ÐÓ°ÉÔ­´´

Take a greasy dish, dip it into a solution of washing-up liquid, and
after some scrubbing the dish should come out clean, leaving dirty dishwater
behind. It sounds simple, but doing the washing-up involves some complicated
chemistry and physics. The washing-up liquid is what is called a surface-active
agent, or ‘surfactant’, which consists of ‘oily’ hydrocarbon molecules with
a water-loving component at one end. In the cleaning process the detergent
molecules combine with the grease to give a suspension of tiny particles
in water – a colloid. The rinse containing the dissolved grease is a complex
mixture of long-chain hydrocarbon molecules and other organic materials.
Understanding how such processes happen is extremely important not just
for formulating better washing-up liquids but also in designing other everyday
consumer products.

Many interesting and useful materials around us, from nylon stockings
to strawberry jam, are made from an irregular arrangement of relatively
large particles, perhaps suspended in a medium of smaller ones. The large
particles might be polymers (long chain molecules) or aggregates of molecules
in a colloidal suspension. Physicists call this ‘soft condensed matter’
to distinguish it from ordered, crystalline materials, in which the position
of particles is fixed and regular.

The design of soft condensed matter with specific properties is one
of the aims of modern technology. For example researchers have developed
polymer fibres for cables that weigh only a fraction of a steel cable and
have a much higher tensile strength. How the bulk material behaves under
external stress depends on the structure and dynamics of the constituent
particles. Understanding, for instance, how the polymer molecules are organised
in the stretched fibre gives us the key to exploring the relationship between
the properties and performance of the material.

The physics of soft condensed matter – some call it ‘sludge science’
– is a rapidly growing area of research. The award of last year’s Nobel
Prize for Physics to Pierre-Gilles de Gennes recognised his theoretical
contributions to this area. De Gennes’ work covered everyday materials such
as the liquid crystals (which are, generally, rigid chain-like molecules),
used in flat television screens, and the polymers used in superglues. Living
matter also belongs to this category. Biological structures include polymeric
molecules such as DNA, and large molecules such as proteins, which are either
suspended in a liquid medium and are free to move in intricate ways or form
organised structures such as membranes.

One of the most important features of such polymeric and colloidal systems
is the way they flow. Liquids like water, made of small molecules, usually
obey what is called ‘Newtonian’ flow if the velocity is not too great and
there are no obstacles in the flow direction. In other words, water flowing
between two parallel plates – one stationary, the other moving with a constant
velocity – produces an even flow with parallel layers, each flowing with
a different velocity. This is called laminar flow. Small molecules move
continuously from one layer to the next, transmitting a momentum across
the flow. Thus more rapid molecules accelerate slow layers and vice versa.
‘Viscous drag’ or the fluid’s viscosity can be interpreted as the retarding
effect of the slow layers on the fast. But the viscosity does not depend
on the velocity of flow.

Polymers and colloids often behave differently. The constituent particles
are relatively large and their structures can be complex. When they move
they tend to interfere with each other causing some elastic behaviour in
addition to the viscous drag. That is why they are often called ‘viscoelastic’.
When a plastic melts, for instance, the cohesional forces of the polymer
molecules will tend to preserve the molecules’ original shape, or ‘conformation’
under flow. You see similar phenomena in colloidal systems, for example,
the dispersions of organic or inorganic particles in liquids such as paints,
cosmetics, pharmaceuticals, ceramics and paper coatings.

The flow of these liquids is ‘non-Newtonian’. Their viscosity changes
as the rate of flow increases. This is because part of the energy from the
applied force might be absorbed by the material itself, resulting in changes
in internal structure. For example, in polymers, the coiled molecular chains
might bend and stretch. Or long rod-like molecules are oriented by the flow
and might slide past each other more freely. Physicists and chemists still
do not completely understand how molecules in these materials behave, but
finding out would be a big step towards understanding the relation between
their structure and dynamics.

To explore solids and liquids at the atomic and molecular level we need
to ‘look’ inside them. For a long time, researchers have been using X-rays
to ‘see’ into crystal-line materials. However, X-rays have their limitations
because they interact with the electron cloud surrounding the nuclei of
atoms. X-rays, therefore, cannot penetrate very deeply into matter.

A better tool for studying colloidal or polymeric materials is another
form of radiation – beams of neutrons. These neutral particles, along with
the positively charged protons, are the constituents of atomic nuclei. Neutrons
can pass through the electron clouds of atoms and interact with their nuclei.
Because the nuclei are very small compared with the surrounding electron
clouds, the neutrons can penetrate quite deeply into colloidal and polymeric
materials before hitting any nuclei in the polymer chains. They are also
more likely to pass straight through the organic liquids (made of small
molecules with few nuclei) in which the materials might be dissolved. Inorganic
materials, such as quartz, which might be used to make the vessels containing
the experimental samples are also fairly transparent to neutrons. In general
neutrons are suitable for studying particles that are between 5 and 500
nanometres across.

Neutrons can be produced in controlled quantities and at specific energies
by nuclear fission in a specially designed nuclear reactor such as that
at the Institut Laue-Langevin (ILL) in Grenoble, France . In a typical experiment
at ILL, a beam of neutrons, all with more or less the same energies, impinges
on a sample. This is called the primary beam. The process is analogous to
firing a series of billiard balls at a collection of moving targets. Most
of the billiard balls, however, do not hit anything and pass straight through.
Others hit a target and are deflected, or ‘scattered’, in various directions
depending on the geometry and mobility of the target. In the case of neutrons,
the targets are the sample’s atomic nuclei. A detector measures how the
intensity and energy of the scattered neutrons vary at different angles
of scattering. Because the scattering angles are usually less than 15 degrees,
these experiments are called small angle neutron scattering (SANS) studies.
The result is a set of data relating the intensity of scat-tered neutrons
to the scattering angle. In general the intensity decreases rapidly as the
scattering angle increases.

What does this information tell us? Let’s take a simple example – a
solution of some polymer molecules in an organic solvent. So far we have
treated neutrons as particles, but a beam of neutrons also behaves as a
wave, and so may be deflected at different nuclei in our large polymer particle.
This interaction gives rise to interference. The more polymer particles
there are, the more likely the neutrons are to be scattered and the greater
the intensity of the scattered neutrons. The intensity measured for each
scattering angle also depends on the size and shape of the individual molecule,
and this gives us information about the structure of the polymer particles.
The larger the scattering angle, the more molecular detail we can see. In
general the SANS technique is suitable for studying molecular structures
that are between 5 and 500 nanometres across.

Obviously the solvent can also scatter neutrons, which means that the
scattering intensity of the particles has to be measured against background
scattering from the solvent. You can think of a neutron passing through
the sample as a car going along a road dotted with small potholes which
tend to make the car veer to one side or the other. The polymer molecules
act like potholes, deflecting the neutrons. If these potholes are deep they
give a large ‘contrast’ against the background scattering from the small
solvent molecules. The greater the contrast, the more scattered radiation
from the polymer particles can be collected out of the way of the primary
beam.

The scattering intensity is also affected by what chemical elements
are present in the particles and hence which nuclei they contain. The probability
of scattering depends on a nuclear property called the ‘scattering length’.
This quantity is different for every element of the periodic table and each
isotope; it does not vary in any regular way with the atomic number. For
example, hydrogen (whose nucleus contains a proton and nothing else has
a scattering length which is very different from that of deuterium, an isotope
of hydrogen whose nucleus contains a neutron as well as a proton. In fact,
it is the largest difference between any two isotopes of an element in the
periodic table. Polymers, colloids and large biological molecules include
a large number of hydrogen atoms, as do many organic solvents. By replacing
some or all of the hydrogen atoms with deuterium (which does not alter the
chemical properties of the polymer) we can increase the contrast between
the solvent and the particles, so making the particles more ‘visible’ to
neutrons. It is also possible to ‘colour’ particular bits of the polymer
with deuterium so that they stand out.

Polymer molecules in solution wriggle around, changing position and
orientation, and altering the direction of their bonds, which changes their
shape. Neutrons will therefore be scattered around the primary beam at a
particular angle in all planes. The scattered radiation is collected by
a two-dimensional detector made up of a large number of small elements that
count neutrons independently. Its sensitive surface is positioned perpendicular
to the primary beam.

In a conventional SANS experiment the sample is at equilibrium – at
the same temperature and pressure as the surroundings, and with a smooth
concentration and a definite, steady chemical composition. However, a system
that is not in equilibrium is much more interesting to researchers working
on industrial processes. They may want to know how a plastic stretches when
being moulded, for example.

Another nonequilibrium process is found in enhanced oil recovery. Here,
oil is extracted from oil-bearing rocks by flooding them with water and
pumping out the mixture of oil and water. But most of the oil tends to remain
coated on the internal surfaces of the rock even at high pumping pressures.
One way to remove the oil more efficiently is to add surfactants to the
water. If a solution of surfactant is at the right concentration, it forms
a colloidal system – a ‘microemulsion’ – with equal volumes of oil and water,
which is then swept out. Another method uses polymer additives of high molecular
weight, such as polysaccharides or poly-acrylamides. In this case, the additives
increase the viscous drag of the water-polymer system compared with that
of the pure oil, helping to flush out the oil more effectively.

At ILL we have been studying these polymers and surfactants under nonequilibrium
conditions by applying an outside force. In such studies we can examine,
for instance, what happens to the polymer’s network of long-chain molecules
as we stretch it and then let it relax back to equilibrium.

An interesting type of experiment is to keep the sample (if it is liquid)
flowing in a steady state. There are several different kinds of flows that
can be studied. The first kind of flow is simply illustrated. Fluid is allowed
to flow from a hole in the bottom of a vessel. This induces ‘elongational’
flow in which gravity causes the velocity of the flow to increase with the
distance from the hole. Another kind of flow is a type of shear flow, called
Poiseuille flow, when a fluid streams through a tube at a constant flow
rate. The velocity at the tube wall is zero reaching a maximum at the centre
of the tube. A second type of shear flow, called Couette flow, takes place
between two parallel plates one of which moves with a constant velocity.
Here the flow gradient is constant.

So what is really happening to the polymer’s molecules? Viscous drag
in the streaming fluid tends to rotate, orient and deform them. The tangled
long-chain molecules will stretch out if the drag is strong enough. Is is
like stirring a saucepan of randomly coiled spaghetti which changes from
tangled but evenly distributed mass to a shape where the strands all tend
to lie in the same direction. Our neutron scattering experiments reveal
that the same thing happens when we make a polymer solution flow in one
direction. A polymer molecule behaves as a coiled sequence of beads connected
by flexible springs. External stress pulls the coils apart while the elastic
properties of polymer chains tend to resist the shape change. If the shear
flow gradient is strong enough the polymer chain segments change their random
orientations and they tend to point in the same direction. As a result the
whole molecule is deformed. This is revealed in measurements of the neutron
scattering intensity which are different in directions perpendicular and
parallel to the flow.

Here we have also a possible clue to understanding an unusual non-Newtonian
phenomenon called drag reduction or the ‘Toms’ effect in pipes which occurs
when the flow of fluids in pipes changes from a smooth, laminar flow to
a turbulent one by increasing the flow velocity. When a Newtonian fluid
like water is pumped at high pressure through a pipe small vortexes form,
which increase the drag and the energy needed to maintain a given flow rate.
The Toms effect might arise when a small amount of a polymer or surfactant
is added to the water: surprisingly, adding small amounts of a polymer or
surfactant seems to modify the turbulence so as to reduce the drag. Additives
are used to reduce drag in pipelines carrying crude oil, so saving on pumping
power, and in sewerage networks in order to reduce the risk of flooding
as sewers become overloaded. Drag-reducing additives could also be used
to increase the ‘squirt’ range of firefighting equipment. Even nature takes
advantage of drag reduction. The Pacific barracuda, a fish that can swim
very rapidly, has a layer of slime covering its body. Under laminar flow
conditions, at low speed, the organic slime does not mix with the water.
At higher speeds the onset of turbulent flow forces the slime to mix with
water close to the fish’s body. The slime solution seems to modify the turbulence
and decrease the friction of water on the body. This reduces the energy
needed to swim fast.

Until recently, no one has really understood how drag reduction in turbulent
flow works. This is because not much is known about the microscopic structure
of additives under nonequilibrium conditions. At ILL we decided to examine
the laminar and turbulent flow of a solution of a surfactant in water using
SANS. The surfactant was a mixture of molecules with hydrocarbon chains
containing between 12 and 14 carbon atoms and a salt, sodium salicylate.
At very low concentrations in water, the surfactant forms rigid aggregates
of rod-like particles, or micelles, about 150 nanometres long and 3 nanometres
in diameter. You might expect the micelles to align when the solution is
flowing smoothly but become randomly oriented when the flow is turbulent.
Our SANS measurements in pipes show that the micelles stay oriented parallel
to the direction of flow after the onset of turbulence, and this property
may be what reduces drag.

Relatively new experiments like these are attracting attention because
they combine ‘classical’ scattering techniques with studies of non-Newtonian
fluid dynamics. There are many other examples where neutron experiments
might help to elucidate nonequilibrium processes involving polymers, including
the injection moulding of plastics, the spinning and stretching of synthetic
fibres, extrusion of plastic tubes and foils, the blowing of bottles and
the expansion of polymer foams. And it is worth pointing out that our work
at ILL illustrates how an analytical tool that is a direct spin-off from
nuclear technology can improve understanding of what happens in flowing
matter, and provide an important benefit to industry.

Peter Lindner is at the Institut Laue-Langevin in Grenoble, France.

* * *

NEUTRONS FOR EUROPEAN RESEARCH

Neutrons are used for experiments in a wide range of research areas
from elementary particle physics, through the physics of condensed matter
to chemistry, biology and material science. The Institut Laue-Langevin (ILL)
in Grenoble, France, is equipped with the most versatile neutron source
for research in the world.

In 1967, West Germany and France agreed to set up the ILL, installing
a 57-megawatt, high flux nuclear reactor as the central research tool. The
UK joined as a third equal partner in 1973. Over the past four years Spain,
Switzerland and Austria have become associate members. About 30 different
instruments are available to some 1500 external users a year, who come mainly
from universities and research centres of the member countries. The ILL
is the only multinational neutron research laboratory in the world. There
are other centres providing neutrons in a number of European countries,
however, that mainly serve their own national scientific communities. These
include ISIS, a pulsed neutron source at Britain’s Rutherford Appleton Laboratory
near Oxford.

The ILL is currently being thoroughly refurbished, following a routine
shutdown last March. A check of the nuclear pile revealed unusual traces
on a baffle designed to reduce turbulence in the heavy water cooling circuit.
The baffle is attached by spacers to another baffle below it, and it eventually
emerged that the lower baffle had become loose and was pushing on the upper
baffle, so producing the damage. The ILL wants to replace the two baffles,
and possibly the heavy water vessel, together with the pile structure. This
would take about two years. The institute is also planning to overhaul the
whole reactor in order to extend its lifetime for another 10 to 15 years.
This second phase of ILL’s existence starts in 1994 at the same time as
the European Synchrotron Radiation Facility (ESRF) will start operation
on the same site. European science will then have two large facilities at
Grenoble for investigating the structure of atoms and molecules, making
it one of the world’s major centres for fundamental research.

Sadly, however, the future of ILL is under threat because the Science
and Engineering Research Council, which is responsible for the UK’s funding
for European research facilities, has decided ‘to seek a reduction in the
level of the UK’s contribution to the ILL, post 1993′.