0)and 40 O C (X). The solid lines represent the best fits of eq 3 to the data, using a least-squares procedure.
the value of AVTO, isothermal volume change upon micellization. The least-squares fit of eq 2 to the data using cmc = 69 mM and = 492 cm3/mol yielded AVTO = 11 f 2 cm3/mol. This volume is to be compared to the reported volume change upon micellization of octyltrimethylammonium bromide (OTAB), "monomer" of 8-6-8: AV&OTAB) = 3.8 f 1cms/mo1.24 Thus, the AVTOvalue for 8-6-8 is about twice that for its monomer. 3. Ultrasonic Absorption Relaxation. The ultrasonic relaxation spectra (plots of alp vs f , have been determined a t concentrations ranging between 50 and 100 mM for 8-3-8 and 60 and 120 mM for 8-64, a t 25 "C.For 8-6-8 some limited measurements were performed at 40 "C. Figure 6 shows typical data for a 100 mM solution of 8-64at 25 and 40 OC. The curves going through the data represent best fits of the relaxation equation (3), with a single relaxation frequency, f R , to the data, using the reported least-squares procedure.3 In eq 3 A is the relaxation amplitude and B a constant generally close to the value of alf2 of the solvent (3)
The 40 "C spectrum clearly shows the expected low frequency plateau (at f
