4894
J . Phys. Chem. 1989, 93, 4894-4898
The present results thus show that isolation of surface Pd atoms can be increased in two ways: (1) decreasing the particle size, Le., producing a high fraction of coordinatively unsaturated Pd atoms; (2) interdispersing foreign atoms in the Pd surface, e.g., forming PdAg alloys or PdSi,. Compound formation is found to be a much more efficient way to isolate Pd atoms. This is in line “ i t h Blyholder’s quantum mechanism calculations;21the extent of back-donation is reduced more strongly by engaging a Pd atom into bonding to an adjacent Si atom than by merely changing its coordination to other Pd atoms. It is important that IR spectroscopy permits detection of Pd silicide formation in an incipient stage where this cannot be easily registered by other instruments. It is also interesting that the departure of the B/L ratio from the “standard” curve is smaller for the sample with high Pd loading. This may be due to surface
enrichment of the surface with Pd. Since Pd, unlike Si, is able to form strong chemical bonds with CO, the surface should become enriched in Pd during exposition to C O because of the usual “chemisorption induced surface segregation” which was proven for numerous alloy systems.)) This surface enrichment of PdSi, in Pd should be more pronounced for particles with larger volume than for small particles for which a large atomic fraction is located at the surface. It is also conceivable that the extent of the interaction between Pd and silica depends on the particle size of the Pd. At present we have no additional data for testing these speculations. Registry No. Pd, 7440-05-3; CO, 630-08-0. (33) (a) Bouwman, R.; Sachtler, W. M. H. J. Caral. 1970, 19, 127. (b) Bouwman, R.; Lippits, G. T. M.; Sachtler, W. M. H. J . C a r d . 1972, 25, 350. (c) Sachtler, W. M. H. Vide 1973, 163, 19.
Photorheological Effects in Micellar Solutions Containing Anthracene Derivatives. A Rheological and Statlc Low Angle Light Scatterlng Study Thomas Wolff,* Claus-Stephan Emming, Thomas A. Suck, and Gunther von Bunau Institut fur Physikalische Chemie, Uniuersitat Siegen, 0 - 5 9 0 0 Siegen, West Germany (Received: September 21, 1988; In Final Form: December 16, 1988/
Viscosities, flow behavior, and masses of micelles in aqueous solutions of cetyltrimethylammonium bromide (CTAB) were determined in the presence of 9-substituted anthracenes. Greatly enhanced viscosities were found when small amounts of nonpolar anthracene derivatives (methyl, ethyl, n-propyl, n-butyl, n-pentyl) are added to pure CTAB solutions while strictly Newtonian flow was observed. Micellar masses exceeding that of pure CTAB micelles by factors of 35-260 were determined. Upon photochemical conversion of 95% of the anthracenes to dimers, viscosities dropped but no significant changes of micellar masses took place. In solutions of CTAB and added 9-anthracenecarboxylic acid non-Newtonian flow behavior, such as rheopexy, viscoelasticity, and thixotropy, was observed being influenced by photodimerization. The rheopectic flow behavior is shown to be connected with low micellar masses
Introduction Flow behavior and viscosity of aqueous micellar solutions are drastically altered by solubilizing but small amounts of suitable aromatic substances that may be called “rheologically active”. Other aromatic substances are “rheologically inactive” and have little influence on the flow properties of these solutions. In previous work it was shown that certain aromatic compounds of the one class may be photochemically transformed into products belonging to the Accordingly these processes are accompanied by “photorheological effects”, such as viscosity changes, transformations of Newtonian into non-Newtonian fluids, and even transitions between different lyotropic liquid crystalline phases.2 Examples are cis-trans isomerizations of some stilbene derivatives) and photodimerizations of some 9-substituted anthracene derivatives in aqueous micellar solutions of the cationic tenside cetyltrimethylammonium bromide (CTAB) and the nonionic tenside Triton X-100.5 It has been previously suggested that the macroscopically observable viscosity effects are due to microscopic changes in size and/or shape of micellar aggregates specifically induced by the (1) Milller, N.; Wolff, T.; von Bilnau, G. J. Photochem. 1984, 24, 37. (2) Wolff, T.; von Bilnau, G. Eer. Bunsen-Ges. Phys. Chem. 1984, 88,
1098. (3) Wolff, T.; von Biinau, G. J. Photochem. 1986, 35, 239. (4) Wolff, T.; Suck, T. A,; Emming, C.-S.; von Bilnau, G. Prog. Colloid Polym. Sci. 1987, 73, 18. ( 5 ) Wolff. T.: Schmidt, F.; von Biinau, G. J. Photochem. Photobiol., A, in press.
0022-3654/89/2093-4894$01 SO10
respective solubilizates. In order to test this assumption directly, we have carried out parallel experiments to obtain macroscopic as well as microscopic data of the same aqueous micellar CTAB solutions before and after photochemical transformation of rheologically active additives solubilized in these systems. The macroscopic flow behavior was examined with a computer-controlled rotating viscometer that discriminates between Newtonian and nowNewtonian fluids. Microscopic information on the size of the micellar aggregates before and after photoconversion was obtained from static low angle scattering of these solutions. Newtonian flow is exhibited by aqueous CTAB solutions containing small amounts of nonpolar 9-substituted anthracenes, Le., 9-methyl-, 9-ethyl-, 9-n-propyl-, 9-n-butyl-, and 9-n-pentylanthracene, all of which are rheologically active inasmuch as their solubilization leads to a drastic increase of the solution visco~ity.~*~ Upon irradiation of these solutions photodimers are formed according to
During photochemical conversion the solution viscosity changes in a specific and complicated way. A particularly large viscosity decrease was found in the cases of n-butyl- and n-pcntylanthracenes which allow an almost complete (>90%) photoconversion without the photodimers precipitating from the micellar 0 1989 American Chemical Societv
Photorheological Effects in Micellar Solutions
The Journal of Physical Chemistry, Vol. 93, No. 12, 1989 4895
TABLE I: Viscosities, q, Molar Masses of Aggregates, M , Second Virial Coefficients, B , Micellar Aggregation Numbers, z, and Average Occupation Numbers, (i), for Aqueous Solutions of CetyltrimethylammoniumBromide ((TAB) Containing Various Amounts of 9-Substituted Anthracene Derivatives (A) at 25 k 0.2 OC q/mPa s hf/lo6 g/mol zc [CTAB]/ 100 X before after before after Bd/ 106 before after solubilized A mmol/L IA1 /[CTABF' irradiation irradiation irradiation irradiation mol L/g2 irradiation irradiation (i)d aoc 0.03 -0.6 0.9280 3-8 120 0.62 3100 1.2 0.935" 3.8 1-4 9-methyl 2.74 3.8 250 1200 16000 0.86 6.1 0.935" 7.4 1-4 5.24 7.4 250 280 7500 0.34 0.946' 2.8 3.8 1-4 9-ethyl 2.91 250 3.8 1210 18600 1.39 0.946" 7.1 1-4 7.4 7.4 7.28 250 70 0.27 1900 0.935" 0.7 1-4 3.8 9-n-propyl 3.42 3.8 250 1050 14000 0.83 7.4 0.954' 5.3 1-4 10.14 7.4 250 3200 120 1.2 1.53 3200 0.957" 1.2 1-4 3.8 9-n-butyl 3.42 1.62 250 3.8 1590 0.69 21600 0.946" 8.3 1-4 7.4 10.19 7.4 250 7200 220 2.6 1.14 0.935' 2.2 6000 1-4 3.8 9-n-pentyl 3.37 2.12 250 3.8 900 1 I600 4.4 1.04 0.939" 1-4 7.4 7.96 250 7.4 11 6 27b 26' 0.007 16 0.005 -1 1.6 20-25 45 9-carboxylic acid OValues at [CTAB] = 2.5 mmol/L. 'Solution behavior non-Newtonian; 7 = 27 mPa s at 9 = 145/s. cAggregation number of CTAB molecules only. dBefore irradiation. cLiterature value" z = 81. 'Values at [CTAB] = 250 mmol/L and 100[A]/[CTAB] = 3.8 from ref 4.
solution. Therefore, these two systems were subjected to light scattering measurements in the irradiated state. On the other hand, CTAB solutions containing polar additives, such as 9-anthracenecarboxylic acid, exhibit non-Newtonian flow behavior that can be changed upon photodimerization of the anthracene derivative. For comparison, further viscometric and light scattering data were obtained from this system and are reported below.
Experimental Section Cetyltrimethylammonium bromide (Merck, p.a., recrystallized two times from a 9:l mixture of acetone and water) and the anthracene derivatives were available from previous investigations.'-' Micellar solutions were prepared by stirring at 30 OC using triply distilled water. If necessary, the solubilization process was accelerated by ultrasonic irradiation. Photochemical reactions were carried out by exposing nitrogen or argon saturated solutions under stirring at 25 OC to the emission of a high-pressure mercury lamp selecting the 366-nm line by an interference filter (Oriel). In order to avoid deposition of irradiated material on windows, the samples were irradiated through the inert gas/liquid interface. Dynamic, Le., absolute viscosities were determined at 25 f 0.2 OC by using a rotating viscometer that has been described elsewhere.4 Under computer control shear rates could be varied between 0.03 and 300 s-I. Flow curves (shear stress vs shear rate) and viscosity vs time diagrams were measured by the same rotating viscometer using Mooney-Ewart geometry (Haake M E 15) with a gap width of 1.5 mm. For recording flow curves, the shear rate was increased linearly from 0 to 145 s-l within 50 s, kept there for 20 s, and decreased linearly from 145 to 0 s-l within 50 s. The viscosities of Newtonian liquids were also measured with Ostwald capillary viscometers. Owing to the small concentrations, solution densities did not differ measurably from that of water. Details of the apparatus for static low angle light scattering have been p ~ b l i s h e d .A ~ Polytec Model PL 710 (unpolarized) He-Ne laser was employed as a source for incident light. The dependence of refractive index on concentration (dnldc, needed for data evaluation) was determined before and after the irradiations by using the apparatus sketched in Figure 1. The liquid of refractive index n was filled into the thermostated cuvette shown (6) Wolff, T.; Muller, N.; von Bunau, G. J . Photochem. 1983, 22, 61. (7) Wolff, T.; Muller, N. J . Photochem. 1983, 23, 131.
Figure 1. Illustration of d n l d c determination; see text.
in the figure (c). The refractive index n is related to the measured quantity h by
h= 1 tan {arcsin ( n sin [65 - [arcsin ( ( l / n ) sin a)]]) - 65
+ a) (2)
The meanings of h, I , and a follow from the illustration in Figure 1. Values ranging from dn/dc = 0.146 mL/g through 0.1 5 1 mL/g were determined depending on the content of solubilizate. The values did not vary significantly upon photodimerization of solubilizates. We avoided the use of an Ab& refractometer because of its small optical path length which might favor orientation of micelles at the glass walls. Solutions for light scattering experiments were prepared with utmost protection from dust and filtered through cellulose nitrate filters with pores of 50-nm diameter (purchased from Sartorius, Gottingen). These solutions were filled into the spectrometer cell at 40 OC and then tempered at 25 OC for several hours until the scattered light intensity was constant.
Results Rheological Measurements. Careful previous experiments had shown that 0.25 M aqueous solutions of cetyltrimethylammonium bromide (CTAB) containing nonpolar anthracene derivatives exhibited Newtonian flow behavior before and after i r r a d i a t i ~ n . ~ This was found as well in the more diluted solutions (1-4 mM CTAB) investigated here since the shear stress T measured with the rotating viscometer turned out to be proportional to the shear rate i.. We therefore used Ostwald capillary viscometers to determine viscosities ( 9 ) of micellar CTAB solutions containing nonpolar anthracene derivatives (A). In Table I 9 values at 2.5 mM CTAB are listed that represent concentration ranges from
4896
The Journal of Physical Chemistry, Vol. 93, No. 12, 1989
j./s-'
Wolff et al.
.
___t
E-
Figure 2. Flow curves (shear stress T vs shear rate i.) at 25 'C for aqueous solutions of 22 mM cetyltrimethylammoniumbromide and 9.9 mM 9-anthracenecarboxylicacid: (a) before and (b) after 5% photo-
conversion. 1 to 4 mmol/L at two different concentration ratios [A]/[CTAB]. (Concentration ranges are required for comparison since other data in Table I are obtained by extrapolation; see below). The values are compared with those of more concentrated solutions (250 mmol/L CTAB) at the same ratio [A]/[CTAB]. The viscosities for each series exceeded that of pure CTAB solutions. In the concentrated solutions q increases in the order methyl < n-propyl < n-butyl while n-pentylanthracene does not continue this order. The diluted solutions are only slightly more viscous than pure CTAB solutions. Upon photoconversion of ca. 90% of n-butyl- and n-pentylanthracene the viscosities drop distinctly. All experiments were carried out at 25 OC, which is near the KRAFFT point of CTAB in water. In order to make sure that the observed effects are not due to beginning precipitation, we repeated some of the experiments at higher temperatures, Le., at 30 and 35 OC, and obtained less pronounced but qualitatively the same viscosity effects. In micellar solutions containing large amounts of aromatic solubilizates the question of whether or not the solutions are in thermodynamic equilibrium always arises. We tried to test this in that we stored samples containing monomeric anthracenes for 3 months and did not observe precipitation. Therefore, the solutions are very stable, if not in equilibrium. This is not always the case after photodimerization: photodimers of short-chain alkyl-substituted anthracenes tend to fall out after some hours in samples of high degrees of photoconver~ion.~ More detailed rheological analyses were necessary in the systems containing the polar compound 9-anthracenecarboxylic acid (9AC). Experiments with more concentrated solutions (150 mM CTAB, 22.5-45 mM 9-AC) had already revealed an extremely complex flow behavior4 varying with the 9-AC content and upon photodimerization; Newtonian, rheopectic, thixotropic, and viscoelastic behavior had been found. The less concentrated solutions (20-25 mM CTAB, 9-1 1.3 mM 9-AC) investigated here for direct comparison with light scattering data exhibited different flow behavior. It can be seen from inspection of the flow curve in Figure 2a that nonirradiated solutions are rheopectic up to a shear rate i. of ca. 40 s-l, then change to thixotropic at higher shear rates, and become Newtonian-like above 100 s-'. In the irradiated samples (only 5% of 9-AC was photoconverted in order to avoid precipitation of photodimers) the feature indicating thixotropy is missing. This complex dependence on shear rate prompted us to investigate the viscosity (shear stress divided by shear rate) at different concentrations as a function of shear time, which can be done at a constant shear rate. Figure 3 shows results at a shear rate y = 65 s-l. The curves for the three lowest concentrations (15, 20, 30 mM CTAB) start with a zero time viscosity qo, and then q increases and levels off after ca. 90 s of shearing. This behavior is typical for rheopexy, a rarely observed flow property. It may be expected when the formation of larger aggregates or the buildup of a structured solution is induced by shearing. At 40 and 60 m M CTAB, solutions do not exhibit a zero time viscosity, q,,, but exhibit a proportionality of shear stress and shear rate for t 0 instead; i.e., the solution does not flow but is merely deformed elastically. This behavior is known as viscoelastic and indicates a gellike structure of the solution at rest. Continued shearing induces plastic flow up to a maximum value of r ) and finally breaks up the solution structure. After shearing 80-90
-
Figure 3. Viscosity qat 25 OC vs shear time t a t a shear rate i. = 65 s-' for aqueous solutions of cetyltrimethylammoniumbromide (CTAB) at various concentrations containing 9-anthracenecarboxylic acid (A); [A]/[CTAB] = 0.4.
s a constant value q m is reached. Higher values of q m and higher maxima appearing at shorter shearing times are obtained when the CTAB concentration is raised from 40 to 60 mM. Viscoelasticity has been observed before in solutions of rodlike surfactant aggregates provided the length of the rods exceeds a value at which free rotation is hindered.8*9 When the CTAB concentration is increased further up to a value of 70 mM, purely thixotropic behavior is observed; Le., 7 decreases from a high value at t = 0 s and levels off at ca. 90 s, which means that minimal shearing times suffice to destroy the solution structure existing at rest. Static Low Angle Light Scattering Experiments. The intensity Z(8) of light scattered from a sample of volume Vis proportional to the intensity Io of the incident light and inversely proportional to the square of the distance r between sample and light detector: (3) 8 is the angle between the directions of incident and scattered light.
C(8) is called the Cabannes factor and can be obtained from the polarization of the scattered light; it depends on the shape of the scattering centers and is unity for spherical particles.1° For given C(0) the "Rayleigh ratio" R(0) is determined by experiment using eq 3. It follows from the theory of light scattering that a particularly simple relation exists between Z(8) and the mean molar mass M of the scattering particles when the scattering angle 8 is sufficiently low (in practice 0 < 5'). When expressing the difference AC = c - CI (4) of total solute (surfactant and solubilizate) mass concentration c and critical micelle concentration c1 (units of g/L) the following relation holds:I0
-KAc - -!-+- 2BAc AR(@ M where A R is the difference of Rayleigh ratios for solution and solvent, B is a virial coefficient measuring deviations from ideality in the osmotic pressure law, and K depends on refractive index n and light wavelength X according to
( L is Avogadro's constant). The determination of micellar masses M from applications of eq 5 requires that special attention be paid to two problems: (i) The micelles, being highly charged particles, can interact elec(8) Hoffmann, H.; Platz, G.; Rehage, H.; Schorr, W. Ber. Bunsen-Ges. Phys. Chem. 1981, 85, 817. (9) Shikata, T.; Hirata, H.; Kotada, T. Langmuir 1987, 3, 1081. (10) Kerker, M. Ed. The Scattering o f f i g h t ; Academic Press: New York, 1969.
Photorheological Effects in Micellar Solutions
.
>21
The Journal of Physical Chemistry, Vol. 93, No. 12, 1989 4897 +
+p
.
a
. L
.
+.+
+.
t
.2
1
.6
+
p1
.
u
?I5 L'
*
1
.2
I
. .
f
I
1
.6
x / g I-!
Figure 4. (a) Differences of Rayleigh factors of solution and solvent, AR, at a scattering angle of 3.1° in aqueous solutions of cetyltrimethylammonium bromide (CTAB) containing 9-ethylanthracene (A) as a function of the concentration Ac = C ~ A -B cmc + CA at 25 OC; [A]/ [CTAB] = 0.038. (b) Ratios K A c / A R of the solutions of (a).
c
.
.
kt
e
.2
4
io IC /g
I
.1,
B
6
.8
il
12
li
t
1-1
Figure 6. (a) Differences of Rayleigh ratios of solution and solvent, AR, at a scattering angle of 3.1° and at 25 OC in aqueous solutions of cetyltrimethylammonium bromide (CTAB) containing 9-anthracenecarboxylic acid (9-AC) as a function of the concentration Ac = cmAB cmc + c ~ . ~ ,[9-AC]/[CTAB] -; = 0.45. +, before; A, after 5% photodimerization. (b) Ratios K A c / A R of the solutions of (a).
According to eq 5 the reciprocal of the intercept is the molar mass, M , of the aggregate. It consists of fractions
J
.2
.4 AC / g I-'
.6
.a
Figure 5. (a) Differences of Rayleigh ratios of solution and solvent, AR, at a scattering angle of 3.1° and at 25 OC in aqueous solutions of cetyltrimethylammonium bromide (CTAB) containing 9-n-butylanthracene (A) as a function of the concentration Ac = CCTAB - cmc + cA; [A]/ [CTAB] = 0.038. 0 , before; 0,after 95% photodimerization. (b) Ratios K A c / A R of the solutions of (a).
trostatically, thereby inhibiting the required free diffusionlo (Le., the positions of the scatterers must be independent so that intensities of scattered light are additive). (ii) Size and shape of micellar aggregates are often strongly dependent on concentration. The former problem was overcome by investigating sufficiently diluted solutions accepting low scattering intensities. Suppressing electrostatic interactions by adding large amounts of electrolyte is unacceptable for our purposes because added electrolyte affects sizes of aggregates. The latter problem restricts the concentration range usable for the determination of aggregate masses. We chose concentration ranges of 1-4 mM CTAB for the samples containing nonpolar anthracene derivatives and 20-25 mM CTAB for 9anthracenecarboxylic acid containing samples. In this case higher solubilizate concentrations are permissible because of the inherently higher electrolyte concentration arising from the solubilization of 9-AC by which HBr is set free.4 Examples of light scattering data and their evaluation for systems containing nonpolar 9-substituted anthracenes are given in Figures 4 and 5. The a parts of the figures show Rayleigh ratio differences, A R , at a scattering angle of 3.1' as a function of concentration difference, Ac. In the absence of significant electrostatic interactions the AR values increase as expected. The b parts of the figures illustrate the determination of the aggregate molar mass. In Figure 5 also the data for the irradiated, less viscous solutions are included. Although the AI? values differ for solutions before and after irradiation, it is seen from Figure 5 that the extrapolated ordinate intercepts coincide.
MCTAB = M[(CCTAB - ci)/AcI and MA= M ( c A / A c ) (7) pertaining to the surfactant and the solubilizate, respectively. With the given molar masses M , for CTAB and M 2 for the solubilizate the quantities z = McTAB/MI and ( i ) = M A / M 2 (8) define an average aggregation number z and an average occupation number ( i ) (number of solubilizate molecules per micelle). Values of M , z, and ( i ) are listed in Table I; their uncertainties depend on the extrapolation procedure to obtain M and amount to f20%. In Figure 6 light scattering data are plotted for micellar solutions of the polar anthracenecarboxylic acid. Above Ac = l l g/L (ca. 30 mM CTAB) the A R values decrease with concentration (Figure 6a), indicating inhibited diffusion of micelles. Therefore, the range Ac > 11 g / L cannot be used for micelle mass determination (cf. problem i above). A linear dependence of the quantity K A c l A R on Ac is found for Ac values between about 10 and 11 g/L (Figure 6b). Below this Ac range deviations from linearity are observed as is expected when the micellar mass depends on concentration; i.e., only apparent micelle masses at a given concentration can be determined (cf. problem ii above). Only the linear part of the curves (Figure 6b) has been extrapolated. The extrapolated value corresponds to the micellar mass in a concentration range in which the solutions behave rheopecticly; see Figure 3. It represents a very low aggregation number ( z = 16; see Table I) which is further decreased by irradiation (z = 11).
Discussion The data presented in Table I demonstrate the most striking result that the addition of 9-alkyl-substituted nonpolar anthracenes may inflate micellar aggregate masses by factors between 35 and 260. Since these solutions show Newtonian flow behavior, we must assume a more or less spherical shape of the micelles. The radii of these spheres exceed the length of a surfactant monomer by far. According to the statistical model of Dill and Flory,I2 it would Lianos, P.;Zana, R.J . Colloid Interface Sci. 1981, 84, 100. (12) Dill, K.A,; Fiory, P.J . Proc. Natl. Acad. Sci. U.S.A. 1981, 78, 676. ( 1 1)
4898
J . Phys. Chem. 1989, 93, 4898-4903
be energetically unfavorable to assume a homogeneous mixture of surfactant ions and nonpolar additives filling up the spherical space. We must, therefore, assume that a substantial part of the solubilizate molecules reside in the center of the micelle which appears to be an aggregate of anthracene derivative molecules surrounded by a surfactant shell. One may regard this as a novel type of an o/w microemulsion in which the anthracenes (although being solids at 25 "C) take the oil part. Rheologically, the systems resemble microemulsions in that they show Newtonian flow behavior, but they differ from those because of their much higher viscosities. Usually the viscosity of a microemulsion is comparable to that of water, which has been ascribed to reversible droplet coale~cence.'~Conversely, the higher viscosities of our systems may indicate a higher aggregate stability as would be expected for anthracene derivative aggregates that are of a pre- or microcrystalline nature as indicated by concentration-dependent features in UV spectra.2 Photoconversion of 90% of the anthracenes causes a large reduction of the macroscopic viscosity while, surprisingly, the aggregate mass is unaffected in the case of the 9-n-butylanthracene and is even increased when the n-pentylanthracene is photodimerized. The aggregate masses and the corresponding viscosities are neither interrelated, nor is there a simple relationship of either quantity to the length of the alkyl chain in the 9-position of the anthracene molecule. We can, therefore, rule out in these cases that the photorheological effects are due to major changes of the size of the entire micelles. Other micellar properties that might contribute to the viscosity of the solutions must change, such as the degree of dissociation and the extent of the electric double layer surrounding the micelles. When the polar 9-anthracenecarboxylic acid is added to CTAB solutions, a completely different situation arises. It had been shown previously for more concentrated solutions (75-250 mM CTAB) that a part of the 9-AC molecules is solubilized in such a way (13) Langevin, D. Acc. Chem. Res. 1988, 21, 255.
that ion pairs of 9-AC anions and cetyltrimethylammonium cations are formed whereby HBr is set free. As a consequence of both the large, strongly binding counterions and the generated electrolyte HBr, long rodlike aggregates are formed that are responsible for the nowNewtonian flow behavior. At the lower concentrations investigated here the observed departure from Newtonian flow has a different cause since many very small and strongly interacting micelles exist in this range. Aggregation is further reduced by photdimerization of the anthracene derivative. Considering the viscosity data, the following microscopic cause of the rheopexy of these solutions emerges: at low shear the viscosity (shear stress 7 divided by shear rate +) is not far from that of water because the small aggregates present behave like dispersed molecules in an ordinary solution. Since the negative slope of the curves KAc/AR vs Ac in Figure 6b reveals attractive interactions of the scattering particles, shearing may induce the growth of aggregates and, at higher shear rates, the formation of larger aggregate structures. Finally, upon further increase of shear rate an equilibrium of buildup and collapse of structures appears as Newtonian flow. Conclusion
Static low angle light scattering experiments have revealed that the photodimerization of some nonpolar anthracene derivatives in aqueous micellar solutions of CTAB is accompanied by highly specific changes of micelle sizes and of macroscopic flow behavior. The polar compound 9-anthracenecarboxylic acid can induce the formation of very small micellar aggregates; photodimerization of this compound leads to further shrinkage of the micelles. This smallness of the micelles is the microscopic cause of the rheopectic flow behavior that is already observed at concentrations as low as 20 mM. Registry No. CTAB, 57-09-0; 9-methylanthracene, 779-02-2; 9ethylanthracene,605-83-4;9-n-propylanthracene,1498-77-7;9-n-butylanthracene, 1498-69-7; 9-n-pentylanthracene, 33576-54-4; 9anthracenecarboxylicacid, 723-62-6.
Role of Defects In Radiation Chemistry of Crystalline Organic Materials. 1. ESR Evidence for Electron Trapping in the Mixed Crystals of Binary n-Alkanes at Low Temperatures Hachizo Muto,* Keichi Nunome, Kazumi Toriyama, and Machio Iwasakit Government Industrial Research Institute, Nagoya, Hirate-cho, Kita- ku, Nagoya, 462, Japan (Received: October 5, 1988; In Final Form: January 20, 1989)
An ESR study has been made on the electron stabilization in the mixed crystals of binary n-alkanes irradiated at 77 and 4 K, as the first step in understanding the role of defects in radiation chemistry. Evidence has been obtained to show that the radiation-induced electrons are trapped in the mixed crystals in contrast to the absence of trapped electrons in the neat n-alkanes, together with the following results on the trapping site. The ESR line width (AHml)of trapped electrons (reflecting the effective defect size) correlates with the chain length difference (An,) between two n-alkanes. An increase of Anc gives a higher yield of the trapped electrons probably because of a stabilization due to an expansion of the defect size. The number of defects accessible to the electrons depends on the crystal structures, being slightly larger in triclinic crystal than in orthorhombic crystals. The effect of deuteriation of the molecules on AH,,, in addition to the above results suggests that the trapping site is a crystalline lattice defect created by inhomogeneous contacts of the different chain length molecules at the layer boundary and that preexisting defects such as voids are necessary for electrons to be trapped in crystalline n-alkanes.
Introduction
The role of trapped electrons in low-temperature organic solids is a fundamental and important subject in radiation chemistry
of organic materials. A large number of works have been rePorted.'" Almost all of them are performed in glassy solids' or ( 1 ) Narayama, M.; Kevan, L.; Samskog, P.-0.; Lund, A,; Kispert, L. D. J . Chem. Phys. 1984, 81, 2297.
'Deceased. 0022-3654/89/2093-4898$01.50/0
0 1989 American Chemical Society
