PickEP: Modeling of Pickering emulsion polymerization

ANR-12-JS09-0007-01

January 2013 – December 2015

Afficher l'image d'origine

 

Partners

 

LAGEP

Dr. Nida Sheibat-Othman, Dr. Yves Chevalier

Laboratoire d’automatique et de génie des procédés (LAGEP, UMR 5007) / UCBL

Dr. Barthélémy Brunier, recruited for PhD in this project

http://www.c2p2-cpe.com/img/charte/logo.gif

Dr. Elodie Bourgeat-Lami

Laboratoire de chimie, catalyse, polymères et procédés (C2P2, UMR 5265) / UCBL

 

Afficher l'image d'origine           Afficher l'image d'origine

 

Objectives

 

Reprinted with permission from Langmuir, 2016, 32, 112−124. Copyright 2016 American Chemical Society.

 

The production of polymer dispersions stabilized by inorganic particles by emulsion polymerization, called here “Pickering emulsion polymerization”, is nowadays recognized as an efficient process allowing the combination of the properties of polymers (e.g. processing, handling, elasticity, etc.) and those of inorganic solids (e.g. mechanical, optical properties) and yielding composite materials with enhanced properties. In Pickering emulsion polymerization, the conventional stabilizer (ionic or nonionic surfactant) is replaced by inorganic solid particles. However, while modeling of conventional emulsion polymerization is well described in the literature, no modeling results were available in the literature at the beginning of this ANR project for Pickering emulsion polymerization. Modeling is important as it allows improving the understanding of the phenomena taking place during the polymerization as well as allowing implementation of process optimization and control.

The two main polymer properties aimed at being modeled in an emulsion polymerization system are the polymer molar mass and the particle size distribution (PSD). The phenomena supposed being influenced by the change in the stabilizer type (clay in this project) are:

-          stabilizer partitioning between the phases

-          monomer transfer, (if the droplet size is importantly affected)

-          polymer particle nucleation

-          particle stabilization/coagulation

-          radical capture efficiency by the polymer particles

These parameters directly affect the PSD and their modelling requires adaptation with regards to change of the stabilizer type.

 


 

 

Results

Clay partitioning

(task 2)

 

 

 

These results have been published in [P1].

In order to illustrate the role of clay and its partitioning between water and the surface of the polymer particles, different analysis techniques were used: transmission electron microscopy (TEM), adsorption measurements using Quartz Crystal Microbalance (QCM-D), clay titration by means of Inductively Coupled Plasma-Atomic Emission Spectroscopy (ICP-AES), and conductometry. Polystyrene latex was produced using potassium persulfate as initiator and small amount of clay Laponite® RDS (BYK Additives Ltd) as stabilizer. Then, different amounts of clay were added to the produced latex to study its partitioning without reaction and derive the adsorption isotherm.

Using pure polystyrene, the three methods demonstrated that non-electrostatic attractions between the clay and polystyrene could overcome electrostatic repulsions in order to allow adsorption. Interestingly, multilayer adsorption of platelets on the polymer surface was demonstrated. A higher adsorption of clay on the polymer particles was measured on uncleaned latex (containing residual electrolyte coming from the initiator) or on cleaned latexes to which KCl was added, indicating an effect of ionic strength on adsorption due to screening of the clay surface charges.

TEM and EDS analyses of clay-armored particles synthesized via surfactant-free emulsion polymerization showed that the clay was adsorbed as multilayers on the polystyrene particles, leading to a thick shell. The formation of multilayer does not require additional salt (multilayer was also observed with QCM in pure water). This suggests that the formation of multilayers involves non-electrostatic contributions. But, adsorption is enhanced by higher ionic strength, showing that electrostatic interactions also matter. Aggregation of the clay platelets takes place specifically on the surface of the polymer particles and did not occur in the bulk aqueous phase. These results show that any added clay adsorbs on the polymer particles and the concentration of clay in the aqueous phase remains constant. This prevents further nucleation of new polymer particles in the suspension during seeded emulsion polymerization experiments. The adsorption behavior was satisfactorily modelled using the BET isotherm, which confirmed the hypothesis of multilayer adsorption with cooperative binding.

 

 

 


 

 

Results

Optimization experiments

(tasks 4 and 6)

 

 

 

Effect of stirring rate. Stirring had no effect on the reaction kinetics between 350 rpm and 500 rpm, with the used impeller, ensuring thus no mass transfer limitations or increased orthokinetic coagulation. This range is comparable to conventional emulsion polymerization and reflects no particularity of Pickering systems to be considered specifically in modelling.

Effect of clay type. Four Laponite® clay types were compared: RDS, XLS, JS and S482 (BYK Additives Ltd) in ab initio semi-continuous emulsion polymerization of styrene using potassium persulfate as initiator.

In terms of adsorption isotherm, it was found that the different clays formed multilayers on the surface of the polymer particles. Similar adsorption isotherms were measured for Laponite RDS and S482 under the considered ionic strength. Lower adsorptions were observed for Laponite XLS and JS.

In terms of polymerization kinetics, the different clay grades did not have the same stabilizing efficiency as it could be observed through the produced number density of particles. The clay Laponite® JS was the less efficient for stabilizing the latex particles. Note that this clay is a synthetic fluorohectorite that contains 4.8% fluorine, and a higher amount of lithium (1.2% vs 0.8% for the other clays) which gives a higher (more negative) surface charge. Laponite RDS and S482 had a similar stabilization behavior: increasing the clay concentration increased the number of polymer particles without levelling off. Laponite XLS showed a similar behavior as the previous two clays at low clay concentrations. This indicates that, for the three Laponite clays, individual platelets had the same surface charge. But, a kind of saturation, in terms of the number of particles, was observed for Laponite XLS at higher concentrations.

Other process parameters were optimised: the effect of the initiator concentration, the initial monomer concentration, the monomer flow rate and radical diffusion limitations were assessed.

 


 

Results

Radical exchange

(task 3)

 

 

 

The proposed radical capture and desorption models in the literature were usually developed for specific systems (specific monomer solubility, range of particle size and types of surfactants) and cannot be generalised to all systems.

Radical desorption. It is admitted that only monomeric radicals, produced by transfer to monomer (or to chain transfer agent), may desorb from the polymer particles as they are smaller than polymeric radicals. Radical exit has been subject to long debates, and the proposed mechanisms gained complexity with time, starting from simple diffusion, next accounting for competitive reactions inside the particles, and finally accounting for competitive radicals in the aqueous phase (fate of exited radicals). But, it was agreed that the stabilizer layer concentration or type would not affect radical exit. This is due to the high diffusivity of the monomeric radicals.

Radical Capture. In emulsion polymerization, a water soluble initiator is usually used in order to enhance the compartmentalization effect that allows producing polymers of high molar masses (by reducing termination). Therefore, the primary radicals are produced in the aqueous phase. These radicals should enter the polymer particles where the polymerization reaction mainly takes place. This becomes possible if the solubility of the radical in water is reduced, due to the formation of short oligomers in water. Then, these radicals transfer to a more hydrophobic phase inside the monomer-swollen particles. Radical capture concerns therefore long enough polymeric radicals . Different mechanisms were considered to describe radical capture and adsorption.

1) diffusion-controlled mechanism, where the capture rate is proportional to the particle diameter. A radical capture efficiency term was later added to this model;

2) collision-controlled mechanism, where radical capture is proportional to the particles surface area.

3) colloidal model assuming radical entry is proportional to the particle diameter.

4) surfactant displacement mechanism, where radical entry is affected by the stabilizer layer

5) propagation mechanism, where radical capture is conditioned by a critical radical length attained after propagation in the aqueous phase.

For conventional surfactants, it is now mostly agreed that the surfactant type or amount on the surface of the polymer particles do not affect radical capture. The mostly admitted mechanism is the propagation mechanism. However, for polymeric surfactants it was observed that the amount of stabilizer might affect radical capture. For the present system involving solid inorganic platelets on the surface of polymer particles, it is thus important to check against the effect of the stabilizer layer on radical capture.

The ab-initio and seeded experiments showed no dependence of radical capture on the concentration of clay. But, there was a clear dependence on the particle size. The propagation model could not describe the experiments satisfactorily, while the diffusional model was found appropriate.

 

Results

Particle nucleation and coagulation

(tasks 2 and 5)

 

 

 

The clay had an important role in stabilizing polymer particles in surfactant-free emulsion polymerization. Smaller particles could be obtained at the same solid content when increasing the clay concentration, which led to a bigger number of clay-stabilized polymer particles at the end of the reaction.

As no surfactant micelles are present, a coagulative-homogenous nucleation mechanism was assumed. DLVO theory was used to calculate the coagulation kernel. The Hamaker constant of the attractive potential, was estimated by provoked coagulation experiments, in the absence of reaction.

The results of the previous tasks allowed the prediction of radical capture/exit as well as clay partitioning among the aqueous and polymer phases. The formation of multilayers of clay on the surface of particles could explain the nonlinear relationship between the total number of particles versus the clay concentration.

Estimation of the stabilization efficiency of the clay was based on experiments involving lower clay concentrations (ensuring less than one layer coverage). Using this information, the effective number of clay platelets contributing to the stabilization of the system was estimated experimentally and a relationship to the clay concentration was proposed. This allowed the development of a predictive model of the rates of particle nucleation and coagulation as a function of the clay concentration.

 

 


 

Publications

 

 

 

 

International scientific journals

[P1]  Barthélémy Brunier, Nida Sheibat-Othman, Yves Chevalier, Elodie Bourgeat-Lami, Partitioning of clays platelets in Pickering emulsion polymerization

Langmuir, 2016, 32, 112−124

DOI: 10.1021/acs.langmuir.5b03576

[P2]  Barthélémy Brunier, Nida Sheibat-Othman, Yves Chevalier, Elodie Bourgeat-Lami, Investigation of Four Laponite Clays in Pickering Emulsion Polymerization

Submitted to Canadian Journal of Chemical Engineering, 2016

[P3]  Barthélémy Brunier, Nida Sheibat-Othman, Yves Chevalier, Elodie Bourgeat-Lami, Radical exchange in Pcikering emulsion polymerization

To be submitted, 2016

[P4]  Barthélémy Brunier, Nida Sheibat-Othman, Yves Chevalier, Elodie Bourgeat-Lami, Particle nucleation in Pickering emulsion polymerization

To be submitted, 2016

Conferences,

Workshops

[C1]   2nd Working Party on Polymer Reaction Engineering

May, 24-26, 2013 (University Hamburg, Germany)

Oral presentation + poster: Investigation of Pickering emulsion polymerization processes

Barthélémy Brunier, Julien Blanquet, Yves Chevalier, Elodie Bourgeat-Lami, Nida Sheibat-Othman

[C2]   Frontiers Of Polymer Colloids 2014

July 20-24, 2014 (Prague, Czech Republic)

Poster: Evaluation of laponite partitioning in Pickering emulsion polymerization

B. Brunier, Y. Chevalier, E. Bourgeat-Lami, N. Sheibat-Othman

[C3]   European Congress of Chemical Engineering (ECCE) 2015

September 27 – October 1 (Nice / France)

Oral presentation: Modelling of Pickering surfactant-free emulsion polymerisation

Barthélémy Brunier, Yves Chevalier, Elodie Bourgeat-Lami, Nida Sheibat-Othman

[C4]  11th Workshop on Polymer Reaction Engineering (PRE)

May, 17-20, 2016 (University Hamburg, Germany)

Poster: Modelling of Pickering emulsion polymerisation

Barthélémy Brunier, Yves Chevalier, Elodie Bourgeat-Lami, Nida Sheibat-Othman

 

PhD thesis

Barthélémy Brunier, « Modeling of Pickering Emulsion Polymerization », defended December 4, 2015, at Université Claude Bernard Lyon 1, Ecole doctorale de chimie, Spécialité génie des procédés.

 


 

View of the project tasks