Langmuir 2008, 24, 4485-4493 4485

Articles
Formation of Single-Phase Microemulsions in Toluene/Water/Nonionic
Surfactant Systems
Albert J. Gotch,*,† Gary W. Loar,‡ Aaron J. Reeder,† and Ellen E. Glista†
Chemistry Department, Mount Union College, Alliance, Ohio 44601, and McGean-Rohco,
2910 HarVard AVenue, CleVeland, Ohio 44105

ReceiVed August 7, 2007. In Final Form: January 31, 2008

Mixtures of toluene and water from 5 to 50% oil fraction and 5 to 25% surfactant by weight were studied. Winsor
Type IV microemulsions were formed in numerous cases. Review of partial ternary phase diagrams for these systems
indicated the area of single-phase microemulsion with toluene could be maximized at an hydrophilic-lipophilic
balance (HLB) of approximately 14.5. Select single-phase samples were further analyzed by surface tension and
dynamic light scattering techniques, which allowed a detailed characterization of the solution equilibrium thermodynamics
and size stability. Particle sizes averaged ∼5 nm and were nearly constant over a wide variety of conditions and for
6-18 months. When benzyl alcohol was used instead of toluene, the optimum HLB for the formation of single-phase
systems was found to have a lower limit of 17. Particle sizes in these systems were <30 nm but showed greater
variability. The decrease in particle size as surfactant concentration increased was determined to be associated with
changes in ethlyene oxide chain conformation. The increase in particle size due to swelling with increased oil concentration
was used to determine the surfactant surface area in the oil phase. A detailed comparison of alkylamine ethoxylate
to octyl- and nonylphenol ethoxylate surfactants in terms of micelle thermodynamics, size, and stability indicate that
the alkylamine-based surfactants are potential candidates for the replacement of nonylphenol-based surfactants in some
systems with a more polar oil phase like benzyl alcohol.

Introduction make microemulsions of wide appeal for a number of practical

Numerous previous studies have detailed the phase behavior
of emulsions, microemulsions, and nanoemulsions.1-4 These
studies indicate that emulsions may be formed between any two
solution components with the proper choice and composition of
the three components; oil, water, and surfactant. Emulsion
containing solutions usually include multiple phases as denoted
by Winsor Type I, II, and III regions of a phase diagram.5 A
thermodynamically stable single-phase microemulsion may also
be formed under the appropriate conditions. These are referred
to as Winsor Type IV systems. Early work indicated that a
cosolvent or cosurfactant was needed to form the Winsor Type
IV system.5,6 However, more recent studies have shown stable,
single-phase microemulsions may also be formed with just three
components.7,8 Microemulsions are further characterized by their
small particle size making them optically translucent or transparent
in contrast to regular emulsions which are cloudy. This
combination of thermodynamic stability and optical transparency
* Corresponding author. E-mail: gotchaj@muc.edu.
† Mount Union College.
‡ McGean-Rohco.
(1) Morrison, I. D.; Ross, S. Colloidal Dispersions: Suspensions, Emulusions,
and Foams; John Wiley & Sons: New York, 2002.
(2) Holmberg, K.; Jonsson, B.; Kronberg, B.; Lindman, B. Surfactants and
Polymers in Aqueous Solution, 2nd ed.; John Wiley & Sons: Chichester, UK,
2003; Chapter 6.
(3) Rosen, M. J. Surfactants and Interfacial Phenomena, 3rd ed.; John Wiley
& Sons: Hoboken, NJ, 2004.
(4) Solans, C.; Kunieda, H. Industrial Applications of Microemulsions, Marcel
Dekker: New York, 1997.
(5) Winsor, P. Trans. Faraday Soc. 1948, 44, 376-398.
(6) Holmberg, K.; Osterberg, E. J. Dispersion Sci. Technol. 1986, 7, 299-
306.
(7) Holmberg, K. Eur. J. Org. Chem. 2007, 731-742.
(8) Zhang, W.; Qiao, X.; Chen, J. Colloids Surf., A 2007, 299, 22-28.
applications in cosmetics,4 organic synthesis,7 pharmaceuticals,9
and nanoparticle formulations.8,10
An important physical property of industrial solutions is low
volatility. This leads to lower emissions, a safer work environment,
and easier disposal of waste. Thus, emulsions based on organic
solvents of low volatility like benzyl alcohol or decane are
preferred to those based on high volatility solvents like toluene
or dichloromethane. One goal of this work was to find a Winsor
Type IV microemulsion system based on benzyl alcohol and to
fully characterize its solution thermodynamics and size stability.
Toluene was used as the volatile analogue because of similarities
in size, shape, and aromaticity.
Some of the most common industrial nonionic surfactants
include ethoxylated nonylphenols with production of ap-
proximately 500 000 tons per year worldwide.11 However,
concerns have arisen due to the endocrine disruption potential
of the natural degradation products of nonylphenol containing
surfactants.12 This has led to severe restrictions on the use of
nonylphenol ethoxylates in Europe.13 In the United States, the
Environmental Protection Agency has recommended limits of
6.6 ppb nonylphenol in fresh water.14 Thus, there is a need to
find suitable alternative surfactants that have the same general
(9) Chan, J.; ElMaghraby, G. M. M.; Craig, J. P.; Alany, R. G. Int. J. Pharm.
2007, 328, 65-71.
(10) Chen, W.; Zhu, Q. Mater. Lett. 2007, 61, 3378-3380.
(11) Ying, G.-G.; Williams, B.; Kookana, R. EnViron. Int. 2002, 28, 215-226.
(12) Gültekin, I.; Ince, N. H. J. EnViron. Man. 2007, 85, 816-832.
(13) Directive 2003/53/EC of the European Parliament Off. J. EU 2003, L178,
24-27.
(14) U.S. Environmental Protection Agency. Fed. Regist. 2006, 71, 9337-
9339.

10.1021/la7024318 CCC: $40.75 © 2008 American Chemical Society

Published on Web 03/20/2008
4486 Langmuir, Vol. 24, No. 9, 2008 Gotch et al.

Table 1. Summary of Surfactant Systems Studied Table 2. Selected Physical Properties of Solvents
surfactant manufacturer HLB hydrophobic group EOs F (g/mL) µ (D)  HLB
Span 80 ICI 4.3 sorbitan monooleate 0 H2O 0.997 1.86 78.5 20
CO210 Ashland 4.6 nonylphenyl 2 BA 1.045 1.7 13.1 3
Plur L62 Rhodia 7 dodecyl Tol 0.866 0.36 2.4 0
Tom 91-2.5 Tomah 8.5 decyl 2.7
Brij 30 Aldrich 9 dodecyl 4 W; and oil, O. Emulsions were prepared at a room temperature of
Tom 91-6 Tomah 12.4 decyl 6
approximately 21 °C and stored in a thermostated room set to 20
CO630 Rhone-Poulenc 13 nonylphenyl 9.5
Var K205 Goldschmidt 13.3 dodecyla 5
( 1 °C. Visual inspection of the samples indicated that equilibrium
CO720 Rhodia 14.2 nonylphenyl 12 was usually reached in 24-72 h. The emulsions were allowed to
CA720 Rhodia 14.6 octylphenyl 12 stand for one week before further observations were recorded to
Surf N150 Huntsman 15 nonylphenyl 15 ensure thermodynamic equilibrium had been attained. Samples were
Var K215 Goldschmidt 16.8 dodecyla 15 generally saved for 6-18 months to monitor the long-term stability
a
of the emulsions.
largest concentration of single hydrocarbon chain, average is C13.4.
Characterization. Digital pictures were generally obtained of
each sample set after approximately 24 h, 7 days, and 6 months.
Phase boundaries were determined by visual observation of a change
properties as the nonylphenols and do not have endocrine in refractive index through a vertical cross-section of the equilibrium
disruption potential. sample. Since nonionic surfactants were used, these emulsions were
We studied a wide range of nonionic surfactants in toluene prepared using a 0.0150 M NaCl solution for later characterization
and water and benzyl alcohol and water mixtures with the goal via conductivity. The conductivity of each equilibrium phase was
of generating thermodynamically stable, ternary component, measured using an Orion 150A+ conductivity meter. Crossed-polar
single-phase microemulsions, i.e., Winsor Type IV systems. observations were taken by placing the sample in front of a sodium
Standard techniques of conductivity and crossed-polar measure- vapor lamp and using a pair of polarizing filters before and after the
ments were combined with surface tension and dynamic light sample rotated by 90° relative to each other. This provided a screening
scattering for further characterization of these systems. Of tool for the presence of ordered phases.17
particular importance was the characterization of the single phase, Microemulsion formation was verified in some of the CA720 and
CO720 systems by using a reverse mixing procedure. This
the size and size distribution of the resulting microemulsions, corresponded to Solans method A, where the surfactant was first
and thermodynamic stability over 6-18 months. During the dissolved in water, rather than oil, and then stirred continuously
course of this study it became clear many of the favorable while oil, rather than water, was added to dilute the sample to the
properties of the nonylphenol ethoxylates could be replicated by desired final composition of the three components.16 Since micro-
surfactants with similar HLB numbers. Provided here is a detailed emulsions are thermodynamically stable, the same equilibrium state
thermodynamic analysis, phase stability, and size distribution is achieved regardless of the mixing order used to prepare the
study of ethoxylated octylphenols, ethoxylated nonylphenols, sample.4,16 These observations in the CA720 or CO720/toluene/
and alkylamine ethoxylates. water test systems coupled with the measured small particle sizes
indicated that our systems are microemulsions rather than na-
Experimental noemulsions. Nanoemulsions are kinetically stable and thus affected
by mixing order as in the Brij 30/dodecane/water system, which was
Materials. Toluene of ACS reagent grade was purchased from also prepared by methods A and B. This latter system produced two
Kodak or Fisher Scientific. NaCl of ACS certified grade was distinctly different equilibrium states confirming nanoemulsion
purchased from Fisher. Water was first deionized then purified to formation.
18.2 MΩ by an Elga Purelab Ultra purification system. Technical Surface tension measurements were made using a DuNuoy
grade benzyl alcohol was supplied by Aceto. The surfactants were tensiometer (Central Scientific Co.) with a Pt-Ir ring. The ring was
of technical grade from a variety of manufacturers as noted in Table washed in dilute aqua regia solution, rinsed with 18 MΩ water, and
1. All materials were used as provided. The hydrophilic-lipophilic flamed between each series of measurements, which was done in
balance (HLB) was obtained directly from company literature. The triplicate. Each measurement was corrected for wetting of the ring
number of ethylene oxide units (EOs) represents the average value using the Harkins-Jordan factor.18,19 The average corrected value
of a distribution of molecules that vary in chain length due to was calculated and plotted via the Gibbs equation to obtain the
limitations in the control of the catalytic addition of these groups.15 critical micelle concentration (cmc) and the surfactant aqueous surface
Emulsion Preparation. The emulsification process was modeled area.
after “method B” used by Solans.16 Water was added using a Razel
Dynamic light scattering (DLS) measurements were performed
A-99 syringe pump at 0.5 mL/min to a previously prepared surfactant
using a Brookhaven Instruments (BI) system including a 35 mW
in oil mixture to achieve the final desired mass % of each component.
HeNe laser light source at 632.8 nm (Melles Griot 25-LHP-928-
The mixture was stirred at a rate of approximately 700 rpm. The
249), a BI-200SM Goniometer, a digital temperature controlled
surfactant concentration is given as a percent by mass relative to the
sample holder at 20.00 ( 0.01 °C (Science Electronics), a BI-9000AT
total mixture. The R value is the percent of mass concentration of
auto correlator, and an EMI 9863 photomultiplier tube. Samples that
the oil relative to the oil + water components of the total emulsion.
were characterized by laser light scattering were drawn from the
initial preparation tube and passed through a 0.2 µm syringe filter.
O
R) × 100% (1) Scatter intensity of laser light was recorded at 45°, 90°, and 135°
O+W for each sample over a nominal time domain of 4 µs to 40 ms.
Collection time was nominally 2-3 min with slits of 400 µm, which
The sum of oil and water components was maintained at a constant allowed sufficient signal for subsequent data analysis. The auto
10.0 g. For convenience, we refer to compositions as R5 meaning correlation function of the scatter intensity of laser light was fit
R ) 5% by mass according to eq 1. In addition, we also use this
notation for the standard total mass percent of surfactant, S; water, (17) Horváth-Szabó, G.; Masliyah, J. H.; Czarnecki, J. J. Colloid Interface Sci.
2001, 242, 247-254.
(15) Nadeau, H. G.; Oaks, D. M.; Nichols, W. A.; Carr, L. P. Anal. Chem. (18) Eastoe, J. Surfactant Chemistry; University of Bristol: Bristol, U.K.,
1964, 36, 1914-1917. 2003.
(16) Forgiarini, A.; Esquena, J.; Gonzalez, C.; Solans, C. Prog. Colloid Polym. (19) Rusanov, A. I.; Prokhorov, V. A. In Interfacial Tensiometry; Mobius, D.,
Sci. 2000, 115, 36-39. Miller, R., Eds.; Elsevier: Amsterdam, 1996.
Single-Phase Microemulsions in Toluene
Figure 1. Selected structures of surfactants used in this study.
Figure 2. Stacked plots of partial ternary phase diagrams for toluene/water/surfactant (left) and benzyl alcohol/water/surfactant (right).
Surfactant HLB increases from bottom to top.
using the non-negatively constrained least-squares (NNLS) method.
Reported values are the average of three measurements with standard
deviations given in parentheses. The theoretical limit of size
discrimination in the Brookhaven system is ∼1 nm and under actual
experimental conditions closer to ∼3 nm.20 Thus, many of the particle
size measurements were near the detection limits of the equipment.
Results
Figure 1 shows select samples of the molecular structures for
the complete range of surfactant materials used in this study. At
5% by weight (S5), any surfactant with an HLB less than 10 gave
a two-phase equilibrium system with toluene/water mixtures.
Thus, these systems were not studied at higher surfactant
concentrations or with the more polar benzyl alcohol oil phase.
(20) Page J. L. University of Akron, Akron, OH. Personal communication
2007.
Langmuir, Vol. 24, No. 9, 2008 4487
For the remaining surfactants with HLBs greater than 10, despite
some variation in the nonpolar group, it became clear that the
dominating factor was the number of ethylene oxide groups
present on the hydrophillic portion of the molecule. Thus, these
samples were studied over a wider range of surfactant concen-
trations and characterized more fully.
Phase Summary/Phase Diagrams. Toluene was the model
oil used in this study. Figure 2 is a stacked plot of the partial
ternary phase diagram based on the HLB number. All toluene
data points were performed in triplicate. In nearly all cases, the
same equilibrium state was observed. When in question, the
majority point was plotted. The two conditions of interest were
low surfactant concentrations and high water concentrations;
thus our range of study included S5-S25 and e R50. For toluene,
the area of one phase is maximized for a surfactant HLB around
14.5. A transparant gel was formed in the four toluene/water
4488 Langmuir, Vol. 24, No. 9, 2008 Gotch et al.

Table 3. Summary of Surface Tension Data Analysis

cmc Σ ∆G°mic ∆G°ads ∆G°ma
surfactant (mol/L) (Å2/molecule) (kJ/mol)a (kJ/mol)b (kJ/mol)c
T916 1.1 × 10-3 92.4 -26.5 -42.1 15.6
K205 5.5 × 10-4 42.3 -28.2 -33.4 5.2
CO720 3.0 × 10-4 111.0 -29.6 -46.3 16.7
CA720 2.7 × 10-4 87.6 -29.9 -43.0 13.1
N150 6.8 × 10-5 87.1 -33.3 -42.1 8.8
K215 1.0 × 10-5 82.3 -38.0 -42.1 4.1
a
∆G°mic based on eq 3. b ∆G°ads based on eq 4. c ∆G°ma ) ∆G°mic
- ∆G°ads.
Figure 3. Sample Gibbs plot for K205 surfactant in water. Each
data point is the average of three trials and is corrected for wetting Two additional thermodynamic parameters may be obtained
of the ring using the Harkins-Jordan factor.18,19 by using a large concentration range that includes data points
above the cmc and those down to infinite dilution. ∆G°mic may
systems near and above the optimum HLB and S25. Thus, even be obtained directly from the cmc as summarized by Langevin22
though the range of study was limited, the formation of gels is and demonstrated by Molyneux et al.23 and Israelechvili24
indicative of additional internal structural changes, which were
cmc
beyond the scope of this study. However, the gel phases form
a natural upper bound that closely matches the cutoff perpendicular
∆Gomic ) RT ln (55.38 ) (3)
to the S axis used in Figure 2.
where we have accounted for the conversion from a molecule
Benzyl alcohol was used to test how polarity might affect the
to a mole basis. This is for the state change (A) f (C) as illustrated
formation of single-phase systems. This represents a change in
in a simplified form in Figure 4. Likewise, Gillap et al.25 have
dipole moment from 0.38 D in toluene to 1.7 D in benzyl alcohol
shown at infinite dilution the Gibbs energy associated with
(BA),21 which exhibits a 5% miscibility in water. While the
forming a monolayer of surfactant molecules at the surface from
dipole moment of BA is just below H2O, the dielectric constants
the solubilized surfactant molecules in solution is given by
of BA and toluene are more similar, as shown in Table 2, reflecting

( )
that BA and toluene are more similar as nonpolar solvents. In - ∂γ
contrast to toluene, BA also has a greater density than water and
o
∆Gads ) - RT ln (4)
∂c2 c2f0
forms the bottom layer when present in excess.
The right side of Figure 2 shows data over the same set of
surfactants except with BA as the oil phase. In this case, each where we have used the relation πi ≡ γoi - γi.
point reflects a duplicate measurement. R values were 5, 10, 20, Thus the natural log of the slope of γ vs c2 yields the Gibbs
30, 40, and 50% rather than the 5% increment used with toluene. energy of adsorption, i.e., state change (A) f (B) in Figure 4.
In terms of phase behavior, BA has a similar increased propensity Then the state change (B) f (C) may be obtained from ∆(∆G°)
to form single-phase systems as HLB increases. However, the ) ∆G°ma ) ∆G°mic - ∆G°ads, also summarized in Table 3.
variation in phase distribution with HLB was less pronounced, Further characterization of these systems was pursued via
and the data implies an HLB of 16.8 is still below the optimum dynamic light scattering (DLS). Simple geometric arguments
value for single phase formation. associated with micelle formation indicate22
Size Characterization. In practical applications, particle size 4
differences can lead to dramatic differences in properties.7-10 NV ) πr3 (5)
3
Thus, one goal of this study was to document the size of the
transparant micelles formed using classic surface tension where N is the number of surfactant molecules in a single micelle,
measurements and combining this with dynamic light scattering V is the volume of the surfactant hydrocarbon tail, and r is the
measurements to obtain a more thorough thermodynamic micelle radius. Likewise, the surface area of a micelle is given
characterization of these systems. by
Figure 3 shows a sample cmc plot for Varonic K205 using the
Gibbs equation: N Σ ) 4πr2 (6)

( )∂γ
∂ln c2 P,T
) -Γ2RT (2)
where γ is the surface tension of a solution of surfactant at a
molar concentration c2, the temperature in Kelvin is T, the excess
surface concentration is Γ2, and R is the universal gas constant.
From the slope of a plot of γ vs ln c2, the excess surface
concentration is obtained, which directly gives the molecular
surface area. The cmc is obtained from the sharp break in the
curve at higher concentrations, signifying that the surface of the
liquid is completely packed with surfactant molecules and further
addition of surfactant leads to micelle formation in the bulk
solution. These results are summarized in Table 3.
(21) CRC Handbook of Chemistry and Physics, 72nd ed.; BocaRaton, FL
1991.
where Σ is the area per surfactant molecule at the micelle surface.
It has been noted that Σ changes with variations in ionic strength
for ionic surfactants and temperature for nonionic surfactants.
Thus Σo is used to denote the area per surfactant molecule in the
minimum-sized micelle geometrically allowed to be formed with
No surfactant molecules. Direct experimental measurement of Σo
using X-ray and neutron scattering22,26 have shown Σ ≈ Σo. This
indicates that simple surface tension measurements can provide
useful data on micelle structure. Thus, combining a measurement
of Σ from the Gibbs plots and a value of r from DLS measurements
(22) Langevin, D. Ann. ReV. Phys. Chem. 1992, 43, 341-369.
(23) Molyneux et al. Trans. Faraday Soc. 1965, 61, 1043-1052.
(24) Israelechvili, J. Intermolecular and Surface Forces, 2nd ed.; Academic
Press: New York, 1992.
(25) Gillap, W. R.; Weiner, N. D.; Gibaldi, M. J. Phys. Chem. 1968, 72,
2218-2222.
(26) Hoffman, H. Progr. Colloid Polym. Sci. 1990, 83, 16-21.
Single-Phase Microemulsions in Toluene Langmuir, Vol. 24, No. 9, 2008 4489

Figure 4. Simplified diagrams of the three main states of surfactant molecules in aqueous solutions: (A) solubilized, (B) surface adsorbed,
and (C) micelle.

summarized in Table 4. Results for N are comparable to reported

work on similar ethoxylated nonionic surfactants.27
Particle size analysis by DLS was extended to a selection of
the single-phase systems that were formed by the six surfactants
in toluene and benzyl alcohol. The goal was to observe how
particle size varied with surfactant concentration, oil concentra-
tion, and oil polarity. Since benzyl alcohol is soluble in water
up to approximately 5% by volume, we chose systems with
higher R values. This led to choosing single-phase systems with
higher surfactant concentrations. Due to the differences in phase
behavior between toluene and benzyl alcohol, a select data set
was chosen to provide meaningful comparisons between these
single-phase systems.
Figure 5. Mean diameters of “empty” micelles formed in the absence In general, the addition of an oil led to swelled micelles in
of oil phase at S5, S15, and S25 from left to right for each surfactant. comparison to the control group micelles formed with just the
Each measurement is the average of three trials. Error bars are for surfactant in water, shown in Figure 5. In toluene, the micelle
1 s. The polydispersity values of all samples were under 0.20 and
the distributions were of single mode except for CO720_S25 and diameters of single-phase systems were all below 10 nm regardless
N150_S15. For the former, the higher polydispersity reflected a of the surfactant or oil concentration. These results are summarized
wider single mode distribution, and for the latter, one of the three in Table 5. Particle sizes are the average of three DLS
measurements was contaminated with dust. Statistical signficance measurements with the standard deviation in parentheses.
within a surfactant set is indicated by * ) 90%, ** ) 95%, and *** Conductivities are reported below each size measurement and
) 99% confidence limit. are given in µS/cm.
Table 4. Summary of Geometric Analysis Benzyl alcohol formed larger micelles for similar conditions
surfactant Σ V N of surfactant, oil, and water concentration in comparison to
in H2O r@5%S (Å) (Å2/molec) (Å3/mic) (molec) the toluene-based systems. These results are summarized in
Table 6.
T916 32(1) 92.4 986 139
K205 35(2) 42.3 496 367
CO720 20(2) 111.0 740 45 Discussion
CA720 18(1) 87.6 526 46
N150 29(1) 87.1 842 121 All phase behavior at each point in each system studied was
K215 28(2) 82.3 768 120 reproduced at least three times for toluene and at least twice for
benzyl alcohol. In most cases, identical phase behavior was
observed at equilibrium. In some cases, discrepancies were
allows one to obtain the average number of surfactant molecules
observed, most notably in T916 with toluene as the oil. In the
in a micelle from eq 6 and the surfactant volume from eq 7
method used here, the surfactant was dissolved in the oil phase
at the final desired concentration and water was slowly stirred
Σr into the mixture. During this process, it was common to observe
V) (7) multiple-phase transitions and large variations in viscosity before
3
reaching the final mixed state. Thus, on occasion, different
equilibrium states could occur even though the mass % of each
Blank DLS measurements were obtained for each surfactant component present was identical. Since this happened so
infrequently, an estimated <30 times out of >1000 data points,
in water at least three times each at S5, S15, and S25; well above
a more sophisticated mixing strategy was not pursued. However,
the cmc’s. The average of three measurements was used in the
the qualitative observations of transient phases forming during
analysis and these are reported in Figure 5. The observed decrease the mixing process suggests that a modified “titration” procedure
in radius as surfactant concentration increases appears to be
associated with surfactant coiling described in more detail below. (27) Li, M.; Rharbi, Y.; Winnik, M. A.; Hahn, K. G., Jr. J. Colloid Interface
The values for V and N are based on the S5 measurements and Sci. 2001, 240, 284-293.
4490 Langmuir, Vol. 24, No. 9, 2008 Gotch et al.

Table 5. Particle Diametersa (nm) by DLS and Conductivitiesb Table 7. Empty Micelle Radii of Various Surfactants vs
(µS/cm) of Toluene in Waterc Concentration
surfactant S15 S20 S25 surfactant formula %Sa r (Å) Å/EO EO structureb ref
R15 R15 R15 R20 R25 T916 C12.4E6 5 32.1 3.4 extended
K205 6.8(0.8) 7.9(0.5) 7.8(0.9) 6.8(0.8)g 7.4(1.1) 15 18.2 1.1 random coil
687f 83f 60f 420f 654 25 19.9 1.4 random coil
CO720 5.2(0.1) 4.5(0.4) 4.5(0.5) 5.4(0.1) 4.8(0.8) Brij 35 C12E23 5 46.1 1.5 random coil 31
2370 720 277 219 189 10 41.4 1.3 random coil 31
CA720 4.7(0.2) 4.4(0.7) 5.5(1.1) 5.0(0.6)g 4.6(0.6)g TX100 C8E10 0.5 37.7 2.9 helix 32
2220 2180 1877 1646 1481 3.5 44.6 3.6 extended 32
N150 2φd NMe 4.7(1.6) 3.3(1.6) gel a
936 716 677 167 Mass %.
b
K215 2φ 2φ gel 2φ 2φ From ref 34.
a
For particle diameters, the average of three trials is reported with agreement with this view as noted by the nearly constant value
standard deviations given in parentheses. b Second row of data for each of ∆Gads. These particular surfactants vary from octylphenyl to
surfactant solution is conductance in µS/cm. c All data reported based dodecyl, which are nearly equally disruptive of the water structure.
on observed single-phase systems. CO720_S25_R20, CO720_S25_R25, Thus, despite these structural variations in the lipophilic portion
and K205_S15_R15 sometimes produced two phases. This was attributed
to these systems being near the phase inversion temperature, since
of the surfactant, ∆Gads ) -43.1 kJ/mol on average with an RSD
temperature changes of 1-2 °C visually produced large variations in of 4.2%. This implies that the specific properties of the lipophile
particle sizes in these particular samples. d 2φ indicates a two-phase have a small moderating effect on the surface energy.29 This is
system formed. e NM ) Not measured. f Birefringent phase present. an important consideration when looking for replacements for
g
These samples produced bimodal size distributions. Approximately nonylphenol ethoxylate surfactants.
85% of the total distribution was from a single mode whose mean value In this study, the hydrophilic portion of the surfactant was
is reported. All other size distributions were single mode and had a
provided almost exclusively by EO groups. Here, the number of
polydispersity <0.25.
EOs effectively moderates the degree of hydrophilicity. In
contrast, the hydrophobic portion of the surfactant was provided
Table 6. Particle Diametersa (nm) by DLS and Conductivitiesb by alkyl and aryl groups. The degree of hydrophobicity is only
(µS/cm) of Benzyl Alcohol in Water weakly moderated by the length of the hydrocarbon chain. The
surfactant S15 S20 S25 alkylamine ethoxylates have similar thermodynamic properties
R10 R10 R10 R20 as the ethoxylated octyl- and nonylphenols. The lower ∆Gma of
CO720 18.6(1.5) 7.4(0.3)*** 5.4(1.4) 27.4(6.5)** the alkylamine ethoxylates indicates that less energy is required
808 696 587 385 to form micelles from the surface adsorbed state; however, their
CA720 14.7(0.7) 6.9(0.2)*** 4.6(0.9)* 23.1(2.3)*** smaller surface area dictates that more material is required to
747 624 512 347 form the micelles as noted in the N values in Table 4.
N150 12.1(2.1) 7.0(0.8)* 5.3(1.0) NMc
965 761 588 373 Empty Micelles. The observed transparency in these systems
K215 6.0(0.5) 4.6(0.8) 4.1(0.8) 11.0(1.0)** suggests that the emulsions formed in these systems are
827 697 583 398 significantly less than 200 nm in diameter. Indeed, detailed DLS
a
For particle diameters, the average of three trials is reported with
measurements show the particle size is actually <10 nm in
standard deviations given in parentheses. The polydispersity values of diameter. Statistical differences at the 90%, 95%, and 99%
all samples were below 0.20, and all samples showed single mode confidence limit in the diameters of empty micelles in Figure 5
distributions. Statistical significance compared to the column on left: are shown with asteriks based on the pooled standard deviation
* ) 90%, ** ) 95%, *** ) 99%. b Second row of data for each surfactant between the two data sets.30
solution is conductance in µS/cm. c NM ) Not measured. There are few reports of variations in micelle size with
surfactant concentration in the absence of a dispersed phase, as
reported here. Tomsic et al.31 reported a decrease in micelle
could lead to a rapid method for screening an entire terniary diameter for Brij 35 (C12E23) from 9.22 nm at S5 to 8.28 nm at
phase diagram. S10. In contrast, Ruiz et al.32 reported a decrease in the micellar
Micelle Thermodynamics. The thermodynamic analysis of diffusion coefficient for Triton X-100 (branched C8φE10) from
these surfactants based on surface tension measurements provide S0.5 to S3.5. This corresponds to an increase in micelle diameter
values that are consistent with previously reported Gibb’s energies from 7.5 to 9.0 nm. These results are summarized with one sample
for similar nonionic surfactants.3 These values indicate that the from this work in Table 7.33
solubilization of the surfactants in water is driven by their It is well-established that EO chains can take multiple
surface active nature, ∆G°ads < 0, and the formation of micelles, conformations including extended, helix, and random coil.34-36
∆G°mic < 0.
Previous studies have shown the dominant driving force in (29) K205 does not fit this trend apparently due to additional structural
differences. K205 is a branched alkylamine having a total of five EO units in two
surface adsorption and micelle formation is due to entropy changes branches. This would lead to a smaller effect on the water structure by the EOs
within the water at the interface with the hydrophobic group.1,28 of K205 in comparison to the straight chain EOs of the other surfactants.
At this interface, the hydrogen bonding is weaker but the water (30) Harris, D. C. QuantitatiVe Chemical Analysis, 6th ed.; W. H. Freeman
& Co.: New York, 2003; Chapter 4.
is more ordered compared to water molecules in the bulk phase. (31) Tomšič, M.; Bešter-Rogač, M.; Jamnik, A.; Kunz, W.; Touraud, D.;
Thus, in going from a solubilized state to a surface adsorbed Bergmann, A.; Glatter, O. J. Colloid Interface Sci. 2006, 294, 194-211.
state (A) f (B), ∆S is positive and the changes are largely (32) Molina-Bolı́var, J. A.; Aguiar, J.; Ruiz, C. C. J. Phys. Chem. B 2002, 106,
870-877.
associated with the lipophilic part of the molecule essentially (33) Mishael and Dubin report for Triton X-100 a micelle diameter of 11.2
going from water solubilized to air. The data in Table 3 are in nm at S0.6 and 13.9 nm at S3.9. Mishael, Y. G.; Dubin, P. L. Langmuir, 2005,
21, 9803-9808.
(34) Tanford, C.; Nozaki, Y.; Rohde, M. F. J. Phys. Chem. 1977, 81, 1555-
(28) Scatena, L. F.; Brown, M. G.; Richmond, G. L. Science 2001, 292, 908- 1560.
912. (35) Olsson, U.; Schurtenberger, P. Langmuir 1993, 9, 3389-3394.
Single-Phase Microemulsions in Toluene
The chain length varies from approximately 3.5 Å/EO in the
fully extended conformation to 1.1 Å/EO in the random coil
conformation with g20 EOs.34 The EO values for T916 micelles
in this work were obtained from the experimental DLS diameters.
First, the diameters were adjusted for the alkyl chain length
obtained from Chem3D energy minimized structures (see below)
and then converted to length in Å/EO. Classification of EO
conformation was based in Figure 5 of Tanford et al.34 The range
of reported values fits nicely with the expected geometric
limitations for these systems.
Some simple molecular dynamics calculations were also done
on the T916 system using Chem3D.37 The gas-phase energy
minimized structure of T916 gave an equilibrated chain length
of 3.16 nm for the fully extended conformer. This is very close
to the DLS measurement for the empty micelle at S5 and
corresponds to 3.4 Å/EO which is nearly identical to the value
given by Tanford. Solvent was added to the computer simulation,
and three energy minimizations were performed with one
surfactant molecule, five toluene molecules, and ninety-four water
molecules. This gave an average chain length of 2.23 nm for the
resulting coiled conformer and corresponds to a value of 1.8
Å/EO; again, in excellent agreement with the value reported by
Tanford. This calculated value also matches well with the DLS
value for the empty T916 micelles in water at S25, as shown in
Table 7 and indicated in Figure 5. Thus, the observed decrease
in empty micelle size as surfactant concentration increases is
likely due to stabilization of the coiled conformer via pairwise
solute-solvent and solute-solute interactions. The reverse trend
observed with TX100 was noted to be associated with degree
of hydration,32 which may be more important at the lower
surfactant concentrations used in their work.
In general, when more surfactant is added to these empty
micelle systems, mass balance dictates that the system must be
forming either larger micelles or more micelles or a combination
of both. Smaller micelle size, as surfactant concentration was
increased, was observed in all six systems as shown in Figure
5. This implies that there must be more micelles forming. A
preliminary study of the sample turbidity at 405 nm using a 96
well plate reader indicated an increase in turbidity as surfactant
concentration increased for the K205, CA720, and K215 samples.
Thus, even though there is a small decrease in micelle size, there
is an increase in the total number of micelles. The low sensitivity
of the instrument coupled with the small path length of the well
plate precluded obtaining definitive supporting data for all six
surfactants. However, it appears quite reasonable that this is
indeed the case in all of the systems studied.
Toluene Systems. When oil is added to an “empty” micelle,
swelling of the micelle is expected to occur resulting in dispersed
droplets of oil,38 as observed in the current study. Comparison
of the empty micelle diameters in Figure 5 and dispersed oil
droplet diameters in Table 5 reveals a swelling of the micelles.
For the CO720 and CA720 samples, the larger diameters are
statistically significant at the 99% level for S15 and at the 90%
level for S25. For K205 and N150, there was no significant
difference for S25 samples. The average size of the micelles
does increase slightly upon the addition of toluene, however the
relatively large standard deviations associated with these particular
samples precludes assigning statistical significance to them.
Within Table 5 are also the following comparisons: micelle
(36) Xu, J.-T., Xue, L.; Fan, Z.-Q.; Wu, Z.-H.; Kim, J. K. Macromolecules
2006, 39, 2981-2988.
(37) Chem 3D, version 9.0; CambridgeSoft Corporation: Cambridge, MA
2004.
(38) Nakajima, H. In Industrial Applications of Microemulsions; Solans, C.,
Kunieda H., Eds.; Marcel Dekker: New York, 1997; Chapter 8.
Langmuir, Vol. 24, No. 9, 2008 4491
size as surfactant concentration increases keeping toluene
concentration constant, micelle size as toluene concentration
increases keeping surfactant concentration constant, and micelle
size with variations of surfactant HLB. Statistical analysis of any
two points in a given row showed no significant differences even
to the 90% confidence limit. There are three possibilities for this
unusual behavior: (i) more micelles are formed, (ii) the micelle
shape is becoming nonspherical, or (iii) a phase transition is
occurring. Turbidity measurements were performed on these
samples using a 96 well plate reader as noted above. Again, the
data showed increased scattering in these systems indicating
more micelles are formed. The dissymmetry factor, Z45°/135°, which
is the ratio of scattered light intensity at 45° to 135°, was also
measured for all DLS samples. It has been well-established that
scattering intensity varies with observation angle and increases
with particle size and shape.39,40 To obtain useful shape
information, the particle size limit40 is approximately a spherical
diameter of 0.1λ, which in our case is 63.3 nm. In this study, all
of the particle diameters were well below this limit, and thus no
shape information was available.
In addition, cross polar and conductivity measurements were
done on each of these samples. An observed changed in
polarization would indicate the presence of a lamellar phase,
while large variations in conductivity are indicative of phase
transitions.41-43 Almost all of the single-phase samples in Tables
5 and 6 were not birefringent, indicating a lack of lamellar phases.
The intial high conductivities of the CO720 and CA720 solutions
indicate that a water continuous phase (L1) is present. The gradual
decrease in conductivity in the CA720 samples as surfactant is
increased with constant oil and as oil is increased with constant
surfactant is indicative of the increased lipophilicity in both cases
with no phase change. This is also consistent with a droplet
model microemulsion phase.44,45 In the case of CO720, there is
nearly a 10× decrease in conductivity as surfactant concentration
is increased from S15 to S25 while oil is held constant at R15.
This indicates an inversion from water continuous (L1) to oil
continuous (L2) phase. Further minor decreases in conductivity
as oil concentration increases from S25R15 to S25R25 as
lipophilicity increases in the continuous phase are expected and
observed. The N150 systems showed an interesting intermediate
conductivity. These have been interpreted as due to bicontinuous
microemulsion (L3) phases.41 The first four K205 samples in
Table 5 are the only ones to show a positive cross-polar response
indicative of a lamellar (LR) phase being present. In this series,
there is also an approximately 10× decrease in conductivity.
This appears to be due to a transition from a lamellar phase at
S15R15 to a reversed lamellar phase as surfactant concentration
is increased. This indicates that the conducting salt water is being
removed from the continuous portion of the solution, possibly
due to vesical formation.42
Thus, three of the four toluene systems show phase changes
occurring within the single-phase regime of their phase diagrams.
The alkylamine ethoxylates with toluene as the oil have similar
micelle sizes as the octyl- and nonylpenol ethoxylates but show
a distinctly different set of phase changes. In contrast, the observed
(39) Kenyon, A. S.; LaMer, V. K. J. Colloid Sci. 1949, 3, 163-184.
(40) Zachariah, M. R.; Chin. D.; Semerjian, H. G.; Katz, J. L. Appl. Opt. 1989,
28, 530-536.
(41) Zourab, S. M.; Miller, C. A. Colloids Surf., A 1995, 95, 173-183.
(42) Beck, R.; Hoffmann, H. Phys. Chem. Chem. Phys. 2001, 3, 5438-5443.
(43) Ozawa, K.; Solans, C.; Kunieda, H. J. Colloid Interface Sci. 1997, 188,
275-281.
(44) Lee, L. T.; Langevin, D.; Wong, K.; Abillon, O. Condens. Matter 1990,
2, SA333-SA338.
(45) Frieberger, N.; Moitzi, C.; de Campo, L.; Glatter, O. J. Colloid Interface
Sci. 2007, 312, 59-67.
4492 Langmuir, Vol. 24, No. 9, 2008
Figure 6. Nakajima plot of radius vs benzyl alcohol oil fraction:
[ CO720, 9 CA720, 2 N150, b K215. Trendline for CA720 is
shown for clarity.
behavior with benzyl alcohol as the oil is much more uniform
as discussed below.
Benzyl Alcohol Systems. Benzyl alcohol was used as the oil
phase because of its lower volatility, structural similarity to
toluene, and increased polarity. Some useful comparisons between
the toluene and benzyl alcohol systems can be made. First, normal
micelle swelling is observed by comparing S25_R0 in Figure 5
to S25_R10 and S25_R10 to S25_R20 in Table 6. This is also
observed in comparing the first three columns of Table 6 right
to left, where the R value is held constant and the surfactant
concentration is increased. When R is converted to %O by mass,
these values become 8.5, 8.0, and 7.5 %O, respectively. Thus,
more benzyl alcohol results in more swelling of the micelles.
Nakajima38 predicts an increase in micelle swelling with
increasing oil content based on the core volume and surface area
of an oil in water micelle.
r) ( ) (
3M
ΓNAFa
ω+
3RM
Γ NAFc
+d ) (8)
In eq 8, adapted from Nakajima, r is the radius of the whole
droplet, d is the thickness of the hydrated hydrophilic head group
of the surfactant, M is the molar mass of the surfactant, Γ is the
area occupied by a single surfactant molecule at the oil-water
interface, NA is Avogadro’s number, Fa and Fc are the densities
of the oil and the alkyl chain of the surfactant, and R is the weight
fraction of the alkyl chain of the surfactant to the total weight
of the surfactant.
From the slope of the plot (Figure 6) in the limit of low oil
concentration, one may obtain the surface area per molecule of
the surfactant in the benzyl alcohol containing micelle. This
parameter is not accessible using surface tension plots due to the
lack of surface activity seen in oil solvent compared to water
solvent.1,3 This analysis requires the radius of the micelle in the
limit of 0% oil, i.e., the empty micelle. However, Figure 5 indicates
this value depends on the surfactant concentration. The values
of S25 from Figure 5 appear to give the best values of Γ in two
of the four surfactants.46 These were found to be 118, 101, 114,
and 92 Å2/molecule for CO720, CA720, N150, and K215,
respectively, which are all larger than their corresponding values
in toluene. In this regard, K215 behaves similarly to CO720 and
CA720.
(46) The limiting value of the empty micelle diameter gives a good result for
Γ for CA720 and N150. This value was not available for K215, however the next
two data points in this particular series are still in the lower linear regime of the
Nakajima plot and also provided a reasonable value for Γ. A discrepency shows
up when some of the values from empty micelles are used in the Nakajima plots
leading to negative slopes. For CO720, the empty micelle diameter at S5 gave
the most reasonable value. We have chosen what appears to be the best values
to report but are cautious that these should be viewed as semiquantitative.
Gotch et al.
Table 8. Micelle Diameter Stabilitya
surfactant S5 S15 S25
CA720 R5 R5 R10 R5 R10
3 months 7.3(0.3) 5.7(0.4) 6.7(0.1) 4.7(0.8) 5.3(0.5)
9 months 5.4(0.1)** 4.2(0.4)* 4.3(0.5)* 2.9(0.3)** 4.2(0.6)
a
Average of three trials is reported with standard deviations given
in parentheses. Statistical significance compared to 3 month samples:
* ) 90%, ** ) 95%, *** ) 99%
Without surfactant present BA has a 5% solubility in water
due to the presence of the polar alcohol group. It is well-established
that micelles are dynamic with surfactant and oil moving in and
out of the micelle into the bulk on a continual basis.8,47 Thus,
it appears that the BA has better accessibility to the polar head
groups and effectively provides a larger channel in the micelle
structure compared to toluene.
In contrast to most of the toluene-based systems, the
conductivities of the benzyl alcohol systems all showed a smooth
trend. This indicates that no phase change occurs over the
composition range studied and is consistent with an increase in
lipophilicity, as observed in the water/toluene/CA720 solutions.
Further, the relative values of the conductances are in the same
range that others have interpreted as a bicontinuous (L3) phase.41,43
In terms of phase stability, the more polar benzyl alcohol-based
systems appear to be better suited for replacing nonylphenol
ethoxylates with alkylamine ethoxylates.
Micelle Stability. One final consideration of this study was
the long-term stability of the single-phase systems. It is very
easy to qualitatively classify an emulsion as opaque, translucent,
or transparent, indicating approximate particle sizes of >200
nm, 50-200 nm, or <50 nm,48 respectively. Almost all of the
transparent systems observed in this study remained completely
transparent in the time frame of 6-18 months. A few systems
were studied quantitatively by DLS for long-term stability and
are summarized in Table 8.
The series of CA720/toluene/water single-phase systems
ranging from S5R5 to S25R10 showed an average decrease in
micelle diameter of -43% between the ages of approximately
3 and 9 months. The differences for four of the five samples
studied are statistically significant at the 90 or 95% confidence
level using pooled standard deviations in the t-test. If variations
between the particle sizes of the individual samples are ignored,
the average diameter at 3 months is 5.9(0.4) nm and decreases
to 4.2(0.4) nm or approximately -29%. While these variations
are large on a relative basis, it is reasonable to conclude that the
particle size for the surfactant/toluene/water systems is ap-
proximately 5 nm and relatively stable over a time frame of
many months.
Conclusions
This study has shown that single-phase oil/water microemul-
sions may be formed using nonionic surfactants in concentrations
e25% by mass with toluene or benzyl alcohol as the oil. At an
optimum HLB of ∼14.5, a toluene/surfactant mol ratio of >10:1
may be achieved in single-phase transparent Winsor Type IV
solutions. The surfactant structure varied between 8 and 12 carbon
chains, with and without a phenyl group, in the hydrophobic
portion and between 0 and 15 EO units in the hydrophyllic portion.
The HLB number was the best single parameter to predict single-
phase microemulsion formation.
(47) Capek, I. AdV. Colloid Interface Sci. 2004, 107, 125-155.
(48) Gee, R. P. Colloids Surf., A 1998, 137, 91-101.
Single-Phase Microemulsions in Toluene
The alklyamine ethoxylate surfactants show similar properties
to the octyl- and nonylphenol ethoxylates with regard to micelle
thermodynamics, size, and macroscopic phase stability and
provide a possible environmental friendly alternative. However,
the specific molecular properties of the oil phase appear to have
an influence on microscopic phase stability, as noted in the
observed variations in conductance of single-phase, toluene-
based, solutions. The more polar benzyl alcohol required a higher
HLB for optimum single-phase stability, whose lower bound
was determined to be HLB g 17. In contrast to toluene, when
benzyl alcohol was used as the oil, both macroscopic and
microscopic phase stability was observed. This indicates alky-
lamine ethoxylates have good potential to replace nonylphenol
ethoxylate surfactants in cases where a more polar oil like benzyl
alcohol is part of the formulation.
Conformational changes in the EO polar head group appear
to play a significant role in the stabilization of these nanosystems
Langmuir, Vol. 24, No. 9, 2008 4493
in both the empty and oil swollen state. A Nakajima plot was
used to analyze the surfactant area in the oil phase, which is
usually impossible to obtain using standard surface tension
measurements and appears to be generally applicable to other
systems. The particle size remained essentially constant at ∼5
nm for the single-phase systems if the oil was toluene between
R5 and R30 and showed long-term stability up to 18 months, the
longest time frame studied. The stable, nanometer sized micelles
formed in these systems suggests the potential use of these
formulations for organic or nanoparticle synthesis.
Acknowledgment. We gratefully acknowledge McGean-
Rohco and Mount Union College for support of this work. Special
thanks to Jon Page for assistance with DLS measurements, Seth
Hennon for assistance with conductivity measurements, and Jenny
Singletonforassistancewithturbiditymeasurements.AJG_Dan7:14.
LA7024318