Experimental set up

Experiments are simulated by a computer model described in [54]. The model out- put is found to provide good agreement with observations obtained via laboratory experimentation and, hence, in this work, assumed to represent the ‘unknown’ un- derlying physical process. All parameters involved in the emulsion copolymerization reaction discussed herein are uniquely identifiable. The main assumptions and char- acteristics of the computer model are summarized as follows [54]:

• Semi-batch seeded copolymerization of a two-monomer system is the simulated process.

• The assumed chemical species are: water, initiator, monomer A, monomer B, a chain-transfer agent (CTA) and the produced copolymer.

• The emulsion is formally divided into three phases: (i) continuous aqueous phase, (ii) monomer droplets, and (iii) polymer particles.

• Polymerization is assumed to proceed only in the polymer particles (except the initiation stage, which proceeds in water).

• The kinetic model is terminal. Both inter-molecular and intra-molecular chain transfer (backbiting) are considered as well as the gel effect. Polymerization kinetics are processed by the population balance of polymer moments.

• The model describes stages II and III of the emulsion polymerization. There- fore, the total number of polymer particles in the reactor is considered constant during the simulation.

• The time-averaged number of radicals in the polymer particles is variable, calculated in the same way as in [53].

• Heat balance of the reaction mixture and of the reactor jacket is included in the model.

• All kinetic constants of the copolymerization reaction simulated by the model are based on data published in the literature.

Key model equations

The key model equations represent: 1. Algebraic equations

• Evaluation of density of the droplet and the polymer phase as well as of the reaction mixture.

• Summation relations (volume additivity of phases). 2. Differential equations

• Mass balance of each monomer in each phase.

Total volume of reaction mixture V consists of the volumes of all present phases (Monomer droplets, Polymer particles and Water):

V =VM+VP +VW. (3.1)

Density of phasej is given by contributions of speciesi: 1 ρj = k X i=1 w_ij ρi (3.2)

withw_ij being weight fraction ofi-th species in phasej and ρi represents density of

pure component i. Total number of components in a given phase is k. Note that monomer droplets consist only of monomers, while the polymer particles contain pure copolymer as well as absorbed monomer species. Density of aqueous phase is assumed to be constant.

Reaction mixture density is calculated from the overall mass balance of the reactor: dρmix

dt V +ρ

mixdV

dt =m˙F, (3.3)

where ˙mF is the mass flowrate of the feed. Molar balance of monomer A in monomer

droplets is implemented as dcM_A dt V M _+cM A dVM dt = ˙ mFwFMw M,F A MA −(1−stage3) ˙nA. (3.4)

The first term on the right hand side of the above equation represents the molar flowrate of monomer A in the feed withw_MF being the weight fraction of the monomer phase in the feed,wM,F_A the weight fraction of monomer A in the monomer phase of the feed, andMA is molar weight of monomer A. Dimensionless switching function

stage3 is based on hyperbolic tangent, so that its value is 0 when the monomer

droplets are present (transfer of monomers from droplets to polymer particles takes place) and 1 if the monomer droplets are depleted in this case the term representing the interfacial transport virtually ‘disappears’ from the model equation. This is a widely accepted method of modifying model equations during integration without the necessity to re-initialize the state variables and restart the integration. The form of switching functionstage3 is

stage3 = 0.5−0.5 tanh

1012 VM −V_thresholdM

with V_thresholdM = 10−9m3. The second term characterizes interfacial transport of monomer A from monomer droplets to polymer particles defined as

˙

nA=KA

w_AP∗−w_AP, (3.6)

withKArepresenting the mass transfer coefficient of monomer A,wP

∗

A andwPAbeing

the equilibrium and actual weight fraction of monomer A in the polymer phase.

Balance of monomer A in the polymer phase also has the term describing the trans- port of monomer into polymer particles, but with positive sign. Moreover, this equation must account for the consumption of monomer A by the chemical reaction RA, which always has a negative absolute value:

dcP_A dt V P _+cP A dVP dt = (1−stage3) ˙nA+RAV P_{. (3.7)}

The model includes balance equations not only for chemical species, but also for all the present phases, which can be demonstrated with the total mass balance of monomer droplets dρM dt V M_+ρM A dVM dt =m˙Fw F M−(1−stage3) ( ˙nAMA+ ˙nBMB+ ˙nCT AMCT A). (3.8)

Conversion is given by a ratio of the amount of reacted monomers and the amount fed into the reactor. Mean particle size is calculated by dividing volume of the polymer phase by number of particles present in the emulsion (which is a model parameter), assuming spherical shape of the particles.

The following are the reaction settings (input variables for the model considered):

• M1

I,MI2,II,EI, andWIare the initial amounts of monomers, initiator, emulsi-

fier and water respectively. T is the reaction temperature,CT Ais the amount of chain transfer agent andP0 is the amount of polymer in seed.

• M_F1, M_F2, EF, IF are the fed amounts of monomers, emulsifier and initiator

respectively. W EF and W IF are emulsifier and initiator solutions in feed.

W IF is split into two parts pW IF and (1−p)W IF (where p ∈(0,1)) and is

added in two stages.

• F_t1, F_t2 are feeding times for adding pW IF and (1−p)W IF respectively and

Pt is the post feeding time. The total reaction duration is the sum of Ft1,Ft2

Figure 3.1: Emulsion copolymerization process scheme example.

The ranges of the variables and constraints are:

1. M_I1 =M_I2 = 1×10−10g and EI =EF = 1×10−4g. These are kept constant

for all recipes.

2. WI,W EF,W IF ∈(0,1500g] and 225≤WI+W EF +W IF ≤1500.

3. II∈(0,0.2WI) and IF ∈(0,0.2W IF).

4. F_t1,F_t2,Pt∈(0,180min] and 30≤Ft1+Ft2 ≤180.

5. T ∈[65,95] in degrees Celsius,CT A∈(0,20g] and P0 ∈[5,30g].

6. M_F1,M_F2 ∈(0,1500g] and 150≤M_F1 +M_F2 ≤1500.

7. The final constraint is that the overall volume of the ingredients must not exceed the capacity of the reactor (3l).