Langmuir 2004, 20, 565-571 565

Synergistic Sphere-to-Rod Micelle Transition in Mixed

Solutions of Sodium Dodecyl Sulfate and Cocoamidopropyl
Betaine
N. C. Christov,† N. D. Denkov,*,† P. A. Kralchevsky,†
K. P. Ananthapadmanabhan,‡ and A. Lips‡
Laboratory of Chemical Physics & Engineering, Faculty of Chemistry, Sofia University,
1164 Sofia, Bulgaria, and Unilever Research U.S., 45 River Road,
Edgewater, New Jersey 07020
Received September 14, 2003. In Final Form: November 18, 2003

Static and dynamic light scattering experiments show that the mixed micelles of sodium dodecyl sulfate
(SDS) and cocoamidopropyl betaine (CAPB) undergo a sphere-to-rod transition at unexpectedly low total
surfactant concentrations, about 10 mM. The lowest transition concentration is observed at molar fraction
0.8 of CAPB in the surfactant mixture. The transition brings about a sharp increase in the viscosity of
the respective surfactant solutions due to the growth of rodlike micelles. Parallel experiments with mixed
solutions of CAPB and sodium laureth sulfate (sodium dodecyl-trioxyethylene sulfate, SDP3S) showed
that the sphere-to-rod transition in SDP3S/CAPB mixtures occurs at higher surfactant concentrations,
above 40 mM. The observed difference in the transition concentrations for SDS and SDP3S can be explained
by the bulkier SDP3S headgroup. The latter should lead to larger mean area per molecule in the micelles
containing SDP3S and, hence, to smaller spontaneous radius of curvature of the micelles (i.e., less favored
transition from spherical to rodlike micelles). The static light scattering data are used to determine the
mean aggregation number and the effective size of the spherical mixed SDS/CAPB micelles. From the
dependence of the aggregation number on the surfactant concentration, the mean energy for transfer of
a surfactant molecule from a spherical into a rodlike micelle is estimated.

1. Introduction Alkylbetaines and their derivatives represent a class of

Mixed solutions of anionic and zwitterionic surfactants zwitterionic surfactants, which are electroneutral inner
are used in various detergent formulations because of their salts in a wide pH range, with a positive charge on the
superior properties in comparison to the properties of nitrogen atom and a negative charge on the carboxyl group.
individual surfactants.1-3 For example, mixed solutions Alkylbetaines and alkylamidobetaines are among the most
often have lower critical micelle concentration (cmc) due widely used cosurfactants in mixtures with anionic
to a strong attraction between the molecules of the anionic surfactants, such as sodium dodecyl sulfate (SDS) and
and zwitterionic surfactants.1-8 The low cmc is beneficial various sodium laureth sulfates (with 2, 3, or 4 ethoxy
in the case of facial cleansers, shampoos, and baby-care groups), especially in shampoos and liquid detergents,
products, because it leads to a reduced irritation action because they have pronounced foam/lather boosting,
of the surfactant solution on the eyes and skin.1-3 In antistatic, and hair-conditioning effects. Betaines also find
addition, the mixed solutions often have improved foaming, industrial applications in textile production (mainly due
emulsifying, and rheological properties, at the same total to their antistatic and softening properties) and as
surfactant concentration, which are important for their emulsifiers and dispersants, for example, for photographic
applications.1-3,8-10 emulsions.1-3
In relation to the above applications, mixtures of
* To whom correspondence should be addressed. Assoc. Prof. alkylbetaines and anionic surfactants have been inves-
Nikolai D. Denkov, Laboratory of Chemical Physics & Engineering, tigated to understand the phenomena that are responsible
Faculty of Chemistry, Sofia University, 1 James Bourchier Ave., for the synergistic action of these systems. Rosen and co-
1164 Sofia, Bulgaria. Phone: (+359-2) 962 5310. Fax: (+359-2) workers4,11,12 characterized the adsorption layers and
962 5643. E-mail: [email protected].
† Sofia University. micelles in mixed surfactant systems in terms of the so-
‡ Unilever Research U.S. called molecular interaction parameters (MIPs). These
(1) Amphoteric Surfactants; Lomax, E. G., Ed.; Surfactant Science authors4,11,12 found that MIPs are strongly negative for
Series, Vol. 59; Marcel Dekker: New York, 1996. mixtures of anionic surfactants and betaines, indicating
(2) Domingo, X. In Amphoteric Surfactants; Lomax, E. G., Ed.; a significant intermolecular attraction (see Figure 1). The
Surfactant Science Series, Vol. 59; Marcel Dekker: New York, 1996;
Chapter 3. molecular interactions of betaines with anionic surfactants
(3) Tsujii, K. Surface Activity: Principles, Phenomena, and Applica- were studied also, in relation to other properties of the
tions; Academic Press: New York, 1998. surfactant mixtures, which are important for their ap-
(4) Rosen, M. J. Surfactants and Interfacial Phenomena, 2nd ed.;
Wiley-Interscience: New York, 1989. plications (e.g., Kraft temperature).5,13-16 Iwasaki et al.8
(5) Holland, P.; Rubingh, D. N. J. Phys. Chem. 1983, 87, 1984. Holland,
P. Adv. Colloid Interface Sci. 1986, 26, 111. (10) Basheva, E. S.; Stoyanov, S.; Denkov, N. D.; Kasuga, K.; Satoh,
(6) Abe, M.; Kato, K.; Ogino, K. J. Colloid Interface Sci. 1989, 127, N.; Tsujii, K. Langmuir 2001, 17, 969.
329. (11) Rosen, M. J.; Zhu, Z. H. J. Colloid Interface Sci. 1984, 99, 427,
(7) Shiloach, A.; Blankschtein, D. Langmuir 1997, 13, 3968. Mulqueen, 435.
M.; Blankschtein, D. Langmuir 2000, 16, 7640. (12) Rosen, M. J.; Zhu, Z. H. J. Am. Oil Chem. Soc. 1988, 65, 663.
(8) Iwasaki, T.; Ogawa, M.; Esumi, K.; Meguro, K. Langmuir 1991, (13) Hines, J. D.; Thomas, R. K.; Garrett, P. R.; Rennie, G. K.; Penfold,
7, 30. J. J. Phys. Chem. B 1998, 102, 8834.
(9) Basheva, E. S.; Ganchev, D.; Denkov, N. D.; Kasuga, K.; Satoh, (14) Tajima, K.; Nakamura, A.; Tsutsui, T. Bull. Chem. Soc. Jpn.
N.; Tsujii, K. Langmuir 2000, 16, 1000. 1979, 52, 2060.

10.1021/la035717p CCC: $27.50 © 2004 American Chemical Society

Published on Web 01/08/2004
566 Langmuir, Vol. 20, No. 3, 2004 Letters

the micelles). The data from the SLS experiments were inter-
preted by using the Zimm plot (see, e.g., refs 17 and 18):

KC 1
[
16 πn
( ) ]
2
) 1+ sin2(θ/2)Rg2 + 2A2C (2)
R(θ) MW 3 λW

where C is the total surfactant concentration in the solution (all

experiments were performed with solutions having concentra-
tions well above the cmc, so that the concentration of the
Figure 1. Schematic presentation of the electrostatic head- monomers, which are not incorporated in micelles, can be
head attraction between SDS and CAPB molecules, incorpo- neglected); R(θ) is the Rayleigh ratio, which is proportional to
rated in a micelle or adsorption layer. the intensity of the scattered light; θ is the scattering angle; Rg
is the radius of gyration of the micelles (too small in the studied
solutions to be reliably measured); λW is the wavelength of the
studied mixtures of SDS and alkylbetaines and found that illuminating light; K is an optical constant,
several properties of the surfactant blends (cmc, micelle
aggregation number, diffusion coefficient, and solubili- 2
4π2 n0 dn 2
zation capacity) strongly depend on the ratio of SDS/
betaine and exhibit maxima or minima when the molar
K)
NA λW4 dC ( ) (3)

fraction of alkylbetaine in the mixture is about 0.6. These

results were explained by strong electrostatic attraction NA is the Avogadro number; n0 is the refractive index of water;
between the SDS and alkylbetaine molecules in the mixed dn/dC is the refractive index increment, which is measured
micelles and adsorption layers. A comprehensive review independently by a Pulfrich refractometer.
From MW, we determined the aggregation number and
on this subject can be found in ref 2.
composition of the spherical micelles, that is, the number of SDS
In the present paper, we describe a systematic light and CAPB molecules per micelle, NSDS and NCAPB, by assuming
scattering study of the sphere-to-rod micelle transition in that the ratio of SDS and CAPB in the micelles is the same as
mixed solutions of cocoamidopropyl betaine (CAPB) and in the total surfactant mixture. The latter assumption is justified
SDS. It is found that the transition in this system occurs insofar as the studied solutions were of concentrations well above
at a rather low total surfactant concentration (≈10 mM) the cmc and the formation of two different micelle populations,
and is related to a sharp increase in the viscosity of the enriched in SDS and CAPB, is not expected in the region of
respective solutions. The data from light scattering are spherical micelles (due to the attraction between the SDS and
used to determine the micelle aggregation numbers and CAPB molecules in the micelles). Thus, NSDS and NCAPB can be
to evaluate the interaction between the SDS and CAPB calculated from the relations NSDS ) (1 - fCAPB)NS and NCAPB )
fCAPBNS, where fCAPB is the molar fraction of CAPB in the
molecules in the rodlike micelles. In addition, the light surfactant mixture and NS is the total aggregation number of
scattering data were used to characterize the micellar the spherical micelles, which is calculated from the Zimm plot:
effective volume fraction in relation to the oscillatory
surface forces, which are known to be operative in thin NS ) MW/[(1 - fCAPB)MSDS + fCAPBMCAPB] (4)
foam and emulsion films formed from micellar solutions.
In eq 4, we have used the fact that the weight-averaged mass of
2. Materials and Methods the micelles can be expressed as MW ) NSDSMSDS + NCAPBMCAPB,
where MSDS or MCAPB is the molecular mass of SDS or CAPB,
2.1. Materials. SDS (99%, a product of ACROS Co., NJ) and respectively.
sodium polyoxyethylene sulfate with three ethoxy groups (SDP3S) Note that the aggregation number of the rodlike micelles, NR,
of commercial name STEOL CS-330 (a product of Stepan Co., was estimated from DLS data by using another procedure,
Northfield, IL) were studied as anionic surfactants. CAPB of explained in section 4 below, because eq 2 is not applicable in
commercial name TEGO Betaine F 50 (a product of Golgschmidt the region of rodlike micelles and MW cannot be determined from
Chemical Co., Hopewell, VA) was used as a zwitterionic sur- the Zimm plot (see section 3.2 for further explanations).
factant. NaCl (a product of Merck KgaA, Darmstadt, Germany) The light scattering measurements were performed at 27 °C
was always added to the surfactant solutions in concentration by means of an instrument 4700C (Malvern Instruments, U.K.),
of 10 mM, as a neutral electrolyte. which is equipped with an argon laser, operating with a vertically
All solutions were prepared with deionized water from a Milli-Q polarized incident beam at λW ) 488 nm.
water purification system (Millipore) in glassware, which was The viscosity of the solutions was measured by means of a
precleaned by sulfochromic acid and rinsed with Millipore water. capillary viscometer thermostated at 27 ( 0.2 °C.
Just before performing the light scattering experiments, all
samples were filtered through a 100 nm filter to remove dust 3. Experimental Results and Discussion
particles (Millex VV, Millipore).
2.2. Methods. Dynamic light scattering (DLS) was used for 3.1. Concentration of Sphere-to-Rod Transition
measuring the translational diffusion coefficient, D, of the and Relation to Solution Viscosity. The DLS mea-
surfactant micelles. Then, the mass-averaged hydrodynamic surements with surfactant mixtures (at fixed ratio SDS/
diameter of the micelles, dh, was calculated by using the Stokes- CAPB) showed a well-pronounced transition in the size
Einstein relation of the mixed micelles at a certain total concentration of
surfactant, CSR. At low surfactant concentrations, the
dh ) kT/(3πηD) (1) hydrodynamic diameter of the micelles was dh ≈ 5 nm for
all studied systems, which corresponds to spherical
where η is the dynamic viscosity of the disperse medium, T is micelles. In Figure 2A, illustrative results are presented
temperature, and k is the Boltzmann constant. for surfactant mixtures with molar fraction of CAPB fCAPB
Static light scattering (SLS) was used to determine the total ) 0.5 and 0.75. After the break point in the curve of dh
mass, MW, and the second virial coefficient, A2, of the spherical
micelles (A2 brings information about the interaction between (17) Mazer, N. A. In Dynamic Light Scattering: Applications of Photon
Correlation Spectroscopy; Pecora, R., Ed.; Plenum Press: New York,
(15) Wüstneck, R.; Miller, R.; Kriwanek, J.; Holzbauer, H.-R. Lang- 1985; Chapter 8.
muir 1994, 10, 3738. (18) Kralchevsky, P. A.; Danov, K. D.; Denkov, N. D. In Handbook
(16) Tsujii, K.; Okahashi, K.; Takeuchi, T. J. Phys. Chem. 1982, 86, of Surface and Colloid Chemistry; Birdi, K. S., Ed.; CRC Press: New
1437. York, 1997, Chapter 11; 2nd ed., 2002, Chapter 5.
Letters Langmuir, Vol. 20, No. 3, 2004 567

Figure 2. (A) Mean hydrodynamic diameter, dh, of the micelles

as determined by DLS in mixed SDS/CAPB solutions at two
different ratios, 1:1 and 1:3. The arrows indicate the break- Figure 3. (A) Viscosity of the surfactant solutions containing
point concentrations in the respective Zimm plots, as determined SDS/CAPB ) 1:1, as a function of the total surfactant
by SLS. (B) Total surfactant concentration at the sphere-to-rod concentration. The arrow corresponds to the sphere-to-rod
transition, CSR, as a function of the molar fraction of CAPB in transition, as determined by light scattering. (B) Dependence
the mixture, fCAPB. The circles correspond to mixtures of CAPB of the hydrodynamic diameter, dh (circles), and of the solution
with SDS, whereas the empty squares are for mixtures with viscosity (empty squares) on the molar fraction of CAPB in the
SDP3S. The curves are guides to the eye. surfactant mixture, fCAPB, at total surfactant concentration C
) 0.1 M.
versus C, the size of the micelles rapidly increases and dh
becomes larger than 10 nm. For comparison, the size of Similar light scattering experiments were performed
the single-component SDS micelles17,19,20 is dh ) 4.8 nm with mixed solutions of CAPB and SDP3S (another anionic
and for the CAPB micelles we measured dh ) 5.2 ( 0.5 surfactant used in practical formulations, which differs
nm in the entire concentration range shown in Figure 2A. from SDS by the presence of three ethoxy groups in the
In Figure 2B, we plot the dependence of the transitional surfactant head). The experiments showed that the sphere-
concentration, CSR, on the molar fraction of CAPB in the to-rod transition in the SDP3S/CAPB mixture occurs at
mixture. One sees that the transition occurs at total higher total surfactant concentrations, CSR > 40 mM, as
surfactant concentrations as low as 10 mM, for CAPB compared to the SDS/CAPB mixtures (see the dotted curve
molar fraction ≈ 0.8. For comparison, the sphere-to-rod in Figure 2B).
transition concentration in solutions of pure SDS occurs The comparison of the results for SDS and SDP3S
at about 250 mM,19 and our own measurements by DLS suggests that the additional ethoxy groups make the
showed that there is no such transition for the studied headgroup of SDP3S bulkier, as compared to SDS. As a
CAPB even at concentrations as high as 500 mM. Hence, result, the mean area per molecule in the micelles
the sphere-to-rod transition in the mixed solutions occurs containing SDP3S is expected to be larger, which should
at total surfactant concentrations that are more than 1 lead to a smaller spontaneous radius of curvature of the
order of magnitude lower than the transition concentra- micelles and, hence, to a less favored transition from
tions for the individual surfactants. spherical to rodlike micelles.
Measurements with mixed surfactant solutions showed 3.2. Aggregation Number, Second Virial Coef-
that their viscosity rapidly increased after the sphere- ficient, Effective Diameter, and Effective Volume
to-rod transition; see Figure 3A. As evidenced by the Fraction of the Spherical Micelles. In the Zimm plot,
comparison of the micelle diameter and solution viscosity obtained by SLS, the transition in the micelle size (and
(see Figure 3B), the observed increase of viscosity is closely the related increase of aggregation number) appears as
related to the increase of the micelle size after the a sharp break point in the data; see the example shown
transition. Note also that the maxima in the curves for in Figure 4 for 0.75 molar fraction of CAPB in the
the micelle size and solution viscosity (Figure 3B) cor- surfactant mixture (similar results were obtained with
respond to the minimum in the curve CSR versus fCAPB the other mixed solutions). At low surfactant concentra-
shown in Figure 2B. tions, the plot of the data has a positive slope; that is, the
second virial coefficient, A2 > 0, indicates a net repulsion
(19) Reiss-Husson, F.; Luzzati, V. J. Phys. Chem. 1964, 68, 3504. between the spherical micelles formed in this concentra-
(20) Rohde, A.; Sackmann, E. J. Colloid Interface Sci. 1979, 70, 494. tion range. The negative slope in the SLS data, after the
568 Langmuir, Vol. 20, No. 3, 2004 Letters

As explained in ref 18, VEFF can be expressed through

the second osmotic virial coefficient measured by SLS:

2
1 A2MW
VEFF ) (6)
4 NA

Further, from VEFF and the aggregation number NS we

estimated the effective volume fraction of the micelles at
the transition concentration,

CSRNA
ΦEFF(CSR) ) FSRVEFF ) VEFF (7)
NS

where FSR is the micelle number concentration at the

transition point.
As seen from Table 1, the effective diameter of the mixed
spherical micelles, dEFF, as determined by SLS, is more
Figure 4. Zimm plot of the mixed solutions of SDS and CAPB
at fCAPB ) 0.75. The dotted lines are guides to the eye. than 2 times larger than the hydrodynamic diameter, dh,
determined by DLS, which is due to the long-ranged
break point, is due to the rapid increase of the micelle electrostatic repulsion between the charged micelles. Note
aggregation number at C > CSR. Note that the data that this difference has a large impact on the effective
interpretation, by means of eq 2, is not justified beyond volume fraction of the micelles in the solution, because
the break point, because eq 2 implies that the micelle size ΦEFF is proportional to (dEFF)3. Therefore, the electrostatic
and shape do not depend on surfactant concentration, repulsion between the mixed micelles significantly in-
which is not the case for C > CSR. creases their effective volume fraction (by more than 10
By using eqs 2-4, one can interpret the experimental times), as compared to the actual volume fraction of the
data from the SLS experiments in the range of low micelles in the solution. As explained in refs 21-25, the
surfactant concentrations (where only spherical micelles effective micelle volume fraction is decisive for the
are present) and determine the micelle aggregation magnitude of the oscillatory structure forces which
number, NS, and the second virial coefficient, A2. The stabilize foam and emulsion films from micellar solutions.
results from this analysis are summarized in Table 1. As Not surprisingly, we found (unpublished results) that foam
seen from the table, the aggregation number of the mixed films formed from mixed SDS/CAPB solutions exhibit a
spherical micelles, NS ≈ 113 ( 5, is larger than that of the pronounced stratification; this would not be the case
micelles consisting of individual components, NSDS ≈ 89 without the strong electrostatic repulsion between the
and NCAPB ≈ 84, respectively. The larger size of the mixed micelles and the ensuing, relatively large values of dEFF
micelles is probably related to the strong attraction and ΦEFF.
between the SDS and CAPB headgroups, which probably The effective volume fraction of the micelles at the
decreases the mean area per molecule and thus favors the concentration of sphere-to-rod transition, ΦEFF(CSR), was
incorporation of more surfactant molecules per micelle. around and below 10% for the mixed solutions with fCAPB
All virial coefficients are positive in the studied systems, > 0.5; see Table 1. The driving energy for the sphere-to-
which indicates a net repulsion between the spherical rod transition in the studied systems is considered in the
micelles. As expected, the magnitude of A2 increases with following section.
the fraction of SDS in the surfactant mixture, because
SDS brings a negative charge to the micelles. The used 4. Interactions between the SDS and CAPB
batch of CAPB contained NaCl, which had a screening Molecules
effect on the surface charge of the micelles (by measuring The sphere-to-rod transition at C ) CSR is driven by the
the electroconductivity of CAPB solutions, we determined difference in the interactions between the surfactant
that the concentration of NaCl in the batch sample was molecules and in the packing of these molecules, in the
7 wt % with respect to CAPB). This screening effect is spherical and rodlike micelles. In this section, we use the
more pronounced at the higher molar fractions of CAPB light scattering data to estimate the difference between
in the mixture, where more NaCl is introduced into the the standard chemical potentials of the surfactant mol-
solutions. Therefore, the reduced virial coefficient at higher ecules incorporated in spherical and rodlike micelles,
CAPB to SDS ratios is related to (1) reduced micelle charge ∆µ0SR ≡ µ0R - µ0S. In other words, ∆µ0SR is the change in the
and (2) more screened electrostatic repulsion between the average interaction energy of a surfactant molecule with
micelles. its neighbors, when the molecule is transferred from a
From the values of A2, one can determine the so-called spherical into a rodlike micelle. From a thermodynamic
effective volume of the micelles, VEFF, which accounts for
viewpoint, ∆µ0SR is the driving energy for the observed
the micelle interactions:18
sphere-to-rod transition. Note that the rodlike micelles
have larger aggregation numbers than the spherical ones,
VEFF )
1
8 ∫0∞ [1 - exp(-W(r)/kT)]4πr2 dr (5)
(21) Kralchevsky, P. A.; Denkov, N. D. Chem. Phys. Lett. 1995, 240,
358.
where W(r) is the interaction energy between two micelles (22) Nikolov, A. D.; Wasan, D. T.; Denkov, N. D.; Kralchevsky, P. A.;
separated at a center-to-center distance, r. For particles Ivanov, I. B. Prog. Colloid Polym. Sci. 1990, 82, 87.
interacting as hard spheres, VEFF ) VS. In contrast, VEFF (23) Wasan, D. T.; Nikolov, A. D.; Kralchevsky, P. A.; Ivanov, I. B.
could be substantially different from the actual micelle Colloids Surf. 1992, 67, 139.
(24) Bergeron, V.; Radke, C. J. Langmuir 1992, 8, 3020.
volume, VS, if long-range interactions between the particles (25) Marinova, K. G.; Gurkov, T. D.; Dimitrova, T. D.; Alargova, R.
are present. G.; Smith, D. Langmuir 1998, 14, 2011.
Letters Langmuir, Vol. 20, No. 3, 2004 569

Table 1. Aggregation Number; Second Virial Coefficient, A2; Sphere-to-Rod Transition Concentration, CSR; Effective
Micelle Volume Fraction at the Transition Concentration Calculated from Equation 7, ΦEFF(CSR); Effective Micelle
Diameter as Calculated from Equation 6, dEFF; and Hydrodynamic Diameter, dh, as Measured by DLS, at Several Molar
Fractions of CAPB in the Surfactant Mixture, fCAPB
aggregation number
fCAPB NSDS NCAPB NS virial coefficient A2 CSR, mM ΦEFF (CSR), % dEFF, nm dh, nm
0a 89 0 89 250 f100 4.8
0.5 58 58 116 0.002 32 35 25 13.9 4(1
0.67 37 73 110 0.001 52 25 11 12.1 4.9 ( 0.5
0.75 28 86 114 0.001 29 17 6.9 11.7 5.3 ( 0.5
0.8b 10 5.8 ( 1
1.0 0 84 84 0.000 56 >500 7.6 5.2 ( 0.5
a The values for pure SDS are calculated by using literature data (refs 17, 19, and 20). b No SLS experiments were performed at fCAPB
) 0.8.

The assumption that the two types of micelles (spherical

and rodlike) are in equilibrium allows us to apply the
thermodynamic theory of self-assembly17,27-30 and to
estimate the difference between the standard chemical
potentials of the surfactant molecules incorporated in
spherical and rodlike micelles, ∆ µ0SR. According to this
theory, the chemical potential of the surfactant molecules
incorporated in noninteracting micelles can be presented
Figure 5. Schematic presentation of the surfactant aggregates as
in the studied solutions, according to the model used to
determine the length, L, and the aggregation number, NR, of kT
the rodlike micelles. In the concentration range below the µS(NS) ) µ0S (NS) + ln(XS/NS) (8a)
sphere-to-rod transition, C < CSR, only spherical micelles with NS
diameter dS are present (region 1). At C g CSR, spherical and
rodlike micelles (possibly of different surfactant composition) kT
are in thermodynamic equilibrium (region 2). In region 2, the µR(NR) ) µ0R (NR) + ln(XR/NR) (8b)
NR
diameter of the spherical micelles is assumed to be the same
as in region 1 (dS ≈ 5 nm) and their concentration to be
approximately constant and equal to CSR. where µS(NS) is the chemical potential of a molecule
incorporated in a spherical aggregate with aggregation
which means that the sphere-to-rod transition is related number NS, and µ0S(NS) is the respective standard part of
to a loss of entropy (due to a decrease in the micelle num- the chemical potential. The quantity XS ) CS/(CS + 55.5)
ber concentration) that should be compensated by a gain is the molar fraction of surfactant incorporated in spherical
in the interaction energy between the molecules; hence, aggregates, normalized by all components in the solution,
∆ µ0SR should be negative. including water (55.5 M is the molar concentration of water
In the region of rodlike micelles, the measured diffusion in the solutions). Respectively, XS/NS is the molar fraction
coefficient, D, is a quantity averaged over all micelles of the spherical micelles in the solution. The subscripts
present in the solution. Furthermore, for the rodlike S and R in eqs 8a and 8b denote quantities referring to
micelles D is averaged over their possible orientations in spherical and rodlike micelles, respectively. In our con-
the solution. In the following consideration, we assume sideration, we do not take explicitly into account the
that the surfactant solution contains two types of micelles, polydispersity of the micelles and associate NS or NR with
spherical and rodlike, which are in thermodynamic the mass-averaged aggregation number of the respective
equilibrium with each other, in the concentration range types of micelles. Hence, XS and XR denote the total molar
beyond CSR (region 2 in Figure 5).26 The surfactant fraction of surfactant that is incorporated into spherical
composition of these two types of micelles is not necessarily and rodlike micelles, respectively.
the same; they can differ in the relative amounts of SDS The assumption that the spherical and rodlike micelles
and CAPB. Indeed, one can expect that the spherical are in equilibrium implies that µR ) µS. From eqs 8a and
micelles in region 2 can be enriched in SDS (in comparison 8b, one deduces the following expression for the difference
with the surfactant composition of the entire solution), between the standard chemical potentials of the surfactant
because the area per molecule in the spherical aggregates molecules in spherical and rodlike micelles:
is larger; as a result, the electrostatic repulsion between
the charged headgroups of SDS in spherical micelles is ∆ µ0RS ≡ µ0R (NR) - µ0S (NS) )
smaller as compared to the repulsion in rodlike micelles
having the same surfactant composition. Since we have kT ln[(XS/NS)1/NS(NR/XR)1/NR] (9)
no methods to determine the exact composition of the
micelles, in the subsequent considerations we use the Therefore, one could estimate ∆ µ0SR if the aggregation
equations for quasi-one-component surfactant aggregates. numbers and the concentrations of the spherical and
This means that the discussed chemical potentials of the rodlike micelles are known. Our SLS experiments showed
surfactant molecules, incorporated in spherical or rodlike
micelles, are averaged over the SDS and CAPB molecules (27) Missel, P. J.; Mazer, N. A.; Benedek, G. B.; Young, C. Y.; Carey,
M. C. J. Phys. Chem. 1980, 84, 1044. Missel, P. J.; Mazer, N. A.; Carey,
present in the respective micelles. M. C.; Benedek, G. B. J. Phys. Chem. 1989, 93, 8354.
(28) Israelachvili, J. N. Intermolecular and Surface Forces; Academic
(26) The coexistence of spherical and rodlike micelles was experi- Press: New York, 1992.
mentally established by SANS measurements in the study by: Nakano, (29) Alargova, R. G.; Danov, K. D.; Kralchevsky, P. A.; Broze, G.;
M.; Matsuoka, H.; Yamaoka, H.; Poppe, A.; Richter, D. Macromolecules Mehreteab, A. Langmuir 1998, 14, 4036.
1999, 32, 697. The possibility for such coexistence is suggested also by (30) Alargova, R. G.; Ivanova, V. P.; Kralchevsky, P. A.; Mehreteab,
the theory of surfactant self-assembly (e.g., refs 27-31 below). A.; Broze, G. Colloids Surf., A 1998, 142, 201.
570 Langmuir, Vol. 20, No. 3, 2004 Letters

that the aggregation number of the spherical micelles in

the studied mixed solutions of SDS and CAPB was
virtually independent of the molar fraction of CAPB in
the mixture, fCAPB, when the latter was in the range
between 0.5 and 0.75 (NS ) 113 ( 5, see Table 1). For that
reason, in the following estimates we assume that NS is
equal to 113 for all mixed solutions that were studied (the
final results were not very sensitive to small changes in
NS). For the molar fraction of the surfactants incorporated
in the spherical micelles, XS, we assume that it remains
approximately constant after the sphere-to-rod transition,
that is, XS ≈ XSR at X g XSR; here X ) C/(C + 55.5) is the
total molar fraction of surfactant in the solution and XSR
is the respective fraction at the transition. Under the latter
assumption, eq 9 can be transformed to read

∆ µ0RS ) kT ln[(XSR/NS)1/NS(NR/(X - XSR))1/NR] (9a)

Therefore, one can estimate ∆ µ0SR from eq 9a if the mean

aggregation number of the rodlike micelles, NR, is known.
To estimate NR, we approximated the shape of the
rodlike micelles by an elongated ellipsoid, whose two small
axes are equal to the diameter of the spherical micelles,
dS ≈ 5 nm; see Figure 5. The long axis of the ellipsoid, L
(the micelle length), can be estimated from the value of
the mean micellar diffusion coefficient, D, which is
measured by DLS. In fact, D is a mass-averaged quantity
over all spherical and cylindrical micelles present in the
solution. Assuming that for C > CSR the concentration of
the spherical micelles remains equal to CSR and their Figure 6. (A) Dependence of the aggregation number of the
diameter remains equal to dS, we calculated the mean rodlike micelles, NR, on the surfactant concentration, (C - CSR),
diffusion coefficient of the rodlike micelles, DR ) (CD - in mixed CAPB/SDS solutions at fCAPB ) 0.5 (circles) and 0.8
CSRDS)/(C - CSR). From DR, we calculated the length of (empty squares). The straight line is a linear fit to the
experimental points. (B) Difference between the standard
the rodlike micelles, L, by using expressions relating the
chemical potentials of a surfactant molecule in rodlike, µ0R, and
diffusion coefficient of an ellipsoid with its dimensions
spherical, µ0S, micelles as a function of fCAPB for C ) 40 mM
(see refs 18 and 29-31 for the respective theoretical (see section 4). The dashed curve indicates the expected
expressions and procedure for analysis of DLS data for
trend of the dependence ∆ µ0RS vs fCAPB.
rodlike micelles). From the dimensions of the rodlike
micelles, dS and L, we estimated the mean volume of the
By using eq 9a, we estimated ∆ µ0SR to be between -0.05
micelles, VRM ≈ (π/6)LdS2, and their mean aggregation
and -0.1 kT per molecule for all SDS/CAPB solutions in
number, NR, at various ratios of SDS and CAPB in the
which the sphere-to-rod transition was observed. The
mixture. In these estimates, we assumed that the molar
ratio SDS/CAPB in the rodlike micelles is the same as magnitude of ∆ µ0SR slightly increased with the raise of the
that in the entire surfactant solution and used VSDS ≈ total surfactant concentration, C. To make a comparison
0.652 nm3 and VCAPB ≈ 0.881 nm3 for the volumes of the between the results for various SDS/CAPB ratios, in
SDS and CAPB molecules, respectively (these volumes Figure 6B we plot the calculated value of ∆ µ0SR at a fixed
were calculated by dividing the volume of the spherical total surfactant concentration, C ) 40 mM. As seen from
micelle formed in the solutions of individual SDS or CAPB Figure 6B, the magnitude of ∆ µ0SR passes through a
by the respective aggregation number, NS). Finally, we relatively flat maximum at fCAPB ≈ 0.7. To draw the dashed
estimated ∆ µ0SR for the solutions containing rodlike curve in Figure 6B, we have used the experimental fact
micelles by means of eq 9a. that the sphere-to-rod transition occurs at much higher
Illustrative results for the increase of NR with the total surfactant concentration when fCAPB < 0.5 and fCAPB
surfactant concentration are shown in Figure 6A for two > 0.85 (that is, the dashed curve indicates the expected
systems, fCAPB ) 0.5 (CSR ≈ 35 mM) and fCAPB ) 0.8 (CSR variation of ∆ µ0SR with fCAPB).
≈ 10 mM). One sees that NR increases linearly with C and The magnitude of ∆ µ0SR (≈0.1 kT) is by no means
reaches almost 103 molecules per micelle at (C - CSR) ) negligible, because one should multiply this value by the
50 mM. Interestingly, we found that the results for these micelle aggregation number, NS or NR (which are of the
two systems fall very close to each other (cf. the circles order 102-103), to estimate the contribution of this term
and the empty squares in Figure 6A), but this could be a to the total energy of micelle formation to be 10-100 kT.
fortuitous coincidence, because the results for the other Such a magnitude of ∆ µ0SR is sufficient to trigger the
SDS/CAPB ratios (not shown on the figure) do not fall on observed sphere-to-rod transition in the studied systems.
the same straight line. The intercept of the linear fit, drawn Indeed, one can make a comparison with the “ladder
through the calculated points in Figure 6A, is 104 ( 10, model” by Missel et al.,27 in which the growth of rodlike
which is very close to the aggregation number of the micelles in one-component surfactant solution is expressed
spherical micelles, as determined by SLS for concentra- through the equilibrium constant, K, which, in our
tions C < CSR. notation, can be expressed as
(31) Alargova, R.; Petkov, J.; Petsev, D.; Ivanov, I. B.; Broze, G.;
Mehreteab, A. Langmuir 1995, 11, 1530. K ≡ exp[NS∆ µ0RS/kT] (10)
Letters Langmuir, Vol. 20, No. 3, 2004 571

For our systems, the exponent, NS∆ µ0RS/kT ≈ 10, is a mixture. The observed micelle transition brings about a
sufficiently large value for triggering the formation of rapid increase of the solution viscosity. From the depen-
rodlike micelles. dence of the aggregation number on the surfactant
In the above estimates, we have neglected the interac- concentration, we estimated the mean energy for trans-
tions between the micelles in the interpretation of the fer of a surfactant molecule from a spherical into a
data from DLS. By comparing theory and experiment, rodlike micelle, ∆ µ0SR, to be of the order of -0.1 kT per
Missel et al.27 have shown that these interactions become molecule.
important only when Parallel experiments with mixed solutions of CAPB and
sodium laureth sulfate (SDP3S) showed that the sphere-
C/C* g 0.45 (11) to-rod transition for this mixture occurs at higher total
surfactant concentrations, above 40 mM. One can explain
where C* is the surfactant concentration at which the the latter result by considering the fact that the additional
mean distance between the micelles is equal to the micelle ethoxy groups make the headgroup of SDP3S bulkier, as
radius of gyration, Rg. In the studied solutions, C was compared to SDS. As a result, the mean area per molecule
always below 0.1 M and the aggregation number varied in the micelles containing SDP3S is expected to be larger,
between ca. 102 and 103 (largest aggregation numbers which should lead to a smaller spontaneous radius of
corresponding to highest concentrations). For cylindrical curvature of the micelles and, hence, to a less favored
micelles, the radius of gyration Rg is given by the transition from spherical to rodlike micelles.
expression (e.g., ref 18) The data for the aggregation number and the second
osmotic virial coefficient of the spherical micelles, mea-
Rg ) (L2/12 + dS2/8)1/2 (12) sured by SLS, were used to calculate the effective micelle
diameter, which includes a contribution from the long-
In our systems, dS ≈ 5 nm and L e 50 nm, which gives range electrostatic repulsion between the micelles. This
Rg e 15 nm. From the experimental data, we estimated effective diameter was found to be about 2 times larger
that C/C* < 0.45 in all studied micellar solutions, which than the hydrodynamic one. The respective large effective
allows us to neglect (as a reasonable first approximation) volume fraction of the micelles promotes pronounced
the interaction between the micelles in the analysis of the colloidal structural forces, created by the repelling mi-
DLS data. celles, in foam and emulsion films formed from mixed
SDS/CAPB solutions (unpublished results).
5. Conclusions
Static and dynamic light scattering experiments showed Acknowledgment. This study was supported by the
that mixed SDS/CAPB micelles undergo a sphere-to-rod Unilever Research Center, Edgewater, NJ. Some of the
transition at rather low total surfactant concentrations, calculations were performed and several of the figures
≈10 mM. For comparison, this transition occurs at 250 were prepared by Dr. S. Tcholakova (University of Sofia);
mM for pure SDS, whereas for CAPB alone, it is above her help is gratefully acknowledged.
500 mM. The minimal transitional concentration was
found at a molar fraction fCAPB ≈ 0.8 in the surfactant LA035717P