C2P2 - Research LCPP


CPE Lyon
43 Bd du 11 Nov. 1918
69616 Villeurbanne cedex

  Chemistry, Catalysis, Polymers and Processes (UMR 5265)



Why is Polymer Reaction Engineering important?

The production of polymers is different from most other chemical production processes in the sense that the structure, properties and value of polymer-based materials are determined in part by the “chemistry” and in part by the “process” used to make them – they are “products” by “process” to cite the popular phrase.  In other words, if we use the same formulation in the feed to 2 different reactors, e.g. a batch stirred tank or a continuous stirred tank reactor, we will not always obtain a polymer with the same final properties at the end of the day. It is therefore critical to be able to understand the impact of the process on product properties (molecular weight and chemical compostion distributions)... but also to understand the impact of how local phenomena influence over all reactor behaviour.

The schema below illustrates this for the example of a olefin polymerisation process.  The local concentration and temperature at the active sites of a supported catalyst will determine the rate of polymerisation (and thus the amount of heat to be removed at the macroscale) and the physical properties of the macromolecules produced there.  The properties will influence the morphology of the particle (and thus the rate at which monomer can arrive at the active sites) and how the particle interacts with its surroundings (is is sticky? is it friable?).  The way in which the particles flow will determine how well we can mix the contents of the reactor, and whether or not we can maintain reactor stability...  and that is only part of the picture!


How do we approach our work?

Joining these two aspects - chemistry and process - is the motivation underpinning our research in polymer reaction engineering. By applying the methods and principles of chemical engineering we seek to understand the relationship between the chemistry and processes of polymerisation and as such "erase" the artificial boundary between the two related disciplines.

We seek to be innovative in the development of new or better processes for the production of polymeric materials, but to do this we need to develop a fundamental understanding of the science and physics that define what happens at the different characteristic length scales of a process.  Ours is an applications-driven approach.  We develop different tools - including mechanistic models - and experimental approaches to either improve or invent processes, and to understand the underlying scientific rules that govern how our products are made.

Finally, one of the biggest advantages of the research environment for process engineering in the LCPP grou is the very strong interaction with our colleagues in the area of polymer chemistry and materials, providing us with a unique working environment.




Some examples of our work in the area of free radical processes.

At the heart of our research is the field of high solid content latex products.  Since we are concerned by applications - and thus need to consider industrial and economic constraints, working with solids contents as high as 75% v/v (all the while keeping the viscosity as low as possible) is of primary concern.  Of course solids contents this high are not possible to attain for all formulations or with all chemistries.  For instance, working with paper coatings and particles necessarily smaller than 200 nm means that we are happy with solids levels of 60% v/v.  The diagram below shows a quick overview of some of the topics that we have worked on over the course of the past few years, and how they relate to each other.



High Solids for Pressure Sensitive Adhesives

To the left, a TEM image of a bimodal high solid content latex made in a one pot process (i.e. with no intermediate seeds).  The key here is to exert fine control over the particle size distribution - too many small particles or small particles that are too small, and we will either get very high viscosities or lose stability.  This means that we need to control the surface coverage of the first population of large particles as they grow, and then to limit the nucleation rate of the second population of smaller particles. Of course the proper choice of initiator to limit the stabilisation of particles generated by homogeneous nucleation is also important.  

Achieving the limit of over 75% v/v solids and viscosities of under 1.5 Pa·s at shear rates of 20 s-1 was made possible through an in-depth experimental and modelling study.  We worked on understanding each of the steps shown in the schema below left in order to develop a protocol (shown below right) that we have applied to many different latex formulations (adhesives, paint binders, PVC...). 

Scale-up of emulsion processes (On-going ANR project)

The logical extension of the fundamental understanding of how particles are created, stabilised and polymerised is to combine that knowledge with advanced modelling techniques like population balance modelling (PBM) and computation fluid dynamics (CFD) to understand how changing length scales changes the critical time scales in these processes.  The images below illustrate this point. To left we can see that the time it takes to mix a tracer uniformly will take much longer in a 10 cubic meter reactor (over 100 times) than in a one litre lab reactor.  To the right, we can see that if we replace the tracer by a coagulating agent, the slow dispersion in the large vessel shows that the consequences can be dramatic in a larger vessel.



Miniemulsions - Innovative approaches to emulsification and applications to hybrids and composites

Miniemulsions are very similar to conventional emulsion polymerisations in many ways.  However there are two major differences:   (1) the stabilisation system contains a hydrophobic molecule in addition to the normal surfactant; (2) even more importantly particles are made by mechanically emulsifying an organic dispersion in water.  It is this second step that is both challenging and significant.  It is challenging because it requires mixing/agitation methods that can reduce monomer droplets to size of around 100-200 nm – in other words we need high shear. Typically this has been done in laboratories with ultrasonic dismembrators.  However this is VERY energy intensive and difficult to scale up.

We have successfully demonstrated that that SMX static mixer elements can be used to effectively and efficiently create polymerisable miniemulsions, especially if we use an oil-soluble acid and water soluble base to generate surfactants in situ (see picture to the left.)  The rapid stabilisation of the droplets generated in this manner due to the acid/base neutralisation reaction that forms the surfactants at the surface of the particle means that this emulsification can be done more rapidly with the in situ surfactant than with preformed surfactant that must diffuse to, then absorb on the droplets.

This approach was then used to develop a continuous process for polymerisation of miniemulsions was studied at the laboratory scale using in situ surfactants with 21 Sulzer SMX mixer elements.  This new approach allowed us to continuously emulsify a monomer phase at a solid content of over 40% v/v and directly polymerise it.


Some examples in the field of olefin polymerisation on supported catalysts.

We saw above, in the schema with the characteristic length scales for olefin polymerisation, that the growing particle can be treated as a sort of filter between what happens in the bulk of the reactor, and what occurs at the active sites.  One of our major contributions has been in the development of specailly adapted lab tools for the study of the critical first moments of the polymerisation - at realistic temperatures, pressures and flow fields.  It turns out that the initial second - or less - of the polymerisation  can be critical to the success of the production run.  We have therefore concentrated a major effort in the area of studying very short reaction times.
In particular, we have developed gas phase and slurry phase stopped flow reactor systems capable of running at over 10 bars of pressure for times as short as 40 ms.  Below is an example of the gas phase stopped flow reactor we have recentily built.
We have used this reactor to demonstrate the role of the process gas in preventing overheating of the reactor. Below left we see the effect of using a gas with a higher heat capacity and thermal conductivity on the outlet gas temperature, and below right we can use an energy balance on the reactor to get a very good estimate of the real particle temperature in the reactor.

Using He (higher cp and kf) leads to lower gas outlet temperatures using the same catalyst system and reaction conditions.


Here, an energy balance on the reactor allows us to directly predict the surface temperature of particles in real time.  This gives us a much better accuracy in terms of studying reaction kinetics than we would have relying on bulk temperature measurements.

To the right is a schema of the stopped flow reactor developed for slurry (and solution) polymerisation.  It can be operated in different ways, and in particular the residence time of the reactor can be adjusted using the pressure drop and tube length.  We have successfully run reactions with residence times as short as 40 milliseconds.  In addition to demonstrating some very unexpected kinetic behaviour - reaction rates during the first second or so are 6 to 10 times higher than those observed in standard reactors - this tool can also be used to look at the impact of catalyst preparation on particle morphology (see just below). 


  Research | Permanent staff | Equipment | Graduate students | Events |
  Open positions | Downloads | Home | Site map | Contact us | Credit |