Characterization of the interaction between a nonionic polymer and a

Anders Carlsson, Gunnar Karlstroem, and Bjoern Lindman. J. Phys. .... Hans Evertsson, Stefan Nilsson, Christina Holmberg, and Lars-Olof Sundelöf. Lan...
2 downloads 0 Views 645KB Size
J . Phys. Chem. 1989, 93, 3673-3677 than that of Fe+ and considerably less than that of Nb'. This is consistent with the behavior of other heterodinuclear cluster ions studied to date. An alkene ligand attached to the cluster greatly enhances the reactivity of the cluster, as expected from previous results on the reactivity of Co2CO'. Preliminary results from our laboratory show heteronuclear cluster carbonyl ions, MM'CO', also to be very reactive toward alkanes in cases where the bare cluster ion is unreactive toward alkanes. Thermochemical values from the reactions with alkenes indicate Do(Nb+-Fe) > 60 kcal/mol and IP(NbFe) < 8.8 eV, in agreement with previous photodissociation results. The oxide chemistry of NbFe' indicates DO(NbFe'4) > 85 kcalfmol, D0(NbFeO3'-H20) > 9 kcalfmol, and DO(NbFeO,+-O) > 119 kcal/mol ( n = 1, 2).

Acknowledgment. Acknowledgment is made to the Division of Chemical Sciences in the Office of Basic Energy Sciences in

3673

the U S . Department of Energy (DE-FG02-87ER13776) for supporting the transition-metal-ion research and to the National Science Foundation (CHE-8612234) for continued support of the Fourier transform mass spectrometry instrumentation. S.W.B. gratefully acknowledges the Purdue Research Foundation for providing fellowship support. Registry No. NbFe', 107474-41-9;NbFeOz', 1 17605-08-0;NbO', 23625-94-7; NbOz+,26317-75-9; NbFeOJt, 117605-09-1;Nb02(H20), 117526-57-5; NbFeO', 117605-07-9; Fe(C0)5, 13463-40-6; Nb', 7782-44-7;ethene, 74-85-1; propene, 115-07-1; 1,3-bu18587-63-8;02, tadiene, 106-99-0; 1-butene, 106-98-9;isobutene, 115-11-7;cis-2-butene, 590-18-1; trans-2-butene, 624-64-6; I-pentene, 109-67-1; 1-hexene, 592-41-6; 3,3-dimethyl-l-butene, 558-37-2; cyclopentene, 142-29-0; benzene, 71-43-2;cyclohexene, 110-83-8;toluene, 108-88-3; l-methylcyclohexene, 59 1-49-1;cycloheptene, 628-92-2; cycloheptatriene, 54425-2; ethylene oxide, 75-21-8.

Characterization of the Interaction between a Nonionic Polymer and a Cationic Surfactant by the Fourier Transform NMR Self-Diffusion Technique Anders Carlsson,+ Gunnar Karlstrom,* and Bjorn Lindman*,' Physical Chemistry 1 and Theoretical Chemistry, Chemical Center, Lund University, P.O.Box 124, S-221 00 Lund, Sweden (Received: April 11, 1988; In Final Form: September 28, 1988)

Self-diffusion measurements have been performed on dodecyltrimethylammonium ions (DoTA') in aqueous solutions of ethyl(hydroxyethy1)cellulose (EHEC) by the Fourier transform NMR self-diffusion technique. The experimental data were analyzed by means of a simple two-site model, and a deduction of the amounts of free and bound DoTA' in the polymer solution was achieved. The approach was shown to be a convenient tool of characterizing polymer-surfactant interactions, and it was demonstrated that either the addition of salt or an increase in temperature promotes the binding of DoTA+ to EHEC. The striking finding of an increased attraction between surfactant and EHEC at elevated temperatures demonstrates that the polymer molecules become increasingly more hydrophobic with increasing temperature, which is referred to as temperature-induced conformational changes in the polymer.

Introduction The interaction between water-soluble nonionic polymers and ionic surfactants has become a field of intense research in recent years (cf. recent review arti~lesl-~).Such studies mainly deal with the action of anionic surfactants, while it is often inferred that cationic species interact comparatively weakly with nonionic polymers. Thus, there are only a limited number of investigations performed on cationic surfactant-neutral polymer systems, where any considerable interaction effects are reported. However, if a rather hydrophobic polymer is used, a more pronounced interaction effect is observed, e.g., for ethyl(hydroxyethyl)cellulose4 and poly(viny1 alcohol) containing 10-1 2% acetate groups.s*6 This has been revealed both by measurements that probe changes in macroscopic behavior, e.g, cloud p ~ i n tand ~ . ~viscosity measurem e n t ~and , ~ by measurements of the surfactant cation activity.6 The Fourier transform N M R pulsed gradient spin-echo (FTNMR-PGSE) technique is well-established in self-diffusionstudies of surfactant solutions in order to characterize, inter alia, surfactant self-association equilibria, counterion binding, solubilization, and hydration.' In this work we apply this technique for monitoring the self-diffusion of the surfactant ion in aqueous solutions of ethyl(hydroxyethy1)cellulose (EHEC) and dodecyltrimethylammonium bromide (DoTAB). Particularly the temperature dependence of the surfactant ion self-diffusion was studied, but also the influence of small amounts of NaCl has been investigated, especially at low DoTAB concentrations. The DoTAB-water

* Author to whom correspondence should be addressed. 'Physical Chemistry 1. 'Theoretical Chemistry. 0022-3654/89/2093-3673$01.50/0

system was used as a reference system. By using a two-site model to interpret the observed self-diffusion coefficients, it was possible to deduce the amount of cationic surfactant bound to EHEC.

Experimental Section Materials. EHEC (Bermocoll E 351 X) was manufactured by Berol Kemi AB, Stenungsund, Sweden (from the same batch previously used4). The viscosity-average molecular weight for this sample was 146 000. The degree of substitution of ethyl groups was 0.9 per anhydroglucose unit, and the molar substitution of ethylene oxide groups was 2.1 per anhydroglucose unit. DoTAB was obtained from Sigma and was always stored in a desiccator. The critical micelle concentration (cmc) of DoTAB in H 2 0 at 25 OC is 0.015 m.8 D 2 0 (99.8 atom % D) was obtained from Norsk Hydro, Oslo, Norway. Hexadecane, used as a reference in the calibration procedure of the self-diffusion experiment, and D. Chem. Ind. 1972, 13, 531. (2) Robb, I. D. In Anionic Surfactant-Physical Chemistry of Surfactant Action; Lucassen-Reynders, E. H., Ed.; Marcel Dekker: New York, 1981; Surfactant Sci. Ser. Vol. 11, Chapter 3. (3) Goddard, E. D. Colloids Surf. 1986, 19, 255. (4) Carlsson, A,; Karlstrom. G.; Lindman, B. Langmuir 1986, 2, 536. (5) Tadros, Th. F. J . Colloid Interface Sci. 1974, 46, 528. (6) Shirahama, K.; Himuro, A.; Takisawa, N. Colloid Polym. Sci. 1987, 265, 96. (7) Lindman, B.; Sderman, 0.; Wennerstrom, H. In Surfactant Solutions: New Methods of Investigation; Zana, R., Ed.; Marcel Dekker: New York, 1987; Surfactant Sci. Ser., Vol. 22, Chapter 6. (8) Mukerjee, P.; Mysels, K. J. Critical Micelle Concentrations of Aqueous Surfactant Systems; National Bureau of Standards: Washington, DC, 1971. (1) Breuer, M. M.; Robb, I.

0 1989 American Chemical Societv

3674

Carlsson et al.

The Journal of Physical Chemistry, Vol. 93, No. 9, 1989

40

1 One - p has e region

sg:~~~!\k,50 20

0

40

6/ms

Figure 1. Normal and spin-echo N M R spectra of 0.0628 m DoTAB in a 0.97% EHEC-D20 solution at 25 OC (nonspinning sample; 10 accumulations).

NaCl were both of analytical grade and used without further purification. Procedures. Desalted and purified EHEC solutions were prepared by dialysis of a 1% (w/w) solution against distilled water for 5 days. (EHEC powder of technical quality contains about 4% (w/w) NaCI.) Hollow fibers made of regenerated cellulose were used as dialyzing membrane (molecular cutoff 6000; Spectrum Medical Industries). After freeze-drying, the polymer was redissolved in D20. The final concentration was 0.97% (w/w). To one portion of the EHEC solution was added 0.015 m NaCI. Samples of DoTAB in 0.97% (w/w) EHEC and DoTAB in D 2 0 (with and without NaC1) were prepared by weighing the components into glass ampules. After homogenization on a Vortex mixer the solutions were equilibrated at room temperature for 1 day. DoTAB and NaCl concentrations are expressed in molal: moles per kilogram of EHEC-D20 solution or D 2 0 . Determination of cloud points (CP) was performed as described p r e v i ~ u s l y .After ~ the samples were heated above the clouding temperature, the C P was taken as the temperature when the last visible sign of clouds disappeared on cooling. The FT-NMR-PGSE technique for measuring self-diffusion coefficients has recently been comprehensively described by Stilbs in ref 9, where all details can be found. The procedure in the present work roughly corresponds to the one in ref 10. The measurements were performed on a Jeol FX-60 FT-NMR spectrometer, operating at a resonance frequency of 60 MHz for the IH nucleus and equipped with a home-built field gradient unit (courtesy of P. Stilbs). The probe temperatures (25,45, and 54 "C) were measured by a copper-constantan thermocouple and adjusted within f 0 . 5 OC. Standard 5-mm N M R tubes were used, and the samples were equilibrated in the probe for 5 min before measurement. The magnetic field was externally locked on the D,O signal. A good signal-to-noise ratio was obtained with a single-pulse sequence except at low surfactant concentrations where 10 spectra were accumulated (pulse repetition time 2.5 s). Spectra for 10-12 different values of t h e duration of the field gradient (6) were recorded for each sample. The range of 6 varied typically from 10-30 ms at low surfactant concentrations t o 10-80 ms at the highest concentration. At least three intensity (0 values were taken for averaging, and the intensity data were fitted by a nonlinear least-squares procedure to the following equation:"

Ii = A exp(-GDis2(A

- 6/3))

(1)

002

004 006 008 DoTAB c o n c / m o l a l

0.10

Figure 2. Cloud point of a 0.97% EHEC-D20 solution versus the DoTAB concentration with and without addition of NaC1: (0) 0 and).( 0.015 m NaCI. The vertical arrows represent clouding occurring above 100 O C , and the horizontal arrows indicate at which temperatures the self-diffusion measurements were performed. Reprinted with permission from ref 22. Copyright 1988 World Scientific Publishing.

coefficient of the ith component, A is the time between the gradient pulses (chosen to be 140 ms according to ref 12), and A is a fitting parameter. G was determined every day in a separate calibration experiment in which either 2% H 2 0 in D 2 0 or neat hexadecane was used. The self-diffusion coefficient of small amounts of HDO m2/s at 25 OC and 3.03 X m2/s at 45 in D 2 0 is 1.90 X OC.13 At 54 "C the self-diffusion coefficient of neat hexadecane m2/s.14 is 7.34 x The self-diffusion measurements are illustrated by 'H N M R spectra of 0.0628 m DoTAB in 0.97% EHEC solution at 25 O C in Figure 1. In the spin-echo spectrum (6 = 0) the -CH2- peak of DoTAB is strongly attenuated compared to the normal spectrum due to relatively short transverse relaxation time of the methylene protons while the uncoupled -N(CH3)3 is essentially unaffected. Therefore, the signal intensity of the latter peak was monitored in all experiments on DoTAB. In order to reduce the intensity of the HDO signal, which can disturb measurement on the -N(CH,), signal at low surfactant concentrations, it is important to use D 2 0 that has been kept airtight to avoid absorption of H 2 0 . The exchange of protons from hydroxyl groups on EHEC at a concentration of 0.97% had no significant effect on the intensity of the HDO signal. This was directly confirmed by freeze-drying an EHEC-D20 sample. After the sample was redissolved in D20, the N M R spectrum had the same appearance as the spectrum of the starting material.

Results and Discussion Presented in Figure 2 are the C P measurements for 0.97% (w/w) EHEC-D20 solutions as a function of the DoTAB concentration. The upper curve represents a salt-free EHEC solution, and the lower one represents an EHEC solution containing 0.015 m NaC1. The latter curve reflects very well the synergistic effect4 observed for EHEC solutions (and for other cellulose ether solutions as well; to be published): When a relatively small amount of salt is present, the addition of ionic surfactant (both cationic and anionic species) initially causes a considerable CP depression. If t h e salt concentration is increased, the depth and width of the C P minimum will be more pronounced. Without surfactant the presence of 0.01 5 m NaCl does not significantly affect the CP. (In order to cause a salting-out effect of 20 OC by NaCl alone, a concentration of 1 m is required.) The upper curve in Figure 2 also possesses a C P minimum, although less pronounced, but the expected result is a more or less immediate elevation of the C P when the ionic surfactant is

Here G is the field gradient strength, Di is the self-diffusion (9) Stilbs, P. Prog. Nucl. Magn. Reson. Spectrosr. 1987, 19, 1 . (IO) Das, K . P.; Ceglie, A,; Lindman, B.; Friberg, S. E. J . Colloid Interface Sci. 1987, 116, 390. (11) Stilbs, P.; Moseley, M. E. Chem. Scr. 1980, 15, 176.

(12) Stilbs, P.; Lindman, B. J . Phys. Chem. 1981, 85, 2587. (13) Mills, R. J . Phys. Chem. 1973, 77, 685. (14) Ertl, H.; Dullien, F.A.L. AIChE J . 1973, 19, 1215. ( 1 5) Jonsson, B.; Wennerstrom, H.; Nilsson, P.-G.; Linse, P. Colloid PoIym. Sci. 1986, 264, 77. (16) Jones, M. N . J . Colloid Interface Sci. 1967, 23, 36.

Nonionic Polymer-Cationic Surfactant Interaction

The Journal of Physical Chemistry, Vol. 93, No. 9, 1989 3675

added in absence of salt. This behavior could be due to impurities of inorganic salt in DoTAB or due to an incomplete dialysis of the starting EHEC solution. However, both explanations could be discarded as DoTAB of high purity was used and as EHEC samples were thoroughly dialyzed. Normally, after dialyzing 5 days or longer, the Na concentration in the EHEC solution is less m NaCl than 0.5 mg/dm3, corresponding to less than 0.03 X (determined by atomic absorption spectroscopy). Instead, this finding can be explained on the basis of the fairly high concentrations of free surfactant ions and counterions (cmc = 0.015 m ) , causing a screening of the repulsive interaction between the charged polymer chains in a manner similar to ordinary elect r o l y t e ~ .(A ~ C P minimum in the absence of NaCl has also been observed for other short-chain ionic surfactants with a cmc of about 0.01 M or above, Le., concentrations similar to those where inorganic electrolytes produce C P minima.4) This leads to a decreased solubility of the polymer-surfactant complexes which is reflected in the decrease of C P at low DoTAB concentrations. If we now consider the lower C P curve again, the addition of salt screens the electrostatic forces further, and the C P depression becomes more drastic. It should be noted that the same C P curves are obtained if H 2 0 is used as solvent instead of heavy water. When more DoTAB is added, the effective charge density on the polymer increases, and since the repulsive electrostatic forces could be expected to grow faster (approximately quadratic) with surfactant concentration than the attractive hydrophobic forces, which grow approximately linear with surfactant concentration, one may expect the cloud point temperature to increase when more surfactant is added. Presented in Figure 3 are self-diffusion coefficient data for the DoTA' ion at 25, 45 and 54 OC. Four systems were investigated: DoTAB in D 2 0 and DoTAB in EHEC-D20, both systems with and without 0.015 m NaC1. At 54 "C only the salt-free systems were studied, since with added NaCl there is a phase separation according to Figure 2. One common feature for all the systems, and irrespective of temperature, is the sharp decrease in the observed self-diffusion coefficient (Dobsd)when the DoTAB concentration is increased above the cmc. This behavior reflects the onset of the strongly cooperative self-association of DoTAB, leading to formation of micelles in D 2 0 and to polymersurfactant complexes in EHEC-D20 solutions. However, there are significant differences, and let us first consider the salt-free systems (open symbols). At 25 "C and at the lowest DoTAB concentration there is no difference in Dobd when EHEC is present or not, but at higher concentrations Dohd is markedly reduced, indicating a strong cooperative interaction between DoTAB and EHEC. When the temperature is increased to 45 "C, the diffusion curves display almost the same characteristics (except for the change of magnitude in Dobsd),but, interestingly, a reduction in Dobsd is also observed at the lowest DoTAB concentration. This also holds at 54 "C. Now let us introduce salt to the systems (cf. the filled symbols in Figure 3). Except at the lowest surfactant concentration, the observed diffusion of DoTA' in D 2 0 is significantly lower than in the salt-free system, at both 25 and 45 "C. This reflects the facilitated micellization of DoTA' in the presence of salt through a more efficient screening of the charged head groups by the added counter ion^.^^*^* However, the most distinguishing feature is the overall reduction in Dobsd when EHEC is present, especially at 25 "C and at low DoTAB concentrations (C0.02 m ) . As a first approximation, the diffusion data can be analyzed according to a two-site model. Dobsdis a population average of the self-diffusion of free DoTA' and DoTA' bound to EHEC, as either monomers or clusters: Dobsd

= PfDf-t PbDb

(2)

Here Pf and Pb are the fractions of free and polymer-bound DoTA', respectively, Df and Db are the self-diffusion coefficients (17) Lindman, B.; Wennerstrom, H. Top. Curr. Chem. 1980, 87, 1. (1 8) Gunnarsson, G.; Jonsson, B.; Wennerstrom, H. J . Phys. Chem. 1980, 84, 31 14.

25

0 1 0

"

0.02

"

"

"

0.04 0.06 0.08 DoTAB conc I molal

"C

"

0.10

10 45

0

0

"C

0.02

0.04 0.06 0.08 DoTAB c o n c / molal

010

002

0.04 006 0.08 DoTAB conc I molal

010

~

0

Figure 3. Observed DoTA* self-diffusion coefficients at three temperatures versus the DoTAB concentration in D20 (circles) and 0.97% EHEC-D,O (squares). The open symbols are for solutions with no salt added, and the filled symbols are for solutions containing 0.015 m NaCI. Reprinted with permission from ref 22. Copyright 1988 World Scientific Publishing.

of free DoTA, and of polymersurfactant complexes, respectively. Since Db