01.06.2013 Views

Solubilization-emulsification mechanisms of detergency

Solubilization-emulsification mechanisms of detergency

Solubilization-emulsification mechanisms of detergency

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

Colloids and Surfaces A: Physicochemical and Engineering Aspects, 74, 169-215 (1993)<br />

Review<br />

<strong>Solubilization</strong>-<strong>emulsification</strong> <strong>mechanisms</strong> <strong>of</strong> <strong>detergency</strong><br />

Clarence A. Miller a,* Kirk H. Raney b<br />

a Department <strong>of</strong> Chemical Engineering, Rice University, P.O. Box 1892, Houston, TX 77251-1892, USA<br />

b Shell Development Co., Westhollow Research Center, P.O. Box 1380, Houston, TX 77251-1380, USA<br />

(Received 14 November 1992; accepted 23 January 1993)<br />

Abstract<br />

The removal <strong>of</strong> oily soils from fabrics having high contents <strong>of</strong> polyester or other synthetic materials<br />

occurs largely by a solubilization-<strong>emulsification</strong> mechanism. A systematic investigation <strong>of</strong> this mechanism<br />

has been conducted during the past several years and is reviewed here. The research has utilized a variety <strong>of</strong><br />

oily soils containing hydrocarbons, triglycerides, and long-chain alcohols and fatty acids and has included the<br />

determination <strong>of</strong> equilibrium phase behavior, the observation <strong>of</strong> dynamic behavior which occurs when<br />

surfactant-water mixtures contact oily soils, and measurement <strong>of</strong> soil removal from polyester-cotton fabrics.<br />

In most cases, pure surfactants and oils have been used for simplicity, but data showing the applicability <strong>of</strong><br />

major conclusions to systems containing commercial surfactants are presented. Because typical anionic<br />

surfactants are too hydrophilic to achieve the desired phase behavior, the work has employed non-ionic<br />

surfactants and mixtures <strong>of</strong> non-ionics and anionics, One major conclusion is that maximum soil removal<br />

usually does not occur when the soil is solubilized into an ordinary micellar solution, but instead when it is<br />

incorporated into an intermediate phase such as a microemulsion or liquid crystal that develops during the<br />

washing process at the interface between the soil and washing bath. Indeed, for hydrocarbon and triglyceride<br />

soils, the washing bath is itself a dispersion <strong>of</strong> a surfactant-rich liquid or liquid crystalline phase in water for<br />

conditions <strong>of</strong> optimum <strong>detergency</strong>, i.e. the temperature <strong>of</strong> the surfactant solution is above - sometimes far<br />

above - its cloud point temperature.<br />

Key words: Detergency; Emulsification; <strong>Solubilization</strong><br />

1. General remarks on <strong>detergency</strong><br />

Fabric <strong>detergency</strong> is a surprisingly complex<br />

process involving interactions among aqueous<br />

detergent solutions, soils, and fabric surfaces.<br />

This Process may occur in an industrial setting<br />

in which large volumes <strong>of</strong> similarly soiled<br />

fabrics are washed, or in a household setting in<br />

which small amounts <strong>of</strong> fabrics containing<br />

differing amounts <strong>of</strong> a wide variety <strong>of</strong> soils are<br />

washed. Because <strong>of</strong> their unique ability to<br />

adsorb at both fabric-water and soil-water<br />

interfaces, surfactants play an essential role in<br />

soil removal processes. To achieve the desired<br />

levels <strong>of</strong> surfactants in the washing solution, the<br />

* Corresponding author.<br />

concentration <strong>of</strong> surfactants (typically nonionic,<br />

anionic, or both) in most liquid and powder<br />

detergent formulations is in the range 10-40%<br />

by weight [1].<br />

Several factors influence the effectiveness <strong>of</strong><br />

surfactants in laundry detergents. The<br />

composition <strong>of</strong> a fabric is important in<br />

determining the mechanism by which soils are<br />

lifted from it. Cotton fabric contains rough and<br />

irregularly shaped hydrophilic fibers [1,2]. In<br />

contrast, synthetic polyester fabric contains<br />

uniform cylindrical fibers which, being <strong>of</strong> a<br />

more hydrophobic nature than cotton, are more<br />

tenaciously covered by oil. The differences in<br />

surface properties <strong>of</strong> the two materials are<br />

demonstrated by the much smaller contact angle<br />

in air formed by water on a cellulose film (32º)


170 C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215<br />

than that by water on a polyester film (79º) [3].<br />

In addition to fiber surface composition and<br />

morphology, the weave <strong>of</strong> the fabric can<br />

influence <strong>detergency</strong>, more loosely woven<br />

fabric typically being easier to clean.<br />

Not surprisingly, the amounts and types <strong>of</strong><br />

soils present on fabric are key factors in<br />

determining the effectiveness <strong>of</strong> a detergent<br />

solution. Oily soils include such common soils<br />

as skin sebum, dirty motor oil, and vegetable oil.<br />

Clay is classified as a particulate soil. Scanning<br />

electron microscopy, both conventional and<br />

environmental, has proven quite useful for<br />

viewing the distribution <strong>of</strong> both oily and<br />

particulate soils within woven fabric samples<br />

[2,4]. While surfactants play the key role in<br />

removing oily and particulate soil, protease<br />

enzymes are commonly present in both powder<br />

and liquid laundry detergents to chemically<br />

break down polymeric protein soil stains such as<br />

blood, egg, and cocoa. Lipase enzymes are also<br />

now being used in detergents to hydrolyze<br />

triglyceride soils and thereby aid the surfactant<br />

in soil removal [5]. Bleaches, both peroxygen<br />

and chlorine-based, decolorize stains such as<br />

those from tea and wine by destroying the<br />

chromophores in the organic molecules<br />

adsorbed to the fiber surfaces [6].<br />

Water hardness and temperature can<br />

pr<strong>of</strong>oundly influence detergent effectiveness. In<br />

hard water, e.g. 300 ppm hardness, calcium and<br />

magnesium ions may precipitate certain<br />

surfactants prior to their being able to act on the<br />

soil. The divalent ions also form complexes<br />

between soils and fabric which increase their<br />

attraction, making the soil more difficult to<br />

remove [1]. Builders such as zeolite, sodium<br />

tripolyphosphate (STPP) and sodium carbonate<br />

are used in powders to negate the effect <strong>of</strong> water<br />

hardness by either precipitating the divalent<br />

ions, as in the case <strong>of</strong> sodium carbonate, or<br />

sequestering the ions from the water, as in the<br />

cases <strong>of</strong> zeolite and STPP. The latter is<br />

particularly effective in this regard. Surfactant<br />

precipitation may occur in cold water for<br />

surfactants with high Krafft points [I]. Also,<br />

wash water temperature, in addition to changing<br />

the performance characteristics <strong>of</strong> the dissolved<br />

surfactants, determines the physical properties<br />

<strong>of</strong> oily soils left on the fabric. As the washing<br />

temperature is reduced, the viscosity <strong>of</strong> the soils<br />

increases or the soils may even solidify, making<br />

them more difficult to remove with the same<br />

level <strong>of</strong> agitation.<br />

Recent trends in washing habits around the<br />

world have made the proper choice <strong>of</strong> a<br />

surfactant system for a detergent formulation<br />

more critical than ever. Greater use <strong>of</strong><br />

temperature-sensitive synthetic fabrics such as<br />

polyester or polyester-cotton blends as well as<br />

energy conservation have led to a world-wide<br />

trend to lower washing temperatures [7].<br />

Phosphate limits or bans have resulted in the use<br />

<strong>of</strong> less effective builders or <strong>of</strong> no builders, so<br />

that surfactants are less protected from the<br />

negative impact <strong>of</strong> water hardness. Also, as a<br />

result <strong>of</strong> the effort to reduce the total amount <strong>of</strong><br />

chemicals released to the environment as well as<br />

the volume <strong>of</strong> packaging materials, detergent<br />

manufacturers are formulating products with<br />

lower dosage requirements. As a result <strong>of</strong> these<br />

trends, the cleaning efficiency required from<br />

surfactant systems is steadily increasing.<br />

2. Mechanisms for removal <strong>of</strong> oily soils<br />

The trends described above can have a<br />

particularly negative impact on the removal <strong>of</strong><br />

Oily soils from synthetic fabrics. Commonly<br />

encountered examples <strong>of</strong> this troublesome<br />

soil-fabric combination are dirty motor oil,<br />

cooking oil, or sebum on 100% polyester or<br />

polyester-cotton blends. Many studies have been<br />

performed to visualize the <strong>mechanisms</strong> by<br />

which such oils are removed from synthetic<br />

fabrics [8-12]. Although in practical situations<br />

the oily soils are trapped in the interstices<br />

between fabric fibers and as thin films along the<br />

fiber surfaces, most work has focused on the<br />

removal <strong>of</strong> oil drops from flat surfaces or<br />

individual fibers with the assumption that<br />

similar removal <strong>mechanisms</strong> would be relevant<br />

to removal from fabric. In fact, good correlation<br />

between Oil removal from flat films and from<br />

chemically similar fabric has been reported [8].<br />

Of relevance to this type <strong>of</strong> study is Young 5<br />

equation, which relates the interfacial tension


C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215 171<br />

between the surfactant solution and oil (γ ow), oil<br />

and solid substrate (γ os) and surfactant solution<br />

and solid substrate (gws) to the equilibrium<br />

contact angle q measured through the soil<br />

cos θ = γ ws − γ os<br />

γ ow<br />

(1)<br />

For quite hydrophilic surfaces like cotton, gws<br />

is smaller than g0s, and a contact angle greater<br />

than go is commonly achieved. Anionic<br />

surfactants which adsorb on the fabric with their<br />

negatively charged head groups oriented toward<br />

the detergent solution are particularly effective<br />

in reducing gws. in this case, the roll-up<br />

mechanism is operative: the water preferentially<br />

wets the fabric, causing the oily stains to be<br />

entirely lifted <strong>of</strong>f the fibers into the washing<br />

solution. This behavior, shown schematicaily in<br />

Fig. 1(b) for soil removal from a flat surface. is<br />

enhanced on cotton fabric due to swelling <strong>of</strong> the<br />

cotton fibers with water which increases the<br />

hydrophilicity <strong>of</strong> the fabric surfaces [9,13].<br />

For low surface energy, i.e. hydrophobic,<br />

materials such as polyester, a contact angle <strong>of</strong><br />

less than 90º is usually observed, and small<br />

portions <strong>of</strong> the oily soil may be removed by<br />

hydraulic currents at the soil-water interface, as<br />

shown in Fig. 1(a). In Fact, if the fabric surface<br />

is initially completely covered by oily soil, no<br />

location is available for the surfactant solution<br />

Fig. 1. Mechanisms <strong>of</strong> liquid soil removal: (a)<br />

<strong>emulsification</strong>; (b) roll-up.<br />

to reach the fiber surface and undercut the soil.<br />

Observation <strong>of</strong> this "necking" or <strong>emulsification</strong><br />

mechanism has been made by many<br />

investigators for mineral oils and mineral<br />

oil-polar soil mixtures on hydrophobic flat films<br />

and fibers [8-12]. Removal in this manner is<br />

enhanced by low interfacial tension at the<br />

oil-water interface which allows the oil film to<br />

be deformed easily to form small emulsion<br />

droplets.<br />

Several factors have been studied with regard<br />

to their effect on the <strong>emulsification</strong> mechanism<br />

for the removal <strong>of</strong> mixtures <strong>of</strong> mineral oil and<br />

polar organic alcohols or acids from polyester<br />

[8-10]. Such model systems, depending on the<br />

ratio <strong>of</strong> the non-polar and polar constituents, can<br />

be considered to be representative <strong>of</strong> sebum<br />

soils from the skin. The rate <strong>of</strong> <strong>emulsification</strong> <strong>of</strong><br />

mineral oil-oleic acid mixtures from polyester<br />

(Mylar) films was found to change as the oleic<br />

acid content was varied [8], Other factors such<br />

as electrolyte concentration and temperature<br />

were also found to have large effects on the rate<br />

<strong>of</strong> soil removal by this mechanism [8,9]. In<br />

some situations, <strong>emulsification</strong> <strong>of</strong><br />

non-polar-polar soil mixtures without external<br />

agitation, i.e. spontaneous <strong>emulsification</strong>, has<br />

been observed [ 13,14]. Emulsification, roll-up,<br />

and other adhesion and detachment phenomena<br />

involving oily soils and solid surfaces are<br />

reviewed in the accompanying paper [ 15].<br />

Another mechanism <strong>of</strong> oily soil removal<br />

involves the formation <strong>of</strong> intermediate phases at<br />

the detergent solution-oil interface<br />

[11,13,14,16]. Apart from the recent studies<br />

described below, this mechanism has most <strong>of</strong>ten<br />

been reported for the removal <strong>of</strong> soils containing<br />

large quantities <strong>of</strong> polar constituents. The<br />

growth <strong>of</strong> liquid crystal occurs in these systems<br />

due to interaction at the interface between the<br />

polar soil constituents and the adsorbed<br />

surfactants. After growing to a sufficient extent,<br />

the intermediate phase is broken <strong>of</strong>f by agitation<br />

and emulsified into the aqueous solution<br />

allowing fresh contact <strong>of</strong> the remaining soil with<br />

the detergent solution.<br />

Direct solubilization <strong>of</strong> oily soils into<br />

surfactant micelles can also occur to a<br />

significant extent if a large excess <strong>of</strong> surfactant


172 C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215<br />

relative to oil is present and if the surfactant is<br />

above its CMC. The solubilization <strong>of</strong> very small<br />

oil drops from polymer fibers has been<br />

visualized for a variety <strong>of</strong> non-polar oils<br />

representative <strong>of</strong> liquid laundry soils [17].<br />

However, soil solubilization rates are <strong>of</strong>ten<br />

enhanced when surfactant-rich phases, either<br />

isotropic or liquid crystalline in nature, are<br />

present in the washing solution. Such phases<br />

exist, for instance, when non-ionic surfactants<br />

are above their cloud points. These phases can<br />

either solubilize oily soils directly or interact<br />

with soil to form intermediate surfactant-rich<br />

phases such as microemulsions containing large<br />

amounts <strong>of</strong> oil. Under favorable conditions, the<br />

intermediate phases can be emulsified into the<br />

washing bath. A detailed discussion <strong>of</strong> this<br />

mechanism <strong>of</strong> soil removal is the subject <strong>of</strong> this<br />

review.<br />

Clear evidence exists that solubilization and<br />

<strong>emulsification</strong> are major factors in removal <strong>of</strong><br />

oily soils from hydrophobic, synthetic fabrics<br />

[18,19]. Unlike roll-up, in which the interaction<br />

<strong>of</strong> the fabric with the oily soil and water is most<br />

critical, the solubilization-<strong>emulsification</strong> mechanism<br />

occurs primarily at the soil-detergent<br />

solution interface and is therefore directly<br />

infiuenced by the phase behavior <strong>of</strong> the<br />

corresponding oil-water-surfactant system. For<br />

example, the formation <strong>of</strong> intermediate liquid<br />

crystalline phases in fatty acid-surfactant-water<br />

systems has been explained by the equilibrium<br />

phase behavior <strong>of</strong> those systems [ 16]. Also,<br />

spontaneous <strong>emulsification</strong> phenomena in<br />

oil-water-surfactant systems have been shown to<br />

be predictable from equilibrium phase behavior<br />

[20]. Therefore, an understanding <strong>of</strong> the phase<br />

behavior in these systems is needed to predict<br />

the effectiveness <strong>of</strong> and/or to optimize detergent<br />

solutions for specific soil compositions and<br />

washing conditions. Available information on<br />

phase behavior is reviewed in the next section.<br />

In subsequent sections, systematic studies <strong>of</strong><br />

dynamic contacting between aqueous surfactant<br />

solutions and oils as well as <strong>detergency</strong> studies<br />

using the same systems are reviewed in order to<br />

further explain the role <strong>of</strong> the solubilization<strong>emulsification</strong><br />

mechanism in practical <strong>detergency</strong><br />

processes.<br />

3. Equilibrium phase behavior<br />

As indicated above, some knowledge <strong>of</strong> the<br />

equilibrium phase behavior <strong>of</strong><br />

soil-water-surfactant systems is needed to<br />

understand solubilization<strong>emulsification</strong><br />

<strong>mechanisms</strong> <strong>of</strong> <strong>detergency</strong>. In this section, we<br />

review such phase behavior with emphasis<br />

given to model systems with well-defined<br />

components. Results are given for one- and<br />

twocomponent oils consisting <strong>of</strong> hydrocarbons,<br />

triglycerides, and/or long-chain alcohols or<br />

acids, representing soils such as lubricating oils,<br />

cooking oils and sebum. In many cases<br />

single-component specific alcohol ethoxylates<br />

are the surfactants. However, most features <strong>of</strong><br />

the behavior described should be applicable to<br />

more complex systems as well, e.g. those<br />

containing multicomponent commercial<br />

surfactants.<br />

3. 1. Phase behavior <strong>of</strong> soil-free washing baths<br />

We begin with the fluids used for washing,<br />

i.e. rather dilute mixtures <strong>of</strong> surfactant and water<br />

with inorganic salts and/or various additives<br />

also present in some cases. As is well known, a<br />

typical hydrophilic surfactant above its Krafft<br />

temperature forms micelles in water at<br />

concentrations above its CMC. If the<br />

temperature or composition <strong>of</strong> a micellar<br />

solution is varied in such a way that the<br />

surfactant becomes less and less hydrophilic, the<br />

separation <strong>of</strong> another phase eventually results.<br />

If. for instance, the temperature <strong>of</strong> a micellar<br />

solution <strong>of</strong> an ethoxylated alcohol is increased, a<br />

second liquid phase begins to appear when the<br />

so-called cloud point temperature is reached. As<br />

Fig. 2 for the n-dodecyl pentaoxyethylene<br />

monoether (C12E5)-water system [21] shows,<br />

the cloud point is a function <strong>of</strong> surfactant<br />

concentration since clouding occurs when the<br />

coexistence curve forming the upper boundary<br />

<strong>of</strong> the aqueous surfactant solution L1 is crossed<br />

during heating. At temperatures well above the<br />

cloud point, the lamellar liquid crystalline phase<br />

La and yet another liquid phase L3, widely<br />

considered to consist <strong>of</strong> bilayers arranged in a<br />

sponge-like structure, are seen in this system for


C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215 173<br />

Fig. 2. Phase diagram <strong>of</strong> C 12E 5-water system [21]. L 1,<br />

L 2, and L 3 denote isotropic liquids; Lα, H 1, and V 1<br />

denote lamellar, hexagonal and viscous isotropic<br />

liquid crystalline phases, respectively. Reprinted with<br />

permission <strong>of</strong> the Royal Society <strong>of</strong> Chemistry.<br />

relatively low surfactant concentrations. As the<br />

particles <strong>of</strong> a liquid crystal do not coalesce as<br />

readily as liquid drops, the dispersions <strong>of</strong> La are<br />

frequently less turbid than those <strong>of</strong> L 3, a<br />

property which can be used to locate phase<br />

transition temperatures at which the La and L 3<br />

phases form [22]. At the highest temperatures<br />

shown in Fig. 2 the surfactant-rich liquid phase<br />

L 2 coexists with water.<br />

For more hydrophilic surfactants such as<br />

n-dodecyl hexaoxyethylene monoether (C 12E 6),<br />

clouding occurs at higher temperatures.<br />

Moreover, the Lα and L 3 phases do not appear<br />

at low surfactant concentrations; the La phase<br />

transforms continuously into L 2, and the cloud<br />

point curve is the only feature <strong>of</strong> this part <strong>of</strong> the<br />

phase diagram (see Fig.3). Phase diagrams for<br />

various binary nonionic surfactant-water<br />

systems are given by Mitchell et al. [23].<br />

Temperature effects are weaker for ionic<br />

surfactants and generally act in the opposite<br />

direction. Since the Debye length, a measure <strong>of</strong><br />

the electric double layer thickness, is<br />

proportional to (kT) 1/2 , where kT is the<br />

characteristic free energy <strong>of</strong> random thermal<br />

Fig. 3. Phase diagram <strong>of</strong> C 12E 6-water system [23].<br />

The symbols for the phases are as in Fig. 2 except<br />

that S is a solid phase and W is a water-rich liquid<br />

phase. Reprinted with permission <strong>of</strong> the Royal<br />

Society <strong>of</strong> Chemistry.<br />

motion, higher temperatures make ionic<br />

surfactant films more hydrophilic, with a greater<br />

tendency to curve toward an oil-in-water<br />

configuration. However, the addition <strong>of</strong><br />

inorganic salts has the opposite effect,<br />

compressing electric double layers and causing<br />

ionic surfactant films to become less<br />

hydrophilic. In some cases, a- second liquid<br />

phase is ultimately formed as salinity increases,<br />

i.e. the behavior is similar to clouding <strong>of</strong><br />

non-ionic surfactant solutions discussed above.<br />

This phenomenon was observed by McBain<br />

many years ago for aqueous soap solutions<br />

[23b]. Another example is shown along the<br />

upper boundary <strong>of</strong> Fig. 4 [24], with NaCl added<br />

to the sodium salt <strong>of</strong> a commercial ethoxylated<br />

sulfate based on a C 12-C 13 alcohol and<br />

containing an average <strong>of</strong> three ethylene oxide<br />

groups (Neodol 23-3S). As Fig. 4 indicates,<br />

multiphase regions involving the lamellar liquid<br />

crystal La are observed at even higher salinities.<br />

In other systems, for instance the Aerosol<br />

OT-NaCl-water system, the first phase formed<br />

upon increasing the salinity is the lamellar liquid<br />

crystalline phase [25,26]. Indeed, such behavior<br />

is typical for anionic surfactant-short-chain<br />

alcohol systems investigated for possible use in


174 C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215<br />

Fig. 4. Phase behavior <strong>of</strong> mixtures <strong>of</strong> C 12E 3, Neodol<br />

23-3S, and NaCl brine, with the temperature and total<br />

surfactant concentration fixed at 30ºC and 5.4 wt.%,<br />

respectively [24]. B, NaCl brine; W(non), weight<br />

fraction <strong>of</strong> non-ionic surfactant in the surfactant<br />

mixture. Reprinted with permission from Dr. Dietrich<br />

Stemkopff Verlag.<br />

enhanced oil recovery, with the L 3 phase found<br />

at still higher salinities [27]. The addition <strong>of</strong><br />

divalent cations (i.e. an increase in hardness)<br />

produces the same effects in these systems,<br />

although generally at lower electrolyte<br />

concentrations [27].<br />

Whether a liquid phase or the liquid crystal<br />

forms upon the addition <strong>of</strong> salt apparently<br />

depends on the relative importance <strong>of</strong> reducing<br />

the effective electrical repulsion between nearby<br />

ions within a micelle and reducing it between<br />

micelles. If the former is the dominant effect,<br />

the micelle shape changes from spherical to<br />

cylindrical to planar as the effective surfactant<br />

head group area decreases, a sequence in<br />

accordance with well-understood surfactant<br />

packing considerations in micelles [28]. The<br />

large, bilayer sheets corresponding to the planar<br />

configuration with nearly equal head and tail<br />

areas arrange themselves into the lamellar<br />

phase. In contrast, if reducing the repulsion<br />

between relatively small micelles is the more<br />

important effect, the micelles eventually<br />

flocculate to form a "coacervate" or micelle-rich<br />

liquid phase.<br />

It appears from the available evidence that<br />

coacervation is the likely outcome <strong>of</strong> increasing<br />

salinity for anionic surfactants that are rather<br />

hydrophilic, while liquid crystal formation is<br />

probable for surfactants whose hydrophilic<br />

characteristics only slightly outweigh their<br />

lipophilic characteristics in the absence <strong>of</strong> salt.<br />

Figure 4 provides an example. The addition <strong>of</strong><br />

less than 20% <strong>of</strong> the non-ionic surfactant<br />

n-dodecyl trioxyethylene monoether (C 12E 3)<br />

reduces the hydrophilic nature <strong>of</strong> the surfactant<br />

mixture sufficiently such that the liquid crystal<br />

phase forms instead <strong>of</strong> a coacervate when the<br />

NaCl concentration is increased. Such behavior<br />

can be explained as follows. Hydrophilic<br />

surfactants such as Neodol 23-3S require large<br />

concentrations <strong>of</strong> salt for the surfactant<br />

aggregates to become planar. Repulsion<br />

between small, nonplanar micelles is apparently<br />

reduced sufficiently so as to produce<br />

coacervation before the salinity increases<br />

enough for planar micelles to form. For less<br />

hydrophilic surfactants, less salt is needed to<br />

produce planar aggregates, and formation <strong>of</strong> the<br />

lamellar phase occurs before coacervation.<br />

The cloud point phenomenon discussed above<br />

for non-ionic surfactants may also be viewed as<br />

coacervation <strong>of</strong> a micellar solution induced by<br />

making the surfactant less hydrophilic. In this<br />

case. raising the the temperature reduces the<br />

interaction between the ethylene oxide chains<br />

and water and thereby reduces the repulsion<br />

between micelles or even reverses it to an<br />

attraction. Of course, the effective area <strong>of</strong> the<br />

ethylene oxide head group is also reduced, and a<br />

change from spherical to cylindrical micelles,<br />

which would facilitate coacervation by an<br />

entropic mechanism, can occur before the cloud<br />

point is reached. Unlike the situation for anionic<br />

surfactants, however, there do not seem to be<br />

any reports <strong>of</strong> the lamellar liquid crystalline<br />

phase forming directly from micellar solutions<br />

Or ethoxylated alcohols before clouding occurs<br />

as the temperature is raised in dilute binary<br />

systems.<br />

The cloud point <strong>of</strong> a non-ionic surfactant<br />

natulrally depends on its structure, increasing


C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215 175<br />

for longer ethylene oxide and shorter<br />

hydrocarbon chains. Shifting the point <strong>of</strong><br />

attachment <strong>of</strong> the ethylene oxide chain from the<br />

end to the central portion <strong>of</strong> the hydrocarbon<br />

chain depresses the cloud point. The addition <strong>of</strong><br />

many common salts, e.g. sodium and potassium<br />

chlorides and sulfates, lowers the cloud point<br />

although the effects are much smaller than for<br />

ionic surfactants. However, some salts cause the<br />

cloud point to increase. The former effect is<br />

generally considered to stem from reduced<br />

hydration <strong>of</strong> the ethylene oxide chains resulting<br />

from competition with the added ions for the<br />

available water molecules. The latter effect<br />

occurs for ions such as I - , SCN - and most<br />

multivalent cations Which break the structure <strong>of</strong><br />

water. For some salts the anion and cation have<br />

opposite effects, with the stronger determining<br />

the direction <strong>of</strong> the cloud point shift. These<br />

effects have been recently discussed by Mackay<br />

[29].<br />

Other additives also influence the cloud point<br />

and other phase boundaries in non-ionic<br />

surfactant-water systems. Long-chain alcohols<br />

make the system less hydrophilic, as Fig. 5<br />

shows for the case <strong>of</strong> n-dodecanol added to<br />

mixtures <strong>of</strong> C 12E 5 and water [30].<br />

Fig. 5. Phase behavior resulting from the addition <strong>of</strong><br />

small amounts <strong>of</strong> n-dodecanol to I Wt'% C 12E 5 in<br />

water [30]; 30 denotes a three-phase region.<br />

Reprinted with permission from Academic Press.<br />

In this case, the cloud point is lowered by some<br />

23ºC when the alcohol content is only 10 wt.%<br />

relative to the surfactant. The temperatures for<br />

the other phase transitions are lowered as well.<br />

For the binary surfactant-water system, the<br />

phase rule constrains the three-phase regions,<br />

e.g. W + L 1 + Lα, to a single temperature, but<br />

the addition <strong>of</strong> the alcohol provides an<br />

additional degree <strong>of</strong> freedom and allows these<br />

regions to span a finite temperature range, as<br />

Fig. 5 indicates.<br />

The additive can, <strong>of</strong> course, be another<br />

surfactant. It is well known that the addition <strong>of</strong><br />

an ionic surfactant greatly increases the cloud<br />

point <strong>of</strong> an ethoxylated alcohol by adding an<br />

electrical repulsion between micelles and<br />

thereby inhibiting coacervation [31]. The<br />

opposite occurs when a second but more<br />

lipophilic non-ionic surfactant is added, e.g.<br />

C 12E 3 to C 12E 6, as shown in Fig. 6 [32]. This<br />

system is particularly interesting because, as<br />

noted previously, the Lα and L 3 phases do not<br />

occur in dilute mixtures <strong>of</strong> water and pure C 12E 6.<br />

However, both these phases appear when only a<br />

few per cent <strong>of</strong> the more lipophilic surfactant<br />

has been added, due to some rather complex<br />

Fig. 6. Phase behavior <strong>of</strong> mixtures <strong>of</strong> C 12E 6 and<br />

C1 2E 3 in water. The total surfactant concentration is<br />

fixed at 1.0 wt.% [32].


176 C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215<br />

phase behavior in the region indicated by the<br />

box. Although details are not given here, an<br />

interesting feature <strong>of</strong> this behavior is the<br />

existence <strong>of</strong> a four-phase region, which is<br />

constrained to a single temperature in this<br />

ternary system by the phase rule and which<br />

involves W, L 1, L 3 and La, phases [32]. Figure 6<br />

also shows a transition from W + L 3 to W + L 2 at<br />

58ºC for the C 12E 3-water system, which was<br />

clearly seen in this study using optical<br />

microscopy but does not appear in the existing<br />

phase diagram [23].<br />

The addition <strong>of</strong> a non-ionic surfactant to an<br />

anionic surfactant makes the surfactant mixture<br />

more lipophilic and causes a shift from L1 to<br />

Lα, to L 3, as shown in Fig. 4. This sequence is,<br />

not surprisingly, consistent with that shown in<br />

Fig. 6 for increase in the C12E3 content <strong>of</strong><br />

C 12E 6-C 12E 3 mixtures at constant temperature.<br />

The same sequence is found when a lipophilic<br />

alcohol is added to an anionic [27] or a<br />

zwitterionic [33] surfactant system.<br />

Although cationic surfactants are rarely used<br />

for cleaning purposes, they are useful for<br />

neutralizing charge build-up on fabric surfaces<br />

and for fabric s<strong>of</strong>tening. As an important trend<br />

in detergent formulation is combining all<br />

ingredients into a single mixture in order to<br />

eliminate the need for the separate addition <strong>of</strong><br />

bleaches, fabric s<strong>of</strong>teners, etc. during the<br />

washing process, it seems useful to include<br />

some information on phase behavior with<br />

cationic surfactants present. Moreover, some<br />

recent work suggests that mixtures <strong>of</strong> cationic<br />

and nonionic surfactants may be useful in<br />

removing oily soils, though not by solubilization<br />

<strong>mechanisms</strong> [34]. Because anionic and cationic<br />

surfactants attract one another, surfactant<br />

aggregation occurs readily with substantial<br />

neutralization <strong>of</strong> charge. When either the<br />

anionic or the cationic surfactant is present in<br />

substantial excess, the result is mixed micelles<br />

but with a much lower CMC than for the<br />

individual surfactants. When the two surfactants<br />

are present in almost equal amounts, the<br />

formation <strong>of</strong> a solid or liquid crystalline phase<br />

can be expected if the hydrocarbon chain<br />

lengths are sufficiently long. In some cases,<br />

vesicles have been found to form spontaneously<br />

upon mixing aqueous solutions with<br />

intermediate ratios <strong>of</strong> anionic and cationic<br />

surfactants [35].<br />

3.2. Phase behavior <strong>of</strong><br />

water-surfactant-hydrocarbon systems<br />

An important feature <strong>of</strong> the phase behavior <strong>of</strong><br />

systems containing water, surfactants, and<br />

hydrocarbon soils is the existence <strong>of</strong><br />

microemulsions, thermodynamically stable<br />

liquid phases containing substantial amounts <strong>of</strong><br />

both water and oil. The formation <strong>of</strong><br />

microemulsions requires that the<br />

surfactant-films which separate oil and water<br />

microdomains be rather flexible, and that the<br />

hydrophilic and lipophilic properties <strong>of</strong> the<br />

surfactant be roughly balanced. However, within<br />

conditions satisfying these overall constraints,<br />

the microstructure is quite sensitive to changes<br />

in the relative strength <strong>of</strong> hydrophilic and<br />

lipophilic interactions. In systems which contain<br />

comparable volumes <strong>of</strong> oil and water and which<br />

are dilute in surfactant but not so dilute as to<br />

preclude aggregation, packing considerations<br />

dictate that oil-in-water microemulsions coexist<br />

with excess oil for hydrophilic surfactants and<br />

water-in-oil microemulsions with excess water<br />

for lipophilic surfactants. Drop size' are <strong>of</strong> the<br />

order <strong>of</strong> 5-50 nm, increasing in size as the<br />

temperature, pressure, or system composition is<br />

changed to shift the surfactant closer to the<br />

condition <strong>of</strong> precise balance between<br />

hydrophilic and lipophilic properties. Very near<br />

this balance. the microemulsion becomes<br />

continuous in both phases and coexists with<br />

both excess water and excess oil. Interfacial<br />

tensions <strong>of</strong> this 'middlephase" microemulsion<br />

with both excess phases are frequently below<br />

0.0 1 mN m -1 - in some systems by an order <strong>of</strong><br />

magnitude or more.<br />

The condition for which the hydrophilic and<br />

lipophilic properties are exactly balanced and<br />

the surfactant films have no spontaneous<br />

tendency 10 curve in either direction has been<br />

called the phase inversion temperature (PIT) or<br />

hydrophile-lipophile balance (HLB) temperature<br />

by Shinoda and Friberg [36] for the case <strong>of</strong>


C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215 177<br />

non-ionic surfactants for which temperature is<br />

usually the variable <strong>of</strong> greatest interest. For<br />

ionic surfactants it is more common to speak <strong>of</strong><br />

"optimal" conditions, e.g. optimal salinity [37].<br />

Whatever one calls it, several criteria have been<br />

used to define the condition for balance in terms<br />

<strong>of</strong> readily measured experimental quantities.<br />

The most common criterion is equal volumetric<br />

solubilization in the microemulsion <strong>of</strong> the oil<br />

and water phases. The differences between the<br />

"optimal" conditions given by this and other<br />

criteria are small for practical purposes and will<br />

be ignored here.<br />

The effects <strong>of</strong> temperature and inorganic salts<br />

on making the surfactant more or less<br />

hydrophilic are basically the same as those<br />

described in the preceding section, and so are<br />

the effects <strong>of</strong> adding alcohols or additional<br />

surfactants, except that one additional factor<br />

must be considered - the relative solubilities <strong>of</strong><br />

the surfactants and additives in the oil phase. It<br />

is the composition <strong>of</strong> the surfactant films<br />

separating oil and water domains that<br />

determines the microstructure <strong>of</strong> the<br />

microemulsion. In a mixture <strong>of</strong> two non-ionic<br />

surfactants the more lipophilic surfactant has a<br />

higher solubility in the oil phase and the<br />

surfactant films are thus more hydrophilic than<br />

the overall surfactant mixture. The magnitude <strong>of</strong><br />

this effect for a given pair <strong>of</strong> surfactants<br />

depends on both the overall surfactant<br />

concentration and the water-to-oil ratio.<br />

Kunieda, Shinoda and co-workers have<br />

developed equations for predicting the<br />

dependence <strong>of</strong> the PIT on system composition<br />

for mixtures <strong>of</strong> two non-ionic surfactants [38]<br />

and for mixtures <strong>of</strong> an anionic and a non-ionic<br />

surfactant [39]. For instance, in the latter case<br />

the following relationship must be satisfied at<br />

the PIT<br />

W n = S sn + S onR ow [(1 - S sn)/(1 - S on)] (X -1) (2)<br />

where W n is the mass fraction <strong>of</strong> non-ionic<br />

surfactant in the overall mixture, S sn is the mass<br />

fraction <strong>of</strong> non-ionic surfactant in the surfactant<br />

films, S on is the mass fraction <strong>of</strong> non-ionic<br />

surfactant in the excess hydrocarbon phase, R ow<br />

is the mass fraction <strong>of</strong> oil in the oil-water<br />

mixture, and X is the total surfactant mass<br />

fraction in the system. The solubility <strong>of</strong> the<br />

anionic surfactant in the excess oil has been<br />

neglected.<br />

It is clear from this equation that a plot <strong>of</strong> Wn<br />

as a function <strong>of</strong> (X -1 - 1) at constant Row and<br />

temperature should yield a straight line from<br />

which values <strong>of</strong> Ssn and Son can be extracted.<br />

Figure 7 shows such plots for mixtures <strong>of</strong> C 12E 3<br />

and Neodol 23-3S at various temperatures along<br />

with the corresponding values <strong>of</strong> Ssn and Son.<br />

The oil phase is n-hexadecane and the aqueous<br />

phase is water containing 1 wt.% NaCl. As<br />

might be expected, nonionic surfactant<br />

solubility in the oil phase S on increases with<br />

increasing temperature. In contrast, the fraction<br />

S sn <strong>of</strong> nonionic surfactant in the films decreases.<br />

Since increasing temperature makes the nonionic<br />

surfactant less hydrophilic, it is reasonable<br />

that less <strong>of</strong> it would be required to achieve the<br />

Fig. 7. PIT results for the C12E3-Neodol 23-3S-1<br />

wt.% NaCl brine-n-hexadecane system [24]; X is the<br />

total surfactant mass fraction in the system.<br />

Reprinted with permission from Dr. Dietrich<br />

Steinkopff Verlag.


178 C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215<br />

balance between hydrophilic and lipophilic<br />

properties at high temperatures.<br />

The PIT is also influenced by the composition<br />

<strong>of</strong> the oil phase, being higher for hydrocarbons<br />

with longer chains. The reason is that their<br />

penetration into the hydrocarbon chain region <strong>of</strong><br />

the surfactant films tends to make the films<br />

curve toward a water-in-oil configuration. Such<br />

penetration is less for longer-chain<br />

hydrocarbons [40], probably due primarily to an<br />

entropic effect [41] except for short-chain<br />

hydrocarbons where energy effects have<br />

recently been shown to be important as well<br />

[42]. Reed and Puerto [43] have developed a<br />

scheme relating optimal conditions to the molar<br />

volume <strong>of</strong> the oil and the solubilization at<br />

optimal conditions.<br />

The other factor mentioned above as being<br />

necessary for microemulsion formation is the<br />

existence <strong>of</strong> flexible films. Films are most rigid<br />

for surfactants having long, straight<br />

hydrocarbon chains. Flexibility can be increased<br />

by promoting less ordered packing in the<br />

hydrocarbon chain region <strong>of</strong> the films, e.g. by<br />

using branched-chain surfactants or mixtures <strong>of</strong><br />

surfactants with different chain lengths, or by<br />

adding short-chain alcohols. Increasing the<br />

temperature also promotes flexibility. Too much<br />

flexibility can be undesirable, however, because<br />

it reduces the solubilization capacity <strong>of</strong> a<br />

microemulsion. However, when the surfactant<br />

films become too rigid, the lamellar liquid<br />

crystalline phase forms. Barakat et al.<br />

determined the conditions in several systems for<br />

which the lamellar phase formed instead <strong>of</strong> a<br />

middle-phase microemulsion because <strong>of</strong> film<br />

rigidity [44]. Hackett and Miller investigated the<br />

detailed phase behavior near the transition [45].<br />

The liquid crystal can also form when the<br />

amount <strong>of</strong> oil or water present becomes too low<br />

[46] or when the surfactant concentration <strong>of</strong> a<br />

middle-phase microemulsion becomes too high<br />

[47].<br />

Kunieda and Shinoda [47] have presented<br />

ternary diagrams at several temperatures ranging<br />

from below to above the PIT for the<br />

C12E5-water-n-tetradecane system. We discuss<br />

below the use <strong>of</strong> such diagrams in interpreting<br />

the dynamic behavior which occurs when<br />

surfactantwater mixtures contact oil.<br />

3.3. Water-non-ionic surfactant- triglyceride<br />

systems<br />

Extensive studies have been made <strong>of</strong><br />

microemulsions in hydrocarbon systems. Much<br />

less information is available for oils which are<br />

liquid triglycerides. An important difference is<br />

that triglycerides such as triolein which are <strong>of</strong><br />

interest for <strong>detergency</strong> are <strong>of</strong> much higher<br />

molecular weight than simple straight-chain<br />

liquid hydrocarbons such as n-hexadecane. The<br />

higher molecular weight makes it much more<br />

difficult to incorporate such triglycerides into<br />

surfactant films, the result being a major<br />

reduction in solubilization in many systems.<br />

Figure 8 shows the general form <strong>of</strong> a<br />

water-nonionic surfactant-liquid triglyceride<br />

ternary diagram [48-50]. Deviations from this<br />

behavior occur at sufficiently high temperatures,<br />

a matter discussed further below. As indicated<br />

on the diagram, the phase designated D' is the<br />

same as that usually called L3 in binary<br />

surfactant-water systems, mentioned in section<br />

3.1. While, like the middle-phase<br />

microemulsions discussed above, the D' phase<br />

coexists with both excess water and excess oil<br />

for suitable overall system compositions, it<br />

differs from the microemulsions in that it can<br />

Fig. 8. Schematic phase diagram for non-ionic<br />

surfactant-water-triolein system at low temperatures<br />

[48]. O denotes a triolein-rich phase.


C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215 179<br />

solubilize only small amounts <strong>of</strong> the oil phase.<br />

Note that solubilization <strong>of</strong> triglyceride in the<br />

lamellar liquid crystalline phase is also low.<br />

Behavior <strong>of</strong> the W + D' + O three-phase<br />

triangle has been studied as a function <strong>of</strong><br />

temperature for pure non-ionic surfactants and<br />

triolein. Figure 9 shows the results for C12E3<br />

[48], a lipophilic surfactant that is already above<br />

its cloud point at 0ºC, well below the<br />

experimental temperature range. The surfactant<br />

content <strong>of</strong> the D' phase increases rapidly with<br />

temperature, the same as for the L3 phase in<br />

binary surfactant-water systems, but<br />

solubilization <strong>of</strong> triolein remains low. The<br />

solubility <strong>of</strong> surfactant in the triolein phase is<br />

substantial - more than 10% by volume even at<br />

the lowest temperature studied (30.5ºC) - and<br />

increases with temperature. For comparison, we<br />

note that the solubility <strong>of</strong> C12E3 in excess<br />

n-hexadecane in equilibrium with<br />

microemulsions at 30ºC is about 3% by volume<br />

(see Fig. 7).<br />

At about 40ºC the rate <strong>of</strong> increase with<br />

temperature <strong>of</strong> surfactant solubility in the<br />

triolein phase increases significantly. Simultaneously,<br />

water solubilization in this phase,<br />

previously rather low, rises dramatically. Such<br />

formation <strong>of</strong> water-in-oil microemulsions is one<br />

way the system can depart at high temperatures<br />

from the behavior shown in Fig. 8.<br />

Another way is illustrated in Fig. 10 by the<br />

Fig. 9. The W + D' + 0 region at several temperatures<br />

in the C12E3-water-triolein system [48].<br />

Fig. 10. The W + D' + O and W + D + O regions at<br />

several temperatures in the C 12E 5-water-triolein<br />

system [48].<br />

corresponding C 12E 5 diagram [48]. Just above<br />

64ºC a phase transformation occurs, and at<br />

higher temperatures the diagram shows a new W<br />

+ D + O three-phase region instead <strong>of</strong> the W +<br />

D' + O region. The D phase is able to solubilize<br />

considerable triolein and is thus more favorable<br />

for <strong>detergency</strong> than D' if it forms as an<br />

intermediate phase during washing. According<br />

to Fig. 10, the composition <strong>of</strong> the D phase shifts<br />

to become richer in oil with increasing<br />

temperature in a manner similar to that seen for<br />

microemulsion systems [47] although the<br />

surfactant concentration <strong>of</strong> about 40% is well<br />

above that observed in typical microemulsions.<br />

Details <strong>of</strong> the phase behavior in the temperature<br />

region where the D phase first appears,<br />

including the existence <strong>of</strong> two four-phase<br />

regions at two closely spaced temperatures, are<br />

given for another system by Kunieda and<br />

Haishima [50].<br />

As indicated above, the inability <strong>of</strong> the<br />

hydrocarbon chain region <strong>of</strong> the surfactant films<br />

to incorporate the large triglyceride molecules is<br />

the chief reason for the poor solubilization.<br />

Similar poor solubilization and phase behavior<br />

have been seen in systems containing the<br />

anionic surfactant Aerosol OT, hydrocarbons<br />

with chain lengths <strong>of</strong> twelve and above, and<br />

NaCl brine [26]. Recently, Binks [51] has<br />

investigated further the phase behavior <strong>of</strong> some<br />

<strong>of</strong> these systems.<br />

If the surfactant films were made more flexi-


180 C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215<br />

ble, i.e. if chain packing in this region were<br />

made more disordered, one might expect<br />

solubilization to increase. Clearly in some<br />

systems such as water-C12E5-triolein (Fig. 10),<br />

increasing temperature provides sufficient<br />

disorder for the D phase to form. In other<br />

systems such as water-C12E3-triolein (Fig. 9), the<br />

D phase does not form at any temperature. One<br />

might expect that adding amphiphilic<br />

compounds with chain lengths different from the<br />

surfactant or with branched chains would<br />

promote less ordered packing and formation <strong>of</strong><br />

the D phase. As Fig. 11 shows, the addition <strong>of</strong><br />

tert-amyl alcohol (TAA) does, in fact, favor<br />

formation <strong>of</strong> the D instead <strong>of</strong> the D' phase in the<br />

water-C12E4-triolein system [52]. In the absence<br />

<strong>of</strong> TAA the D phase is seen only over a range <strong>of</strong><br />

about 0.2ºC near 55ºC, and not at all for the<br />

somewhat lower temperatures <strong>of</strong> Fig. 12 [48].<br />

For the purposes <strong>of</strong> improving <strong>detergency</strong>, the<br />

use <strong>of</strong> TAA to promote solubilization <strong>of</strong><br />

long-chain liquid triglycerides at relatively low<br />

temperatures is not attractive because its high<br />

solubility in water requires that it be used at<br />

Fig. 11. Partial phase diagram <strong>of</strong> the C 12E 4-tertiary<br />

amyl alcohol-water-triolein system: surfactant<br />

content, 16 wt.%; equal volumes <strong>of</strong> water and triolein<br />

[52]; LC denotes the lamellar or Lα phase.<br />

Fig. 12. Partial phase diagram <strong>of</strong> the C 12E 4-watertriolein-n-hexadecane<br />

system [48]. W. and 0. denote<br />

water-continuous and oil-continuous microemulsions.<br />

rather high concentrations. The same<br />

disadvantage applies to the use <strong>of</strong> hydrotopes, a<br />

possibility considered by Friberg and Rydhag<br />

[53]. Alander and Warnheim [54] managed to<br />

solubilize a mixture <strong>of</strong> medium-chain<br />

triglycerides (C 8-C 10) in aqueous solutions <strong>of</strong> a 1<br />

:2 mixture <strong>of</strong> sodium oleate and n-pentanol at<br />

25ºC, but they had less success with long-chain<br />

triglycerides.<br />

Recent results suggest that the use <strong>of</strong><br />

doublechain surfactants with varying chain<br />

lengths, e.g. secondary alcohol ethoxylate<br />

surfactants, instead <strong>of</strong> straight-chain surfactants<br />

may prove effective for forming the D phase<br />

and thereby solubilizing reasonable amounts <strong>of</strong><br />

triolein at temperatures suitable for warm water<br />

washing [55]. Here, too, the basic idea is that<br />

solubilization should be improved for surfactant<br />

films with disordered packing in the<br />

hydrocarbon chain region.<br />

3.4. Mixtures <strong>of</strong> hydrocarbons and triglycerides<br />

Microemulsion formation is easier in mixed<br />

soils <strong>of</strong> hydrocarbon and triglyceride than for


C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215 181<br />

pure triglyceride soils. In the mixed soil<br />

systems, hydrocarbon molecules presumably<br />

penetrate the surfactant films, allowing oil<br />

droplets or oil microdomains <strong>of</strong> other shapes to<br />

form. Triglyceride and hydrocarbon are jointly<br />

solubilized within the microdomains. Figure 12<br />

shows phase behavior in the water-n-dodecyl<br />

tetraoxyethylene monoether (C12H4)-triolein-n hexadecane system for a particular overall<br />

surfactant concentration and equal volumes <strong>of</strong><br />

oil and water phases [48]. The transition from<br />

the existence <strong>of</strong> a middle-phase microemulsion<br />

(D phase) in hydrocarbon-rich systems the D'<br />

phase in triolein-rich systems is clear. Note that<br />

the sequence <strong>of</strong> phases seen with increasing<br />

temperature over one temperature range for pure<br />

triolein is, omitting the excess oil phase, Lα, L2 + D'. D', W + D'. The same sequence occurs in<br />

,he oil-free system at modest surfactant<br />

concentrations. That is, the sequence <strong>of</strong> phases<br />

with triolein present is the same as that when it<br />

is absent, except that an excess oil phase is<br />

present and the transition temperatures are<br />

somewhat lower. This behavior is to be<br />

expected when solubilization is low.<br />

At intermediate oil compositions, D phase<br />

formation can be promoted by increasing either<br />

the temperature or the hydrocarbon content <strong>of</strong><br />

the oil. By making the surfactant less<br />

hydrophilic, increasing the temperature<br />

promotes a surfactant film configuration where<br />

the hydrocarbon chains diverge, and thereby<br />

facilitates the solubilization <strong>of</strong> triolein. The<br />

transition between the W + D' + O and W + D +<br />

0 regions occurs by means <strong>of</strong> a narrow<br />

four-phase region W + D'+ D + O. Table I gives<br />

compositions <strong>of</strong> the four-phase region at one<br />

temperature. Clearly, solubilization <strong>of</strong> both<br />

hydrocarbon and triglyceride is greater in the D<br />

phase. Moreover, hydrocarbon is solubilized in<br />

preference to triolein in both phases. Similar<br />

phase behavior has been reported for another<br />

triglyceride [49].<br />

Figure 13 shows interfacial tensions measured<br />

with the spinning drop apparatus at 30ºC for 1<br />

wt.% C12E4 with various mixtures <strong>of</strong><br />

n-hexadecane and triolein [48]. For the pure<br />

hydrocarbon the tensions drops after 10 minutes<br />

Table 1<br />

Compositions in volume fractions <strong>of</strong> four coexisting<br />

phases <strong>of</strong> the C12E4-water-triolein-n-hexadecane<br />

system at 39.2ºC [48]<br />

to about 0.005 mN m -1 , which is a reasonable<br />

value for a system near its PIT. In contrast, the<br />

tension reached after some 3 h is about 0.2 mN<br />

m -1 for pure triolein. The higher tension is<br />

expected in view <strong>of</strong> the low solubilization <strong>of</strong><br />

triolein. Mixed oils have intermediate values <strong>of</strong><br />

interfacial tension.<br />

Fig. 13. Interfacial tensions (IFT) at 30ºC for 1 Wt'%<br />

C 12E 4 with various mixtures <strong>of</strong> triolein and<br />

n-hexadecane [48].


182 C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215<br />

3.5. Water-surfactant-polar soil systems<br />

Hydrocarbons and triglycerides are frequently<br />

referred to as non-polar soils, while long-chain<br />

fatty acids and alcohols are termed polar soils.<br />

In this section we consider only the case <strong>of</strong> pure<br />

polar soils.<br />

Ekwall has studied the phase behavior <strong>of</strong><br />

many anionic surfactant-water-polar soil<br />

systems [56]. Figure 14 is a partial ternary<br />

diagram at 50ºC for the water-sodium<br />

octylsulfonate-n-hexanol system, which has<br />

been investigated in recent years by Kunieda<br />

and Nakamura [57]. A matter <strong>of</strong> interest for later<br />

sections <strong>of</strong> this paper is that the region <strong>of</strong><br />

coexistence <strong>of</strong> L1 and L2 phases near the<br />

water-hexanol axis is bounded by a three-phase<br />

triangle involving these phases and the lamellar<br />

liquid crystal La. In other cases, including the<br />

water-sodium octanoate-n-decanol system that<br />

was studied extensively by Ekwall's group,<br />

careful examination <strong>of</strong> the dilute region reveals<br />

that the D' (or L3) phase replaces Lα, in the<br />

three-phase triangle which terminates the L1-L2 region [58]. In this system a second three-phase<br />

triangle D'-Lα-L2 also exists, the arrangement<br />

Fig. 14. Phase behavior <strong>of</strong> the sodium<br />

octylsulfonate-water-nhexanol system in the dilute<br />

region at 50ºC [57]. Reprinted with permission <strong>of</strong> the<br />

American Chemical Society.<br />

being similar to that shown in Fig. 8 for<br />

triglyceride systems. In almost all diagrams <strong>of</strong><br />

this type involving relatively long-chain<br />

compounds, the lamellar phase and its<br />

associated multiphase regions are prominent<br />

although other liquid crystalline phases may be<br />

present as well at high surfactant concentrations<br />

[56].<br />

Kunieda and Nakamura [57] showed that the<br />

addition <strong>of</strong> NaCl to the system <strong>of</strong> Fig. 14 caused<br />

the D' phase to appear and the dilute portion <strong>of</strong><br />

the phase diagram to resemble Fig. 8. The<br />

higher salinity is likely to make the<br />

surfactant-alcohol bilayers more flexible and<br />

enables them to assume the locally saddleshaped<br />

configuration necessary for formation <strong>of</strong><br />

the sponge-like microstructure <strong>of</strong> the D' phase<br />

for slightly lipophilic conditions. Apparently<br />

only the lamellar structure is possible for more<br />

rigid bilayers.<br />

When the surfactant is non-ionic, the same<br />

behavior, i.e. the appearance <strong>of</strong> the D' phase and<br />

a shift from a diagram resembling Fig. 14 to one<br />

resembling Fig. 8, can be effected by increasing<br />

the temperature, as Kunieda and Miyajima [591<br />

showed for the water-n-dodecyl octaoxyethylene<br />

monoether (C12E8)-n-decanol system.<br />

Here. too. the ability to form the D' phase at<br />

temperatures above about 14ºC is probably the<br />

result <strong>of</strong> increased flexibility <strong>of</strong> the<br />

surfactant-alcohol bilayers.<br />

3.6. Mixtures <strong>of</strong> non-polar and polar soils<br />

As mentioned in section 3.2, the addition <strong>of</strong><br />

rather lipophilic amphiphilic compounds reduce,<br />

the PIT <strong>of</strong> non-ionic surfactant-water-hydrocarbon<br />

systems. Long-chain alcohols and<br />

(undissociated) fatty acids ate <strong>of</strong> this type, as<br />

Fig. 15 shows for the addition <strong>of</strong> oleyl alcohol<br />

to systems containing water, n-hexadecane, and<br />

several non-ionic -surfactants [12]. Note that 5%<br />

oleyl alcohol in the system by oil phase reduces<br />

the PIT in the C12E6 about 35ºC. Similar results<br />

were found by the same authors for oleic acid.<br />

When enough polar soil is present that the<br />

system is above its PIT but the surfactant is


C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215 183<br />

Fig. 15. PIT values for non-ionic surfactant-water-nhexadecane-oleyl<br />

alcohol systems [12]. Reprinted<br />

with permission <strong>of</strong> the American Oil Chemists'<br />

Society.<br />

below its cloud point, one might expect that<br />

phase behavior in the dilute region would be<br />

similar to that described in the preceding<br />

section, i.e. the L 1-L 2 region would terminate in<br />

a three-phase region involving either the D' or<br />

the La phase. Experiments at 30ºC with C 12E 7<br />

and oils having ratios <strong>of</strong> n-hexadecane to oleyl<br />

alcohol <strong>of</strong> 3/1 and 1/1, respectively, showed that<br />

such behavior did, in fact, occur, with the third<br />

phase being the lamellar liquid crystal Lα [60].<br />

However, in contrast to the situations shown in<br />

Fig. 8 and 14 where the oil phase solubilizes<br />

modest amounts <strong>of</strong> water, L 2 phases in these<br />

systems extended to compositions containing up<br />

to 75-80% water which were in equilibrium with<br />

aqueous micellar solutions. Evidently, the<br />

presence <strong>of</strong> hydrocarbon and <strong>of</strong> the double bond<br />

in the alcohol chain makes the surfactant films<br />

sufficiently flexible so that the L 2 phase can<br />

invert continuously and become water<br />

continuous, ultimately reaching compositions<br />

comparable to those <strong>of</strong> the D' phase in systems<br />

such as that shown in Fig. 8. The relevance <strong>of</strong><br />

this behavior to <strong>detergency</strong> is discussed later.<br />

When the long-chain alcohol is mixed with a<br />

liquid triglyceride instead <strong>of</strong> with a<br />

hydrocarbon, multiphase regions containing the<br />

D' phase are prominent, as is shown in Fig. 16<br />

[61] for the C 12E 6-water-triolein-oleyl alcohol<br />

system. The sequence <strong>of</strong> phases observed with<br />

increasing temperature in Fig. 16 for oils having<br />

oleyl alcohol contents exceeding about 20% is<br />

the same as was found for the water-non-ionic<br />

surfactant-triolein systems discussed in section<br />

3.3, as may be seen, for instance, along the<br />

right-hand boundary <strong>of</strong> Fig. 12 for water-C 12E 4-<br />

Fig. 16. Partial phase diagram <strong>of</strong> C12E6-water-triolein-oleyl alcohol system with 10 wt.% surfactant, 45 wt.%<br />

water, and 45 wt.% mixed oil [61]. The symbol IV denotes the four-phase region W + D'+ D + O.


184 C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215<br />

triolein. The D phase is seen, however, for<br />

surfactant concentrations well above that <strong>of</strong> Fig.<br />

16. Note that the amounts <strong>of</strong> oleyl alcohol<br />

needed to depress the temperatures <strong>of</strong> the<br />

various phase boundaries are much greater than<br />

are shown in Fig. 15 for hydrocarbon systems.<br />

A likely explanation is that much <strong>of</strong> the oleyl<br />

alcohol is dissolved in the bulk triglyceride<br />

phase, leaving relatively little alcohol in the<br />

surfactant films which, to a large extent,<br />

determine the basic phase behavior by<br />

controlling the aggregate shape.<br />

4. Diffusion path analysis<br />

Being <strong>of</strong> rather short duration and typically<br />

involving small quantities <strong>of</strong> soils, <strong>detergency</strong><br />

processes are strongly influenced by dynamic,<br />

diffusional phenomena which occur on a<br />

microscopic scale. Oily soil removal, in<br />

particular, depends on phase transitions which<br />

occur at the oil-washing solution interface. The<br />

preceding section described equilibrium phase<br />

behavior in both water-surfactant systems,<br />

representing the washing solution, and<br />

oil-water-surfactant systems, As demonstrated<br />

below, such equilibrium phase behavior can be<br />

combined with the theory <strong>of</strong> diffusion processes<br />

to interpret certain dynamic behavior such as<br />

intermediate phase formation and spontaneous<br />

<strong>emulsification</strong> that occurs during detergent<br />

processes.<br />

A mathematical technique called diffusion<br />

path analysis has been used with success in<br />

predicting certain dynamic phenomena in<br />

multicomponent solid and liquid systems when<br />

two phases not in equilibrium are brought into<br />

contact with one another [62-64]. Essentially, a<br />

time-invariant path <strong>of</strong> compositions can be<br />

plotted across an equilibrium phase diagram<br />

by-solving component transport equations with<br />

certain assumptions and boundary conditions.<br />

First, convection in the system from any source<br />

is assumed to be negligible. Second, the two<br />

phases are assumed to be semiinfinite in extent.<br />

This assumption simplifies the mathematical<br />

analysis and is probably reasonable at least for<br />

short times after contacting. Third, diffusion <strong>of</strong><br />

each species is assumed to be dependent only on<br />

its own concentration gradient with a uniform<br />

diffusion coefficient in each phase. Also, local<br />

equilibrium is assumed at all interfaces which<br />

form; this means that the compositions at the<br />

interfaces are defined by equilibrium tie lines.<br />

Diffusion path analysis is most conveniently<br />

applied to three-component, i.e. ternary,<br />

systems. In this situation, the phase diagram can<br />

be represented in the form <strong>of</strong> a two-dimensional<br />

triangle as, for instance, in Fig. 8, with<br />

two-phase regions shown as regions <strong>of</strong> varying<br />

shape containing equilibrium tie lines and<br />

three-phase regions represented as triangles in<br />

which the compositions <strong>of</strong> the equilibrium<br />

phases are shown as the vertices. The analysis in<br />

this case consists <strong>of</strong> solving in each phase the<br />

following transport equations for two <strong>of</strong> the<br />

species<br />

(∂wi/∂t) = Di(∂ 2 wi/∂x 2 ) i = 1,2 (3)<br />

where x is the distance from the initial surface<br />

<strong>of</strong> contact, t is time and wi and D i are the mass<br />

fraction and diffusivity <strong>of</strong> species i,<br />

respectively. The value <strong>of</strong> W3 for the third<br />

species is found by invoking Σw i = 1. The<br />

semi-infinite phase assumption allows transformation<br />

<strong>of</strong> the above equations to ordinary<br />

differential equations in the similarity variable<br />

η i = [x/(4D it) 1/2 ]. Integration yields the following<br />

error function solutions for the diffusion path<br />

segment in each phase<br />

w i = A i + B i erf η I i = 1,2 (4)<br />

where A i and B i are constants which are<br />

evaluated from the boundary conditions. Since<br />

η i varies from - ∞ to + ∞ at each value <strong>of</strong> time,<br />

the set <strong>of</strong> compositions given by Eq. (4) is<br />

independent <strong>of</strong> time although the position x <strong>of</strong> a<br />

specific composition does vary with time. It is<br />

<strong>of</strong>ten convenient to plot the compositions or<br />

"diffusion path" directly on the equilibrium<br />

phase diagram.<br />

In evaluating A i and B i for phases in contact.<br />

iteration on a tie line is performed until the<br />

individual species mass balances at the interface<br />

are satisfied. In addition to obtaining the path <strong>of</strong><br />

compositions that forms between the initial


C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215 185<br />

phases, one may also calculate the relative<br />

velocities <strong>of</strong> all interfaces, and therefore the<br />

growth rates <strong>of</strong> intermediate phases. More<br />

detailed descriptions <strong>of</strong> the mathematical<br />

analysis and the specific error function solutions<br />

can be found elsewhere for a ternary system<br />

forming a single interface [63] or two or more<br />

interfaces [65].<br />

The utility <strong>of</strong> diffusion path theory in ternary<br />

liquid systems was first shown for predicting the<br />

occurrence <strong>of</strong> spontaneous <strong>emulsification</strong> in<br />

alcohol-water-oil systems [63]. Specifically,<br />

when an alcohol-oil mixture denoted d in Fig.<br />

17 is brought into contact with water,<br />

spontaneous <strong>emulsification</strong> <strong>of</strong> oil drops in the<br />

water phase is observed. In this situation, the<br />

construction <strong>of</strong> a diffusion path between the<br />

initial compositions shows the formation <strong>of</strong> an<br />

interface with equilibrium compositions b and c<br />

connected by the tie line represented by the<br />

broken line. Spontaneous <strong>emulsification</strong> in the<br />

aqueous phase can be explained by the passage<br />

<strong>of</strong> that path segment from b to W through the<br />

corner <strong>of</strong> the two-phase envelope, thereby<br />

predicting the formation <strong>of</strong> small drops <strong>of</strong><br />

oil-alcohol mixture below the interface.<br />

Experiments showed that, in the absence <strong>of</strong><br />

interfacial turbulence, interfacial displacement<br />

Fig. 17. Schematic diffusion path in alcohol(A)water(W)-oil(O)<br />

system showing supersaturation<br />

leading to spontaneous <strong>emulsification</strong>.<br />

is proportional to the square root <strong>of</strong> time, as<br />

predicted by the theory [65,66]. This diffusion<br />

mechanism <strong>of</strong> spontaneous <strong>emulsification</strong> is<br />

distinct from other modes <strong>of</strong> spontaneous<br />

<strong>emulsification</strong> in which interfacial instability<br />

results in the mechanical dispersion <strong>of</strong> one<br />

phase in another [67].<br />

Diffusion path analysis was later applied to<br />

oil-water-surfactant systems [20,64,68]. In these<br />

cases, the use <strong>of</strong> pseudoternary phase diagrams<br />

was required. For example, commercial<br />

surfactants are almost always complex mixtures<br />

containing numerous species <strong>of</strong> surfactants.<br />

Rather than solving the diffusion equations for<br />

each species, one can sometimes combine all<br />

surfactant components together and treat them<br />

as a pseudocomponent. Mixtures <strong>of</strong><br />

hydrocarbons can also be considered as<br />

pseudocomponents. Although diffusion path<br />

studies are typically performed when<br />

single-phase systems are originally present, the<br />

ability to calculate diffusion paths in which one<br />

<strong>of</strong> the initial compositions is a stable dispersion<br />

<strong>of</strong> one phase in another, e.g. a liquid crystalline<br />

dispersion, has also been demonstrated [64].<br />

5. Dynamic contacting studies<br />

Direct observation <strong>of</strong> the dynamic phenomena<br />

that occur when non-equilibrated liquid phases<br />

are brought into contact can be made in various<br />

ways. On a macroscopic scale, a liquid can be<br />

gently placed on top <strong>of</strong> another liquid in a tube,<br />

and rather large-scale phenomena can be<br />

observed. This simple technique was used in the<br />

early studies <strong>of</strong> spontaneous <strong>emulsification</strong> in<br />

oil-water-alcohol systems [63] and has been<br />

used with surfactant systems to monitor the<br />

formation <strong>of</strong> microemulsion and liquid<br />

crystalline phases between oil and surfactant<br />

solutions [68-71]. In these cases, the oil is<br />

gently layered on top <strong>of</strong> the aqueous phase, and<br />

dynamic phenomena are observed without<br />

magnification. Crossed polarizers aid in the<br />

identification <strong>of</strong> birefringent liquid crystalline<br />

phases. A shortcoming <strong>of</strong> this technique is the<br />

inability to observe events which occur<br />

immediately after the contacting <strong>of</strong> the two


186 C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215<br />

phases. Also, the experimental scale increases<br />

the possibility <strong>of</strong> convection occurring due to<br />

the formation <strong>of</strong> adverse density gradients.<br />

Nevertheless, intriguing phenomena including<br />

oscillations in the interfacial position and abrupt<br />

changes in the interfacial velocity have been<br />

observed by Friberg and co-workers [70,71] in<br />

systems in which the lamellar liquid crystal is<br />

present, although generally at rather long times<br />

(a few days) after initial contact. For various<br />

reasons, diffusion path theory is not applicable<br />

for describing such phenomena, as these<br />

workers have pointed out.<br />

To facilitate the observation <strong>of</strong> dynamic<br />

phenomena at short times after contact, a<br />

microscopic technique was developed. A key<br />

aspect <strong>of</strong> the technique is the use <strong>of</strong> rectangular<br />

glass capillaries, having a path width 200 gm,<br />

tohold the sample [68]. Figure 18 shows a<br />

diagram <strong>of</strong> the sample cell that is 50 min in<br />

length. After the aqueous phase is imbibed<br />

about half way into the capillary, that end is<br />

Fig. 18. Rectangular glass capillary cell used in<br />

vertical-stage contacting experiments.<br />

sealed by a resin curable by ultraviolet light. The<br />

oil phase is then injected into the other end by<br />

use <strong>of</strong> a syringe. Initially, the resulting diffusion<br />

phenomena were observed using a conventional<br />

microscope with the cell in a horizontal<br />

configuration. However, due to the density<br />

difference between the two phases, overriding <strong>of</strong><br />

the oil over the surfactant solution occurred,<br />

causing distortion <strong>of</strong> the interface and<br />

complicating the interpretation <strong>of</strong> the results<br />

[68].<br />

A vertical-stage microscope was then<br />

designed that allowed the cell to be placed in a<br />

vertical configuration in a controlled<br />

temperature environment. Details <strong>of</strong> the<br />

microscope and contacting technique are given<br />

elsewhere [20]. In this configuration in a stable<br />

region <strong>of</strong> contact between the two initial phases<br />

can be maintained, allowing easy viewing. The<br />

vertical-configuration microscope was<br />

subsequently improved and equipped with a<br />

video imaging system [48,72]. The use <strong>of</strong> video<br />

taping and image analysis allows detailed<br />

review <strong>of</strong> phenomena which may be missed<br />

initially in real time, and improved<br />

determination <strong>of</strong>, for example, the velocities <strong>of</strong><br />

interfaces and rates <strong>of</strong> formation <strong>of</strong> intermediate<br />

phases.<br />

The vertical-contacting technique was first<br />

used to study the dynamic contacting in<br />

water-anionic surfactant-oil systems<br />

representative <strong>of</strong> those used in enhanced oil<br />

recovery processes [20]. Intermediate phase<br />

formation and spontaneous <strong>emulsification</strong><br />

experimentally observed in a brine-petroleum<br />

sulfonate-hydrocarbon system were found to be<br />

predictable from calculated diffusion paths<br />

based on relevant phase diagrams [64]. Widely<br />

varying but predictable phenomena were found<br />

as the salinity <strong>of</strong> the aqueous phase was varied.<br />

6. Diffusional phenomena in detergent systems<br />

6.1. Water - alcohol ethoxylate - hydrocarbon<br />

systems<br />

Studies <strong>of</strong> diffusional phenomena in systems<br />

having direct relevance to <strong>detergency</strong> processes


C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215 187<br />

have recently been performed. Experiments<br />

were designed to investigate the effects <strong>of</strong><br />

changes in temperature on the dynamic<br />

phenomena which occur when aqueous<br />

solutions <strong>of</strong> pure non-ionic surfactants contact<br />

hydrocarbons such as tetradecane and<br />

hexadecane [18,72]. These oils can be<br />

considered to be models <strong>of</strong> non-polar soils such<br />

as lubricating oils. The dynamic contacting<br />

phenomena, at least immediately after contact,<br />

are representative <strong>of</strong> those which occur when a<br />

detergent solution contacts an oily soil on a<br />

synthetic fabric surface. The following is a<br />

summary <strong>of</strong> the observed behavior interpreted<br />

through the use <strong>of</strong> schematic diffusion paths.<br />

Detailed phase behavior in such systems has<br />

been reported previously and was used in<br />

construction <strong>of</strong> the diffusion paths [47].<br />

With C12E5 as the non-ionic surfactant at a 1<br />

wt.% level in water, quite different phenomena<br />

were observed below, above, and well above the<br />

cloud point when tetradecane or hexadecane<br />

was carefully layered on top <strong>of</strong> the aqueous<br />

solution. Below the cloud point temperature <strong>of</strong><br />

31ºC, very slow solubilization <strong>of</strong> oil into the<br />

one-phase micellar solution was observed. An<br />

interesting phenomenon was observed at 20ºC<br />

involving a "volcanolike" instability which<br />

caused flow <strong>of</strong> the aqueous solution to the oil<br />

interface. This flow column, which is believed<br />

to have resulted from an adverse density<br />

gradient within the aqueous phase, is shown in<br />

Fig. 19 [72]. The upper tip <strong>of</strong> the column was<br />

observed to oscillate, probably due to gradients<br />

in interfacial tension along the oil-water<br />

interface (Marangoni flow). Of more importance<br />

to a <strong>detergency</strong> process, the schematic diffusion<br />

path shown in Fig. 20(a) explains why no<br />

intermediate phase formed between the water<br />

and oil. Also, due to the low solubility <strong>of</strong> oil in<br />

the dilute aqueous surfactant solution in this<br />

region <strong>of</strong> the ternary phase diagram, it predicts<br />

the quite slow solubilization <strong>of</strong> oil into the<br />

surfactant solution. At temperatures just below<br />

the cloud point temperature, an intermediate<br />

phase depleted in surfactant did form between<br />

the micellar solution and the oil. The schematic<br />

diffusion path in this case is shown in Fig.<br />

20(b). Once again, instabilities in the aqueous<br />

Fig. 19. "Volcano" instability in C 12E 5-water-ntetradecane<br />

system at 20ºC. The image is out <strong>of</strong> focus<br />

to allow observation <strong>of</strong> refractive index variations<br />

[72]. Reprinted with permission <strong>of</strong> Academic Press.<br />

Fig. 20. (a) Diffusion path well below the cloud point<br />

showing no intermediate phase formation; (b)<br />

diffusion path slightly below the cloud point showing<br />

the formation <strong>of</strong> intermediate phase W [72].<br />

Reprinted with permission <strong>of</strong> Academic Press.<br />

ous phase occurred, in this case due to a density<br />

difference between the original micellar solution<br />

and the intermediate phase, causing flow <strong>of</strong><br />

surfactant solution to the oil interface.<br />

Nevertheless, at all temperatures studied below<br />

the cloud point, only very slow solubilization <strong>of</strong><br />

oil into the surfactant solution was observed.<br />

At 35ºC, which is just above the cloud point, a<br />

much different behavior was observed. The<br />

surfactant-rich L1 phase separated to the top <strong>of</strong><br />

the aqueous phase prior to contacting by<br />

hexadecane.


188 C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215<br />

Upon addition <strong>of</strong> the oil, the drops <strong>of</strong> the L1 phase rapidly solubilized the hydrocarbon to<br />

form an oil-in-water microemulsion containing<br />

an appreciable quantity <strong>of</strong> hydrocarbon. After<br />

depletion <strong>of</strong> the larger surfactant-containing<br />

drops, a front developed as smaller drops were<br />

incorporated into the microemulsion phase. This<br />

behavior is shown schematically in Fig. 21.<br />

Unlike the experiments carried out below the<br />

cloud point temperature, an appreciable<br />

solubilization <strong>of</strong> oil was observed in the time<br />

frame <strong>of</strong> the study, as indicated by upward<br />

movement <strong>of</strong> the oil-microemulsion interface.<br />

Similar phenomena were observed with both<br />

tetradecane and hexadecane as the oil phases.<br />

When the temperature <strong>of</strong> the system was<br />

raised to just below the phase inversion<br />

temperatures <strong>of</strong> the hydrocarbons with C12E5 (45ºC for tetradecane and 50ºC for hexadecane),<br />

two intermediate phases formed when the initial<br />

dispersion <strong>of</strong> L1 drops in the water contacted<br />

the oil. One was the lamellar liquid crystalline<br />

phase Lα (probably containing some dispersed<br />

water). Above it was a middle-phase<br />

microemulsion. In contrast to the studies below<br />

the cloud point temperature, appreciable<br />

solubilization <strong>of</strong> hydrocarbon into the two<br />

intermediate phases, shown 4 min after<br />

contacting in Fig. 22, was observed. A diagram<br />

<strong>of</strong> the phenomena observed is shown in Fig. 23.<br />

A similar progression <strong>of</strong> phases was found at<br />

35ºC using n-decane as the hydrocarbon. At this<br />

temperature, which is near the phase inversion<br />

Fig. 21. Schematic diagram showing conversion <strong>of</strong><br />

the L 1 phase into an oil-in-water microemulsion at<br />

temperatures above the cloud point [72]. The symbol<br />

me denotes a microemulsion. Reprinted with<br />

permission <strong>of</strong> Academic Press.<br />

Fig. 22. Video frame showing intermediate phases 4<br />

min after contact in C 12E 5-water-n-tetradecane<br />

system at 45ºC near the PIT [72]. Reprinted with<br />

permission <strong>of</strong> Academic Press.<br />

Fig. 23. Schematic diagram showing the conversion<br />

<strong>of</strong> L 1 phase into a middle-phase microemulsion and a<br />

liquid crystal dispersion, for the experiment depicted<br />

in Fig. 22 [72]. Reprinted with permission <strong>of</strong><br />

Academic Press.<br />

temperature <strong>of</strong> the water-C12E5-decane system,<br />

the existence <strong>of</strong> a two-phase dispersion <strong>of</strong> Lα<br />

and water below the middle-phase<br />

microemulsion was clearly evident.<br />

To compare the rate <strong>of</strong> tetradecane<br />

solubilization at temperatures below and near<br />

the phase inversion temperature, verticalcontacting<br />

experiments were performed between<br />

the L1 phase and oil at 40 and 48ºC. In both<br />

cases, the surfactant-rich phase was the<br />

equilibrium phase that separated from a 10%<br />

aqueous solution at that temperature. At 40ºC a<br />

liquid crystalline phase and an oil-in-water


C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215 189<br />

micro emulsion formed between the L1 and oil<br />

phases. At 48 º C, a similar phase progression<br />

was observed, with a middle-phase microemulsion<br />

forming in place <strong>of</strong> the oil-in-water<br />

microemulsion. As shown in Fig. 24, plots <strong>of</strong><br />

the position <strong>of</strong> the oil-micro emulsion interface<br />

versus the square root <strong>of</strong> time were straight lines<br />

in both cases indicating diffusion-controlled<br />

mass transfer. In Fig. 24, an arbitrary constant<br />

has been added to each position for ease <strong>of</strong><br />

comparison, i.e. x = 0 is not the initial contact<br />

position. At any given time, the relative slopes<br />

<strong>of</strong> the two lines are indicative <strong>of</strong> the relative<br />

rates <strong>of</strong> oil solubilization. In this case, the rate <strong>of</strong><br />

solubilization near the PIT is 2.2 times greater<br />

than at 40ºC.<br />

Well above the phase inversion temperatures<br />

for both hexadecane and tetradecane, 1 wt.%<br />

C12E5 exists as a dispersion <strong>of</strong> liquid crystal Lα<br />

in water. Rapid movement <strong>of</strong> the liquid crystal<br />

to the oil occurred upon contacting, causing<br />

extensive spontaneous <strong>emulsification</strong> <strong>of</strong> water in<br />

the oil phase. Eventually a layer depleted in<br />

liquid crystal formed near the oil interface.<br />

Figure 25 shows the spontaneous <strong>emulsification</strong><br />

that occurred. Interpretation <strong>of</strong> the phenomena<br />

Fig. 24. Variation <strong>of</strong> oil- microemulsion interface<br />

with time at 40ºC and 48ºC following contact <strong>of</strong> the<br />

L 1 phase with oil for the C 12E 5-water-n-tetradecane<br />

system.<br />

Fig. 25. Video frame showing spontaneous<br />

<strong>emulsification</strong> observed 18 min after initial contact in<br />

the C 12E 5-water-n-hexadecane system at 60ºC [72].<br />

Reprinted with permission <strong>of</strong> Academic Press.<br />

by diffusion path analysis indicated that the oil<br />

was being converted into a waterin-oil<br />

microemulsion at this high temperature; this<br />

means that very little solubilization <strong>of</strong> oil into<br />

the aqueous phase was taking place. The<br />

spontaneous <strong>emulsification</strong> occurred due to the<br />

passage <strong>of</strong> the oil-phase diffusion path segment<br />

across the twophase water-micro-emulsion<br />

region, as shown in Fig. 26.<br />

Experiments similar to those described above<br />

were also performed using C12E4 as the<br />

surfactant [18]. This surfactant is more hydrophobic<br />

than C12E5 and therefore has a lower<br />

Fig. 26. Schematic diffusion path for the experiment<br />

depicted in Fig. 25. Point d represents the<br />

composition <strong>of</strong> the initial water-surfactant mixture;<br />

HC, S, and tie denote hydrocarbon, surfactant, and<br />

microemulsion. The last is oil-continuous in this<br />

case. Some multiphase regions are not shown [72].<br />

Reprinted with permission <strong>of</strong> Academic Press.


190 C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215<br />

cloud point (7ºC) and PIT with hexadecane<br />

(30ºC). Contacting experiments were performed<br />

below, at, and above the PIT. In contrast to<br />

C12E5, the structure <strong>of</strong> the 1 wt% C12E4 solution<br />

was a lamellar liquid crystalline dispersion in<br />

water at all the temperatures studied. With<br />

regard to the intermediate phases which formed<br />

at the different temperatures, the following<br />

results were obtained. Below the PIT, an<br />

oil-in-water microemulsion formed between the<br />

lamellar liquid crystalline dispersion and oil. At<br />

the PIT, the behavior was dependent upon the<br />

concentration <strong>of</strong> liquid crystalline material at the<br />

initial oil-surfactant solution interface. At the<br />

low concentration <strong>of</strong> liquid crystal present at a<br />

1% surfactant level, no intermediate phase<br />

formation was observed when the aqueous<br />

dispersion was contacted with hexadecane. To<br />

provide a higher level <strong>of</strong> liquid crystal at the<br />

initial point <strong>of</strong> contact, La drops were allowed to<br />

cream to the air-water interface prior to the<br />

contacting studies. These drops converted to<br />

concentrated La domains upon heating to 30ºC.<br />

With this initial aqueous phase structure, the<br />

formation <strong>of</strong> a middle-phase microemulsion<br />

layer was observed upon contacting with oil.<br />

Finally, if a pure lamellar liquid crystalline<br />

phase containing 25% surfactant was contacted<br />

with oil, swelling <strong>of</strong> the liquid crystalline phase<br />

was observed, as shown in Fig. 27.<br />

An interpretation <strong>of</strong> these results was made by<br />

comparing the qualitative diffusion paths,<br />

shown in Fig. 28, resulting from the different<br />

aqueous starting compositions [18]. The<br />

differences in behavior for the dilute (C),<br />

concentrated (B) and pure liquid crystal (A)<br />

cases were attributed to the relatively high<br />

solubility <strong>of</strong> C12E4 in the oil phase at this<br />

temperature, which influenced whether the<br />

diffusion path passed above or below the<br />

various three-phase regions that are present on<br />

the phase diagram.<br />

At temperatures <strong>of</strong> 40-50ºC, which are above<br />

the PIT, behavior similar to that for C12E5 was<br />

observed. Dissolution <strong>of</strong> the lamellar liquid<br />

crystalline phase into the oil resulted in the<br />

formation <strong>of</strong> a water-in-oil microemulsion and<br />

spontaneous <strong>emulsification</strong> <strong>of</strong> water in the oil<br />

phase.<br />

Fig. 27. Video frame showing swelling <strong>of</strong> the<br />

lamellar liquid crystalline phase 43 min after initial<br />

contact for the C 12E 4-water-n-hexadecane system at<br />

the PIT <strong>of</strong> 30ºC [18], Reprinted with permission <strong>of</strong><br />

Academic Press.<br />

Dynamic contacting studies were also<br />

performed with hydrophobic additives<br />

combined with C 12E 5 [30]. C 12E 3 and<br />

n-dodecanol were added to C 12E 5 in proportions<br />

to yield cloud points near that <strong>of</strong> C 12E 4<br />

(approximately 7ºC). This study was conducted<br />

Fig. 28. Schematic diffusion paths representing<br />

different beha- vior observed at different surfactant<br />

concentrations for the C 12E 4-water-n-hexadecane<br />

system at the PIT <strong>of</strong> 30ºC [18] The symbol (mp µe)<br />

denotes a middle-phase microemulsion. Reprinted<br />

with permission <strong>of</strong> Academic Press.


C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215 191<br />

ducted to determine whether formation <strong>of</strong><br />

intermediate microemulsion phases containing a<br />

high proportion <strong>of</strong> oil could be obtained at<br />

temperatures lower than those for C 12E 5 alone.<br />

However, despite exhibiting phase behavior in<br />

the absence <strong>of</strong> oil similar to that <strong>of</strong> C 12E 4, the<br />

C 12E 5-C 12E 3 mixture behaved in a way<br />

intermediate to the behavior seen with C 12E 4 and<br />

C 12E 5 upon being contacted with hexadecane.<br />

Also, the microemulsion phases formed at<br />

various temperatures when the C 12E 5-dodecanol<br />

system contacted oil were essentially unchanged<br />

from those seen in the C 12E 5 system without any<br />

additive present. For example, rather than<br />

forming a middle-phase microemulsion with<br />

hexadecane at 30ºC, the two systems formed<br />

oil-in-water micro-emulsions. The contacting<br />

temperature had to be increased to 40ºC in the<br />

case <strong>of</strong> the C 12E 5-C 12E 3 System and 50ºC in the<br />

case <strong>of</strong> the C 12E 5-dodecanol system before<br />

middlephase microemulsion formation was<br />

observed. These differences between the two<br />

systems were attributed to differences in<br />

partitioning <strong>of</strong> the additive and the more<br />

water-soluble C 12E 5 between the oil and the<br />

microemulsion phases. The observed differences<br />

between the two additive systems would be less<br />

if smaller quantities <strong>of</strong> oil relative to the<br />

surfactant solution had been present.<br />

6.2. Water-alcohol ethoxylate-triglyceride<br />

(+ hydrocarbon) systems<br />

Triolein is a pure triglyceride suitable for use<br />

as a model for kitchen soils such as vegetable<br />

oils. Dynamic contacting studies similar to those<br />

described above for hydrocarbons were<br />

performed with triolein using aqueous solutions<br />

containing the three alcohol ethoxylates C 12E 3,<br />

C 12E 4 and C 12E 5 [48]. As described in the phase<br />

behavior section, ternary triolein - water-non-<br />

ionic surfactant systems exhibit different phase<br />

behavior than those containing straight-chain<br />

hydrocarbons. Specifically, the large size <strong>of</strong> the<br />

triolein molecules inhibits solubilization and<br />

formation <strong>of</strong> microemulsion phases.<br />

At low temperatures, schematic phase<br />

behavior like that shown in Fig. 8 is observed in<br />

which two three-phase regions are present in the<br />

ternary diagram. At these temperatures, the<br />

surfactantwater mixture is a dispersion <strong>of</strong> the<br />

liquid crystal La in water. When this dispersion<br />

is contacted with triolein, a water layer forms<br />

between the liquid crystal and the oil, and<br />

extensive spontaneous <strong>emulsification</strong> occurs in<br />

the oil phase [48]. This behavior can be<br />

explained in terms <strong>of</strong> a diffusion path which<br />

passes below the bottom three-phase region in<br />

Fig. 8. Spontaneous <strong>emulsification</strong> occurs in the<br />

oil phase due to passage <strong>of</strong> that diffusion path<br />

segment across the two-phase water-oil region.<br />

In general, insufficient surfactant is available at<br />

the interface to form an intermediate L 3 or D'<br />

phase due to the high solubility <strong>of</strong> non-ionic<br />

surfactants in triolein at these conditions. At still<br />

higher temperatures where C 12E 3 and C 12E 4 are<br />

in the form <strong>of</strong> aqueous dispersions <strong>of</strong> L 3 (greater<br />

than 30ºC for C 12E 3 and 55ºC for C 12E 4), similar<br />

behavior occurs except that even more vigorous<br />

spontaneous <strong>emulsification</strong> is observed.<br />

C 12E 5 exhibited behavior at approximately<br />

65ºC quite comparable to that observed with<br />

hydrocarbon systems near the PIT. In fact, as<br />

shown in Fig. 10, the D phase in that system<br />

contains almost equal volumes <strong>of</strong> triolein and<br />

water at 65ºC. When a concentrated lamellar<br />

liquid crystalline dispersion contacted triolein at<br />

that temperature, the D phase formed, first as<br />

lenses along the water-oil interface, then as a<br />

continuous layer. A schematic diffusion path<br />

corresponding to this behavior is shown in Fig.<br />

29.<br />

The dynamic contacting <strong>of</strong> C 12E 4. solutions<br />

with oil mixtures containing varying proportions<br />

<strong>of</strong> triolein and hexadecane was also studied<br />

[48]. For 3/1 hexadecane-triolein mixtures,<br />

behavior comparable to that found with the pure<br />

hydrocarbon system was obtained. Similarly,<br />

3/1 triolein-hexadecane systems behaved in the<br />

contacting experiments like the pure triglyceride<br />

system. However, contacting <strong>of</strong> the concentrated<br />

liquid crystalline dispersion at 38ºC with a 1/1<br />

mixture <strong>of</strong> triolein and hexadecane resulted in<br />

the formation <strong>of</strong> transient middle-phase<br />

microemulsion droplets. The failure to form a


192 C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215<br />

Fig. 29. Diffusion path corresponding to observed<br />

behavior for the C 12E 5-water-triolein system at<br />

64.5ºC [48].<br />

complete microemulsion layer during the course<br />

<strong>of</strong> the experiment probably resulted from the<br />

high solubility <strong>of</strong> the surfactant in the oil<br />

mixture.<br />

The latter experiment was repeated with<br />

tertiary amyl alcohol (TAA) added to the<br />

concentrated La dispersion at three different<br />

TAA/C12E4 ratios: 0.05, 0.10, and 0.20 by<br />

weight [52]. All three experiments showed the<br />

rapid formation <strong>of</strong> an intermediate phase,<br />

presumably the D or microemulsion phase. The<br />

higher the ratio <strong>of</strong> TAA to surfactant, the faster<br />

was the formation <strong>of</strong> the intermediate phase as a<br />

continuous layer. As indicated in the previous<br />

paragraph, the formation <strong>of</strong> a continuous<br />

middle-phase microemulsion layer did not occur<br />

in the absence <strong>of</strong> TAA.<br />

7. Oil drop-contacting experiments<br />

As discussed in the preceding section,<br />

contacting experiments using vertically oriented<br />

cells, and their interpretation using diffusion<br />

path theory, have provided a fundamental<br />

understanding <strong>of</strong> the conditions for the<br />

occurrence <strong>of</strong> intermediate phase formation and<br />

spontaneous <strong>emulsification</strong> for a variety <strong>of</strong><br />

systems containing water, pure surfactants, and<br />

non-polar oils. However, when either surfactant<br />

mixtures, e.g. the C12E5-additive systems<br />

discussed above, or commercial surfactants that<br />

are themselves complex mixtures, are used, the<br />

conditions for which intermediate phases form<br />

during a vertical cell-contacting experiment are<br />

frequently rather different from those expected<br />

during washing experiments in the same system<br />

at the same temperature. The reason is<br />

differential partitioning <strong>of</strong> different surfactant<br />

species into the oil phase. As explained<br />

previously in section 3, the extent <strong>of</strong> differential<br />

partitioning depends on both the overall<br />

surfactant concentration in the aqueous phase<br />

and the water-to-oil ratio (see Eq. (2)). In<br />

particular, the surfactant remaining after some<br />

partitioning into the oil has occurred, the factor<br />

which largely controls whether an intermediate<br />

phase will form, is less hydrophilic for a<br />

washing experiment in which the water-to-oil<br />

ratio is very large than for a contacting<br />

experiment in which the volumes <strong>of</strong> oil and<br />

water are comparable. It should be noted that<br />

this problem does not arise for systems<br />

containing mixtures <strong>of</strong> anionic surfactants and<br />

hydrocarbon oils, because none <strong>of</strong> the individual<br />

surfactant species has appreciable solubility in<br />

the hydrocarbon phase.<br />

When the oil consists <strong>of</strong> one or more polar<br />

components or <strong>of</strong> both polar and non-polar<br />

components, there is another limitation <strong>of</strong> the<br />

vertical cell-contacting technique that is<br />

significant even when pure surfactants are used.<br />

Basically, both experiment and diffusion-path<br />

theory yield information on the behavior <strong>of</strong> the<br />

system during the early stages <strong>of</strong> contact when<br />

the oil phase remains at its initial composition<br />

except in a region near the surface <strong>of</strong> contact.<br />

However, when the volume <strong>of</strong> the oil phase is<br />

small, the diffusion <strong>of</strong> polar material out <strong>of</strong><br />

and/or diffusion <strong>of</strong> surfactant into the oil can<br />

cause the composition <strong>of</strong> the entire oil phase to<br />

vary with time, i.e. no location exists where the<br />

oil has its initial composition. This effect is<br />

especially important when inverse micelles or<br />

other aggregates form in the oil that have mixed<br />

films <strong>of</strong> surfactant and polar compounds. As a<br />

result, situations occur in which an intermediate<br />

phase does not form on initial contact but de-


C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215 193<br />

Fig. 30. Schematic illustration <strong>of</strong> contacting<br />

experiment in which a small oil drop is injected into<br />

an aqueous surfactant solution.<br />

velops later in the experiment when the oil<br />

composition becomes suitable. Examples <strong>of</strong><br />

such behavior are discussed below.<br />

An oil drop-contacting technique was<br />

developed in which the water-to-oil ratio is<br />

large, as in practical washing situations. As<br />

shown in Fig. 30, a drop <strong>of</strong> oil, usually some<br />

10-100 mm in diameter, is injected into a<br />

horizontal rectangular glass cell by means <strong>of</strong> a<br />

very thin hypodermic needle. The cell, which is<br />

400 Rm thick, is inside a thermal stage modified<br />

to enable the drop to be observed by<br />

videomicroscopy from the moment <strong>of</strong> injection<br />

[24,60]. Since the drop must be viewed through<br />

the surfactant solution in which it is immersed,<br />

this technique works best when the surfactant<br />

solution is below its cloud point temperature.<br />

However, some experiments have been<br />

successfully carried out in which the initial<br />

surfactant solution was a dispersion <strong>of</strong> the<br />

lamellar liquid crystal in water, as discussed<br />

below. It is noteworthy that no similar limitation<br />

exists for the vertical cell technique, and indeed<br />

almost all <strong>of</strong> the experiments described above<br />

for non-polar oils were conducted above the<br />

cloud point temperature.<br />

7. 1. Experiments with surfactant mixtures and<br />

nonPolar oils<br />

Mixtures <strong>of</strong> anionic and non-ionic surfactants<br />

are now almost universally used in liquid<br />

detergents for laundry applications since they<br />

are more effective than anionics alone for<br />

washing synthetic fabrics at low temperatures.<br />

The oil dropcontacting technique was used to<br />

determine whether an intermediate<br />

microemulsion phase would form near the PIT<br />

for these mixed surfactant systems with<br />

hydrocarbon soils [24] in a manner similar to<br />

that described above for pure non-ionic<br />

surfactants with the vertical cell technique.<br />

The pure non-ionic surfactant C12E3 and the<br />

commercial anionic surfactant Neodol 23-3S<br />

were used in this study, i.e. the same<br />

combination as discussed above in the phase<br />

behavior section. The use <strong>of</strong> a commercial<br />

mixture rather than a pure anionic surfactant had<br />

minimal effect on the differential partitioning<br />

since all the individual anionic species in the<br />

mixture had very low solubilities in<br />

n-hexadecane, the hydrocarbon used. Because<br />

the volume <strong>of</strong> the oil drop injected was small, it<br />

dissolved little non-ionic surfactant, and the<br />

relevant PIT was that for which the surfactant<br />

composition Ssn in the films within the<br />

microemulsion phase was the same as the<br />

overall surfactant composition in the system.<br />

Data on Ssn for this system when the aqueous<br />

phase contains 1 wt.% NaCl are given in Fig. 7.<br />

As may be seen from Fig. 4, the initial washing<br />

bath, i.e. the oilfree mixture <strong>of</strong> the surfactant<br />

and a 1 wt.% NaCl solution, forms a dispersion<br />

<strong>of</strong> the lamellar liquid crystal in brine for<br />

surfactant compositions equal to the relevant<br />

values <strong>of</strong> Ssn. Contacting experiments were conducted for a<br />

surfactant mixture containing 78 wt.% <strong>of</strong> the<br />

nonionic surfactant [24]. According to Fig. 4,<br />

the relevant PIT is 30ºC. At 25ºC, no<br />

intermediate phase was observed and the drop<br />

diameter did not decrease appreciably with time.<br />

Thus, solubilization <strong>of</strong> hydrocarbon by the<br />

liquid crystalline phase was very slow at this<br />

temperature below the PIT. At 30ºC, an<br />

intermediate microemulsion phase was<br />

observed. Its volume continued to increase until<br />

the oil phase disappeared. At 40ºC, the drop<br />

diameter increased with time, the expected<br />

behavior above the PIT as the oil phase takes up<br />

surfactant and water. The liquid crystalline<br />

particles surrounding the drop made it difficult<br />

to discern whether spontaneous <strong>emulsification</strong>


194 C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215<br />

in the oil occurred under these conditions, as<br />

would be expected based on observations above<br />

the PIT described above for the vertical cell<br />

experiments. The conclusion reached from these<br />

experiments is that mixed surfactant systems<br />

behave in basically the same manner as pure<br />

surfactant systems for hydrocarbon oils,<br />

provided that the PIT used to interpret the<br />

behavior is as defined above.<br />

Valuable results were also obtained by the oil<br />

drop technique for the C12E4-water-triolein-TAA system. As discussed in Section 3, the addition<br />

<strong>of</strong> TAA reduced the temperature at which the D<br />

phase formed in this system and also promoted<br />

formation <strong>of</strong> the D instead <strong>of</strong> the D' phase. The<br />

latter has considerably less ability to solubilize<br />

triolein than the former. When a drop <strong>of</strong> pure<br />

triolein was injected into an alcohol-free<br />

mixture <strong>of</strong> C12E4 and water at 50ºC, spontaneous<br />

<strong>emulsification</strong> was observed within the drop as<br />

well as the growth <strong>of</strong> small myelinic figures at<br />

the drop surface [52]. The drop volume<br />

decreased slowly with time, an indication that<br />

some triolein was being solubilized into the<br />

liquid crystalline particles present initially. This<br />

behavior is, as expected, the same as was seen<br />

with the vertical cell technique for the same<br />

system and temperature.<br />

In contrast, with TAA added to the<br />

surfactantwater mixture in an amount equal to<br />

20 wt.% <strong>of</strong> the surfactant, an intermediate liquid<br />

phase, presumably the D phase, formed when a<br />

triolein drop was injected at the same<br />

temperature. As shown in Fig. 31, the drop<br />

shrank considerably during the first few minutes<br />

<strong>of</strong> the experiment as triolein was solubilized into<br />

the D phase. After about 5 min, the triolein drop<br />

had disappeared completely. The contacting<br />

experiments thus confirmed the conclusion<br />

reached from the phase behavior results that<br />

TAA promoted the formation <strong>of</strong> the D phase,<br />

which is capable <strong>of</strong> solubilizing considerable<br />

triolein.<br />

7.2. Experiments with pure long-chain alcohols<br />

It can be shown using diffusion path theory<br />

that when the initial surfactant concentration in<br />

Fig. 31. Video frames showing the dynamic behavior<br />

<strong>of</strong> a drop <strong>of</strong> triolein contacted with 5 Wt-% C 12E 4<br />

with added TAA at 50-C (TAA/C12E7=0.20).<br />

Obtained from Ref. 52.<br />

an aqueous micellar solution is sufficiently low,<br />

no intermediate phases form upon initial contact<br />

<strong>of</strong> this solution with a long-chain alcohol.<br />

However, an intermediate phase does form<br />

when the surfactant mass fraction exceeds a


C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215 195<br />

critical value ws* given by the following<br />

equation [73]<br />

w s* = w sB F 1(w oC) + w sC F 2 (w oC) (5)<br />

Here w sB and w sC are the surfactant mass<br />

fractions at the L 1 and L 2 ends <strong>of</strong> the limiting tie<br />

line forming one boundary <strong>of</strong> the two-phase<br />

region for these phases, and w oC is the oil mass<br />

fraction at the L 2 end. The quantity w sB is <strong>of</strong>ten<br />

called the "limiting association concentration"<br />

or LAC [56]. The functions F 1 and F 2 are<br />

defined as follows<br />

where D o and D s are the diffusivities <strong>of</strong> the oil<br />

and the surfactant in the L2 phase, D's is the<br />

diffusivity <strong>of</strong> the surfactant in the L1 phase and<br />

hoC is the (constant) value <strong>of</strong> the similarity<br />

parameter [x/(4D.t) 1/2 ] at the L2 end <strong>of</strong> the<br />

limiting tie line (see discussion <strong>of</strong> diffusion path<br />

analysis above).<br />

Vertical cell-contacting experiments gave<br />

results in reasonable agreement with Eq. (3) for<br />

the sodium octanoate-n-decanol-water system<br />

[73]. One might expect that <strong>detergency</strong> would<br />

be improved when the intermediate phase - in<br />

this case the lamellar liquid crystal - is formed<br />

since more <strong>of</strong> the alcohol is solubilized. Indeed,<br />

Kielman and van Steen [11] observed such<br />

behavior in the potassium octanoate-n-decanolwater<br />

system.<br />

The focus <strong>of</strong> this section is the time-dependent<br />

behavior <strong>of</strong> the system if a drop <strong>of</strong> alcohol is<br />

injected into a solution whose surfactant<br />

concentration is below the critical value. The<br />

question to be answered is whether an<br />

intermediate phase will form at some time after<br />

initial contact.<br />

A series <strong>of</strong> such experiments was performed<br />

for the C 12E 5-water-oleyl alcohol system at 30ºC<br />

[74]. The L 1-L 2 coexistence curve and some<br />

interfacial tensions between these two phases<br />

are shown in Fig. 32. Note that the coexistence<br />

curve terminates at a point F corresponding to a<br />

water content <strong>of</strong> about 70 wt.%, which is one<br />

vertex <strong>of</strong> the L 1-Lα-L 2 three-phase triangle.<br />

Video frames taken at various times during an<br />

experiment in which the surfactant<br />

concentration was 1 wt.% are shown in Fig. 33.<br />

Note that the drop, initially some 70 mm in<br />

diameter, swells as it takes up water and<br />

surfactant. After about 23 min, the lamellar (La)<br />

phase begins to develop as myelinic figures<br />

which grow into the aqueous solution.<br />

Eventually, nearly all the alcohol is converted to<br />

liquid crystal.<br />

The order <strong>of</strong> magnitude <strong>of</strong> the time required<br />

for diffusion within the drop is the ratio <strong>of</strong> the<br />

square <strong>of</strong> its radius to the diffusion coefficient.<br />

As this time is much less than that <strong>of</strong> the<br />

experiment, a quasi-steady state scheme may be<br />

used to model drop behavior. Basically, this<br />

means that drop composition may be viewed as<br />

TIE -LINE INTERFACIAL TENSION (dyne/cm)<br />

1 2.52<br />

2 1.03<br />

3 0.25<br />

4 0.03<br />

Fig. 32. Partial ternary phase diagram for the<br />

C 12E 5-water- alcohol system at 30ºC showing the<br />

L 1-L 2 coexistence curve and the limiting tie line EF.<br />

The oil drop composition varies as indicated by the<br />

arrows. The interfacial tensions are between<br />

pre-equilibrated phases for the four tie lines shown<br />

[74]. Reprinted with permission <strong>of</strong> Plenum Press.


196 C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215<br />

Fig. 33. Video frames showing dynamic behavior following contact <strong>of</strong> 1.0 wt.% solution <strong>of</strong> C 12E 5 with a drop <strong>of</strong><br />

pure oleyl alcohol at 30ºC [74]: (a) alcohol drop about 2 min after injection; (b) the same, about 20 min later; (c)<br />

initial formation <strong>of</strong> lamellar phase about I min later; (d) growth <strong>of</strong> myclinic figures into the surrounding aqueous<br />

phase about 3 min later; (e) almost complete conversion <strong>of</strong> alcohol into liquid crystal about 8 min later. Reprinted<br />

with permission <strong>of</strong> Plenum Press.<br />

moving along the coexistence curve in the<br />

direction <strong>of</strong> the arrows in Fig. 32. Since the<br />

long-chain alcohol has low solubility in the<br />

aqueous phase, the drop volume increases<br />

during this process as its contents <strong>of</strong> water and<br />

surfactant increase. The result is swellingas


C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215 197<br />

observed during the experiment. When the drop<br />

composition reaches point F at the end <strong>of</strong> the<br />

coexistence curve, the driving force for<br />

diffusion <strong>of</strong> the surfactant and water into the<br />

drop remains. gut because the drop cannot<br />

remain in the L2 region, according to the phase<br />

diagram, the lamellar phase begins to form.<br />

A theory has been developed which predicts<br />

that the time t from the start <strong>of</strong> the experiment<br />

until the liquid crystal starts to form is given by<br />

the following expression [60]<br />

2<br />

t = Ks (R0 /Dsws∞) (6)<br />

Here R 0 is the initial radius <strong>of</strong> the drop, D s and<br />

are the diffusivity and bulk concentration <strong>of</strong> the<br />

surfactant in the aqueous solution and K s is a<br />

constant that depends only on the shape <strong>of</strong> the<br />

coexistence curve and the location <strong>of</strong> point F.<br />

The predicted proportionality between t and Ro<br />

2 has been confirmed by experiment (see Fig.<br />

34), as has the inverse relationship between t and<br />

the bulk surfactant concentration w s∞. With a<br />

Fig. 34. Plot <strong>of</strong> the square root <strong>of</strong> the time t required<br />

to initiate liquid crystal formation as a function <strong>of</strong><br />

initial drop size for the system <strong>of</strong> Fig. 32. The<br />

surfactant concentration in the aqueous phase is 1.0<br />

wt.% [74]. Reprinted with permission <strong>of</strong> Plenum<br />

Press.<br />

value <strong>of</strong> 0.514 for K. calculated using the phase<br />

behavior <strong>of</strong> Fig. 32, Eq. (6) and the measured<br />

values <strong>of</strong> t were used to estimate Ds. A value <strong>of</strong><br />

about 4 x 10 -11 m 2 s -1 was obtained, which is<br />

reasonable for a micellar solution.<br />

A similar equation has been developed for the<br />

case <strong>of</strong> a uniform layer <strong>of</strong> oil on a flat solid<br />

surface immersed in a stirred aqueous surfactant<br />

solution [74]<br />

t = K p (h 0d/D sw s∞) (7)<br />

where ho is the initial thickness <strong>of</strong> the oil layer<br />

and d is the thickness <strong>of</strong> the diffusion boundary<br />

layer adjacent to the oil. Like Ks in Eq. (6), Kp depends only on the shape and terminal point <strong>of</strong><br />

the coexistence curve between the L1 and L2 phases.<br />

For the C12E8-water-n-decanol system at<br />

temperatures above 14ºC, the three-phase<br />

triangle bounding the L1-L2 region has D'<br />

instead <strong>of</strong> Lα as the additional phase [59]. When<br />

a contacting experiment was conducted at 27ºC<br />

in this system, with a drop initially about 60 µm<br />

in diameter, a liquid intermediate phase<br />

developed after about 14 min and surrounded<br />

the initial alcohol drop [55]. As the intermediate<br />

phase grew during the next 8 min to a diameter<br />

<strong>of</strong> about 83 mm, the alcohol drop shrank slightly<br />

to a diameter <strong>of</strong> about 83 µm. Presumably, this<br />

growth process could be analyzed using a<br />

"shrinking core" model with the quasi-steady<br />

state approximation. However, as the available<br />

data on phase behavior [59] do not include<br />

coexistence curves for the L1-D' and L2-D' two-phase regions, it is not currently possible to<br />

make quantitative comparisons between<br />

predictions <strong>of</strong> the analysis and the experimental<br />

results.<br />

7.3. Experiments with mixtures <strong>of</strong> hydrocarbons<br />

and long-chain alcohols<br />

More interesting for <strong>detergency</strong> applications<br />

than pure alcohols are mixtures <strong>of</strong> polar and<br />

nonpolar oils, which are representative <strong>of</strong><br />

sebum-like soils. Here we discuss a series <strong>of</strong><br />

experiments in which drops <strong>of</strong> various mixtures<br />

<strong>of</strong> n-hexadecane and oleyl alcohol, typically


198 C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215<br />

containing at least 50 wt.% hydrocarbon, were<br />

injected into dilute aqueous solutions <strong>of</strong> pure<br />

non-ionic surfactants at temperatures below<br />

their cloud points [60]. PIT measurements for<br />

these systems were given previously (Fig. 15).<br />

A general pattern <strong>of</strong> behavior was observed.<br />

At temperatures below the PIT, as given by Fig.<br />

15, no intermediate phase was seen at any time<br />

during the experiment and the drop volume<br />

decreased very slowly with time owing to<br />

solubilization <strong>of</strong> the alcohol and hydrocarbon<br />

into the micellar solution. Above the PIT, the<br />

behavior was similar to that described in the<br />

preceding section for the C12E5-water-oleyl alcohol system. That is, the drop would swell,<br />

and, at a particular time, the lamellar liquid<br />

crystalline phase could be seen growing as<br />

myelinic figures. However, the myelinic figures<br />

were shorter, smaller in diameter and more<br />

numerous, and they grew faster than in the pure<br />

alcohol system. The differences can be seen by<br />

comparing Fig. 35 for a drop initially containing<br />

85 wt.% hydrocarbon and 15 wt.% alcohol<br />

immersed in an aqueous solution containing<br />

0.05 Wt-% C12E8, with Fig. 33 for the pure<br />

alcohol system. The low surfactant<br />

concentration for the mixed oil experiment is<br />

typical <strong>of</strong> those used in household laundry<br />

processes.<br />

Coexistence curves for the L1-L2 region were<br />

determined at 30ºC for systems containing<br />

n-dodecyl heptaoxyethylene monoether (C12E7) and oils with 50% and 75% n-hexadecane,<br />

respectively [60]. In both cases the curves<br />

extended to water contents <strong>of</strong> 75-80 wt.%.<br />

Contacting experiments for various initial drop<br />

sizes and surfactant concentrations in the latter<br />

system confirmed that the dynamic behavior<br />

was consistent with Eq. (6). Here, too, liquid<br />

crystalline intermediate phases were seen for<br />

drops in contact with solutions containing only<br />

0.05 wt.% surfactant.<br />

It was also noted that the rate <strong>of</strong> swelling <strong>of</strong><br />

the drops increased markedly as the limiting tie<br />

line was approached. Since the coexistence<br />

curves exhibited nearly constant alcohol-tosurfactant<br />

ratios near the limiting tie lines, this<br />

behavior was expected as little surfactant must<br />

diffuse into the drop for it to experience a<br />

substantial increase in volume. The basic<br />

analysis leading to Eq. (6) confirmed<br />

quantitatively that rapid swelling should occur<br />

for these conditions.<br />

Values <strong>of</strong> the parameter K s were found from<br />

the phase behavior data to be 0.313 and 0.0516<br />

for the two systems. That is, for a given initial<br />

drop size and surfactant concentration,<br />

formation <strong>of</strong> the liquid crystal occurred more<br />

rapidly in the system containing less alcohol,<br />

presumably because less surfactant had to<br />

diffuse into the drop to balance hydrophilic and<br />

lipophilic properties <strong>of</strong> the surfactant films -to<br />

the extent that conditions favorable for<br />

formation <strong>of</strong> the lamellar phase were created.<br />

Indeed, a general result <strong>of</strong> the contacting<br />

experiments with various surfactants and oils<br />

was that more time was required for liquid<br />

crystal formation when the system was made<br />

less hydrophilic by increasing temperature or<br />

the alcohol content <strong>of</strong> the drop or by reducing<br />

the ethylene oxide chain length <strong>of</strong> the surfactant<br />

[60].<br />

One may ask why the intermediate phase<br />

formed during the contacting experiments<br />

should be the lamellar liquid crystal instead <strong>of</strong> a<br />

microemulsion, in cases where the drops<br />

contained substantial amounts <strong>of</strong> hydrocarbon.<br />

After all, diffusion <strong>of</strong> surfactant into the drop<br />

should cause the surfactantalcohol films within<br />

it to become more hydrophilic, as indicated<br />

above. Eventually, one might expect film<br />

composition to approach that corresponding to<br />

the PIT at the experimental temperature and<br />

thereby cause formation <strong>of</strong> a middle-phase<br />

microemulsion. The answer is that the drop<br />

itself becomes a microemulsion as it takes up<br />

water and surfactant. The occurrence <strong>of</strong> low<br />

interfacial tensions supports this conclusion. For<br />

instance. a tension <strong>of</strong> about 0.03 mN m -1 was<br />

measured by the spinning drop technique about<br />

50 min after the start <strong>of</strong> the experiment for an<br />

oil phase initially containing 75% n-hexadecane<br />

and 25% oleyl alcohol and an aqueous solution<br />

containing 0.1 wt% C 12E 7 at 30-C [60].<br />

The coexistence curve for this system<br />

indicates that the water-to-hydrocarbon ratio<br />

during an oil drop contacting experiment varies


C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215 199<br />

Fig. 35. Video frames showing dynamic behavior following the contact <strong>of</strong> 0.05 Wt'% C12E8 solution with a drop<br />

<strong>of</strong> 5.67/1 n-hexadecane-oleyl alcohol at 50ºC [60]: (a) oil drop about 13 min after injection; (b) the same about 8<br />

min later; (c) growth <strong>of</strong> myelinic figures less than 6 s later; (d) further growth <strong>of</strong> myelinic figures into the aqueous<br />

phase 12 s later. Reprinted with permission <strong>of</strong> the American Chemical Society.<br />

from its initial value <strong>of</strong> zero to about 5 when<br />

the drop composition eventually reaches the<br />

end <strong>of</strong> the coexistence curve (corresponding to<br />

point F <strong>of</strong> Fig. 32). No intermediate phase<br />

forms until point F is reached because the<br />

hydrocarbon content does not exceed the<br />

solubilization limit <strong>of</strong> the microemulsion.<br />

However, at point F, where the<br />

hydrocarbon-to-amphiphile ratio has dropped<br />

to about 1.5, the microemulsion structure<br />

apparently cannot be sustained and an<br />

intermediate lamellar phase begins to form.<br />

That it would form at such ratios <strong>of</strong> the various<br />

components and when the composition <strong>of</strong> the<br />

surfactant-alcohol films is approximately<br />

balanced between the hydrophilic and<br />

lipophilic properties is generally consistent<br />

with phase behavior results for other systems<br />

reported by Ghosh and Miller [46]. Indeed,<br />

lamellar phases with even higher oil contents<br />

have been reported near the PIT <strong>of</strong> the<br />

C 12E 4-water-n-hexadecane system [75].


200 C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215<br />

When the alcohol content <strong>of</strong> the oil is low and<br />

the temperature is near or only slightly above<br />

the PIT, only a small amount <strong>of</strong> surfactant need<br />

diffuse into the drop in order for hydrophilic and<br />

lipophilic properties <strong>of</strong> the surfactant-alcohol<br />

films formed there to be almost balanced. In this<br />

case, the hydrocarbon solubilization limit is<br />

exceeded and drops <strong>of</strong> oil form within the<br />

original drop, which has become a middle-phase<br />

microemulsion. Such behavior was observed,<br />

for example, when a drop containing 90 wt.%<br />

n-hexadecane was injected into a solution<br />

containing 0.05 wt.% C12E7 at 40ºC, which is<br />

near the PIT for an oil <strong>of</strong> this composition. As<br />

the oil content <strong>of</strong> the microemulsion decreased<br />

during the experiment owing to this spontaneous<br />

<strong>emulsification</strong>, and as the surfactant content<br />

continued to increase, a point was eventually<br />

reached when the hydrocarbon-to-amphiphile<br />

ratio was too low to sustain the microemulsion<br />

structure and the lamellar phase again developed<br />

as myelinic figures [76].<br />

Finally, it is noteworthy that when drops <strong>of</strong><br />

oil containing 75% n-hexadecane and 25% oleyl<br />

alcohol were contacted with a surfactant<br />

solution containing 0.05 Wt-% C12E7 at<br />

temperatures above 40ºC, it appeared that the<br />

first intermediate phase formed was a liquid,<br />

presumably D' [76]. That is the system<br />

apparently experienced between 30. and 400 the<br />

transition discussed in the phase behavior<br />

section above which D' replaces La as the third<br />

phase in the three-phase region bounding the<br />

region <strong>of</strong> coexistence between La and L2.<br />

Detailed phase behavior at 40º was not<br />

determined, however. The intermediate D' phase<br />

in the contacting experiment was later converted<br />

to liquid crystal (myelinic figures).<br />

7.4. Experiments with mixtures <strong>of</strong> triolein and<br />

longchain alcohols<br />

Drops containing various mixtures <strong>of</strong> triolein<br />

and oleyl alcohol were injected into 0.1 wt.%<br />

solutions <strong>of</strong> C 12E 6 at 40ºC, which is about 10º<br />

below the cloud point temperature <strong>of</strong> dilute<br />

solutions <strong>of</strong> this surfactant [61]. As discussed in<br />

section 3 and shown in Fig. 16, the sequence <strong>of</strong><br />

phases seen with increasing temperature in this<br />

system is similar to that found with pure triolein,<br />

the various transition temperatures being lower<br />

for higher oleyl alcohol contents. For drops<br />

containing less than 25 wt.% oleyl alcohol, a<br />

scarcely perceptible change in drop size with<br />

time was observed, an indication that the oil was<br />

being slowly solubilized into the surfactant<br />

solution. As indicated above, similar behavior<br />

was seen for hydrocarbon- alcohol drops below<br />

the PIT. Figure 16 confirms that the system is<br />

quite hydrophilic under these conditions.<br />

A liquid intermediate phase, presumably D',<br />

formed after a few minutes for a drop containing<br />

50 wt.% oleyl alcohol (Fig. 36(b)). Careful<br />

study <strong>of</strong> the image at various depths <strong>of</strong> focus<br />

revealed that this phase did not surround the<br />

original drop but instead formed a drop in<br />

contact with the original drop but larger in<br />

diameter, as the figure shows. Later, myelinic<br />

figures began to grow outward into the aqueous<br />

solution from the D'phase (Fig. 36(c)).<br />

Ultimately, what was left <strong>of</strong> the original oil drop<br />

- its volume was only about a third <strong>of</strong> the initial<br />

value - became detached from the intermediate<br />

liquid phase and showed no further changes<br />

with time.<br />

Although this system has four components,<br />

with the result that its phase behavior cannot be<br />

adequately described by a triangular diagram<br />

similar to Fig. 32, an explanation can be given<br />

for the dynamic behavior observed. As in the<br />

hydrocarbon-alcohol systems, surfactant<br />

diffuses into the drop, and the surfactant-alcohol<br />

films which form there become more<br />

hydrophilic with time. Eventually, the boundary<br />

<strong>of</strong> the L1-L2 coexistence region is reached, and<br />

an intermediate D' phase forms as surfactant<br />

continues to diffuse into the drop. As this new<br />

phase solubilizes little triolein, the<br />

alcohol-to-triolein ratio in the original drop<br />

decreases greatly. However, the D' phase is<br />

itself in contact with the surfactant solution and<br />

accordingly experiences an increase in<br />

surfactant content with time. When the<br />

hydrophilic and lipophilic properties <strong>of</strong> its<br />

surfactant - alcohol films are balanced, the D'


C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215 201<br />

Fig. 36. Video frames from an experiment in which a drop containing equal amounts <strong>of</strong> triolein and oleyl alcohol<br />

was injected into a 0.1 Wt'% C 12E 6 solution at 40ºC [61]. (a) Shortly after injection; (b) approximately 9 min later<br />

(the intermediate phase has formed); (c) about 5 min later (the intermediate phase has grown, myelinic figures are<br />

starting to form); (d) about I min later (a drop <strong>of</strong> unsolubilized oil separates from the intermediate phase); (e) the<br />

drop is never completely solubilized. Reprinted with permission <strong>of</strong> Marcel Dekker publishers.


202 C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215<br />

phase, whose films are known to be slightly<br />

lipophilic, cannot persist, and the lamellar phase<br />

forms as myelinic figures. Since the triolein-rich<br />

drop does not dissolve, it may well be that the<br />

system enters a four-phase region at the time the<br />

myelinic figures develop.<br />

For a drop containing 75% oleyl alcohol<br />

behavior was more complex. Two intermediate<br />

liquid phases formed at various times during the<br />

experiment and greater solubilization was<br />

observed [61]. Ultimately, the liquid crystalline<br />

phase was seen as well.<br />

When the drop is pure oleyl alcohol, no D'<br />

phase is seen, and the first intermediate phase is<br />

the lamellar liquid crystal. It develops as many<br />

short myelinic figures, the appearance being<br />

more similar to that <strong>of</strong> Fig. 35 for<br />

hydrocarbon-oleyl alcohol drops than to that <strong>of</strong><br />

Fig. 33 for the C 12E 5-water-oleyl alcohol system.<br />

8. Fabric <strong>detergency</strong> test methods<br />

Laboratory evaluations <strong>of</strong> laundry <strong>detergency</strong><br />

can range from use <strong>of</strong> an actual washing<br />

machine with a capacity <strong>of</strong> several gallons to<br />

use <strong>of</strong> a Terg-O-Tometer mini-laundry machine.<br />

The Terg-0-Tometer is a washing machine<br />

which replicates on a small scale the cleaning<br />

action <strong>of</strong> an agitatortype washing machine [77].<br />

The capacity <strong>of</strong> a single Terg-O-Tometer pot is<br />

approximately 11. The temperature can be<br />

controlled quite accurately through the use <strong>of</strong> a<br />

water bath surrounding a bank <strong>of</strong> four or six<br />

pots. The small size <strong>of</strong> a Terg-O-Tometer allows<br />

rapid comparative studies <strong>of</strong> detergent<br />

formulations under the same washing<br />

conditions. Also, formulations containing<br />

expensive ingredients, e.g. pure alcohol<br />

ethoxylates, or experimental surfactants <strong>of</strong><br />

limited quantity, can be evaluated at reasonable<br />

costs. For these reasons, the use <strong>of</strong><br />

Terg-O-Tometer machines in detergent research<br />

laboratories has become quite common.<br />

The measurement <strong>of</strong> soil removal from fabrics<br />

can be performed in several ways [78]. The<br />

most straightforward technique involves visual<br />

inspection for determination <strong>of</strong> cleanliness.<br />

More sophisticated techniques involve the use<br />

<strong>of</strong> spectrophotometry, chemical analyses, and<br />

radiotracer methods. As with visual inspection,<br />

the measurement <strong>of</strong> light reflectance from fabric<br />

does not quantify actual soil removal but<br />

depends on factors such as the distribution <strong>of</strong><br />

soil on the fabric, and the particle sizes.<br />

Analyses <strong>of</strong> weight gain or loss through the<br />

chemical extraction <strong>of</strong> soils from the fabric as<br />

well as radiotracer methods both yield<br />

measurements <strong>of</strong> actual soil removal and do not<br />

depend on the change <strong>of</strong> appearance <strong>of</strong> the<br />

fabric. Although not representative <strong>of</strong> real-world<br />

visualization <strong>of</strong> cleanliness, both techniques<br />

provide complementary information to<br />

reflectance measurements and are more useful<br />

for mechanistic studies.<br />

Specific advantages <strong>of</strong> the radiotracer<br />

<strong>detergency</strong> method are given elsewhere [79,80].<br />

This method makes use <strong>of</strong> mildly radiolabeled<br />

soils for highly sensitive quantitative<br />

determination <strong>of</strong> soil removal by simple<br />

radiochernical techniques. Several radioisotopes<br />

may be used as labels. Generally, polar oily<br />

soils such as alcohols and acids are tagged with<br />

small amounts <strong>of</strong> carbon-14 labeled material.<br />

The use <strong>of</strong> two labels in a soil such as artificial<br />

sebum, which contains both polar and non-polar<br />

components, allows the removal <strong>of</strong> both types <strong>of</strong><br />

soil to be monitored simultaneously. The recipe<br />

for such a labeled artificial sebum has been<br />

given elsewhere [80]. Radiolabelled particulate<br />

and protein soils have also been developed and<br />

utilized [80].<br />

The Shell Development Company has utilized<br />

a radiotracer <strong>detergency</strong> method for many years<br />

to evaluate laundry detergents and to study the<br />

fundamental nature <strong>of</strong> oily soil removal [81,82].<br />

In the case <strong>of</strong> oily soils, a 4-in square fabric<br />

swatch is soiled with 28 mg <strong>of</strong> radiolabeled oil.<br />

The soil is applied to the fabric in a toluene<br />

solution, which is allowed to air dry. The swatch<br />

is washed under controlled conditions in a<br />

Terg-0-Tometer. Aliquots <strong>of</strong> wash water are<br />

removed for radioactive counting, with the level<br />

<strong>of</strong> soil remaining on the swatch determined by<br />

difference. In addition to the sebum oil<br />

described above, oily soils which have been<br />

commonly utilized include tritium-labelled<br />

hexadecane (cetane), 1/1 hexadecane-squalane


C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215 203<br />

(a C 30 branched hydrocarbon) and triolein.<br />

Also, hydrocarbon-fatty acid (or alcohol) soils<br />

<strong>of</strong> varying composition have been studied in<br />

which both carbon- 14 and tritium labels are<br />

utilized to allow the discrimination <strong>of</strong> removal<br />

<strong>of</strong> both components. The following are results<br />

<strong>of</strong> fundamental <strong>detergency</strong> studies using some<br />

<strong>of</strong> these soils, which were performed to allow<br />

correlation to the phase behavior and dynamic<br />

contacting studies described above.<br />

9. Fabric <strong>detergency</strong> - non-polar<br />

hydrocarbon soils<br />

Results <strong>of</strong> radiotracer <strong>detergency</strong> studies <strong>of</strong><br />

removal <strong>of</strong> hexadecane from 65/35 permanent<br />

press polyester-cotton fabric are shown in Fig.<br />

37 [18]. This soil can be considered a model for<br />

non-polar hydrocarbon-based soils such as<br />

lubricating oils. The washing solutions<br />

contained 0.05 wt.% surfactant, resulting in a<br />

fabric-to-soil weight ratio <strong>of</strong><br />

Fig. 37. Removal <strong>of</strong> n-hexadecane from 65/35<br />

polyester-cotton fabric using 0.05 wt.% aqueous<br />

solutions <strong>of</strong> C 12E 4 and C 12E 5 [18]. The arrows show<br />

the cloud point temperatures <strong>of</strong> the surfactants.<br />

Reprinted with permission <strong>of</strong> Academic Press.<br />

40/1 and a surfactant-to-soil ratio <strong>of</strong> 9/1.<br />

Triethanolamine (50 ppm) was included as a<br />

buffer, and no water hardness was present. The<br />

washing time in the Terg-0-Tometer was 10<br />

min.<br />

Of particular interest in these studies was the<br />

effect <strong>of</strong> temperature and non-ionic alcohol<br />

ethoxylate surfactant structure on the levels <strong>of</strong><br />

soil removal. As shown in Fig. 37, a strong<br />

dependence on temperature was observed with<br />

the highest levels <strong>of</strong> soil removal occurring<br />

almost 200C above the cloud point <strong>of</strong> the<br />

washing solution, a temperature regime in which<br />

the washing solution structure is a liquid<br />

crystalline dispersion. In fact, the optimum<br />

<strong>detergency</strong> temperature (ODT) in each case<br />

occurred very near the PIT <strong>of</strong> the<br />

water-surfactant-hexadecane system. Although<br />

not shown in Fig. 37, very poor <strong>detergency</strong><br />

occurred in the temperature range <strong>of</strong> interest<br />

with C 12E 3 as the surfactant. Also, when a 1/1 by<br />

weight mixture <strong>of</strong> hexadecane and squalane was<br />

used as the soil, a similar peak in performance<br />

was found for each surfactant near its PIT for<br />

the mixed soil. These PITs were approximately<br />

7ºC higher than the corresponding values for<br />

hexadecane alone, i.e. 37 and 58ºC versus 30<br />

and 52ºC. The equivalence <strong>of</strong> the PIT and ODT<br />

in this type <strong>of</strong> <strong>detergency</strong> system has been<br />

confirmed by the work <strong>of</strong> Schambil and<br />

Schwuger [22] as well as Solans et al. [83].<br />

Also, the same correspondence has been found<br />

for the removal <strong>of</strong> hydrocarbon by the same<br />

surfactant from 100% polyester fabric [84]. The<br />

high levels <strong>of</strong> soil removed near the PIT can be<br />

attributed to the ultralow interfacial tensions<br />

achieved near that temperature, and to the high<br />

rates <strong>of</strong> oily soil solubilization into<br />

middle-phase microemulsions, as was visualized<br />

in the dynamic contacting studies described<br />

previously.<br />

When the detergent properties <strong>of</strong> mixtures <strong>of</strong><br />

C 12E 5 with the hydrophobic additives C 12E 3 and<br />

n-dodecanol (C 12E 0) were studied, optimum<br />

<strong>detergency</strong> was found at temperatures lower<br />

than the ODT for C 12E 5 alone but somewhat<br />

higher than the ODT for C12E4. Detergency<br />

data for the two additive systems, which<br />

exhibited the same cloud point as C 12E 4, are


204 C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215<br />

Fig. 38. Removal <strong>of</strong> n-hexadecane from 65/35<br />

polyester-cotton fabric using 0.05 wt.% aqueous<br />

solutions <strong>of</strong> a 90/10 blend <strong>of</strong> C 12E 5 and n-dodecanol<br />

[30]. Reprinted with permission <strong>of</strong> Academic Press.<br />

shown in Figs. 38 and 39 [30]. The optimum<br />

temperatures correspond quite closely to the<br />

phase inversion temperatures extrapolated to the<br />

low soil-to-surfactant ratios used in the<br />

<strong>detergency</strong> studies. In this regard, C 12E 3 was<br />

somewhat more effective than dodecanol in<br />

lowering the optimum <strong>detergency</strong> temperature.<br />

Of interest here is that the <strong>detergency</strong> results<br />

Fig. 39. Removal <strong>of</strong> n-hexadecane from 65/35<br />

polyester-cotton fabric using 0.05 wt.% aqueous<br />

solutions <strong>of</strong> a 60/40 blend <strong>of</strong> C 12E 5 and C 12E 3 [30].<br />

Reprinted with permission <strong>of</strong> Academic Press.<br />

differed from the behavior observed in the<br />

verticalcontacting studies in which, as discussed<br />

above, somewhat higher temperatures were<br />

required for fastest oil solubilization due to<br />

partitioning <strong>of</strong> the additive into the oil.<br />

Practical <strong>detergency</strong> applications utilize<br />

commercial alcohol ethoxylate surfactants<br />

which contain a broad range <strong>of</strong> species having<br />

varying hydrophobe lengths and levels <strong>of</strong><br />

ethylene oxide, Detergency studies were<br />

performed in the manner described above for the<br />

single ethoxylate and ethoxylate-additive<br />

systems using commercial ethoxylates based on<br />

a blend <strong>of</strong> predominantly normal C 12-C 13<br />

alcohols and containing an average <strong>of</strong> 3, 4 and 5<br />

mol <strong>of</strong> ethylene oxide, respectively [85]. These<br />

materials are denoted N23-3, N23-4 and N23-5.<br />

Table 2 compares the ODTs for these systems to<br />

those <strong>of</strong> the corresponding specific alcohol<br />

ethoxylate systems having the same average<br />

structure. The ODTs <strong>of</strong> the commercial<br />

materials and their molar-average equivalents<br />

are the same in all three cases. The ODTs <strong>of</strong> the<br />

commercial systems, in fact, match the PITs for<br />

those systems with hexadecane extrapolated to<br />

very low levels <strong>of</strong> oil. The PIT data are shown<br />

in Fig. 40. Shown for comparison are the PIT<br />

vs. soil-surfactant ratio plots for C 12E 4 and<br />

C 12E 5. These are flat, indicative <strong>of</strong> the fact that<br />

the PITs for ternary specific alcohol ethoxylatewater-hydrocarbon<br />

systems are independent <strong>of</strong><br />

the oil-surfactant ratio. Bercovici and Krussman<br />

Table 2<br />

Correlation <strong>of</strong> PIT to optimum <strong>detergency</strong><br />

temperature ODT<br />

Surfactant PIT (ºC) ODT (ºC)<br />

Specific ethoxylate<br />

C 12E 5 52 50<br />

C 12E 4 31 30<br />

C 12E 3 < 20 < 20<br />

Broad-range Ethoxylate<br />

N23-5 51 a 50<br />

N23-4 26 a 30<br />

N23-3 < 20 a < 20<br />

The data shown are for cetane removal from 65/35<br />

polyester-cotton with 0.05% surfactant [85].<br />

a Evaluated at cetane-surfactant ratio → 0.


C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215 205<br />

Fig. 40. PIT as a function <strong>of</strong> n-hexadecane(cetane)surfactant<br />

weight ratio for commercial (broad-range)<br />

and specific alcohol ethoxylates [85].<br />

[86] have shown that the addition <strong>of</strong> long-chain<br />

alcohols to commercial alcohol ethoxylates, as<br />

described above for the specific ethoxylate<br />

C 12E 5, reduces the PIT and optimum <strong>detergency</strong><br />

temperature for nonpolar soil removal by those<br />

surfactants. As with the specific ethoxylate and<br />

surfactant- additive systems, optimum soil<br />

removal with commercial nonionic surfactants<br />

occurs near the extrapolated PIT because the<br />

balanced surfactant system provides high rates<br />

<strong>of</strong> oil solubilization and low interfacial tensions.<br />

The effects <strong>of</strong> the addition <strong>of</strong> anionic<br />

surfactants to non-ionics on the relationship <strong>of</strong><br />

soil removal to washing temperature have also<br />

been investigated [87]. In this case, the same<br />

commercial sodium alcohol ethoxysulfate<br />

mentioned in the phase behavior section<br />

(Neodol 23-3S) and the sodium salt <strong>of</strong> C 12 linear<br />

alkylbenzenesulfonate (denoted C 12LAS) were<br />

used as the anionic surfactants. As noted<br />

previously, C 12E 3 itself is not effective at<br />

removing hexadecane from 65/35 polyestercotton<br />

at temperatures between 20 and 70ºC.<br />

This result can be attributed to C 12E 3 being too<br />

hydrophobic. However, the addition <strong>of</strong><br />

appropriate amounts <strong>of</strong> hydrophilic anionic<br />

surfactant was found to improve its performance<br />

in a 1% NaCl solution, as shown in Figs. 41 and<br />

42. The total surfactant concentration in the<br />

Fig. 41. Removal <strong>of</strong> n-hexadecane (cetane) by<br />

C 12E 3-N23-3S blends [87]. Reprinted with<br />

permission <strong>of</strong> the American Oil Chemists' Society.<br />

Fig. 42. Removal <strong>of</strong> n-hexadecane by C 12E 3-C 12LAS<br />

blends [87]. Reprinted with permission <strong>of</strong> the<br />

American Oil Chemists' Society.<br />

washing solution in these experiments was<br />

0.05% by weight, the same as in the non-ionic<br />

surfactant studies discussed above.<br />

Optimum <strong>detergency</strong> temperatures were<br />

increased to approximately 30ºC by the<br />

substitution for C 12E 3 <strong>of</strong> 22% and 24% N23-3S<br />

and LAS, respectively, and to approximately<br />

50ºC by the substitution <strong>of</strong> 33% and 38%<br />

N23-3S and LAS. Consistent with these results<br />

was the finding <strong>of</strong> optimum <strong>detergency</strong> at 35ºC<br />

with a 76.5/23.5 mixture <strong>of</strong> C 12E 3 and N23-3S<br />

[24]. Thus, in contrast to the hydrophobic


206 C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215<br />

additives such as dodecanol described above,<br />

the addition <strong>of</strong> anionic surfactants increases the<br />

temperature for optimum <strong>detergency</strong>. However,<br />

as found for the hydrophobic additives, the ODT<br />

still corresponds closely to the PIT at a low<br />

oil-surfactant weight ratio. In the case <strong>of</strong> the<br />

anionic surfactants, the ratio <strong>of</strong> anionicto-nonionic<br />

surfactant yielding phase inversion at a<br />

given temperature was determined rather than<br />

the PIT for a given surfactant composition, as<br />

the oil-surfactant ratio was varied [87]. The<br />

phase inversion composition plots determined<br />

by electrical conductivity measurements and<br />

used for comparison with the <strong>detergency</strong> data<br />

are shown in Figs. 43 and 44 [87]. These results<br />

are consistent with those <strong>of</strong> other phase behavior<br />

studies with the same system (see Fig. 7). At a<br />

given temperature, the extrapolated phase<br />

inversion composition at a very low<br />

soil-surfactant ratio matches the detergent<br />

composition yielding optimum <strong>detergency</strong> at<br />

that temperature. The high levels <strong>of</strong> soil removal<br />

can be explained by the solubilization <strong>of</strong> oil into<br />

a middle-phase microemulsion at the optimum<br />

conditions, as described in the oil<br />

dropcontacting section above.<br />

In summary, for the removal <strong>of</strong> n-hexadecane<br />

soils by a variety <strong>of</strong> pure non-ionic surfactant<br />

systems, non-ionic surfactant-hydrophobic addi-<br />

Fig. 43. Phase inversion compositions for<br />

C 12E 3-N23-3S system [87]. Reprinted with<br />

permission <strong>of</strong> the American Oil Chemists' Society.<br />

Fig.44. Phase inversion compositions for<br />

C 12E 3-C 12LAS system [87]. Reprinted with<br />

permission <strong>of</strong> the American Oil Chemists' Society.<br />

tive systems, and non-ionic-anionic surfactant<br />

systems, optimum <strong>detergency</strong> can be related to<br />

the PIT. More specifically, the optimum<br />

<strong>detergency</strong> temperature is the extrapolated PIT<br />

at low oil-surfactant ratios where the<br />

composition <strong>of</strong> the surfactant films matches the<br />

composition <strong>of</strong> detergent in the washing<br />

solution. In all cases the surfactant solutions at<br />

the optimum conditions are dispersions <strong>of</strong><br />

surfactant-rich phases, most commonly lamellar<br />

liquid crystals in water.<br />

10. Fabric <strong>detergency</strong> - triglycerides and<br />

hydrocarbon-polar soil mixtures<br />

Radiotracer <strong>detergency</strong> studies <strong>of</strong> triolein soil<br />

removal from 65/35 permanent press<br />

polyester-cotton fabric have been performed<br />

under the same conditions as. described in the<br />

previous section [48]. The triolein, which<br />

models more complex triglyceride mixtures<br />

such as those found in cooking oils, was tagged<br />

with tritium-labeled triolein. Figure 45 shows<br />

the results for triolein removal using 0.05%<br />

solutionso <strong>of</strong> C 12E 3, C 12E 4, and C 12E 5. For C 12E 3,<br />

triolein removal is very low at all temperatures<br />

studied. This result is consistent with that found<br />

for the removal <strong>of</strong> hexadecane by the same<br />

surfactant. C 12E 4 provides higher and almost<br />

constant soil removal between 25 and 50 ºC.


C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215 207<br />

Fig. 45. Removal <strong>of</strong> triolein from 65/35 polyestercotton<br />

fabric by washing with pure non-ionic<br />

surfactants [48].<br />

Above 50ºC, <strong>detergency</strong> drops <strong>of</strong>f sharply. C 12E 5<br />

yielded optimum <strong>detergency</strong> at still higher<br />

temperatures, with a rather sharp peak in<br />

performance occurring at 65ºC.<br />

The optimum <strong>detergency</strong> temperature ranges<br />

for both C 12E 4 and C 12E 5 are higher than the<br />

ODTs found with hexadecane soil. At 65ºC with<br />

C 12E 5, optimum <strong>detergency</strong> corresponds to the<br />

temperature at which the D phase forms in that<br />

system, i.e. the PIT. With C 12E 4, no sharp peak<br />

in performance is noted because the D phase<br />

forms in that system only for a very narrow<br />

temperature range [48]. The decrease in<br />

<strong>detergency</strong> at high temperatures (above 50ºC for<br />

C 12E 4 and 65ºC for C 12E 5) corresponds to the<br />

regime where the surfactant solubility in the<br />

triolein phase increases significantly and a<br />

water-in-oil microemulsion forms, i.e. where<br />

little if any solubilization <strong>of</strong> oil into the washing<br />

solution occurs.<br />

Detergency studies have also been performed<br />

in which TAA was added to the washing<br />

solution at levels up to 20% relative to the<br />

surfactant [52]. This study was performed to<br />

determine if the addition <strong>of</strong> TAA could increase<br />

triolein removal by promoting the formation <strong>of</strong><br />

the D phase during the washing process. The<br />

levels <strong>of</strong> soil removal were found by extracting<br />

the washed fabric swatches (both 100%<br />

polyester and 65/35 polyester-cotton) with<br />

hexane. For C 12E 4, triolein removal was poor at<br />

50ºC and was unaffected by the addition <strong>of</strong><br />

TAA. Soil removal was much greater at 35ºC,<br />

with the addition <strong>of</strong> 20% TAA providing only a<br />

slight improvement when compared with<br />

<strong>detergency</strong> with C 12E 4 alone. The lack <strong>of</strong><br />

significant performance enhancement was<br />

attributed to the high water solubility <strong>of</strong> TAA,<br />

which limits its effectiveness in highly dilute<br />

surfactant solutions. Formation <strong>of</strong> the D phase<br />

in the oil drop-contacting studies (see Fig. 31)<br />

apparently occurred due to the higher<br />

concentration <strong>of</strong> surfactant and TAA used in<br />

that study.<br />

A similar dependence <strong>of</strong> soil removal on<br />

temperature was observed when the same<br />

surfactants were used to remove a soil mixture<br />

containing 1/1 hexadecane-triolein. In this<br />

study, the triolein was labeled with tritium and<br />

the hexadecane was labeled with carbon-14 to<br />

allow independent measurement <strong>of</strong> the removal<br />

<strong>of</strong> the two components. The results <strong>of</strong> that study<br />

are shown in Fig. 46. When compared with the<br />

results in Fig. 45, somewhat higher removal<br />

levels <strong>of</strong> triolein were found from the soil<br />

mixture. Also <strong>of</strong> interest is that removal <strong>of</strong> hexa-<br />

Fig. 46. Removal <strong>of</strong> triolein and n-hexadecane from<br />

65/35 polyester-cotton fabric by washing with pure<br />

non-ionic surfactants [48]. Soil contains 50 wt.%<br />

triolein.


208 C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215<br />

decane was consistently higher than that <strong>of</strong><br />

triolein at all temperatures. The regions <strong>of</strong><br />

optimum <strong>detergency</strong> occur near the PITs <strong>of</strong> the<br />

surfactants with the oil blend. The preferential<br />

removal <strong>of</strong> hexadecane can be attributed to the<br />

higher solubility <strong>of</strong> hexadecane in the<br />

middle-phase microemulsion D (see Table 1).<br />

The removal <strong>of</strong> blends <strong>of</strong> hexadecane with<br />

oleyl alcohol and oleic acid by 0.05% solutions<br />

<strong>of</strong> alcohol ethoxylates has also been studied<br />

[12]. These binary soil mixtures are models <strong>of</strong><br />

sebum-like soils containing both non-polar and<br />

polar fractions. In general, somewhat more<br />

hydrophilic surfactants are required to obtain<br />

high removal <strong>of</strong> these soils compared to the<br />

optimum surfactants for nonpolar hexadecane<br />

removal. Results for the removal <strong>of</strong> hexadecane<br />

from 9/1 and 3/1 blends <strong>of</strong> hexadecane and oleyl<br />

alcohol are shown in Figs. 47 and 48. N25-9 is a<br />

broad-range commercial ethoxylate based on a<br />

predominantly linear C 12-C 15 alcohol and<br />

containing an average <strong>of</strong> nine ethylene oxide<br />

(EO) units. In the case <strong>of</strong> the 9/1 soil at 20ºC,<br />

<strong>detergency</strong> was found to increase with<br />

decreasing EO content <strong>of</strong> the pure non-ionic<br />

surfactants, while the reverse trend was true at<br />

the higher temperatures. For both soils,<br />

optimum <strong>detergency</strong> occurred over a<br />

temperature range below the cloud points <strong>of</strong> the<br />

surfactants, behavior which contrasts markedly<br />

Fig. 47. Removal <strong>of</strong> n-hexadecane for mixed 9/1<br />

n-hexadecane-oleyl alcohol soil [12]. Reprinted with<br />

permission <strong>of</strong> the American Oil Chemists' Society.<br />

Fig. 48. Removal <strong>of</strong> n-hexadecane for mixed 3/1<br />

n-hexadecane-oleyl alcohol soil [12]. Reprinted with<br />

permission <strong>of</strong> American Oil Chemists' Society.<br />

markedly with the results discussed above for<br />

hydrocarbon and triolein soils. The cleaning<br />

efficiency <strong>of</strong> the commercial broad-range<br />

ethoxylate is almost identical at all temperatures<br />

to that <strong>of</strong> the specific ethoxylate (C 12E 8) having<br />

nearly the same cloud point temperature,<br />

indicating, once again, that surfactant<br />

partitioning effects in the <strong>detergency</strong> studies are<br />

negligible.<br />

The effect <strong>of</strong> the soil composition on soil<br />

removal can be seen more clearly in Fig. 49.<br />

While the <strong>detergency</strong> peaks near 80ºC with the<br />

pure hexadecane soil, good <strong>detergency</strong> is<br />

obtained at much lower temperature with the<br />

Fig. 49. Effect <strong>of</strong> oleyl alcohol content on<br />

n-hexadecane removal [12]. Reprinted with<br />

permission <strong>of</strong> the American Oil Chemists' Society.


C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215 209<br />

mixed soils. Maximum <strong>detergency</strong> is reached<br />

near 40ºC with 10% oleyl alcohol in the soil<br />

while high levels <strong>of</strong> cetane removal are attained<br />

down to 20ºC with 25% oleyl alcohol present.<br />

The temperatures at the lower limit <strong>of</strong> the<br />

plateaus in <strong>detergency</strong> are the respective phase<br />

inversion temperatures. For C 12E 7, PIT values<br />

can be obtained from the curves in Fig. 15.<br />

Above the PIT, soil removal remains at a high<br />

level until the cloud point <strong>of</strong> the surfactant is<br />

reached. As implied in the oil drop-contacting<br />

section, the formation <strong>of</strong> an intermediate liquid<br />

crystalline phase plays the key role in the<br />

soilremoval process in the plateau regime.<br />

The relative amount <strong>of</strong> polar soil also affects<br />

the rate <strong>of</strong> soil removal. Despite being removed<br />

at a comparable level to that <strong>of</strong> the 3/1<br />

cetane-oleyl alcohol soil after 10 min at 50ºC,<br />

the 9/1 soil is actually removed much more<br />

quickly, as indicated in Fig. 50. This very fast<br />

removal can be attributed to that soil having a<br />

PIT <strong>of</strong> approximately 50ºC with C 12E 7. A 1/1<br />

cetane-oleyl alcohol soil is removed very slowly<br />

at 50ºC, although a high level <strong>of</strong> removal is<br />

ultimately attained. Removal <strong>of</strong> the hexadecane<br />

soil reaches a plateau at a lower level since the<br />

washing temperature is well below the optimum<br />

<strong>detergency</strong> temperature for its removal by C 12E 7,<br />

i.e. approximately 80ºC.<br />

Fig. 50. Kinetics <strong>of</strong> n-hexadecane removal for mixed<br />

n-hexadecane-oleyl alcohol blends [12]. Reprinted<br />

with permission <strong>of</strong> the American Oil Chemists'<br />

Society.<br />

Similar results to those described for<br />

hexadecane-oleyl alcohol soils were obtained<br />

for hexadecane-oleic acid soils having the same<br />

ratio <strong>of</strong> nonpolar and polar constituents. These<br />

studies were performed in the absence <strong>of</strong><br />

triethanolarnine to insure that the oleic acid was<br />

in the non-ionized state. Also, the oleic acid was<br />

tagged with 14C-labeled material to allow<br />

measurement <strong>of</strong> its removal as well as that <strong>of</strong><br />

hexadecane. The results for 3/1 cetane-oleic acid<br />

soil are shown in Fig. 51. High levels <strong>of</strong> soil<br />

removal were found over a wide temperature<br />

range from 20ºC to the cloud point <strong>of</strong> the<br />

surfactant. As with the oleyl alcoholcontaining<br />

soils, optimum removal was found between the<br />

PIT and cloud point temperature. Interestingly,<br />

the data show that the removal <strong>of</strong> oleic acid<br />

paralleled that <strong>of</strong> cetane, although at levels<br />

10-20% higher. This behavior results from the<br />

higher ratio <strong>of</strong> oleic acid to hexadecane found in<br />

the intermediate liquid crystalline phase<br />

compared to that in the original soil.<br />

Presumably, preferential removal <strong>of</strong> oleyl<br />

alcohol would have been found in the<br />

hexadecane-oleyl alcohol experiments if the<br />

alcohol had been radioactively tagged.<br />

11. Discussion<br />

As shown above, maximum removal <strong>of</strong> liquid<br />

hydrocarbon soil from synthetic fabrics in non-<br />

Fig. 51. Oil removal for mixed 3/1 n-hexadecaneoleic<br />

acid soil [12]. Reprinted with permission <strong>of</strong> the<br />

American Oil Chemists' Society.


210 C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215<br />

ionic and anionic-non-ionic surfactant systems<br />

occurs near an appropriately defined PIT because<br />

solubilization <strong>of</strong> hydrocarbon into intermediate<br />

microemulsion or liquid crystalline phases<br />

is high and interfacial tensions are low. The low<br />

tensions facilitate <strong>emulsification</strong> <strong>of</strong> the intermediate<br />

phases into the agitated washing bath.<br />

Hydrocarbon soil removal is low at<br />

temperatures significantly below the PIT<br />

because, as is well known, the amount <strong>of</strong> oil that<br />

can be solubilized by microemulsion phases<br />

under these conditions is well below the<br />

solubilization capacity <strong>of</strong> middlephase<br />

microemulsions near the PIT, and interfacial<br />

tensions are significantly higher. Moreover, the<br />

initial rate <strong>of</strong> solubilization is less than that<br />

found near the PIT, as shown in Fig. 24 above.<br />

At temperatures significantly above the PIT,<br />

the explanation for the poor <strong>detergency</strong> is<br />

different. In this case both surfactant and water<br />

diffuse into the oil phase, so that it actually<br />

increases in volume instead <strong>of</strong> being solubilized.<br />

Moreover, extensive spontaneous <strong>emulsification</strong><br />

<strong>of</strong> water in the oil occurs, as seen, for example,<br />

in Fig. 25. Solans and Azemar [19] reported that<br />

polyester-cotton fabric washed at temperatures<br />

above the PIT <strong>of</strong>ten experienced a net gain in<br />

weight, a result consistent with the above<br />

statements. They further observed that<br />

"aggregates", which apparently contained both<br />

oil and surfactant, remained on the fabric<br />

surfaces after washing. They suggested that<br />

additional water drops might be incorporated<br />

into the initial spontaneously formed emulsion<br />

as a result <strong>of</strong> agitation during the washing<br />

process, the result being an emulsion with a<br />

viscosity so high that it is not easily removed.<br />

Indeed, they proposed that the disperse phase<br />

content could become so large that a so-called<br />

high internal phase ratio (HIPR) emulsion could<br />

be formed on the fabric with the drops <strong>of</strong> water<br />

becoming polyhedral and the emulsion<br />

developing a yield stress.<br />

Yang and Rathman [88] confirmed that<br />

increases in weight occurred for polyestercotton<br />

fabrics washed above the PIT, and they<br />

found a surfactant-to-oil ratio <strong>of</strong> about 0.5 in<br />

material extracted from the washed fabric in one<br />

system.<br />

They also found that clean fabric became<br />

soiled if it was washed with soiled fabric above<br />

the PIT. Thus, redeposition <strong>of</strong> soil, probably in<br />

the form <strong>of</strong> an oil-continuous emulsion or<br />

microemulsion, must be considered a significant<br />

factor in the washing process under these<br />

conditions.<br />

In the studies discussed so far, the effects <strong>of</strong><br />

electrolytes on <strong>detergency</strong> performance were not<br />

investigated. Recent results have shown that the<br />

optimum <strong>detergency</strong> temperature and the PIT<br />

are equivalent but lower when salts such as<br />

sodium citrate, a common builder for liquid<br />

detergent formulations, are present in appreciable<br />

amounts in the washing solution [89]. Similar<br />

effects with other salts would be expected.<br />

For the hydrocarbon soils and surfactants <strong>of</strong><br />

interest for <strong>detergency</strong>, the PIT is invariably<br />

above the surfactant cloud point, so that the<br />

washing bath is a dispersion <strong>of</strong> the<br />

surfactant-rich liquid L1 or the lamellar liquid<br />

crystalline phase La in water. In most contacting<br />

experiments near the PIT an intermediate<br />

microemulsion phase was observed near the<br />

surface <strong>of</strong> contact. However, as mentioned<br />

above in connection with the vertical cellcontacting<br />

experiments for C12E4 and n-hexadecane<br />

at 30ºC, no microemulsion phase was<br />

seen when the dispersion <strong>of</strong> the lamellar phase<br />

was sufficiently dilute. In this case, small<br />

particles <strong>of</strong> liquid crystal striking the oil-water<br />

interface apparently dissolved in the oil,<br />

although it is likely that tiny drops <strong>of</strong> microemulsion<br />

quickly formed and dissolved during<br />

the process. However, if only a small quantity <strong>of</strong><br />

oil had been present in this experiment, as in<br />

practical washing situations, it would have<br />

rapidly become saturated with surfactant by<br />

taking up such particles, and subsequent<br />

particles striking the interface would have<br />

produced an intermediate microemulsion phase.<br />

The experiment where the pure lamellar phase<br />

was contacted with oil for the same system at<br />

the same temperature is also <strong>of</strong> interest. Here<br />

considerable oil was solubilized directly by the<br />

liquid crystal although, as the dark region in Fig.<br />

27 and the diffusion path <strong>of</strong> Fig. 28 indicate,<br />

some microemulsion did form as a dispersion<br />

with the liquid crystal at a location away from


C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215 211<br />

the oil-liquid crystal interface. This experiment<br />

demonstrates that the formation <strong>of</strong> an<br />

intermediate phase is not necessary for high<br />

solubilization if sufficient quantities <strong>of</strong> a<br />

surfactant-rich phase are present initially.<br />

Indeed, such solubilization into the lamellar<br />

phase was apparently responsible for the<br />

moderate removal <strong>of</strong> pure triolein from<br />

polyester-cotton fabrics by C 12E 4 at temperatures<br />

between about 25 and 50ºC (Fig. 45). The poor<br />

<strong>detergency</strong> performance at high temperatures in<br />

this case is, as with hydrocarbon soils, the result<br />

<strong>of</strong> conversion <strong>of</strong> the oil phase into a water-in-oil<br />

microemulsion accompanied by spontaneous<br />

<strong>emulsification</strong> <strong>of</strong> the water there.<br />

If two different types <strong>of</strong> hydrocarbon soils are<br />

present, a useful strategy with a single pure<br />

surfactant would be to wash near the higher <strong>of</strong><br />

the two PIT values and rinse near the lower.<br />

Experiments have shown that substantial soil<br />

removal does take place when washing occurs<br />

above the PIT and rinsing near the PIT [83,88],<br />

as would be the case in this example for the soil<br />

with the lower PIT. One reason this method<br />

works is that substantial surfactant remains on<br />

fabric washed above the PIT, as mentioned<br />

above, much <strong>of</strong> it probably dissolved in the<br />

unremoved soil. If there are multiple<br />

hydrocarbon soils and if a commercial<br />

surfactant or surfactant mixture is used, washing<br />

should again start at the highest PIT. In this<br />

case, however, it would be best to decrease<br />

temperature with time during the latter stages <strong>of</strong><br />

washing or during rinsing, so that the PIT values<br />

for the various soils would all be reached at<br />

some time during the process. Note that during<br />

rinsing the PIT values will not be the same as<br />

those for the initial washing bath owing to<br />

differences in surfactant composition and<br />

surfactant-to-oil ratio during washing and<br />

rinsing. The concept <strong>of</strong> designing a process so<br />

that the PIT is achieved at some time as the<br />

system is gradually made more hydrophilic is<br />

related to certain strategies that have been<br />

proposed for enhanced oil recovery processes,<br />

e.g. variation <strong>of</strong> the injected salinity [90]. The<br />

concept <strong>of</strong> washing at high and rinsing at low<br />

temperatures to improve <strong>detergency</strong> was<br />

proposed by Rubingh and Stevens [91] although<br />

their explanation did not involve the PIT and the<br />

formation <strong>of</strong> intermediate phases.<br />

Achieving excellent <strong>detergency</strong> for pure<br />

liquid triglyceride soils and synthetic fabrics is<br />

difficult because the large triglyceride molecules<br />

are not readily solubilized. Indeed, substantial<br />

solubilization apparently requires conditions<br />

where an intermediate D phase forms. With<br />

straight-chain ethoxylated alcohols, this seems<br />

not to be possible except at undesirably high<br />

temperatures, e.g. about 65ºC with C12E5 and<br />

triolein (Fig. 10). Adding a short-chain alcohol<br />

does seem to make the surfactant films more<br />

disordered and thereby promote D phase<br />

formation at slightly lower temperatures (Fig.<br />

11), but the relatively high solubility <strong>of</strong> the<br />

alcohols in water makes them ineffective for this<br />

purpose when the dilute surfactant solutions <strong>of</strong><br />

interest for <strong>detergency</strong> are used. As indicated<br />

previously, preliminary results with secondary<br />

alcohol ethoxylate surfactants, whose<br />

double-tailed structure with different chain<br />

lengths also creates disordered films, look more<br />

promising as a means <strong>of</strong> forming the D phase at<br />

temperatures existing during warm and cold<br />

water washing [55]. Further study <strong>of</strong> these<br />

systems is in progress. Of course, the D phase<br />

could also be formed by using washing baths<br />

containing solubilized hydrocarbon (see Fig.<br />

12), but this scheme seems less attractive than<br />

using the secondary alcohol ethoxylate<br />

surfactants.<br />

Results presented above demonstrate that<br />

liquid mixtures <strong>of</strong> hydrocarbons and long-chain<br />

alcohols or undissociated fatty acids can be<br />

removed effectivcly at temperatures below the<br />

cloud point <strong>of</strong> the surfactant but equal to or<br />

above the PIT <strong>of</strong> a system whose excess oil<br />

phase coexisting with the microemulsion has the<br />

same composition as the soil. Under these<br />

conditions, the soil takes up surfactant and water<br />

from the washing bath, as shown in the oil<br />

drop-contacting experiment <strong>of</strong> Fig. 35, the<br />

results being the lowering <strong>of</strong> the interfacial<br />

tension between the soil and surfactant solution<br />

and the eventual formation <strong>of</strong> an intermediate<br />

lamellar liquid crystalline phase as filaments or<br />

myelinic figures. These, presumably, are broken<br />

<strong>of</strong>f and dispersed into the washing bath as a


212 C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215<br />

result <strong>of</strong> agitation. Note that Fig. 48 and 49<br />

show no significant change in <strong>detergency</strong><br />

between 40 and 60ºC for C 12E 7 and the soil<br />

containing 75% cetane even though the<br />

contacting experiments mentioned above for this<br />

system indicated that an intermediate D' phase<br />

formed first and only later the liquid crystal<br />

formed. Below the PIT no intermediate phase<br />

was seen in all the systems, the drops were<br />

slowly solubilized into micelles in the washing<br />

bath, and soil removal was lower.<br />

For all the systems discussed above for which<br />

both videomicroscopy observations and soil<br />

removal measurements have been performed,<br />

intermediate phase formation and low interfacial<br />

tensions leading to good <strong>detergency</strong> occurred<br />

when the hydrophilic and lipophilic properties<br />

<strong>of</strong> the surfactant or surfactant -alcohol films <strong>of</strong><br />

the intermediate phase were approximately<br />

balanced. This balance may be present initially,<br />

as in the experiments with hydrocarbon soils<br />

near the PIT, or it may develop during washing<br />

as a result <strong>of</strong> mass transfer, as in the<br />

experiments with mixtures <strong>of</strong> hydrocarbons and<br />

long-chain alcohols. Note that if mixed<br />

surfactants and lower surfactant-to-oil ratios had<br />

been used in the hydrocarbon washing<br />

experiments, mass transfer effects would most<br />

likely have been important there as well, i.e. the<br />

system could have moved closer to or further<br />

from the PIT over time. A similar conclusion on<br />

the need for balance was reached by Malmsten<br />

and Lindman [92], although without the proviso<br />

that a system not balanced initially may achieve<br />

balance during washing. It should also be<br />

recognized that, even when balance is achieved,<br />

the intermediate phase formed must have the<br />

capability <strong>of</strong> solubilizing considerable soil, a<br />

property clearly lacking in some <strong>of</strong> the<br />

non-ionic surfactant-triolein systems discussed<br />

above.<br />

If a system contains soils with several<br />

different mixtures <strong>of</strong> hydrocarbons and<br />

long-chain polar compounds on the fabrics, an<br />

effective strategy would seem to be to choose a<br />

washing temperature and surfactant so that the<br />

temperature is initially above the PIT <strong>of</strong> all soil<br />

compositions. The surfactant must be<br />

hydrophilic enough to be somewhat below its<br />

cloud point but lipophilic enough to meet the<br />

above PIT criterion. Mass transfer during the<br />

washing process will bring the various soils to a<br />

balanced condition (though at different times) at<br />

which they can be removed by some<br />

combination <strong>of</strong> intermediate phase formation<br />

and <strong>emulsification</strong> promoted by low interfacial<br />

tensions. This strategy is, <strong>of</strong> course, similar to<br />

that suggested above for multiple hydrocarbon<br />

soils except that no variation <strong>of</strong> temperature is<br />

required.<br />

For hydrocarbon soils the phase behavior <strong>of</strong><br />

the surfactant-water mixture in the initial<br />

washing bath did not seem to have a significant<br />

influence on <strong>detergency</strong>. Maximum soil removal<br />

occurred at the system PIT whether the washing<br />

bath consisted <strong>of</strong> a dispersion <strong>of</strong> the L 1 phase or<br />

<strong>of</strong> the Lα phase. In contrast, soil removal<br />

dropped sharply when the cloud point<br />

temperature <strong>of</strong> the surfactant solution was<br />

reached for the mixed hydrocarbon-alcohol soils<br />

above their PITs. It is likely that water formed<br />

as in the intermediate phase for temperatures<br />

above the cloud point temperature and hindered<br />

the mass transfer necessary to achieve the<br />

balanced condition discussed above. That is, the<br />

diffusion path is most likely similar to that <strong>of</strong><br />

Fig. 26 except that the surfactant-rich phase is<br />

L 1 instead <strong>of</strong> La.<br />

In the oil drop-contacting experiments it was<br />

not unusual for 20 min or even more to elapse<br />

before liquid crystal formation began. Such long<br />

times are unsatisfactory for washing processes.<br />

It should be kept in mind, however, that the<br />

contacting experiments were deliberately<br />

conducted with no imposed mixing to facilitate<br />

the interpretation <strong>of</strong> the observations in terms <strong>of</strong><br />

diffusion theory. In actual washing situations,<br />

mixing would greatly speed up mass transfer,<br />

and liquid crystal could be expected to form<br />

after much shorter times. Further discussion,<br />

including some quantitative estimates <strong>of</strong> the<br />

enhancement in mass transfer rates, may be<br />

found in the original papers [60,74]. When<br />

mixing occurs, the times required for liquid<br />

crystal formation were found to be reasonable<br />

for practical processes, a conclusion confirmed<br />

by the excellent soil removal obtained in the<br />

washing experiments described above.


C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215 213<br />

In the preceding section, it was mentioned<br />

that the time required for removal <strong>of</strong> mixed<br />

hydrocarbon-long-chain alcohol soils increases<br />

when the initial state <strong>of</strong> the system is further<br />

above the PIT. This result is consistent with the<br />

intuitive expectation that surfactant would have<br />

to diffuse into the oil for a longer time before<br />

liquid crystal formation occurred for systems<br />

further above the PIT as well as with the<br />

quantitative values <strong>of</strong> the parameter K s in Eq.<br />

(6) found for soils with different compositions.<br />

It is also noteworthy that in some <strong>of</strong> the<br />

contacting experiments, convection arose<br />

spontaneously near the interface, an effect that<br />

would also speed up mass transfer. The<br />

Marangoni flow mentioned in connection with<br />

Fig. 19 for the C12E5-water-n-tetradecane<br />

system at 20ºC is one example, although it<br />

occurred for a situation when soil removal is<br />

minimal owing to a low solubilization capacity<br />

<strong>of</strong> the surfactant solution. Perhaps more relevant<br />

to <strong>detergency</strong>, vigorous convection was<br />

sometimes seen during the early stages <strong>of</strong><br />

formation <strong>of</strong> a non-wetting intermediate phase,<br />

i.e. one that forms preferentially as lenses rather<br />

than as a continuous layer. Intermediate phase<br />

formation in the C 12E 5-water-n-tetradecane<br />

system at temperatures just below and above the<br />

cloud point is an example. If non-uniformities in<br />

intermediate phase thickness develop<br />

immediately following the initial contact when<br />

the phase is a still a thin liquid film, disjoining<br />

pressure effects will promote flow from thin to<br />

thick portions <strong>of</strong> the film, thus increasing the<br />

discrepancies in thickness. Diffusion processes<br />

will continue to form more <strong>of</strong> the intermediate<br />

phase in the thin regions, and disjoining<br />

pressure gradients will continue to drive the<br />

newly formed material to thicker regions.<br />

No soil removal experiments are currently<br />

available for the systems discussed above<br />

containing mixed triolein-oleyl alcohol soils and<br />

a pure nonionic surfactant. The contacting<br />

experiments indicate, however, that the first<br />

intermediate phase formed is D', which<br />

solubilizes little triolein although readily<br />

incorporating alcohol. It may be that some <strong>of</strong> the<br />

same ideas suggested above to improve<br />

<strong>detergency</strong> with pure triolein, e.g. use <strong>of</strong><br />

secondary alcohol ethoxylate surfactants, will be<br />

needed to achieve high degrees <strong>of</strong> removal <strong>of</strong><br />

the triolein portion <strong>of</strong> these mixed soils. Of<br />

course, liquid triglyceride soils normally contain<br />

some fatty acids formed by triglyceride<br />

hydrolysis. In addition, lipase enzymes are<br />

incorporated into some detergent formulations<br />

to promote the breakdown <strong>of</strong> triglycerides into<br />

monoglycerides, diglycerides and fatty acids.<br />

Increasing the pH can convert these acids to<br />

soaps, which could well improve <strong>detergency</strong>.<br />

This behavior is currently being studied by the<br />

contacting techniques discussed above.<br />

12. Conclusions<br />

<strong>Solubilization</strong>-<strong>emulsification</strong> is a key<br />

mechanism in the removal <strong>of</strong> oily liquid soils<br />

from polyester and polyester-cotton fabrics.<br />

Systematic studies, discussed above, utilizing<br />

several model soils indicate that solubilization<br />

into an intermediate phase formed during<br />

washing is generally much more extensive and<br />

rapid than solubilization into a micellar solution.<br />

Accordingly, knowledge <strong>of</strong> the phase behavior<br />

<strong>of</strong> surfactant-soil-water systems is needed to<br />

make a rational choice <strong>of</strong> optimum surfactant<br />

compositions and washing conditions. With<br />

hydrocarbon soils, for instance, washing at<br />

temperatures near the PIT is best. The<br />

intermediate phase is a microemulsion although<br />

the lamellar liquid crystalline phase may form as<br />

well if it is not already present in the initial<br />

washing bath. For commercial non-ionic<br />

surfactants and their mixtures with anionic<br />

surfactants, the appropriate PIT in the usual case<br />

<strong>of</strong> washing with a high surfactantto-oil ratio is<br />

that for which the composition <strong>of</strong> the surfactant<br />

films in the middle-phase microemulsion is that<br />

<strong>of</strong> the surfactant mixture in the initial washing<br />

bath.<br />

For mixed soils containing hydrocarbons and<br />

long-chain alcohols or fatty acids, good soil<br />

removal is achieved at temperatures above the<br />

PIT but below the cloud point <strong>of</strong> the surfactant.<br />

In this case the PIT is evaluated at high<br />

surfactant-to-oil ratios as above with the<br />

additional condition that the excess oil phase in<br />

equilibrium with the microemulsion has the


214 C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215<br />

initial soil composition. The lamellar liquid<br />

crystal is the intermediate phase. It begins to<br />

grow as myelinic figures at a time after initial<br />

contact between surfactant solution and soil<br />

which can be predicted in simple cases using<br />

diffusion theory and certain limited information<br />

on system phase behavior.<br />

In both these situations, the intermediate<br />

phase, or phases, start(s) to grow when the films<br />

<strong>of</strong> surfactant and long-chain polar compound, if<br />

present, are roughly balanced with respect to<br />

hydrophilic and lipophilic properties. This<br />

balance, which may be present initially or may<br />

occur at some later time during the washing<br />

process as a result <strong>of</strong> mass transfer, facilitates<br />

<strong>emulsification</strong> because interfacial tensions,<br />

including those involving intermediate phases,<br />

are low.<br />

Long-chain liquid triglycerides are less<br />

readily solubilized than the hydrocarbons <strong>of</strong><br />

interest in <strong>detergency</strong> owing to their higher<br />

molecular volumes. The best prospect for<br />

removing such triglycerides seems to be to<br />

establish conditions under which an<br />

intermediate D phase will develop during<br />

washing. This phase, which is closely related to<br />

the middle-phase microemulsion, can be<br />

expected to form if sufficient hydrocarbon is<br />

present with the triglyceride in the initial soil.<br />

Otherwise, its formation can be promoted by<br />

using mixtures <strong>of</strong> surfactants with varying<br />

hydrocarbon chain lengths, so that surfactant<br />

films with rather disordered packing in the chain<br />

region are present.<br />

Acknowledgments<br />

The research at Rice University discussed in<br />

this paper was supported by the National<br />

Science Foundation, Shell Development<br />

Company, and the Clorox Corporation.<br />

References<br />

1 J. Falbe (Ed.), Surfactants in Consumer Products:<br />

Theory, Technology and Application, Springer-Verlag,<br />

New York, 1986.<br />

2 S.K. Obendorf and J.J. Webb, Text. Res. J.,57 (1984)<br />

557<br />

3 T. Fort, Jr., H.R. Billica and T.H. Grindstaff, J. Am. Oil<br />

Chem. Soc., 45 (1968) 354.<br />

4 K.H. Raney and J.H. Rask, Proc. 3rd CESIO World<br />

Surfactants Congress, London, 1992, Vol. D, p. 101.<br />

5 H. Malmos, Chem. Ind. London, March (1990) 183.<br />

6 E. Kissa, H.M. Dohner, W.R. Gibson and D. Strickman,<br />

J. Am. Oil Chem. Soc., 68 (1991) 532.<br />

7 E. Smulders and P. Krings, Chem. Ind. London, March<br />

(1990)160.<br />

8 K.W. Dillan, E.D. Goddard and D.A. McKenzie, J. Am.<br />

Oil Chem. Soc., 56 (1979) 59.<br />

9 K.W. Dillan, E.D. Goddard and D.A. McKenzie, J. Am.<br />

Oil Chem. Soc., 57 (1980) 230.<br />

10 O.S. Cha, J. Korean Soc. Text.Eng.Chem., 23 (1986) 18<br />

11 H.S. Kielman and P.J.F. van Steen, Surface Active<br />

Agents, Soc. <strong>of</strong> Chem. Ind., London, 1979, p. 191<br />

12 K.H. Raney and H.L. Benson, J. Am. Oil Chem. Soc.,<br />

67 (1990)722.<br />

13 T. Fujii, Hyomen, 18 (1980) 88.<br />

14 D.G. Stevenson, in K. Durham (Ed.), Surface Activity<br />

and Detergency, Macmillan, London, 1961, Chapter 6,<br />

15 B.J. Carroll, Colloids Surfaces A: Physicochem. Eng.<br />

Aspects, 74 (1993) 131-167.<br />

16 A.S.C. Lawrence, Nature, 183 (1959) 1491.<br />

17 B.J. Carroll, J. Colloid Interface Sci., 79 (1981) 126.<br />

B.G.C. O'Rourke, A.J.I. Ward and B.J. Carroll, J.<br />

Pharm. Pharmacol., 3 (1987) 865.<br />

18 K.H. Raney, W.J. Benton and C.A. Miller, J. Colloid<br />

Interface Sci., 117 (1987) 282.<br />

19 C. Solans and N. Azemar, in S. Friberg and B. Lindman<br />

(Eds.), Organized Solutions, Marcel Dekker, New<br />

York, 1992, Chapter 19.<br />

20 K.H. Raney, W.J. Benton and C.A. Miller, in D.O.<br />

Shah (Ed.), Macro- and Microemulsions, ACS Symp.<br />

Ser. 272, American Chemical Society, Washington,<br />

DC, 1985, p. 193.<br />

21 R. Strey, R. Schomacker, D. Roux, F. Nallet and U.<br />

Olsson, J, Chem. Soc., Faraday Trans., 86 (1990) 2253.<br />

22 F. Schambil and M.J. Schwuger, Colloid Polym. Sci.,<br />

265 (1987) 1009.<br />

23 (a) D.J. Mitchell, G.J.T. Tiddy, L. Waring, T. Bostock<br />

and M.P. McDonald, J. Chem. Soc., Faraday Trans. 1,<br />

79 (1983)975.<br />

(b) J.W. McBain, quoted in H.R. Kruyt (Ed.), Colloid<br />

Science, Vol. 2, Elsevier, Amsterdam, 1949, p. 701.<br />

24 F. Mori, J.C. Lim and C.A. Miller, Prog. Colloid<br />

Polym. Sci., 82 (1990) 114.<br />

25 K. Fontell, in K.L. Mittal (Ed.), Colloidal Dispersions<br />

and Micellar Behavior, ACS Symp. Ser. 9, American<br />

Chemical Society, Washington, DC, 1975, p. 270.<br />

26 O. Ghosh and C.A. Miller, J. Phys. Chem., 91 (1987)<br />

4528.<br />

27 C.A. Miller, O. Ghosh and W.J. Benton, Colloids<br />

Surfaces.19 (1986) 197.<br />

28 J. Israelachvili, D. Mitchell and B. Ninham, J. Chem.<br />

Soc. Faraday Trans. 2, 72 (1976) 1525.<br />

29 R.A. Mackay, in M.J. Schick (Ed.), Nonionic<br />

Surfactants Physical Chemistry, Marcel Dekker, New<br />

York, 1987. Chapter 6.<br />

30 K.H. Raney and C.A. Miller, J. Colloid Interface Sci.,<br />

119 (1987) 539.<br />

31 W.N. Maclay, J. Colloid Sci., 11 (1956) 272.


C.A. Miller and K.H. Raney/Colloids Surfaces A: Physicochem. Eng. Aspects 74 (1993) 169-215 215<br />

32 W.J. Benton and C.A. Miller, unpublished results, 1987<br />

33 C.A. Miller, M. Gradzielski, H.H. H<strong>of</strong>fmann, U.<br />

Kramer and C. Thunig, Colloid Polym. Sci.,83 (1990) 29<br />

34 D.N. Rubingh, in D.N. Rubingh and P.M. Holland<br />

(Eds.), Cationic Surfactants Physical Chemistry, Marcel<br />

Dekker, New York, 1991, Chapter 10.<br />

35 E.W. Kaler, A.K. Murthy, B.E. Rodriguez and J.N.<br />

Zasadzinski, Science, 245 (1989) 1371.<br />

36 K. Shinoda and S. Friberg, Emulsions and<br />

<strong>Solubilization</strong>, Wiley, New York, 1986.<br />

37 R.L. Reed and R.N. Healy, in D.O. Shah and R.S.<br />

Schechter (Eds.), Improved Oil Recovery by Surfactant<br />

and Polymer Flooding, Academic Press, New York,<br />

1977, p. 383 ff.<br />

38 H. Kunieda and K. Shinoda, J. Colloid Interface Sci.,<br />

107 (1985) 107.<br />

39 H. Kunieda, K. Hanno, S. Yamaguchi and K. Shinoda,<br />

J. Colloid Interface Sci., 107 (1985) 129.<br />

40 R. Aveyard, B. Binks, P. Cooper and P. Fletcher, Adv.<br />

Colloid Interface Sci., 33 (1990) 59.<br />

41 S. Mukherjee, C.A. Miller and T. Fort, Jr., J. Colloid<br />

Interface Sci., 91 (1983) 223.<br />

42 D.G. Peck, R.S. Schechter and K.P. Johnston, J. Phys.<br />

Chem., 95 (1991) 9541.<br />

43 R.L. Reed and M. Puerto, Soc.Pet.Eng. J.,23 (1983) 669<br />

44 Y. Barakat, L. Fortney, C. LaLanne-Cassou, R.<br />

Schechter, W. Wade, U. Weerasooriya and S. Yiv, Soc.<br />

Pet. Eng. J., 23 (1983) 913.<br />

45 JL. Hackett and C.A. Miller, SPE Reservoir Eng., 3<br />

(1988) 791.<br />

46 0. Ghosh and C.A. Miller, J. Colloid Interface Sci., 116<br />

(1987) 593.<br />

47 H. Kunieda and K. Shinoda, J. Dispersion Sci.<br />

Technol., 3 (1982) 233.<br />

48 F. Mori, J.C. Lim, O.G. Raney, C.M. Elsik and C.A.<br />

Miller, Colloids Surfaces, 40 (1989) 323.<br />

49 H. Kunieda, H. Asaoka and K. Shinoda, J. Phys. Chem.,<br />

92 (1988) 185.<br />

50 H. Kunieda and K. Haishima, J. Colloid Interface Sci.,<br />

140 (1990)383.<br />

51 B.P. Binks, paper presented at 9th Int. Symp. on<br />

Surfactants in Solution, Varna, Bulgaria, 1992.<br />

52 J.C. Lim, C.A. Miller and C. Yang, Colloids Surfaces,<br />

66 (1992)45.<br />

53 S. Friberg and L. Rydhag, J. Am. Oil Chem. Soc., 48<br />

(1971) 113.<br />

54 J. Alander and T. Warnheim, J. Am. Oil Chem. Soc., 66<br />

(1989) 1661.<br />

55 T.Tungsubutra and C.A. Miller, unpublished results,<br />

1992<br />

56 P. Ekwall, in G. Brown (Ed.), Adv. Liq. Cryst., 1<br />

(1975) 1.<br />

57 H. Kunieda and K. Nakamura, J. Phys. Chem., 95<br />

(1991) 8861.<br />

58 W.J. Benton and C.A. Miller, in K.L. Mittal and B.<br />

Lindman (Eds.), Surfactants in Solution, Vol. 1, Plenum<br />

Press, New York, 1984, p. 205ff.<br />

59 H. Kunieda and A. Miyajima, J. Colloid Interface Sci.,<br />

129 (1989)554.<br />

60 J.C. Lim and C.A. Miller, Langmuir, 7 (1991) 2021.<br />

61 T. Tungsubutra and C.A. Miller, in S. Friberg and B.<br />

Lindman (Eds.), Organized Solutions, Marcel Dekker,<br />

New York, 1992, Chapter 7.<br />

62 J.S. Kirkaldy and L.C. Brown, Can. Metall. Q., 2<br />

(1963) 89.<br />

63 K.J. Ruschak and C.A. Miller, Ind. Eng. Chem.<br />

Fundam., 11 (1972) 534.<br />

64 K.H. Raney and C.A. Miller, AlChE J., 33 (1987) 1791.<br />

65 K.H. Raney, Ph.D. Thesis, Rice University, Houston,<br />

TX, 1985.<br />

66 C.A. Miller, Colloids Surfaces, 29 (1988) 89.<br />

67 J.T. Davies and E.K. Rideal, Interfacial Phenomena,<br />

2nd edn. New York, Academic Press, 1963, Chapter 8.<br />

68 W.J. Benton, C.A. Miller and T. Fort, Jr., J. Dispersion<br />

Sci. Technol., 3 (1982) 1.<br />

69 A. Kotidou and J. Shaeiwitz, J. Colloid Interface Sci.,<br />

92 (1983)265.<br />

70 S. Friberg, Z. Ma and P. Neogi, in D.H. Smith (Ed.),<br />

Surfactant-Based Mobility Control, ACS Symp. Ser.<br />

373, American Chemical Society, Washington, DC, p.<br />

108ff.<br />

71 Z. Ma, S. Friberg and P. Neogi, AlChE J., 35 (1989)<br />

1678.<br />

72 W.J. Benton, K.H. Raney and C.A. Miller, J. Colloid<br />

Interface Sci., 110 (1986) 363.<br />

73 J.C. Lim and C.A. Miller, Prog. Colloid Polym. Sci., 83<br />

(1990)29.<br />

74 J.C. Lim and C.A. Miller, in K.L. Mittal and D.O. Shah<br />

(Eds.), Surfactants in Solution, Vol. 11, Plenum Press,<br />

New York, 199 1, p. 49 1 ff.<br />

75 N. Moucharafieh, S. Friberg and D. Larsen, Mol. Cryst.<br />

Liq. Cryst., 53 (1979) 189.<br />

76 J.C. Lim and C.A. Miller, unpublished results, 1990.<br />

77 W.S. Gilman, J. Am. Oil Chem. Soc., 66 (1989) 6.<br />

78 E. Kissa, in W.G. Cutler and E. Kissa (Eds.),<br />

Detergency: Theory and Technology, Marcel Dekker,<br />

New York, 1987, Chapter 1.<br />

79 W.T. Shebs, in W.G. Cutler and E. Kissa (Eds.),<br />

Detergency: Theory and Technology, Marcel Dekker,<br />

New York, 1987, Chapter 3.<br />

80 N.E. Prieto, J. Am. Oil Chem. Soc., 66 (1989) 10.<br />

81 J.C. Illman, T.B. Albin and H. Stupel, J. Am. Oil<br />

Chem. Soc., 49 (1972) 217.<br />

82 H.L. Benson and C.L. Merrill, Chem. Times Trends, 8<br />

(1985)29.<br />

83 C. Solans, N. Azemar, J. Parra and J. Calbet, Proc.<br />

CESIO 2nd World Surfactants Congress, Paris, 1988,<br />

Vol. 2, p. 421 ff.<br />

84 K.H. Raney, unpublished results, 1988.<br />

85 K.H. Raney, paper presented at Annual Meeting <strong>of</strong> Am.<br />

Oil Chem. Soc., Baltimore, 1990.<br />

86 R. Bercovici and H. Krussman, Tenside Surf. Deterg.,<br />

27 (1990)8.<br />

87 K.H. Raney, J. Am. Oil Chem. Soc., 68 (1991) 525.<br />

88 C. Yang and J. Rathman, personal communication.<br />

89 N. Azemar, 1. Carrera and C. Solans, Proc. CESIO 3rd<br />

World Surfactants Congress, London, 1992, p. 55ff.<br />

90 R.C. Nelson, Soc. Pet. Eng. J., 22 (1982) 259.<br />

91 D. Rubingh and T. Stevens, U.S. Patent 4 248729, 1981<br />

92 M. Malmsten and B. Lindman, Langmuir, 5 (1989) 1105

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!