| dc.description.abstracteng | In this work, I studied the impact of surfactants on interfacial phenomena out of equilibrium.
I investigated three different effects, (1) the stabilisation of interfaces upon
surfactant adsorption (adsorption kinetics), (2) the interaction of the surfactant with the
interior of the droplet (surfactant synthesis) and (3) the interaction of the surfactant with
the exterior of the droplet (interfacial flows).
Adsorption kinetics of surfactants are of relevance for the stabilisation of interfaces against
coalescence and transport. I investigated the adsorption kinetics of the acidic surfactant
KrytoxFSL to the interface of water-in-fluorinated-oil emulsions using droplet-based microfluidics.
The surfactant, which is soluble inside the fluorous continuous phase, adsorbs
to the interface of the aqueous droplet after droplet production. Upon adsorption the surfactant
deprotonates causing a change in the pH inside the aqueous phase containing a
buffer. I measured this pH change with a pH dependent dye using a fluorescence setup.
The age of the droplet can be related to the distance from the production unit on chip.
The pH change provides information on the adsorption kinetics of the surfactant.
The pH change is faster and more pronounced the higher the initial surfactant concentration inside the fluorous phase. This relation was determined for surfactant concentrations
above the CMC (4 µmol/L). For the highest concentration investigated in this study
(C = 0.69 mM, 0.07 w% in oil), the pH change induced through the adsorption is even
larger than 1 and is obtained within the order of seconds. The equilibrium change in
proton concentration inside the aqueous phase equals the surfactant concentration that
was initially inside the fluorous phase. The pH change results from the adsorption of the
acidic surfactant to the interface, its deprotonation, the subsequent formation of the salt
with a counter ion from the aqueous phase (sodium ion) and the desorption of this surfactant
from the interface. Thus, the transfer of protons and other ions across the interface is
related to their partitioning between the two phases. The adsorption / desorption process
is quantitatively linked with the kinetics of partitioning.
At small length scales, the adsorption process of the surfactant to the interface is slower
than the dffusion of surfactant molecules towards the interface, causing the adsorption
process to be rate limiting. Thus, the kinetics of the adsorption process of surfactants
are accessible using droplet-based microfluidic experiments. To measure the adsorption
kinetics, I used the pH change as an indicator for the amount of adsorbed acidic surfactant
to the initially empty, and at a later stage partially and then fully, covered interface. Describing
the surfactant adsorption with a simple first order Langmuir model fails, because
the rate of adsorption does not scale linearly, but rather quadratically with the surfactant
concentration. The adsorption rate constant can be obtained applying the second
order adsorption model developed here with the proviso that the equilibrium surfactant
coverage is independently determined through standard measurements. Additionally,
the adsorption model gives information on the timescale of the adsorption process. The
pH change and the partitioning take place with a timescale of the order of one second.
Therefore, phase partitioning across the interface is always in equilibrium for any process
occurring at larger timescales. At smaller timescales the dynamics of surfactants at the
interface have to be taken into account. The timescale of partitioning (and pH change) increases with the radius of the droplet and decreases with the typical speed, which depends on the adsorption constant, the concentration of the
surfactant and the free sites at the interface. The timescale of partitioning is at
least one or two orders of magnitude larger than the timescale for the stabilisation of the
interface obtained from coalescence experiments. Combining these results
it is found that the stabilisation of two interfaces against coalescence is only possible for
surfactant concentrations close to or larger than the CMC.
The maximum interfacial coverage is accessible from standard bulk interfacial tension
measurements by applying a model for the adsorption kinetics. The Langmuir and the
second order adsorption model provide the same value for the maximum interfacial coverage (8 µmol/m<sup>2</sup>). Hence, bulk interfacial tension techniques cannot distinguish
between different adsorption isotherms which is in contrast to the microfluidic method
developed in this study.
As demonstrated, the pH change is due to an acidic surfactant (Krytox) adsorbing to the
interface at any time during the experiment. Small amounts of residual acid in another
surfactant lead to a pH change of the droplet. Accordingly, acid free surfactants are crucial for the use in emulsion
stabilisation. The typical surfactant used in droplet-based microfluidics is usually synthesised
from such an acidic surfactant. Hence, a high purity of the synthesised
surfactant is required. The product quality depends on the residual acidic surfactant.
Additionally, this impurity (acidic surfactant) needs to be quantified for a reliable use of
the synthesised surfactants for the stabilisation of emulsions. In this work, I improved the
synthesis showing the crucial role of inert atmospheric conditions. I characterised the
purity of the synthesised surfactant using infrared spectroscopy and partitioning experiments
between an aqueous phase and a fluorous phase. The partitioning experiments
are bulk measurements and determine the amount of water soluble dye (Rhodamine 6G)
extracted towards a fluorous phase in which the surfactant is dissolved. The amount
of extracted dye correlates directly with the amount of residual carboxylic acid (acidic
surfactant) present inside the surfactant. I found that my synthesised surfactants show
a partitioning corresponding to less than 1 w% contamination of carboxylic acid. The
purity of the final product is therefore of the order of 99 w%. With the improvement of
the synthesis and the possibility to characterise the amount of residual carboxylic acid, a
more reliable production of surfactants for emulsion stabilisation becomes feasible.
Adsorption of surfactants to interfaces influence the behaviour of droplets in emulsions.
Out-of equilibrium dynamics at interfaces can lead to the Marangoni effect, which couples
inhomogeneities of the surfactant distribution at the interface to interfacial flows and can
lead to self-propulsion of droplets. The self-propulsion of droplets is directly linked to the
dynamics of surfactants at interfaces and thus to the adsorption kinetics of surfactants.
To obtain further insight into the relation between self-propulsion and surfactant adsorption,
I studied interfacial properties of aqueous droplets in a squalane continuous phase.
These droplets show biomimetic behaviours (ongoing experiments) such as swimming,
deformation (ameoba-like), division or microemulsification of the droplet or formation
of a shell around the droplet. These different behaviours depend on the concentration of
the solutes inside the aqueous phase and on the concentrations of the
surfactant oleic acid inside the oil phase. In this system, the movement of the droplet is
likely related to the pH of the droplet and the pKa of the oleic acid. The
characterisation of the temporal evolution of the interfacial tension and its dependence
on the concentration of the oleic acid, including the determination of the CMC, is of great
relevance for the understanding of these behaviours. These measurements are challenging,
due to the impurity present in the oil from the supplier. Purifications upon filtering
over Celite® or Alumina do not remove all impurities. Therefore, the determination of
the CMC, 40 µmol/L, is only a rough estimate on the preliminary data.
In additional experiments, I developed a microfluidic device to study the evolution of the
droplet velocity upon surfactant adsorption. Due to height variations of the microfluidic
channels, no final conclusion on the Marangoni effect was obtained so far.
These two ongoing experiments are promising to gain insight into the interplay of the
interfacial flow and the adsorption kinetics of surfactants. | de |