Modeling the dynamics of structuration of a porous
organic membrane - application to water treatment
The porous organic
membranes are developed by a demixing system polymer / solvent
(or polymer / solvent / non-solvent) inducing the formation of a phase
rich in polymer (membrane) around a phase very dilute in polymer (the
removal of solvent). The basic mechanisms that control the formation of
pores are still only partially understood and rarely perfectly
controlled. On an industrial scale, membrane manufacturing is mostly
done through trial and error loops, the operating parameters being
adjusted gradually and empirically. The formalization of the mechanisms
through systemic models is even less developed, although it allows to
slide towards a more rational and predictive of the development.
The aim of the thesis is to improve understanding of the mechanisms of
phase separation by spinodal decomposition, in order to better control
the formation of porous organic membranes. A modeling approach will be
coupled with an experimental approach to complete this project.
Beyond the fundamental nature of the project, it fits into a broader
context of wastewater reuse (irrigation, watering public ...) that
requires the use of innovative membranes for controlled selectivity:
the retention of small organic molecules -resistant membranes requires
that the distribution of pore sizes is perfectly controlled.
Phase separation by spinodal decomposition phenomenon is described by
the Cahn-Hilliard equation [1-4]. Although its expression is relatively
simple, its resolution is complex because it incorporates many
parameters. To simplify the modeling, the first phase of work will be
to choose a process of phase separation and a relatively simple system
polymer / solvent model. In this context, the TIPS process-UCST (Upper
Critical Solution Temperature) will be retained because it was the
subject of many studies and scientific literature is abundant about it
. It is based on a phase separation induced by a decrease in
temperature and allows to work with a binary polymer / solvent. The
results of numerical simulation will be compared with experimental
measurements by light scattering and / or light microscopy according to
the size of structures.
A second step, given the experimental results, refine the original
model proposed by Cahn and Hilliard. Two improvements will be tested:
the first will include a more complete description of diffusion
phenomena considering the fact that the representation of Flory-Huggins
free energy of polymer-solvent mixture is often inaccurate, especially
for water soluble polymers . The second enhancement is to describe more
accurately the convective and the couplings involved in the
viscoelastic phase separation through the velocity field . The
numerical simulation model of phase separation will be done via the
computer code COMSOL Multiphysics ®, which allows the resolution of
partial differential equations by the finite element method. Model
validation will be done through studies of light scattering at small
angles that can record the time evolution of the structure factor, to
ultimately obtain information on the evolution of the characteristic
size of the structure that separates . When the size becomes too
large for the light scattering to be possible, a study by light
microscopy and image processing will be undertaken.
J.W. and J. E. Hilliard, Free energy of nonuniform system. I.
free energy, J. Chem. Phy. 28 (1957) 795.
K-W. D., P. K. Chanb, X. Feng, Morphology development and
of the phase-separated structure resulting from the thermal-induced
separation phenomenon in polymer solutions under a temperature
Eng. Sci. 59 (2004) 1491
B., A. C. Powell, Phase field simulations of early stage structure
formation during immersion precipitation of polymeric membranes in 2D
and 3D, J.
Membr. Sci. 268 (2006) 150
H., T. Araki, Viscoelastic phase separation in soft matter:
Numerical-simulation study on its physical mechanism, Chem. Eng. Sci.
D., W. B. Krantz, A. R. Greenberg, R. L. Sani, Membrane formation via
thermally induced phase separation (TIPS): Model
validation, J. Membr. Sci. 279 (2006) 50
M.; Guenoun, P. ; Beysens D. ; Delsanti M. ; Petitjeans
P. ; Kurowski P. Transient surface tension in miscible fluids.
82 (2010) 041606.
2) Nanoimprint: Towards ordered
with large aspect ratio
Copolymers on surfaces provide easily nanostructures thanks to their
microphase separation.(see Fig.1) We can also guide these
nanostructures perpendicular to the substrate and organize them in the
plane by techniques of nanoimprinting (NIL, see principle Fig.2). In
this thesis we now propose to control the growth of structure in the
third dimension to produce structures with large aspect ratio which can
be used as deep molds or for optical applications.
Fig.1: Lamellar phase
oriented perpendicular to the substrate but
disordered in the plane. The orientation perpendicular to the substrate
is described in
24 Pages: 9609-9612 Published: DEC 22 2009.
NIL is able to orientate the lamellae either perpendicular to the mold
grooves, by flow effects, or along the mold grooves
by surface field effect.
3) Copolymers self-assembled in the
third dimension, application to metamaterials
The phase separation of copolymers can
generate nanometer-sized structures that can be directed from a
surface. We can thus access to these structures and consider them as
molds of high spatial resolution for future applications in
nanoelectronics. The possibility of obtaining periodic nanoscale
structures also opens the way for the making of optical metamaterials.
We have recently shown (Macromolecules, 2009) that cylinders of
diameter 40 nm or strips of same period could be erected on a surface
with a suitable surface treatment. However the film thickness of
copolymer that we can organize remains limited to about fifty
nanometers. More generally, difficulties arise when one tries to
control self-assembly of three-dimensional structures. A first problem
is how to go from one point to another in space along a single
continuous phase. However, such a topological control, would enable new
electronic architectures and simplify the integration of self assembled
systems into devices with aspect ratio larger than 1 (height>
width of the pattern created). By etching these structures and a
subsequent metallization one could also create metamaterials, periodic
structures of sub optical wavelength to provide, among others,
properties of nanocapsules for ultrasound imaging
The ultrasound contrast
agents (ACU) are nanocapsules which, when injected into the blood, can
improve the contrast of ultrasound clinics. They can also be used as
drug carriers by delivering the active ingredient in diseased tissue,
thereby enhancing efficacy while reducing its effect on healthy organs.
The ACU studied in the laboratory consists of a shell surrounding a
liquid core. In order to optimize their use in imaging and therapeutic
applications, we want to evaluate and understand the factors that
contribute to the improvement of their properties. This improvement is
through better modeling of the mechanical behavior of objects is
essential to understand the behavior in the physiological flow. We can
make this model only if significant mechanical parameters are known.
For this purpose the mechanism of the nanocapsules will be studied by
atomic force microscopy (AFM). Indeed an AFM lever is not only a sensor
to produce an image at the atomic scale, but also a sensor to record
nano forces – of order nanoNewton or less.
In this mode the AFM nanoindentation tip is lowered onto the
nanocapsule at a fixed speed which causes the indentation of the
sample. The deflection (bending) of the lever depends on the hardness
of the sample: a harder sample causes a larger deflection. Its bending
is proportional to the force that the AFM tip exerts on the sample.
From the deflection curves the indentation is deduced as well as the
relationship between force and indentation if the response of the lever
on a rigid surface is known. For this we first compare the bending of
the lever on an infinitely rigid substrate such as silicon. Modeling of
the curve force - indentation will then test different mechanical
models of the nanocapsules. The aim is to go back to their Young's
modulus, taking into account the possible adhesion of the tip to the
particles. We will pay a particular attention to plastic effects that
may occur, as well as viscoelastic effects that occur if the results
depend on the speed of approach.
We can extend the thesis work to nanoparticles that serve as drug
carriers. These nanoparticles, except for the constraints associated
with a chemical transport of sufficient quantities of the drug and a
very gradual release of the latter, also undergo mechanical stress in
blood flow and the crossing of cellular barriers. Their mechanical
properties are also crucial to quantify their optimization. In this
case, the geometry is generally more complex than a simple compound ACU
shell surrounding a liquid core. Moreover surface modifications can be
made to modulate the adhesion of the particles.