COLLOIDS AND ELSEVIER
Colloids and Surfaces A: Physicochemicaland Engineering Aspects 132 (1998) 203-212
A
SURFACES
A small-angle neutron scattering study of the structure of graphitized carbon black aggregates in Triton X- 100/water solutions Vasil M. Garamus
a,., Jan Skov Pedersen b
a Frank Laboratory of Neutron Physics, Joint Institute for Nuclear Research, Dubna, Moscow Region, 141980, Russia b Department of Solid State Physics, Riso National Laboratory, Roskilde, Denmark
Received 21 October 1996; accepted 5 June 1997
Abstract
The structure of graphitized carbon black (CB) aggregates dispersed in water solutions with a non-ionic surfactant are studied by small-angle neutron scattering using contrast variation by heavy/light water mixing. The addition of CB to Triton X-100/water mixtures shifts the critical micelle concentration to a lower value. The CB aggregates have a fractal structure and the apparent fractal dimension is lower near the match point (75% heavy water). The scattering data are modelled using fractal-like aggregates (CB + surfactant), and voids in the CB particles and micelles. The data are fitted simultaneously for three different contrasts. The fractal dimension is found to be larger than 3 with the maximum size of the fractal aggregate being around 150-200 A. The primary CB particles have a broad size distribution with an average size of about 30-80 .~. The surfactant coverage of the CB particles is 8% and is constant with varying CB and surfactant concentration. The volume fraction of the voids does not exceed 1% of the CB. The micelle structure is found to be the same as in surfactant/water solutions. © 1998 Elsevier Science B.V. Keywords: Graphitized carbon black aggregates; Micellar solutions; Small-angle neutron scattering
1. Introduction
Studies of the adsorption of non-ionic surfactants are important in the development of the physicochemical background for adsorption purification of waste water and the stabilization of particles in dispersed systems. Graphitized carbon black (CB), having a homogeneous surface with a known crystal and chemical structure, can serve as a model object for studying the adsorption of different surfactants. The effects of adsorbed layers on the stability of dispersions, as well as on the * Corresponding author. Fax: (+ 7) 09621 65882; e-mail:
[email protected] 0927-7757/98/$19.00 © 1998 Elsevier Science B.V. All rights reserved. PII S0927-7757 (97)00185-4
electric properties of the interfaces, can be studied using systems consisting of CB particles and ionic and non-ionic surfactants. Extensive studies concerning the influence of polymers on the stability of solid dispersions have been reported [1]. M a n y investigations have dealt with adsorption studies at the solid-liquid interface of a variety of uncharged polymers [2]. In particular, it has been found that polycyclic aromatic hydrocarbons are adsorbed strongly by CB due to the attraction of van der Waals forces [3]. Graphitized CB has been studied [4] by gas adsorption and liquid immersion techniques; from the results, it has been inferred that the graphon surface can be described as consisting of basal
204
V.M. Garamus, J.S. Pedersen / Colloids Surfaces A: Physicochem. Eng. Aspects 132 (1998) 203-212
planes of graphite. The surface was shown to be highly homogeneous with barely any ionizable sites on polar groups. This work was confirmed by an X-ray photoelectron spectroscopy analysis [51. The aim of the present study is to determine the structural characteristics of CB particles stabilized by micellar Triton X-100 in water solutions. We used the small-angle neutron scattering (SANS) technique, which is a method that is well-suited to structural analysis. This technique has been used in some recent studies of CB/polymer compositions, which reported a fractal-like structure of the CB aggregates with large inner inhomogeneities [6,7].
2. Experimental Graphitized CB was provided by the Institute of Technical Carbon (Omsk, Russia). The Brunauer-Emmett-Teller (BET) area, calculated from N 2 adsorption, was 105m 2g-1. The 1,1,3,3-tetramethylbutylphenol ethylene oxide CsHlvC6H4(OC2H4),OH (Triton X-100), where n = 9 or 10, was obtained from Rohm and Haas, USA. The critical micellar concentration (CMC) in aqueous solution, measured by the surface tension method [8], was equal to 0.24 mmol 1-1 (corresponding to 0.15 g l-l). Doubly distilled light water, and heavy water with 98 mol% of deuterium were used in preparing the samples. The samples were prepared by weight, and an ultrasonic device was used to facilitate the dispersion. All samples were stored for 24 h before the scattering experiments were performed. A set of SANS experiments was performed at the DR3 reactor at Riso National Laboratory, Denmark [9]. The range of scattering vectors q from 0.003 to 0.5 A- 1 was covered by four combinations of neutron wavelength (3, 8 and 10 .~) and sample-to-detector distances (from 1 to 6 m). The wavelength resolution was 18% (full-width-at-halfmaximum value). The samples were kept in quartz cells (Hellma) with a path length of 1 or 2 mm. The raw spectra were corrected for backgrounds from the solvent, sample cell, and other sources by conventional
procedures [10]. The two-dimensional isotropic scattering spectra were azimuthally averaged, converted to an absolute scale, and corrected for detector efficiency by dividing by the incoherent scattering spectra of pure water [ 11 ]. Furthermore, the scattering intensity was normalized by dividing by the concentration of CB. Another set of SANS experiments was performed at the " M U R N " time-of-flight SANS spectrometer on the IBR-2 pulsed reactor at the Frank Laboratory of Neutron Physics, Joint Institute for Nuclear Research, Dubna, Russia [12]. The range of scattering vectors q from 0.01 to 0.4 A-1 was covered with a sample-to-detector distance of 6 m. The neutron wavelength interval was from 0.7 to 5 A. The intensities, which were recorded by a one-dimensional detector, were background corrected by subtracting the scattering from the solvent. Calibration of the neutron intensity was performed using a vanadium standard. Throughout the data analysis, corrections were made for instrument smearing [9,13]. For each instrumental setting, the ideal scattering curves were smeared by the appropriate resolution function when the model scattering intensity was compared to the measured intensity by means of the least squares method. The parameters in the models were optimized by conventional least squares analysis and the errors of the parameters were calculated by conventional methods [14]. Transmission electron microscope measurements were performed with a Philips CM-10 microscope at the Catholic University of Leuven, Leuven, Belgium. The microscope was operated at an accelerating voltage of 60 kV. The solutions were partially cleared from the surfactant by dialysis. The samples were placed on a grid (160 lines per 1 cm) and dried. The magnification varied from x 10 000 to x 73 000.
3. Results A set of experiments varying the CB concentration (10, 25 and 50g1-1) and Triton X-100 concentration (1.83 to ll.4g1-1) using three different contrasts (50, 75 and 100% D20) was performed. The variation of the surfactant concen-
V.M. Garamus, J.S. Pedersen / Colloids Surfaces A: Physicochem. Eng. Aspects 132 (1998) 203 212
tration was related to the CB concentration with the aim to keeping the surfactant/CB ratio constant between different solutions. This was done in order to distinguish both the dependence of the CB/surfactant structure on the CB concentration and on the surfactant concentration. The average neutron scattering length densities of CB and Triton X-100 are 6.5x 101°cm -2 and 0.6 × 101° cm -z respectively. That is why, we supposed that we could eliminate the scattering contribution from CB at 100% D20 (neutron scattering length density 6.33 x 101° cm -2) and Triton X-100 at 50% D20 (neutron scattering length density 2.59 x 101° c m 2) solvents. A typical log-log plot of the neutron scattering cross-section dS(q)/df2 as a function of the length of the scattering vector q is shown in Fig. 1. At
10 6
•
100% D20
D
7 5 % D20
•
50% D20
105 '
"7
10 4
'
E (9
O"
10 a .
M
low tl one observes a power-law behaviour with indications of a cross-over to a constant value at the lowest q values. At higher q, one observes a shoulder for 75 and 100% D20 that can be associated with the Triton X-100 micelles. The behaviour at low q is characteristic of fractal objects. This suggestion is in agreement with the transmission electron microscopic experiments (Figs. 2 and 3). These data show that the structures consist of smaller, almost spherical, units that are connected in an open ramified structure--which looks like a fractal aggregate [15]. For all contrasts, the measured cross-section at low q (q<0.01 ,~-1) follows approximately a power law: d S ( q ) / d f 2 ~ q -°, where D can be considered to be the apparent fractal dimension. At larger q (q >0.02 ~-1), the measured points fall on a straight line only for the 50% DzO data; in the case of the 75% and 100% D20 data, one observes a shoulder originating from the micelles and the cross-section follows a Guinier-like law [16] behaviour, d_r(q)/dI2= dZ'(0)/dg2 exp(-q2R~/3), where Rg is the radius of gyration of the particle/micelle. The determination of D from the slope of the log-log plots of dL'(q)/d(2 at low q does not always provide correct results for the fractal dimension. In particular, the existence of cut-off lengths occurring in real fractal objects may lead to wrong estimates of the fractal dimension. In spite of this we performed fits to the low-q part of the data with the power law. dZ(q)/dI2 ~ q - O
10 2 •
101
1oo 0.001
0.01
0.1
q, A "I
Fig. 1. Measured n e u t r o n scattering cross-sections for mixtures with 10 g 1 - 1 of CB and 1.83 g 1 - 1 of Triton X-100 in different light/heavy water mixtures and model fit (solid line) by Eq. (6) (100% and 75% D 2 0 data are multiplied by 100 and 10 respectively, for clarity).
205
( 1)
We note that this can only be considered as an empirical parameterization of the data and that D should be considered only as an apparent fractal dimension (as previously mentioned above). Results obtained using Eq. ( 1 ) are given in Table 1. One observes a stronger dependence of the apparent D on contrast than on the concentration of CB and Triton X-100. The lowest values of D are found in 75% D20. In the case of 50% D20, the cross-section locally follows a power law with different values of D (Fig. 4). Four different regions can be distinguished, although some of them are quite narrow in q. The D values are given in Table 2. The variation of D is larger than statistical errors, so
206
V.M. Garamus, J.S. Pedersen / Colloids Surfaces A: Physieoehem. Eng. Aspects 132 (1998) 203-212
7~
Fig. 2. The transition electron microscopy images of CB/Triton X-100 dispersions; CB 10 g 1 1 and Triton X-100 1.83 g 1-1 in light water; x 73 000 magnification.
Fig. 3. The transition electron microscopy images of CB/Triton X-100 dispersions; CB 10 g 1-1 and Triton X-100 1.83 g 1-1 in light water; x 15 500 magnification.
the differences are significant. Some of the regions are quite narrow, and the differences in the slope could be due to cross-overs between different
regions related to the characteristic length scales of the objects, i.e. from the region of the whole object to the primary particle size.
V.M. Garamus, J.S. Pedersen / Colloids Surfaces A." Physicochem. Eng. Aspects 132 (1998) 203 212
207
Table 1 Results for the apparent fractal dimension D obtained from power law fits using Eq. (1), fitting the range of q from 0.003 to 0.01 .~ '
The data in 100% D20 can be modelled in the interval from 0.02 to 0.5 ~,-1 by
Concentration CB/ Triton X- 100 (gl -:)
D (100% D20 )
(75% D20 )
10/1.83 10/3.18 10/4.68 25/4.36 25/5.7 25/7.8 50/8.57 50/9.9 50/11.4 20/0.0 1/0.14
2.92+0.03 3.04_+0.04 2.96_+0.03 3.1 _+0.2
2.29_+0.03 2.47_+0.02 2.35 _+0.04 2.51 _+0.05
where A/q ° is the scattering from fractal objects, BF2(q,v,R) is the scattering form factor of a rotational ellipsoid with the semi-axes (R,R,vR), which describes the micellar contribution. The dependences of B, R and v on the Triton X-100 concentration at different CB concentrations are given in Table 3. The micellar parameters stay constant, which agrees with the results [17] for Triton X-100/water solutions. The number of micelles increases with varying surfactant concentrations. We also performed measurements for CB/Triton X-100/water mixtures with a concentration of Triton X-100 that is less than the CMC for the pure Triton X-100/water solution. The scattering data could be fitted by expression in Eq. (2) with v = l , i.e. by the form factor of a sphere. We obtained the values: B=(5.2-1-0.8)x 10-3cm ' and R = 25 + 5 ,~. The magnitude of the radius is very close to the average radius of the Triton X- 100 micelles in the Triton X- 100/water solutions, where the surfactant concentration is larger than the CMC. We can conclude that the addition of CB particles shifts the CMC of Triton X-100 in water to a lower value. The SANS data obtained for a dry CB powder show a typical power law behaviour of the crosssection (Fig. 5). The data were analysed using Eq.(1) and the value obtained tbr D was 3.72_+0.03, which is characteristic for an object with a fractal surface.
2.28 _+0.03 2.17 _+0.06 2.93 -t-0.04 2.8 _+0.1 2.43 _+0.25
2.41 -+0.04 2.42 _+0.04 2.43 _+0.06 2.5 _+0.2
3.2_+0.1
. . . . . . . 10 4
(50% D20 )
I
"
o o 10 3 '
II B
=
a
°° o
~ Qo
%
•
,
10 2 ' Od
E "o
101,
10 o ,
(2)
% o
\
4. Discussion and modelling (b)
1 0 "1
1 0 -2
dZ'(q)/dI2 = Aq ° + BF2(q,v,R) + C
. . . . . . .
i
. . . . . . . .
I
0.1
0.01
q , A "1 Fig. 4. Measured neutron scattering cross-sections for CB/Triton mixtures; CB/Triton X-100 concentrations (a) 10/1.83 g 1-1 and (b) (divided by 10) 50/8.57 g 1-1, 50% D20.
As discussed above, the shapes of the scattering curves (Fig. 1) point towards the presence of different kinds of inhomogeneities in the samples. At small q values, the main contribution is from the fractal-like CB structures. We assume that the CB particles with adsorbed Triton X-100 molecules also form fractal aggregates. The reasons for this assumption are: (i) the CB particles themselves have a fractal-like structure in CB/polymer systems [7]; (ii) our electron microscopy observations
V.M. Garamus, J.S. Pedersen/ Colloids Surfaces A." Physicochem. Eng. Aspects 132 (1998) 203-212
208
Table 2 "Local" apparent fractal dimensions obtained using Eq. (1) for 50% DzO data varying the q fitting range Concentration CB/ Triton X-100 (g 1-1)
D 0.003-0.01 ,~-1
0.01 0.027 ,~-1
0.027-0.07 ~-1
0.07-0.5 A -1
10/1.83 50/8.57
2.4+0.1 2.4+0.1
3.20___0.15 3.4±0.I
2.5±0.2 2.20±0.15
3.2±0.4 3.2±0.1
Table 3 Simple power forfractals and ellipsoid formicelles (Eq. (2) Concentration Triton X-100 (g 1- 1)
B (x 102cm -1)
R (A)
0.14 1.83 3.18 4.68
0.52±0.08 1.97±0.05 2,99±0.06 3.70±0.06
25±5 23.3±0.2 23.1±0.2 23.3±0.1
100
.
.
.
.
.
.
.
.
1 2.7±0.1 2.7±0.1 2.5±0.1
=
10.
"7,
oE
0.1
(Figs. 2 and 3) of the CB used in the present work. In addition, the presence of a power-law behaviour at low q in the SANS data for CB in water and Triton/water solutions points towards the presence of fractal-like objects. (In the case where pure water is used as the solvent, one cannot obtain stable dispersions.) The intermediate and high-q parts of the scattering curves are affected by the Triton X-100 micelles, which gives rise to a shoulder in the scattering data. That this contribution originates from the micelles can be concluded from the fact that the parameters of the particles obtained by modelling the data by a power law plus ellipsoidal particles are the same as in the case of the Triton X-100/water systems. In addition, the scattering length density of the particles is close to those of the Triton X-100 micelles [17] without the CB particles. The third expected contribution to the scattering intensity is from voids in the CB particles. The presence of voids was detected in other studies [18]. It is, of course, a complication that the voids also have to be considered, as it introduces additional fitting parameters. However, our analysis has shown that it is not possible to model the recorded scattering data without including the contribution from the voids. For a dilute population of CB objects suspended in a medium of scattering-length density p,, the position-dependent scattering length density p(r) within a particle can be described by p(r) = p, + ((p> - p,) + ~(r)
.
0.03
.
.
0.06
.
(4)
i
0.1
0.5
q,A 1
Fig. 5. Measured neutron scattering cross-section of dry CB powder and fit (solid line) by the simple power law (Eq. (1).
where (p> is the average of p(r) within the particle shape, and ~(r) describes the internal fluctuations of the scattering length density around an average value, i.e. ~ ( r ) = p ( r ) - ( p > . The average contrast
V.M. Garamus,J.S. Pedersen/ ColloidsSurfaces A: Physicochem.Eng. Aspects 132 (1998) 203 212 is the difference between an average object and the solvent scattering length density, A p = ( p ) - p s . The condition Ap = 0 is called the contrast match point. The scattering intensity is proportional to the square of the scattering amplitude (the Fourier transform of Eq. (4), and, hence, it can be written as an equation quadratic in Ap. For a dilute population of objects, the intensity is given by
dZ(q, Ap)/dY2 = Ap2 (d£'(q)/d(2),~ + Ap(dX(q)/dK2)o~¢ + (dZ(q)/dO)¢
(5)
where (dX(q)/d(2)o, is the scattering intensity due to the shape of the sample volume that excludes solvent, (dL'(q)/d(2)~ is the scattering intensity from the internal structure of the solvent-excluding parts, and (dZ(q)/dg2)o~, is the scattering due to the correlation between the shape and the internal structure of the solvent-excluding part of the sample. The observed fractal dimension (Table 1) at low q depends on the deuterated fraction of the water. For 100% deuterated water, one can observe the power laws d,S(q)/dQ~q-Z95; at 75%: dZ(q)/ dQ,-~q-23; and at 50%: dZ(q)/dg2~q -25. Such a large effect shows that there is a very strong fluctuation of the scattering length density inside the CB particles (cf. Eq. (4). The lower value of the observed fractal dimension in 75% deuterated water shows that it is close to the contrast match point of the CB-Triton X-100 particles: the scattering is dominated by internal fluctuations of the scattering length density. The radii of the primary particles making up the fractal aggregates can be estimated from the interval in which the exponent of the power law changes from a value characteristic of a mass fractal aggregate to an exponent for a surface fractal [19]. The scattering data from CB/Triton X-100 in the present work display such a crossover region (Table 2 and Fig. 3). The fact that the transition occurs in essentially the same interval of q for all samples indicates that all of these aggregates are composed of primary particles of similar radius [19]. The radius can be estimated to be approximately 40 ,~.
209
The scattering intensity of the solutions was modelled by dZ(q)/d£2 = A ~( ~2 (q, ( R ),a,L ) ) S(q,d,~ )
+ A2AaptrF~el(q,a,v) + A1A3pZ~p(q,r) + A4
(6)
where the first term is the contribution from the CB aggregates with an adsorbed Triton X-100 layer. We assumed that the fractal-like CB aggregates consist of connected polydisperse primary particles. The organization of the primary particles within a cluster is described by the fractal structure factor S(q,d,~). The scattering from the primary particles is described by (~bZ(q,(R),a,L)), which is the scattering from a system of polydisperse two-layer spheres (CB core and Triton X-100 layer) with an average core radius ( R ) and a size polydispersity a. We used a Gaussian distribution of CB core radii:
f(R) = ~
l
exp
[
2
2~
]
(7)
Note that the negative-R part of this distribution is omitted. The form factor of a primary particle with radius R is ~2 (q,R,L) = [V(R)F(R,q) (Pcb Ptr) -
-
+~V(R + L)F(R + L,q)(ptr-ps)] 2 (8) where L is the Triton X-100 layer thickness, Peb, Ptr and ps are the scattering length densities of the CB core, Triton X-100 and solvent (mixture of heavy and light water) respectively. The parameter ~ is the fraction of the CB surface covered by surfactant molecules (0 < ~ ~<1), F(r,q) is the scattering form factor of a sphere of radius r, and V(r) is the volume of such a sphere. The final contribution from the primary particles was obtained by a numerical calculation of ('
(q~2(q'(R)'a'L)) = J
-5, ~b2(q,R,L)f(R) dR (9) The term S(q,d,~) in Eq. (6) is the structure
V.M. Garamus, J.S. Pedersen/ Colloids Surfaces A: Physicochem. Eng. Aspects 132 (1998) 203-212
210
factor of a fractal [20]
Dr(D - 1) x sin[(D- 1) tan -1 (q()]
S(q)-- 1 +
[ 1],o-1,,2 (qR)°[_ 1 _]
(10)
+ (q~)2
where R' is an average radius of the primary particles, ~ is the size of aggregates, D is the fractal dimension, and F(x) is the gamma function. The second term in Eq. (6) describes the scattering from the Triton X-100 micelles, which is modelled by an ellipsoid of rotation, i.e. ~(q,a,v) is the scattering form factor of an ellipsoid of rotation with the semiaxis (a,a,va). The factor A2g is the contrast of Triton X-100 in the solvent, i.e.
(Ptr- p,)2. The third term in Eq. (6) describes the scattering from the voids in the CB particles, which are assumed to be spherical, and ~p(q,r) is the form factor of a sphere of radius r. The parameters A1, A2, AIA3 are the prefactors connected with the number and volume of CB particles, Triton X-100 micelles, and voids respectively. The factor A4 describes the residual incoherent background in the data. The scattering data were fitted simultaneously for two or three samples with different contrasts
but with the same CB and Triton X-100 concentrations. The results of the fitting are presented in Fig. 1 and Table4. Table4 also contains the volume fraction fm of the micelles as well as the effective average radius of primary particles R'=(5/3)l/2Rg, where Rg is calculated for a polydisperse system Rg=3(R8)/5(R6). The volume Vm=47t/3va 3 of a micelle is also given. The values obtained by direct modelling for the fractal dimension are quite different from those determined by the simple fractal expression of Eq. (1). The values of D are larger than 3, which points to the fact that this surface is so porous and rough that it practically fills the volume space [21]. One can, therefore, assume that the volume and surface terms play equally important roles in the thermodynamic equation for the system [22]. The slight decrease of the prefactor of the term for the fractal objects with varying CB concentrations can be explained as cluster-cluster overlap that leads to lower values of the observed prefactors and slope (fractal dimension). This is a screening effect similar to the one observed for semi-dilute polymer solutions. The size of the fractal aggregates is around 150-200,~. The volume fraction of the voids in the CB particles does not exceed 1%. The decreasing average radius indicates a greater
Table 4 Results obtained by fitting Eq. (6). The parameters are defined in the text Parameter
A1 (g2 cm-1) O (R> (A) (~,) R' (]k) ((,&) A2 ( 10- 2cm- 1) a (,&) v Vm (,~3× 105) fm X 10 -3 A 3 X 10 -3 r (,~,) L (A)
CB/Triton X-100 concentrations 10/1.83
10/3.18
10/4.68
50/8.57
50/9.9
9.21 +0.05 3.61 _0.01 -(19.7+0.1) 142 __+2 70+2 151.3+0.5 1.89 + 0.02 25.5+0.1 2.52 _+0.01 1.75+0.05 1.1 +0.1 8.5±0.2 15.9+0.1 0.076 + 0.001 32.8+0.1
29.8+0.1 3.42+0.01 - ( 3 4 + 1) 80.4 ___1.5 31 _ 1 187.9+0.6 2.74 + 0.03 24.9+0.1 2.27 + 0.01 1.46+0.04 1.9+0.2 2.5+0.1 14.3+0.1 0.078 _+0.001 27.1 +0.1
24.3+0.1 3.45+0.01 - ( 3 4 + 1) 80.1 + 1.4 31 + 1 154.5+0.5 3.49 + 0.04 25.7+0.1 2.10 + 0.01 1.49+0.04 2.3+0.2 6.4+0.2 17.6+0.1 0.092 + 0.002 15.50+0.08
3.36___0.01 3.05+0.01 -(70+2) 81.3 + 1.4 20.2+0.8 199.0+0.6 1.593 + 0.02 29.0+0.1 1.64 + 0.01 1.67+0.04 1.0+0.1 9.5+0.3 17.9+0.1 0.075 + 0.001 5.60+0.05
4.31 +0.01 3.01 +0.01 -(70+2) 78.1 + 1.3 19.2+0.8 199.6+0.6 2.89 + 0.03 26.7+0.1 1.67 _ 0.01 1.33+0.03 2.2+0.2 10.0+0.4 19.2+0.1 0.074 + 0.001 3.02+0.05
V.M. Garamus, J.S. Pedersen / Colloids Surfaces A." Physicoehem. Eng. Aspects 132 (1998) 203-212
repulsive pressure in the system with larger surfactant concentration. The repulsive pressure is due to the forces which work against the aggregation of the CB/Triton X-100 particles that were formed when the ultrasonic device was used. The surface coverage of the CB particles by the surfactant molecules is 8% and remains constant with varying CB and surfactant concentrations. The thickness of the Triton X-100 layer absorbed by the CB particles does not exceed the length of the Triton X-100 molecule and decreases with increasing Triton X-100 concentration. These results show that monolayer adsorption takes place, and they agree with the observed effect of decreasing intensity in the 100% D20 solvent (Fig. 6). At this contrast, one can expect the main scattering contribution to be from the Triton X- 100 and the one order of magnitude decrease in the 10 4
211
scattering intensity could be connected with a change in the parameters of the Triton X-100 layers. The systematic variation in layer thickness can be connected with the increasing total surface of CB as we obtained that the average radius of primary particle becomes smaller. The difference can be explained by a change in the direction of the adsorbed Triton X-100 molecules from perpendicular to the surface (32 A) to parallel (3 A). As mentioned above, the addition of CB particles shifts the CMC of Triton X-100 in water to a lower value. It is known that different additives change the CMC. For example, the lyotropic salts decrease the CMC values of non-ionic surfactants [23]. In the present work, the hydrophobic dispersion of CB decreases the CMC, i.e. the addition of CB is equivalent to a decrease of effective hydrophilicity of the non-ionic surfactant.
. . . . . . . . . . . . . . . . . . . .
5. Conclusions
103.
"7
102= 0"
\\
"0
101
Y
10 0
'
0.001
'
'
'
' ' ' ' 1
0.01
0.1
q, A -1 Fig. 6. Measured neutron scattering cross-sections of CB/Triton mixtures: (a) 10/1.83gl -1 and (b) 50/8.57gl -1 in 100% D20.
The CB powders dispersed in Triton X-100/ water mixtures form fractal-like aggregates with a fractal dimension larger than 3. The aggregate size is 150-200 ,~. The primary particles constituting the fractals have a broad size distribution and an average size of about 20-70 .~. The CB particles contain voids with a volume fraction that does not exceed 1%. The coverage of the CB surface by Triton X-100 molecules remains constant in the range of CB and Triton X-100 concentrations investigated. The width of the Triton X-100 layer confirms the observation of monolayer adsorption. The presence of the CB particles shifts the CMC of the non-ionic surfactant to a lower value. The structure of the micelles was found to be the same as in pure Triton X-100/water mixtures. There is a significant volume fraction of micelles in the CB/Triton X-100/water mixtures even at surfactant concentrations close to the CMC. The constant degree of coverage of the CB particle surface with various surfactant concentrations shows that there is a limit to the adsorption (purification) capacity of the CB sorbents. The size of the CB aggregates is appropriate for use in connection with separating membranes.
212
V.M. Garamus, J.S. Pedersen/ Colloids Surfaces A: Physicochem. Eng. Aspects 132 (1998) 203-212
Acknowledgment T h e help o f D r . - I n . A . M . A e r d t s a n d P r o f e s s o r H. R e y n a e r s in the p e r f o r m a n c e o f the e l e c t r o n m i c r o s c o p y is g r a t e f u l l y a c k n o w l e d g e d .
References [1] Th.F. Tadros, Adv. Colloid Interface. Sci. 12 (1980) 141. [2] A.J. Groszeck, Trans. Faraday Soc. 7 (1975) 109. [3] K.M.G. van Dolsen, Ph.D. Thesis, University of Southern California, 1970. [4] G.D. Parfitt, E.J. Willis, J. Phys. Chem. 68 (1984) 1780. [5] S. Bettarini, L. Pedocchi, G. Rovida, G. Gobrielli, unpublished results, 1990. [6] L. Salome, J. Phys. II (France) 3 (1993) 1647. [7] R.P. Hjelm Jr.,, W.A. Wampler, P.A. Seeger, M. Gerspacher, J. Mater. Res. 9 (1994) 3210. [8] V.N. Moraru, F.D. Ovcharenko, L.I. Kobylinskaya, T.V. Karmazina, Colloid J. Russ. Acad. Sci. Engl. 46 (1984) 1148. [9] J.S. Pedersen, J. Phys. IV (Paris) Colloq. 8 (3) (1993) 491. [10] J.P. Cotton, in: P. Lindner, T. Zemb (Eds.), Neutron, X-Ray and Light Scattering: Introduction to an
Investigative Tool for Colloidal and Polymeric Systems, North-Holland, Amsterdam, 1991. [11] G.D. Wignall, F.S. Bates, J. Appl. Crystallogr. 20 (1986) 28. [12] Yu.M. Ostanevich, Makromol. Chem. Macromol. Symp. 15 (1988) 91. [13]J.S. Pedersen, D. Posselt, K. Mortensen, J. Appl. Crystallogr. 23 (1990) 321. [14] B.R. Bevington, Data Reduction and Error Analysis for Physical Sciences, McGraw-Hill, New York, 1969. [15] B.B. Mandelbrot, in: L. Pietronero, E. Tosatti (Eds.), Fractals in Physics, North-Holland, Amsterdam, 1986, pp. 3-28. [16] A. Guinier, G. Fournet, Small Angle Scattering of X-Rays, Wiley, New York, 1955. [17] L.A. Bulavin, V.M. Garamus, T.V. Karmazina, S.P. Shtanko, Colloid J. Russ. Acad. Sci. Engl. 57 (1995) 902. [18] W.M. Hess, C.R. Hurd, in: J.-B. Donner, R.C. Bansal, M.Z. Wang (Eds.), Carbon Black, Dekker, New York, 1993, pp. 91 106. [19] A.J. Hurd, D.W. Schaefer, J.E. Martin, Phys. Rev. A 35 (1987) 2361. [20] J. Teixeira, in: H.E. Stanley, N. Ostrovsky (Eds.), On Growth and Form, Nijhoff, Dordrecht, 1986. [21] P. Pfeifer, D. Avnir, D. Farin, J. Stat. Phys. 36 (1984) 699. [22] P. Pfeifer, D. Avnir, J. Chem. Phys., 79 (1983) 3558; erratum, J. Chem. Phys. 80 (1984) 4573. [23] M. Kahlweit, R. Strey, D. Haase, J. Phys. Chem. 89 (1985) 163.