J. Phys. Chem. 1992,96,8979-8982 (22) Diehl, P.; Kcllerhals, H.; Lustig, E. NMR Basic Principles and Progress; Diehl, P., Fluck, E., Kosfeld, R., Eds.; Springer-Verlag: Berlin,
1912. (23) Aelion, R.; Loebcl, A.; Eirich, F. J. Am. Chem. Soc. 1950,72,5705.
8979
(24) Babushkin, V. I.; Matveyev, G. M.; Mchedlov-Petrossyan, 0. P. Thermodynamics of Silicates; Springer-Verlag: Berlin, 1985. (25) Matsuyama, 1.; Satoh, S.; Katsumoto, M.; Susa, M. J . Non-Crys?. Solids 1991, 135, 22.
Electrochemlcal Measurements on the Blndlng of Sodium Dodecyl Sulfate and Dodecyltrimethylammonium Bromide wlth a- and B-Cyciodextrlns W. M. Z. Wan Yunus, J. Taylor, D. M. Bloor, D. G.Hall, and E. Wyn-Jones* Department of Chemistry and Applied Chemistry, University of Salford, Saljord M5 4WT, U.K. (Received: April 8, 1992; In Final Form: June 8, 1992)
The binding of ionic surfactants (S)to a-and 8-cyclodextrins (CD) has been investigated using surfactant-selectiveelectrodes. These electrochemical measurements have shown that S(CD) and S(CD)2inclusion complexes are formed between sodium dodecyl sulfate and both a-and /3-cyclodextrinsand also between dodecyltrimethylammonium bromide and a-cyclodextrin. On the other hand, the cationic surfactant only forms a 1:l complex with 8-cyclodextrin. From the data the equilibrium binding constants for the formation of each of the complexes have been evaluated.
Introduction Cyclodextrins are cyclic carbohydrates consisting of 6, 7, or 8 glucose units respectively called a-,8-,and 7-cyclodextrin.' To a first approximation they can be regarded as cylinders with a hydrophilic exterior and a hydrophobic interior. In aqueous solutions, the insertion of a hydrophobic guest into the cyclodextrin molecule results in complexation in which no covalent bonds are formed.14 Surfactants are ideal guests which allow a systematic study of complexation with cyclodextrins since both their hydrophobic and hydrophilic moieties can be systematically changed.s-'8 In this paper, we report a study of the equilibrium properties of complexes formed between a-and 8-cyclodextrins with the surfactants sodium dodecyl sulfate (SDS) and dodecyltrimethylammonium bromide (DTAB). A literature survey reveals that, after an initial period that p r o d u d sometimes ambiguous and inconclusive information,s-18it is now generally regarded that the existence of 1:l and 2:l cyclodextrin/surfactant complexes has been confirmed. Despite this progress, there has been a paucity of data concerning the determination of the actual individual equilibrium binding constants for both steps in the formation of these complexes. Indeed, as far as we are aware, only one publication has been rep~rted'~ in which both equilibrium constants have been evaluated. The main reason concerning the lack of quantitative information on these systems is that most of the techniques that have been used for studying surfactant cyclodextrin complexes in the past have used indirect methods or techniques that essentially measure the macroscopic properties of the solution (e.g., conductivity). As we have shown previously on systems involving cationic surface active drugs and a-cyclodextrins, one of the key parameters that is required in order to understand binding of this kind is to be able to measure the monomer guest concentration in a formulation consisting of cyclodextrin and guest."*' Recently we have been extremely successful in the application of surfactant-selective membrane electrodes to investigate the equilibrium properties of aqueous solutions of surfactants containing various additives.'g-21 The advantage of these electrodes is that monomer concentration of surfactants can be monitored directly in these various formulations. We report here on studies involving emf measurements of surfactant electrodes selective to SDS and DTAB in their complexation with both u- and B-cyclodextrins.
Experimental Section 1. Electrodes. The surfactant-selectivemembrane electrodes used in the present work were constructed using a method which has been described previously.22-2sThe membrane comprises a specially conditioned poly(viny1 chloride) and a commercially available polymeric plasticizer. For the anionic surfactants the poly(viny1 chloride) used in the present work contains positively charged groups; for the cationics the PVC contains negatively charged groups. In order to make membranes selective to SDS and DTAB, the respective poly(viny1 chlorides) are neutralized by the oppositely charged surfactant ions before use. For SDS the monomer surfactant activities in various solutions can be obtained from emf measurements from the following cell
i
surfactant (SDS) test solution containing ommercial bromide selective electrode a constant amount of selective electrode cyclodextrin and [mol dm-3] sodium bromide
*Author to whom correspondenceshould be addressed.
0022-3654/92/2096-8979$03.00/00 1992 American Chemical Society
*,
8980 The Journal of Physical Chemistry, Vol. 96, No. 22, 1992 250
-
1% -I
-50
lo4
10.I
lo3
Wan Yunus et al. 1
102
IO2
10.2
10.1
Figure 1. Plot of emf versus total concentration for SDS in a-cyclodextrin at 25 OC. Concentration of a-cyclcdextrin: (0)0.0 mol dm-); (A) 0.75 X lo-) mol dm-); (0)1.5 X lo-) mol dm-’.
x I8
04 0
0.1
0.2
0.3
0.5
0.4
0.6
0.7
0.8
0.9
1
r
Figure 2. Scatchard plot for DTAB in 0.8 X lo-) mol dm-) fi-cycle dextrin at 25 OC.
Typical data for such experiments are shown in Figure 1. In all cases the electrodes displayed almost perfect Nernstian measurements with slopes of 59 mV/decade. From these data it is possible to evaluate the surfactant monomer concentration ml at each total concentration of surfactant CI for which the measurements were taken. 3. Treatment of Experimental Results. In the first instance we checked whether the data were consistent with only the formation of a 1:l complex. We can test whether the complexation between surfactant (S)and cyclodextrin (CD) only results in a 1:l complex (i.e., the first step (eq 1) in the scheme below) by using the Scatchard equation in the form r / m l = K - Kr where r is concentration of surfactant complexed to cyclodextrin/total concentration of cyclodextrin = (C, - ml)/C,. If r / m l is plotted against r then for a 1:l complex a straight line should result with an intercept on the ‘r’ axis of 1 and slope K . In the present work the behavior only occurs between DTAB and @-CDas shown in Figure 2. In all other cases the data indicate that more than one cyclodextrin molecule binds to a single surfactant ion forming 1:1 and 2 1 complexes according to the scheme K
S + CD d S(CD) S(CD)
+ CD 2S(CD)2
IO’
100
TOM SDS ConcMwtion x lo3 [mol
~ o t a lSDS conmwtim x lo3 [mol dm.31
(1)
(2) This is also in agreement with some recent reports on cyclodextrin/surfactant complexes.+18 As we have said previously, the pertinent question that needs to be addressed in this area is to estimate the equilibrium constants K I and K2 associated with the above scheme. Before proceeding, we make the following assumptions:
IO‘
102
dm“1
Figure 3. Plot of concentration of SDS monomer (mi) versus total concentration of SDS (C,) in a-cyclodextrin. Concentration of a-cyclodextrin: (0)0.75 X lo-) mol d d ; ( 0 ) 1.5 X 10-3mol dm-’; (0)4.0 X lo-) mol dm-); (A) 8.0 X lo-) mol dm-).
(i) Cyclodextrin does not associate in water; i.e., it exists exclusively in monomer form. (ii) There is no interaction between the cyclodextrins and other species in the system, Le., Na+ and Br- ions. Although the cyclodextrins have been known to complex with ionic species, we found no evidence of such association in the present work. This was checked by measuring the emf of a sodium and bromide electrode against a calomel reference electrode. (iii) Each complex only contains one surfactant ion. (iv) The activity coefficients for ionic species of the same valency are the same and hence cancel. In a typical electrochemical experiment we keep the total concentration of cyclodextrin, C,, constant, vary the surfactant concentration C1,and measure each correspding value of monomer surfactant m, with the surfdctant electrode. In schemes 1 and 2 let m, be the monomer (Le., uncomplexed) concentration of cyclodextrin and m’and m”be the equilibrium concentration of the complexes S(CD) and S(CD)2, respectively. Therefore
+
Since CI- ml = m’ m” and C, - m, = m’ + 2m”, then m’ = 2(CI - m l ) - (C, - mc) (4) m” = (C, - m,) - (C, - m , ) (5) Algebraic manipulation of eqs 3, 4, and 5 lead to the following cubic equation in m,: m:KIK2 m,2(KI - K,K2C,- 2C1KIK2)+ m,(l K I C l - K,C,) - C, = 0 (6) and m, follows from
+
+
m , = (C, - C, - m,)/(KIK2m? - 1) (7) In the present analysis we have used a least mean squares computer fitting program using K , and K2 as adjustable parameters and the criteria that we have employed for the “goodness of fit” in the difference between m,calculated via eqs 6 and 7 using known K , and K2and the measured m, value. The procedure
is repeated for all surfactant concentrations measured during a particular experiment and for all the separate experiments in which the total cyclodextrin concentration was varied. Typical fits are shown in Figures 3 and 4. In all cases the experimental data could be adequately described by the twestep mechanism (3). This, however, docs not preclude the existence of further steps in the mechanism; it simply means that in the present &cumstan- the goodness of fit docs not justify further steps being considered. On the other hand, if the fitting procedure was inconsistent with the data we recognize that further analysis involving a different mechanism is necessary. It is very difficult to estimate the errors and uncertainties in the values of K , and K2 determined using the iterative least mean
The Journal of Physical Chemistry, Vol. 96, No. 22, 1992 8981
Binding of Ionic Surfactants to Cyclodextrins
I .4
I f 0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
m, I lo3 [mol dm‘31
Figure 4. Plot of concentration of SDS monomer (ml) versus total concentration of SDS (C,) in 8-cyclodextrin. Concentration of 8-cyclodextrin: ( 0 )0.79 X lo-’ mol dm-’; (0)6.44 X lo-’ mol dm-’; (A) 10.3 X lo-’ mol dm-’.
TABLE I: Equilibrium Constants for SDS and DTAB in a- and &Cy~lode~Mn st 25 O C
a-cyclodextrin K’
(mol-’ dm’) SDS DTAB
2 1 w 17000
8-cyclodextrin
K2 (mol-’ dm’) 18000” 1000
K’ (mol-’ dm’) 2 1 000b 18100
K2
(mol-’ dm’) 21ob
“Quoted stability constant*(KIK2)= 4.9 X lo8 mol-’ dm6. bQuoted K , = 25600 mol-’ dm’; K2 = 200 mol-’ dm3. literature squares p r d u r e described above. The difficulties always arise when dealing with coupled stepwise equilibria. Certainly the uncertainty in K 1is less than for K2;the difficulty, however, is to fix a number to this uncertainty. Hence the results quoted in Table I refer to the best set of KIand K2that fit all the data using a conservative estimate that K 1is accurate to about f25% and K2 to *SO%. At this stage, it is of interest to consider the nature of the complexes and the possible factors which influence their formation. In the absence of direct structural information we consider the following: (i) First, the structures of the cyclodextrins are such that the interior of the cavity is lined with -CHI groups which provide a relative hydrophobic environment thus favoring interaction with the hydrocarbon chain of the surfactant. It is generally regarded that the inclusion of a polar ionic head group in the cavity is unfavorable especially at the concentrations used in this work. X-ray crystallographic studies of some a-CD complexes also confirm t h i ~ . ~ ~In? terms ~ ’ of relative dimensions the internal diameter and depth of the a-CD molecule are 4.5 and 6-7 A, tively, whereas both structural parameters have a value of 7 for 8-CD. Then the volume of the a-CD and 8-CD cavities are 95-1 10 and 270 A3, respectively. The length (L)and volume (V)of a fully extended hydrocarbon chain C,,H2,,+lare estimated to be L (A) = 1.5 + 1.265(n - 1)
‘=r
V (A3) = 27.4
+ 26.9(n - 1)
If a kink is created by gauche/gauche bonds in a hydrocarbon chain the all trans len h is decreased by 1.25 A and the volume is increased by 20-50 per kink. On the basis of the arguments put forward by Park and SongI5 in dealing with geometrical Considerationsof inserting a hydrocarbon chain into a cyclodextrin cavity it follows that 8-CD will accommodate 4 carbon atoms of an all trans chain and 8 carbon atoms of a chain with two kinks. On the other hand, the inclusion of a hydrocarbon chain into the a-CD cavity produces a much tighter fit with a maximum of four carbon atoms for a fully extended chain and 3 carbon atoms for a chain with one kink. When considering sodium dodecyl sulfate these numbers should be decreased by 1 since the ether oxygen
pi‘,
Figure 5. Plot of ratio of complexed a-cyclodextrin to complexed SDS (C,- m,)/Cl - mI versus uncomplexed a-cyclodextrin concentration ( m J , Concentration of a-cyclodextrin: (0) 0.75 X lo-’ mol dm-’; (A) 1.5 X lo-) mol dm-’.
linkage usually behaves like an extra methylene group.IS Therefore, on geometrical consideration alone there are no reasons why both SDS and DTAB should not form 1:l and 1:2 complexes with both cyclodextrins. However, DTAB only form a 1:l complex and therefore other considerations, notably steric and dipolar repulsions, must be considered. When the 1: 1 complex is formed the cyclodextrin molecule will presumably be located in a position along the hydrocarbon chain that results in minimum interactions. The formation of a 2:1 complex will involve a second CD molecule stacking “end on” with the first CD creating a channel to accommodate the hydrocarbon chain. In some cases this might be accompanied by the first CD molecule being pushed nearer the charged hydrophilic head group, creating repulsive dipolar interactions not only between one end of the first CD molecule and the bulky hydrophobic head group but also between the two CD molecules on the chain in the complex. On the basis of geometrical considerations there is always a tight fit when a hydrocarbon chain is inserted into an CY-CD cavity. This of course would enhance the hydrophobic interaction. On the other hand, a fully extended chain has considerable freedom of movement in 8-CD. If one or two kinks are inserted into the chain, the fit gets tighter as expected. In DTAB/&CD the repulsive forces dominate to an extent that the 1:2 complex is unfavorable. The uncertainties in the values of the respective equilibrium constants preclude any further detailed discussion of the balance of forces in these complexes. (ii) It has now been established that once inclusion of surfactant occurs, a considerable amount of counterion bindng takes place, leading to the formation of ion-pair inclusion complexes.28 It has been brought to our attention29 that the formation of the 2:l complex could occur as a result of a second cyclodextrin molecule including the ion pair and terminating the possibility of further inclusion complexes-a result that is consistent with the present and other experimental observations.28At this stage the above suggestion connected with ion pairs raises two queries: (a) &CD has been found not to form a 2:l complex with DTAB. On the other hand, this is the case with CY-CD, which has a smaller cavity. If the addition of the second CD was associated with the head group, then both a- and 8-CD should show identical stochiometry. (b) K2has been found to decrease substantially with decreasing chain length for a homologous surfactant series.’s This observation is not expected if K2 is purely associated with a head-group inclusion process involving ion-pair formation. However, as we have said previously, the answer to these questions can only be solved by direct structural studies. Alternatively, more systematic studies involving variation in chain length and head group of the surfactants would also be useful. Recently, Park and Songls who investigated a series of sodium alkyl sulfates binding with 8-CD using a fluorescence probe method found that their value of K2was dependent on the concentration of 8-CD used. In an attempt to explain this discrepancy
J. Phys. Chem. 1992, 96, 8982-8988
8982
they suggested other complexation products such as a dimer of the 1:l complex formed with or without a surfactant monomer. Another explanation has invoked a 3: 1 complex. We do not believe that such complexes are formed in this work. For example, if we plot the ratio (bound cyclodextrin)/(bound surfactant) against free cyclodextrinas shown in Figure 5 it is clear that the maximum value of this ratio is always less than 2 thus indicating that only 1:1 and 1:2 complexes are formed.
Acknowledgment. W.M.Z.W.Y. thanks the UK Commonwealth Scholarship Commission for a Commonwealth Fellowship. We also thank Unilever Ltd. for financial support and a maintenance grant (J.T.). Registry No. SDS-aCD (1:2), 143545-39-5;SDS-@CD (1:2), 143545-40-8;DTAB-aCD (1:2), 143547-76-6;DTAB-@CD ( l : l ) , 143509-50-6.
References and Notes (1) Bender, M. L.; Koyugana, M. Cyclodextrin Chemistry; SpringerVerlag: New York, 1987. (2) Fendler, J. H. Membrane Mimetic Chemistry; Wiley-Interscience: New York, 1982;pp 194-201. (3)Griffiths, D. W.; Bender, M. L. Adu. Catal. 1973, 23, 209. (4)Cramer, E.;Saenger, W.; Spatz, H-Ch. J . Am. Chem. SOC.1967.89, 14. ( 5 ) Okubo, T.; Kitano, H.; Tse, N. J . Phys. Chem. 1976, 80, 2001. (6) Satake, I.; Ikenoue, T.; Takeshita, T.; Hayakawa, K.; Maeda, T. Bull. Chem. SOC.Jpn. 1985,58, 2146. (7)Satake, I.; Yoshida, S.; Hayakawa, K.; Maeda, T.; Kusomoto, Y. Bull. Chem. SOC.Jpn. 1986, 59, 3991.
(8) Hersey, A.; Robinson, B. H.; Kelly, H. C. J . Chem. Soc., Faraday Trans. I 1986, 82, 1271. (9) Georges, J.; Desmettre, S. J . Colloid Interface Sei. 1987, 118, 192. (10)Palepu, R.; Reinsborough, V. C. Can. J . Chem. 1988, 66, 325. (1 1) Jobe, D. J.; Verrall, R. E.; Palepu, R.; Reinsborough, V. C. J . Phys. Chem. 1988, 92, 3582. (12) Palepu, R.; Richardson, J. E.; Reinsborough, V. C. Langmuir 1989, 5, 218 (13)Okubo, T.; Maeda, Y.; Kitano, H. J . Phys. Chem. 1989, 93, 3721. (14)Saint Aman, E.;Serve, D. J . Colloid Interface Sei. 1990, 138, 365. (15) Park, J. W.; Song,H. J. J . Phys. Chem. 1989, 93,6454. (16)Smith, V. K.; Ndou, T. T.; de la Pena, M. A.; Warner, I. A. J . Inclusion Phenom. Mol. Recognit. Chem. 1991, 10, 471. (17) Lavandier, C. D.; Pelletier, M. P.; Reinsborough, V. C. Aust. J . Chem. 1991, 44,457. (18) Palepu, R.; Reinsborough, V. C. Can. J . Chem. 1989, 67, 1550. (19) Takisawa, N.; Hall, D. G.; Wyn-Jones, E.; Brown, P. J . Chem. Soc., Faraday Trans. I 1988.84, 3059. (20) Thomason, M. A.; Mwakibete, H.; Wyn-Jones, E. J . Chem. Soc., Faraday Trans. 1990,86, 15 1 1 . (21) Mwakibete, H.; Bloor, D. M.; Wyn-Jones, E. J . Inclusion Phenom. Mol. Recomit. Chem. 1991. 10. 497. (22) P a k e r , D. M.; Hall, D.’G.; Wyn-Jones, E. J . Chem. SOC.,Faraday Trans. I 1988, 84, 773. (23)Takisawa, N.;Brown, P.; Bloor, D. M.; Hall, D. G.; Wyn-Jones, E. J . Chem. Soc.. Faradav Trans. I 1989. 85. 2099. (24) Kelly, G.; Takkawa, N.; Bloor, D. M.; Hall, D. G.; Wyn-Jones, E. J . Chem. Soc., Faraday Trans. 1 1989,85, 4321. (25) Davidson, C.J. Ph.D. Thesis, University of Aberdeen, 1983. (26) Harata, K. Bull. Chem. SOC.Jpn. 1916, 49, 2066. (27)Harata, K. Bull. Chem. SOC.Jpn. 1976, 49, 1493. (28)MacPherson, Y. E.;Palepu, R.; Reinsborough, V. C. J . Inclusion Phenom. Mol. Recognit. Chem. 1990, 9, 137. (29)We thank the reviewer for this suggestion.
Light-Induced Bubble-Stripe Transitions of Gaseous Domains in Porphyrin Langmuir Monolayers Mitsuru Yoneyama,* Akiteru Fujii, Shuichi Maeda, and Tetsuo Murayama Mitsubishi Kasei Corporation, Research Center, 1000 Kamoshida-cho, Midori- ku, Yokohama 227, Japan (Received: April 27, 1992; In Final Form: August 3, 1992)
We report the observation of bubble-to-stripe shape transitions of gaseous domains in Langmuir monolayers of 544-Ndodecylpyridiniumy1)-10,15,20-tri-ptolylporphyrinmixed with arachidic acid or 4-octyl-4’-(3-carboxytrimethyleneoxy)ambemne at the air-water interface. Monolayer textures are visualized by fluorescence microscopy with excitation of the porphyrin molecules by a focused beam. The same light is also utilized to expand the monolayer locally to induce shape transitions. Under continuous illumination, gaseous bubble domains emerge in the monolayer, which grow in size and then suddenly elongate into stripe structures at specified subphase pH. The critical bubble size characterizing the transition is found to increase with increasing pH, indicating the importance of electrostatic interactions in favoring stripe patterns.
Introduction Langmuir monolayers at the air-water interface afford a rich variety of model systems in which fundamental properties of two-dimensional pattern formation can be studied.’V2 The nature of patterns, i.e., lateral distributions of finite domains corresponding to solid, liquid, or gas, depends on the particular film-forming molecules and also on the thermodynamic-state variables. These nonuniformities can be easily visualized by fluorescence microscopy,) making it possible to correlate the growth, shapes, and sizes of domains directly with experimental conditions such as surface pressure, subphase temperature, and pH. Up-to-date various complex domain structures have been observed and characterized with fluorescence m i c r o s ~ o p y . ~Es-~~ pecially interesting is the observation of shape transitions between circular and noncircular domains. A host of theoretical works have focused on these shape transitions, ranging from simple transitions such as circular to elliptical domain^^^-^^ and square to rectangular domainsZoto more generalized transitions between regularly undulating shapes.2i*22Also, related theoretical descriptions have been developed that deal with phase transitions
between different infinite arrays of two-dimensional domain^.^^-^ All of these models build the physical mechanism for organization of domains on a combination of short-range attractions and long-range dipolar repulsions. Thus, they can be applied in principle to a wide variety of domains regardless of the nature of their phases. Most of the noncircular domains that have been experimentally observed have been concerned with s ~ l i d - l i q u i dor ~ ~liquid-liq~ uid13J6phases; it is only recently that the existence of elliptical or stripe structures in liquid-gas coexistence regions has come to our kn~wledge.~’-~~ Broadening the compass of monolayers that can show anisotropic gaseous domains should therefore facilitate a systematic characterization of domain shape transitions as well as a more profound understanding of the underlying physics and chemistry. In a previous paper,29we reported briefly the formation of gaseous stripe patterns in Langmuir monolayers composed of a surface-active pyridiniumylporphyrin and arachidic acid over a narrow range of subphase pH. There, the monolayer contained an excess amount of arachidic acid, leading to the three-phase
0022-365419212096-8982%03.00/00 1992 American Chemical Society
