This article was downloaded by: [University of Stellenbosch]

On: 26 July 2013, At: 04:11

Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

Polymer Reaction Engineering

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/lree20

Some Key Factors in Emulsion Polymerization Process

Development
a a a a
J. Meuldijk , M. F. Kemmere , S. V.W. de Lima , X. E.E. Reynhout , A. A.H. Drinkenburg
a b
& A. L. German
a
Process Development Group, Department of Chemical Engineering and Chemistry,
Eindhoven University of Technology, P.O. Box 513, 5600 MB, Eindhoven, The Netherlands
b
Laboratory of Polymer Chemistry, Department of Chemical Engineering and Chemistry,
Eindhoven University of Technology, P.O. Box 513, 5600 MB, Eindhoven, The Netherlands
Published online: 15 Feb 2007.

To cite this article: J. Meuldijk , M. F. Kemmere , S. V.W. de Lima , X. E.E. Reynhout , A. A.H. Drinkenburg & A. L. German
(2003) Some Key Factors in Emulsion Polymerization Process Development, Polymer Reaction Engineering, 11:3, 259-276, DOI:
10.1081/PRE-120023902

To link to this article: http://dx.doi.org/10.1081/PRE-120023902

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the
Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and Francis shall not be liable for
any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of
the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
 POLYMER REACTION ENGINEERING
Vol. 11, No. 3, pp. 259–276, 2003

Some Key Factors in Emulsion Polymerization

Process Development
Downloaded by [University of Stellenbosch] at 04:11 26 July 2013

J. Meuldijk,1,* M. F. Kemmere,1 S. V. W. de Lima,1

X. E. E. Reynhout,1 A. A. H. Drinkenburg,1
and A. L. German2

1
Process Development Group and 2Laboratory of Polymer Chemistry,
Department of Chemical Engineering and Chemistry, Eindhoven
University of Technology, Eindhoven, The Netherlands

ABSTRACT

Emulsion polymerisations are mostly carried out as (semi-)batch

processes in stirred tanks. In ab-initio emulsion polymerisation the
product properties in terms of particle concentration and particle size
distribution strongly depend on the course of the nucleation stage. The
course of the nucleation stage is strongly related to the quality of
emulsification and the temperature of the reaction mixture. The influence
of temperature on the particle size distribution was investigated for
completely isothermal as well as for non-isothermal operation. Reaction

*Correspondence: J. Meuldijk, Process Development Group, Department of Chem-

ical Engineering and Chemistry, Eindhoven University of Technology, P.O. Box 513,
5600 MB, Eindhoven, The Netherlands; Fax: 31-40-246-3966; E-mail: c.j.w.v.os-
rovers@tue.nl.

259

DOI: 10.1081/PRE-120023902 1054-3414 (Print); 1532-2408 (Online)

Copyright D 2003 by Marcel Dekker, Inc. www.dekker.com
 260 Meuldijk et al.

calorimetry has been chosen as a tool to define an operating window for

control of the nucleation stage in emulsion polymerisation. It was
demonstrated that reaction calorimetry is a very poweful tool to define
operational details for complete control of the nucleation process in
emulsion polymerisation in stirred tanks accurately and quickly. Col-
loidal stability of the latex is an important issue during the stage of
particle growth by simultaneous polymerisation and monomer absorption
from the monomer droplets. It was demonstrated that a proper recipe for
the production of a colloidally stable latex can be developed in a short
time with reaction calorimetry.
Downloaded by [University of Stellenbosch] at 04:11 26 July 2013

INTRODUCTION

Emulsion polymerisation processes are frequently used for the

production of rubbers, coatings and many other products. The usually
slightly or moderately water-soluble monomers are mainly present in the
monomer droplets which are dispersed in the aqueous phase. The actual
free radical polymerisation reaction is generally initiated by a water-soluble
initiator and proceeds in the latex particles which are much smaller than the
monomer droplets. The product is a colloidally stable dispersion of sub-
micron particles in water.
The rate of polymerisation in emulsion polymerisation is a function of
the propagation rate coefficient (kp), the monomer concentration in the
particles (CM,p), the average number of growing chains per particle (ñ), the
particle concentration per unit volume of the aqueous phase (Nparticles), the
volume of the aqueous phase (Vwater) and Avogadros number (NAv):

kp CM;p 
nNparticles Vwater
Rp ¼ ð1Þ
NAv

Note that CM,p is constant in the presence of monomer droplets.

The industrial production of latex products is usually carried out in
(semi-) batchwise operated stirred tanks. The outcome of batch emulsion
polymerisation processes in terms of e.g. the particle concentration and the
particle size distribution, strongly depends on the equipment, e.g. impeller
type, the number and size of the baffles, the axial position of the impeller and
the ratio of impeller diameter and tank diameter. In addition operational var-
iables, e.g. impeller speed and temperature, are important, see Poehlein (1997).
The present paper defines key operational factors which govern the
particle concentration and the particle size distribution of a batch emulsion
polymerisation process. The combination of reaction calorimetry and
transmission electron microscopy proves to be very powerful in emulsion
 Emulsion Polymerization Process Development 261
Downloaded by [University of Stellenbosch] at 04:11 26 July 2013

Figure 1. Schematic representation of the flow pattern in a stirred tank equipped

with a six bladed turbine impeller with 4 baffles or with a 45
pitched downflow
impeller with 4 baffles.

polymerisation process development. This is demonstrated by the emulsion

polymerisation of styrene in stirred tanks equipped with 4 baffles, a six
bladed disc turbine impeller or a pitched six bladed downflow impeller, see
Figure 1 for the characteristic flow patterns.

DEVELOPMENT OF BATCH PROCESSES

FOR EMULSION POLYMERISATION:
MAIN ASPECTS TO BE CONSIDERED

Figure 2 shows the conversion time history for an emulsion poly-

merisation with a sparsely water soluble monomer and micellar nucleation
(Smith and Ewart, 1948). In the beginning of the polymerisation emul-
sification and nucleation govern the course of the process. The specific area
of the monomer-water phase boundary has to be sufficiently large to pro-
vide a negligible resistance against monomer transport from the monomer
phase via the aqueous phase to the growing particles. In the case of
negligible mass transfer resistances the actual rate of polymerisation is only
determined by the intrinsic rate coefficients of all fundamental reaction
steps involved, and by the occurring phase equilibria, i.e. monomer parti-
tioning. It should be noted that insufficient emulsification leads to mass
transfer limitation which means that the observed rate of polymerisation is
at least in part determined by mass transfer from the monomer phase to the
reacting polymer particles. This paper demonstrates that proper emulsifica-
tion is a prerequisite for good control of the nucleation stage, the course of
the polymerisation process and the product properties in terms of con-
version, particle concentration and particle size distribution as well.
 262 Meuldijk et al.
Downloaded by [University of Stellenbosch] at 04:11 26 July 2013

Figure 2. Schematic representation of the various important issues in ab-initio batch

emulsion polymerisation with sparsely water soluble monomers, see Kemmere (1999).

Besides emulsification, temperature plays an important role in the

nucleation process. The influence of temperature as well as the temperature
time history for non isothermal operation on the particle concentration and
particle size distribution will be discussed.
After the stage of particle nucleation, the particles grow by simul-
taneous polymerisation in the particles and transfer of monomer from the
monomer droplets to the particles, i.e. interval II. Figure 3 shows the rate
of polymerisation time history and the particle concentration time history
for simple Smith – Ewart case 2 kinetics, i.e. after the stage of particle
nucleation Nparticles is independent of conversion, and ñ is independent of
particle size (Smith and Ewart, 1948).

Figure 3. Rate of polymerisation and particle concentration as a function of time in

the case of complete colloidal stability in interval II. Possible gel effects in interval III
are not shown.
 Emulsion Polymerization Process Development 263

As a consequence of particle growth the particle volume fraction

increases and with that the (apparent) viscosity of the reaction mixture, see
e.g. Mayer et al. (1994), Poehlein (1997) and Kemmere et al. (1998a). Partial
heat transfer coefficients decrease considerably with increasing particle
volume fraction (Kemmere et al., 2000) and depending on the recipe there is
a risk of temperature runaway. The risk of runaway can be ruled out by
limiting the rate of polymerisation for example by working under partial
swelling conditions: a semi-batch operation with a controlled monomer feed
rate. In such operation there are no monomer droplets. Note that for high
Downloaded by [University of Stellenbosch] at 04:11 26 July 2013

solids emulsion polymerisations the choice of the reactor configuration and

operational conditions are crucial for thermal control of the process.
In interval II the total surface area of the particles increases. As a
consequence the fractional surface coverage (q) of the particles with
emulsifier decreases and with that the repulsive forces between the particles
also decrease. Occasionally particle coagulation may occur in interval II
because repulsive forces become too small to compensate the attractive Van
der Waals forces, see e.g. Hiemenz (1986), Kemmere et al. (1998b) and
references cited in that paper. Colloidal stability will be lost if the fractional
surface coverage of the particles with emulsifier falls below a critical value
(qcrit). Coagulation of particles continues until qcrit is reached again. This
controlled process of growing and coagulating particles continues until the
monomer droplets have disappeared, or in the case of a semi-batch
operation the monomer feeding has stopped. Particle growth by simultan-
eous reaction and monomer absorption stops and the fractional surface
coverage of the particle surface with emulsifier remains approximately
constant in the remaining part of the polymerisation.
For emulsion polymerizations using ionic surfactants the amount of
electrolyte coming from e.g. the initiator, the surfactant and the pH buffer,
is an important factor for colloidal stability. The electrostatic repulsion
forces between the particles strongly decrease with an increasing electrolyte
concentration, see e.g. Hiemenz (1986). As a consequence more surfactant
per unit of particle surface area is required for colloidal stability at higher
electrolyte concentrations; the value of qcrit strongly increases with the
electrolyte concentration.
Particle coagulation leads to less particles and therefore the rate of
polymerisation decreases for emulsion polymerisations where the average
number of growing chains per particle is independent or slightly dependent
of the particle size.
Loss of colloidal stability may result in off-spec products and
troublesome operation. Sometimes controlled coagulation can be used to
control particle size. Some attention will be paid to controlled coagulation
and fouling.
 264 Meuldijk et al.

REACTION CALORIMETRY

The Mettler –Toledo RC1e reaction calorimeter measures the rate of

heat production by reaction (Qr) as a function of time. The total heat
produced during the reaction is:
Z 1
Total heat ¼ Qr ðtÞdt ð2Þ
0
Downloaded by [University of Stellenbosch] at 04:11 26 July 2013

From the amount of monomer (recipe) and the heat of reaction, the
calorimetric conversion (Xcal) time history can be calculated. The
calorimetric conversion is defined to be one when heat production stops.
The calorimetric conversion is equal to the monomer conversion for
situations where heat production stops at the moment that all monomer is
converted into polymer. Figure 4 demonstrates how heat production (or
consumption) by reaction as measured by the RC1e can be transformed into
conversion time histories and heat production as a function of conversion.

Figure 4. Heat production rate Qr as a function of time and conversion time history
as derived from calorimetric data.
 Emulsion Polymerization Process Development 265

Excellent papers about the application of reaction calorimetry in

emulsion polymerisation research are amongst others published by
Urretabizkaia et al. (1993) and Varela de la Rosa et al. (1999a – c).

EMULSIFICATION

Emulsification of styrene and vinyl acetate without polymerisation is

investigated by means of visualisation experiments. The reaction mixture is
Downloaded by [University of Stellenbosch] at 04:11 26 July 2013

assumed to be properly emulsified if no clear liquid layers can be

distinguised, see Figure 5 experiment D. The impeller speed at which the
clear liquid layer has just disappeared is called Nvis.
General conclusions from the visualisation experiments are (Kemmere
et al., 1999a):

. Emulsification of styrene water mixtures requires more energy per

unit of mass than emulsification of vinyl acetate water mixtures,
see also Fontenot and Schork (1993).
. The monomer to water volume ratio has no significant influence on
Nvis for monomer to water volume ratios between 0.3 and 1.
. Emulsifier considerably reduces the value of Nvis for emulsifier
concentrations below the CMC. Above the CMC there is no
significant influence.
. Impellers perform poorly when positioned at the monomer water
interface.
. For the same reactor geometry turbine impellers ask less energy
than pitched blade impellers.

Figure 5. Visualisation experiments to determine Nvis necessary for sufficient

emulsification. Scale: 10 dm3, six bladed rushton turbine impeller, 4 baffles,
dimpeller = 0.33Dvessel, no emulsifier added. monomer: styrene. A: 100 rpm, B: 150
rpm, C: 200 rpm and D: 320 rpm = Nvis.
 266 Meuldijk et al.

. A constant power input into the monomer water mixture per

unit of mass due to stirring is a safe scale-up rule for proper
emulsification.

NUCLEATION

Emulsification and Nucleation

Downloaded by [University of Stellenbosch] at 04:11 26 July 2013

Figure 6 shows the influence of the impeller speed on the conversion

time histories, the development of the volume averaged particle size with
time and the particle size distribution of the final product for the ab-initio
emulsion polymerisation of styrene. The results in Figure 6 demonstrate
that for impeller speeds of 250 and 275 rpm the reaction apparently stops at
a conversion of about 0.4 and 0.7, respectively. For these impeller speeds

Figure 6. Conversion time history, volume averaged particle size (dp,v) as a function
of reaction time and particle size distribution of the final latex product of ab-initio
emulsion polymerisations of styrene at various impeller speeds: 5: 250; .: 275; 5:
360 = Nvis; ^: 500; 4: 800 rpm. Reaction volume: 1.85 dm3; dimpeller = 0.33Dvessel.
Temperature: 50
C. Initial concentrations (kmol/m3water): styrene: 3.2; sodium
persulphate: 0.010; sodium dodecyl sulphate: 0.020; sodium carbonate: 9.4510 3.
 Emulsion Polymerization Process Development 267

the reaction mixture is poorly emulsified. A styrene layer is still present on

top of the reaction mixture. Participation of this layer in the reaction seems
negligible. For impeller speeds of 360 and 800 rpm complete conversion is
reached, and the particle size distributions are approximately the same. The
results point out that the polymerisation obeys intrinsic polymerisation
kinetics for impeller speeds N Nvis!
Because of the mass transfer limitation for N < Nvis the volumetric
growth rate of the particles is lower than for N Nvis. As a consequence
the consumption of emulsifier by adsorption onto the surface of the growing
Downloaded by [University of Stellenbosch] at 04:11 26 July 2013

particles becomes less. So more micellar emulsifier is available for particle

nucleation over a longer period of time. Consequently, the stage of particle
nucleation for N < Nvis lasts longer than for N Nvis, resulting in a
broader particle size distribution and a higher particle concentration.
The total amount of heat produced by reaction for situations with poor
emulsification is less than the total heat produced by a similar reaction, i.e.
the same recipe, with proper emulsification, see Figure 7. The results
obtained with reaction calorimetry (Kemmere et al., 2001) clearly
demonstrate that for impeller speeds lower than 400 rpm significantly less
heat is generated by the polymerisation than for impeller speeds of 425 rpm
and higher. For impeller speeds lower than 400 rpm the results point out
that the polymerisation reaction has apparently stopped at a monomer
conversion level significantly lower than one. The emulsification is poor for
impeller speeds lower than 400 rpm; part of the monomer is in a separate
layer on top of the reaction mixture and does not participate in the reaction.
For impeller speeds 425 rpm the total heat production by reaction

Figure 7. Ab-initio emulsion polymerisation of styrene in a Mettler-Toledo RC1e

(HP60 reaction vessel). Heat production as a function of time (left) and the fractional
calorimetric conversion (right) for different impeller speeds. Six bladed turbine
impeller; dimpeller = 0.33Dvessel. Temperature: 50
C. Monomer weight fraction in the
recipe 0.25 kg/kgreaction mixture. Other initial concentrations: sodium persulphate:
0.010; sodium dodecyl sulphate: 0.010; sodium carbonate: 910  3. –—: 300 rpm;
-----: 375 rpm; – – – : 425 rpm; ——: 700 rpm. Proper emulsification for impeller
speeds 425 rpm!
 268 Meuldijk et al.

remains the same within experimental error. Approximately complete

monomer conversion is obtained at complete calorimetric conversion. For
impeller speeds higher than 425 rpm the reaction mixture is obviously
properly emulsified.
The emulsion polymerisation experiments in the RC1e performed with
different impeller speeds demonstrate that with only a few experiments,
optimal operational conditions can be chosen for proper emulsification
during the emulsion polymerisation process.
Downloaded by [University of Stellenbosch] at 04:11 26 July 2013

Isothermal Processes at Different Temperatures

The course of the reaction expressed by the heat production rate time
history, the particle concentration and the particle size distribution strongly
depend on the reaction temperature, see Figure 8. In Figure 8 the results are
collected for the emulsion polymerisation of styrene performed isothermally

Figure 8. Isothermal ab-initio emulsion polymerisation of styrene in a Mettler

Toledo RC1e (HP60 reactor). Influence of reaction temperature on the heat production
as a function of time (top) and the particle size distribution of the final product
(bottom). Initial concentrations (kmol/m3water): styrene: 0.644; sodium persulphate:
0.014; sodium dodecyl sulphate: 0.036. A: temperature 40
C, particle concentration:
1.31021 1/m3water. B: temperature 50
C, particle concentration: 2.11021 1/m3water. C:
temperature 75
C, particle concentration: 5.21021 1/m3water.
 Emulsion Polymerization Process Development 269

at 40, 50 and 75
C. The results demonstrate that the particle concentration
strongly increases with temperature. There are two counteracting effects
which ultimately govern the influence of temperature on the final particle
concentration: the influence of temperature on the radical formation by
thermal dissociation of the initiator and the influence of temperature on
the rate of polymerisation. The radical production rate and by that the
nucleation rate of the available micelles increase strongly with temperature.
On the other hand the rate of polymerisation and as a consequence the
particle growth rate also increase with temperature. When the particles
Downloaded by [University of Stellenbosch] at 04:11 26 July 2013

grow faster, the particle surface increases also faster and colloidal stability
asks for more emulsifier adsorption per unit of time. An enhanced rate of
polymerisation has a negative effect on the amount of nucleated particles
because less emulsifier micelles are available for the nucleation process.
The influence of temperature on the final particle concentration is domi-
nated by the enhanced rate of radical production. The increase of the rate of
propagation with temperature for vinyl monomers is much less than the
increase of the rate of radical production. This different behaviour is due to
the differences in activation energy: 140 kJ/mol for thermal dissociation of
persulfate initiators (Rawlings and Ray, 1988) and e.g. 32.3 kJ/mol for the
propagation rate coefficient of styrene (Manders et al., 1996). Higher
temperatures result in narrower particle size distributions, because higher
temperatures are accompanied by a smaller elapse of time between the start
and the end of the nucleation process.

Influence of Starting Temperature and Heating Rate in

Non-isothermal Processes

In a large scale ab-initio emulsion polymerisation usually all ingre-

dients, including the initiatior, are supplied to the reactor and the reaction
mixture is heated up to the desired temperature. The initial temperature of
the reaction mixture and the heating rate are important factors for the
particle concentration and the particle size distribution at the end of the
nucleation stage and in the final product, see Figure 9.
The results with the RC1e, see Figure 9, demonstrate that the heating
rate has a large influence on the course of the emulsion polymerisation
process in terms of polymerisation rate. However, for the same starting
temperature but different heating rates (experiments A and B) the widths of
the final particle size distribution are only slightly different. Note that the
particle concentration for experiment A, heating from 50
C to 90
C in 15
minutes, is considerably larger than for experiment B, where the reaction
mixture is heated from 50
C to 90
C in 90 minutes. Obviously the thermal
initiator decomposition rate is much higher and provides considerably more
 270 Meuldijk et al.
Downloaded by [University of Stellenbosch] at 04:11 26 July 2013

Figure 9. Ab-initio emulsion polymerisation of styrene in a Mettler Toledo RC1e

(HP60 reactor). Influence of starting temperature and heating rate on the heat
production as a function of time (top) and the particle size distribution of the final
product (bottom). Initial concentrations (kmol/m3water): styrene:0.644; sodium
persulphate: 0.014; sodium dodecyl sulphate: 0.036. A: starting temperature 50
C,
final temperature 90
C, reached in 15 minutes, particle concentration: 3.21021
1/m3water. B: starting temperature 50
C, final temperature 90
C, reached in 60
minutes, particle concentration: 2.11021 1/m3water. C: starting temperature 30
C, final
temperature 90
C, reached in 90 minutes, particle concentration: 2.11021 1/m3water.

radicals in the nucleation stage in experiment A than in experiment B. As a

consequence the radical flux to micelles is considerably larger in experi-
ment A than in experiment B, resulting in a higher particle concentration in
experiment A than in experiment B. For experiment C, heating from 30
C
to 90
C in 90 minutes, the particle size distribution is considerably broader
than for the experiments A and B. In experiment C the nucleation stage
lasts relatively long as a result of the relatively low radical formation rate
as well as the low polymerisation rate in the initial phase of the reaction,
where the temperature is low. As a consequence a rather broad partice size
distribution appears in experiment C.
The experimental results as collected in Figure 9 demonstrate that the
starting temperature as well as the heating rate may have a considerable
influence on the particle concentration and/or the particle size distribution of
the final product. With the RC1e it is perfectly possible to accurately adjust
 Emulsion Polymerization Process Development 271

the heating rate. Therefore reaction calorimetry together with the determina-
tion of the particle concentration and the particle size distribution of the final
product enable a rather quick development of an optimal procedure for the
start of an emulsion polymerisation process leading to specific product
properties in terms of particle concentration and particle size distribution.

COAGULATION
Downloaded by [University of Stellenbosch] at 04:11 26 July 2013

The influence of the electrolyte concentration on the heat production

rate time history and the rate of polymerisation as a function of time for the
seeded emulsion polymerisation of styrene is shown in Figure 10.
The results in Figure 10 demonstrate that for conversions below 0.5 the
rate of polymerisation for CNa+ = 0.38 kmol/m3 is significanly lower than
for CNa+ = 0.05 kmol/m3. The reason for this is controlled coagulation in
interval II for CNa+ = 0.38 kmol/m3. The number of particles decreases
with more than a factor of two and the mean particle size in the final latex
product increases considerably on increasing CNa+ from 0.05 kmol/m3 to
0.38 kmol/m3. Also the particle size distribution changes from a distribution
with tailing at the small particle size side of the distribution into a particle
size distribution with tailing at the large particle size side. The much higher
rate of polymerisation at higher conversion for the recipe with CNa+ = 0.38
kmol/m3 originates from the gel effect, which becomes more pronounced
for larger particles.
Experimental work on seeded emulsion polymerisation of styrene
reported by Kemmere et al. (1998b, 1999b) pointed out that the impeller
speed, the impeller type and diameter as well as the scale of operation do
not have a significant influence on qcrit, the fractional surface coverage of
the particles with emulsifier just sufficient to prevent coagulation. Therefore
there is no significant influence of these parameters on the colloidal stablity
of an emulsion polymerisation reaction mixture.
For monomer to water ratios up to one and small to moderate
electrolyte concentrations, colloidal stability does not depend on the ratio
of the volume of the monomer swollen particle phase and the volume of
the water phase in the reaction mixture (Kemmere et al., 1999b). In
emulsion polymerisation processes with controlled coagulation in interval
II Brownian coagulation dominates over shear coagulation, because the
polymer particles are much smaller than the smallest eddies in the flow
field, i.e. the Kolmogorov microscale of isotropic turbulence. The results of
Kemmere et al. (1998b, 1999b) seem to be in contradiction with the results
of Lowry et al. (1984, 1986). However, Lowry and coworkers reported the
results of ab-initio experiments in which perhaps some droplet nucleation
has taken place, resulting in very few particles with a diameter considerably
 272 Meuldijk et al.
Downloaded by [University of Stellenbosch] at 04:11 26 July 2013

Figure 10. Seeded emulsion polymerisation of styrene carried out in a Mettler

Toledo RC1e reaction calorimeter, operated in the isothermal mode (50
C).
Emulsifier used: sodium dodecyl sulphate. Influence of the cation (Na+) con-
centration on the heat production time history (top left), on the rate of poly-
merisation as a function of conversion (top right) and the particle size distribution
of the final latex product (bottom). Weight fraction styrene: 0.15. Initial fractional
surface coverage of the monomer swollen particles with emulsifier: 0.8. Particle
diameter of the unswollen polystyrene seed: 41 nm. Cation concentration adjusted
with sodium chloride. ———: CNa + = 0.05 kmol/m3water; – – – – – : CNa + = 0.38
kmol/m3water. A: CNa + = 0.05 kmol/m3water, particle concentration: 0.421021
1/m3water, particle diameter, dp,v = 87 nm, fractional surface coverage with
emulsifier: 0.31. B: CNa + = 0.38 kmol/m3water, particle concentration: 0.181021
1/m3water, dp,v = 120 nm, fractional surface coverage with emulsifier: 0.38 ffi qcrit.

larger than one micrometer. Lowry and coworkers also reported the pres-
ence of large flocs. Coagulation of such large particles or flocs with the
submicron latex particles proceeds mainly by shear coagulation, which can
explain the observed influence of stirring characteristics.
It should be noted that high solids recipes are more sensitive to fouling
than low solids recipes. The occurence of fouling was observed by a
tremendous increase of the power input into the reaction mixture due to
stirring, see Figure 11 (Kemmere et al., 1998a; Meuldijk et al., 1998). The
power input into the reaction mixture due to stirring (P) was derived from the
 Emulsion Polymerization Process Development 273
Downloaded by [University of Stellenbosch] at 04:11 26 July 2013

Figure 11. Volume averaged energy dissipation per kg reaction mixture based on
torque measurement as a function of conversion for a seeded emulsion polymerisation
of styrene. Volume of the reaction mixture: 7.48 dm3. Temperature: 50
C. Monomer
weight fraction in the recipe 0.5 kg/kgreaction mixture. Tank (volume 10 dm3) equipped
with a six bladed turbine impeller, dimpeller = 0.33Dvessel, tank height = tank
diameter, clearance of the impeller from the tank bottom 0.5tank height.

measured torque on the impeller shaft (T), see Eq. 3. In Figure 11 the
volume averaged power input per unit of mass has been plotted as a function
of conversion. Note that the increase of the power input due to stirring
started for monomer conversions larger than 0.6. The increase of the power
input due to stirring obviously results from an increase of the apparent
impeller diameter by deposition of polymeric material on the impeller
blades. Eq. 3 gives the power input due to stirring (P) as a function of the
dimensionless power number (NP), the density of the reaction mixture (r),
the impeller speed (Nimp) and the impeller diameter (dimp).

P ¼ 2pNimp T ¼ NP rN3imp d5imp ð3Þ

Eq. 3 demonstrates that a small increase of dimp by polymer deposition

has a large influence on P!

CONCLUDING REMARKS

Reaction calorimetry together with an accurate method for measuring

the particle size distribution such as transmission electron microscopy,
 274 Meuldijk et al.

has been demonstrated to be a very powerful tool for quick process

development of complex chemical processes such as emulsion polymerisa-
tion. The course and the outcome of the nucleation process is very sensitive
to operation and scale. Proper emulsification is very important for a good
control of the particle size distibution in the final product. In addition the
particle concentration and the particle size distribution are very sensitive to
the operating temperature for isothermal operation. The particle concentra-
tion and the particle size distribution are in many cases stronly dependent
on the starting temperature and the heating rate if the ab-initio emulsion
Downloaded by [University of Stellenbosch] at 04:11 26 July 2013

polymerisation process is started at a lower temperature than the tempe-

rature at which the main part of the polymerisation proceeds.
Reaction calorimetry enables a quick definition of an operating window
for complete control of the nucleation process.
The risk for loss of colloidal stability can easily be mapped by a few
emulsion polymerisation experiments in a reaction calorimeter. To avoid
floc formation during emulsion polymerisation, the radical polymerisation
should strictly be restricted to the latex particles. Droplet polymerisation
should be avoided.
A sudden increase of the power input to the reaction mixture due to
stirring during the reaction points to fouling.

ACKNOWLEDGMENTS

The authors wish to thank NeoResins BV, Waalwijk, The Netherlands

and the Foundation Emulsion Polymerization (SEP) for financial support of
this work.

REFERENCES

Fontenot, K., Schork, F. J. (1993). Sensitivities of droplet size and stability in

monomeric emulsions. Ind. Eng. Chem. Res. 32:373 –385.
Hiemenz, P. C. (1986). Principles of Colloid and Surface Chemistry. New
York: Marcel Dekker.
Kemmere, M. F. (1999). Batch emulsion polymerization: a chemical engi-
neering approach. Ph.D. Thesis, Eindhoven University of Technology.
Kemmere, M. F., Meuldijk, J., Drinkenburg, A. A. H., German, A. L.
(1998a). Rheology and flow during high solids emulsion polymeriza-
tion of styrene. Polym. React. Eng. 6(3 & 4):243 – 268.
Kemmere, M. F., Meuldijk, J., Drinkenburg, A. A. H., German, A. L.
 Emulsion Polymerization Process Development 275

(1998b). Aspects of coagulation during emulsion polymerization of

styrene and vinyl acetate. J. Appl. Polym. Sci. 69:2409– 2421.
Kemmere, M. F., Meuldijk, J., Drinkenburg, A. A. H., German, A. L.
(1999a). Emulsification in batch emulsion polymerization. J. Appl.
Polym. Sci. 74:3225 – 3241.
Kemmere, M. F., Meuldijk, J., Drinkenburg, A. A. H., German, A. L.
(1999b). Colloidal stability of high solids polystyrene and polyvinyl
acetate latices. J. Appl. Polym. Sci. 74:1780– 1791.
Kemmere, M. F., Meuldijk, J., Drinkenburg, A. A. H., German, A. L. (2000).
Downloaded by [University of Stellenbosch] at 04:11 26 July 2013

Heat transfer in batch emulsion polymerization. Polym. React. Eng.

8(3):271 – 297.
Kemmere, M. F., Meuldijk, J., Drinkenburg, A. A.H., German, A. L.
(2001). Emulsification in batch emulsion polymerization of styrene
and vinyl acetate: a reaction calorimetric study. J. Appl. Polym. Sci.
79:944– 957.
Lowry, V., El-Aasser, M. S., Vanderhoff, J. W., Klein, A. (1984). Mechanical
coagulation in emulsion polymerization. J. Appl. Polym. Sci. 29:3925–
3935.
Lowry, V., El-Aasser, M. S., Vanderhoff, J. W., Klein, A., Silebi, C. A.
(1986). Kinetics of agitation induced coagulation of high solid latexes.
J. Colloid Interface Sci. 112(2):521– 529.
Manders, B. G., Chambard, G., Kingma, W. J., Klumperman, B., van Herk,
A. M., German, A. L. (1996). Estimation of activation parameters for
the propagation rate constant of styrene. J. Polym. Sci., A, Polym.
Chem. 34:2473 –2479.
Mayer, M. J. J., Meuldijk, J., Thoenes, D. (1994). Emulsion polymerization
in various reactor types: recipes with high monomer contents. Chem.
Eng. Sci. 49(24B):4971 –4980.
Meuldijk, J., van den Boomen, F. H. A. M., Kemmere, M. F., Wijers, J. G.
(1998). Scale sensitivity of emulsion (Co)polymerization. Entropie
212/213:13 –17.
Poehlein, G. W. (1997). Reaction engineering for emulsion polymerization.
In: Asua, J. M., ed. Polymeric Dispersions: Principles and Applica-
tions. Kluwer Academic Publishers, pp. 305– 330.
Rawlings, J. B., Ray, W. H. (1988). The modeling of batch and continuous
emulsion polymerization reactors. Part II: comparison with experi-
mental data from continuous stirred tank reactors. Polym. Eng. Sci.
28:257– 274.
Smith, W. V., Ewart, R. H. (1948). Kinetics of emulsion polymerization. J.
Chem. Phys. 16:592 – 599.
Urretabizkaia, A., Sudol, E. D., El-Aasser, M. S., Asua, J. M. (1993).
 276 Meuldijk et al.

Calorimetric monitoring of emulsion copolymerization reactions. J.

Polym. Sci., A, Polym. Chem. 31:2907 –2914.
Varela de la Rosa, L., Sudol, E. D., El-Aasser, M. S., Klein, A. (1999a).
Emulsion polymerization of styrene using reaction calorimeter I: above
and below critical micelle concentration. J. Polym. Sci., A, Polym.
Chem. 37:4054– 4065.
Varela de la Rosa, L., Sudol, E. D., El-Aasser, M. S., Klein, A. (1999b).
Emulsion polymerization of styrene using reaction calorimeter II:
importance of maximum in rate of polymerization. J. Polym. Sci., A,
Downloaded by [University of Stellenbosch] at 04:11 26 July 2013

Polym. Chem. 37:4066 – 4072.

Varela de la Rosa, L., Sudol, E. D., El-Aasser, M. S., Klein, A. (1999c).
Emulsion polymerization of styrene using reaction calorimeter III:
effect of initial monomer/water ratio. J. Polym. Sci., A, Polym. Chem.
37:4073 –4089.

Received September 19, 2002

Accepted January 22, 2003