Professional Documents
Culture Documents
A R T I C L E I N F O A B S T R A C T
Article history: The emulsion polymerization of vinyl acetate monomer (VAM), ethylene and N-methylol acrylamide
Received 13 June 2009 (NMA) was carried out in a high pressure stirred reactor using an Rushton flat blade impeller at a
Accepted 18 December 2009 rotational speed of 300 rpm under a semi-batch polymerization process, which was subsequently grown
by the continuous addition of monomers (VAM, ethylene, and NMA). The particle diameters of the
Keywords: samples withdrawn during polymerization were determined by using capillary hydrodynamic
Emulsion polymerization fractionation (CHDF). Also, the number of particles was calculated from the data obtained by CHDF.
Particle distribution
Coagulum formation was studied. It was found that coagulum formation increased with increasing
Vinyl acetate-ethylene copolymer
Waterborne
conversion and auto-accelerated from ca. 20% conversion. The definition of coagulum in this research
Latex was presented. The coagulum formation pattern on the impeller was shown. Also, the flow pattern in the
reactor under the same geometrical configuration and operating conditions was simulated.
ß 2010 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V.
All rights reserved.
1. Introduction Bonardi et al. [3–7] found that the NMA units were preferentially
located on the particle surface as strongly adsorbed or grafted
In emulsion polymerization, mixing, which results in the bulk polymer. This study was applied of the approach method of the
movement of the fluid, is very important for maintaining the agitation effects in emulsion copolymerization of n-butyl methac-
homogeneity of the reaction mass. Good mixing is essential to rylate and N-methylol acrylamide [8] at the vinyl acetate, ethylene
prevent high local concentrations of added surfactants and water- and N-methylol acrylamide polymerization. Agitation may affect
soluble monomers in the case of a semi-batch reaction where the degree of incorporation of the water-soluble NMA monomer in
certain components of the reaction are added continuously. These the polymer particles. Therefore, emulsion properties such as
effects are especially important in large volume reactors. Also, cross-link density and morphology can be affected by the agitation
mixing is essential to promote the mass transfer of monomer when conditions [9]. The purpose of this study are to research for trend
the monomer has relatively high water solubility. The shear particles and grits by agitation and to be used the different impeller
component of agitation is another important factor that strongly and rpm to check for the trend of particles and grits.
influences coagulum formation in high-solids emulsion [1]. Shear
can accelerate coagulation through the development of fluid 2. Experimental
velocity gradients, which can increase the force of collisions
between particles present in the reactor as well as the frequency of 2.1. Materials
the collisions.
Emulsion copolymers of vinyl acetate-ethylene emulsion are 15 ppm mono-methyl ether of hydroquinone (MEHQ) inhibitor
widely used in adhesive, coating and nonwoven application was removed from vinyl acetate monomer (VAM, Samsung BP) by
industry however this emulsion has a little lower water resistance, passing the monomer through an inhibitor-removal column
so that various applications want grafting polymer like as N- (Sigma–Aldrich). Ethylene monomer (SK-Chemical) was used as
methylol acrylamide (NMA), it is widely used not only coating received. N-methylol acrylamide (NMA, Cytec) was received as a
application but also adhesive application [2]. NMA acts as a cross- 48 wt% solution in water and was inhibited with 5 ppm of MEHQ.
linking agent in the synthesis of reactive emulsion. For example, Sodium lauryl sulfate (SLS, Dongnam chemical), potassium
persulfate (KPS, Decksung chemical), and sodium bicarbonate
(Decksung chemical) were used as surfactant, initiator, and buffer,
* Corresponding author. Tel.: +82 051 629 6451. respectively. All of the chemicals except VAM were used as
E-mail address: wonki@pknu.ac.kr (W.-K. Lee). received.
1226-086X/$ – see front matter ß 2010 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.
doi:10.1016/j.jiec.2010.01.042
432 Y.-H. Choi, W.-K. Lee / Journal of Industrial and Engineering Chemistry 16 (2010) 431–436
Fig. 1. Schematic diagram of the experimental setup used to study the effect of agitation during emulsion copolymerization.
2.2. Emulsion synthesis and reaction kinetics NaHCO3, and DI water was added together into the reactor at the
beginning of the feeding stage reaction. Ethylene was fed using a
The emulsion was being synthesized in an 18 l high pressure compressor from a bombe, precision syringe pump of NMA at a rate
and temperature-controlled stirred reactor with two baffles. A of 3 ml/min. The aqueous feed, which was composed % of DI water/
stirrer, which maintains a constant speed irrespective of changes in NMA solution/SLS (18.31/5.03/0.55), it was using a syringe pump
the viscosity of the reaction fluid, was used to provide agitation. at a rate of 0.29 ml/min. The reactions were carried out using an
The reactor lid had a sampling port for the injection of the initiator 10 cm Rushton agitator at a rotational speed of 300 rpm. About
and withdrawal of samples from a region next to the impeller 20 ml of emulsion sample were withdrawn from the reactor at
blade. The schematic diagram of the experimental setup is shown 30 min intervals during the feeding stage. Instantaneous conver-
in Fig. 1. sion was calculated to be greater than 94% from about 30 min into
The recipe was used for the emulsion copolymerization of VAM, the reaction onward. Hence, the semi-continuous batch polymeri-
ethylene and NMA is given in Table 1. The final solids content is ca. zation was carried out under monomer-starved conditions.
55.61%. A mixture of VAM, SLS, NaHCO3, and deionized (DI) water
was heated to the reaction temperature (80 8C) with nitrogen 3. Results and discussion
bubbling. The seeding stage reaction was continued for 60 min
after injection of the initiator solution. The mixture of KPS, Fig. 2 shows the overall conversion as a function of time for the
feed stage reactions using the recipe shown in Table 1. The
Table 1 beginning of the feed stage monomer conversion corresponded to
VAM/ethylene/NMA emulsion polymerization recipe. the conversion of the seed stage at 60 min. The conversions were
Stage Ingredient Mass (%)
Table 2
Comparison of the particle size and number of particles of emulsion Run Number 2
prepared using the recipe given in Table 1.
Dw Dn PDI
Fig. 3. Evolution of particle diameter (Dn) as a function of feed time in the semi-
determined by gravimetry. There was good agreement among the batch emulsion copolymerization of VAM/ethylene/NMA.
conversion-time profiles for three runs of the reaction, which
indicates good reproducibility.
particles with relatively large sizes. Therefore, the number of the
3.1. Particle size analysis and evolution of particle diameter particles calculated based on the measured size can be artificially
smaller than the actual values.
The particle size and size distribution in the final emulsion Fig. 3 shows the evolution of particle size as a function of feed
obtained at the end of the seed stage as well as samples taken at time in the semi-batch emulsion copolymerization of VAM/
intervals of 30 min during the feed stage were measured by ethylene/NMA. The data obtained from CHDF is compared with
capillary hydrodynamic fractionation (CHDF 1100, Matec Applied the theoretical curve, which is calculated from Eq. (2).
Sciences). From Table 3 and Fig. 3, it can be seen that the calculated
Table 2 shows the weight-average diameter (Dw), the number- diameter of the particles based on the seed particle size measured
average diameter (Dn), the polydispersity index (PDI (Dw/Dn)), and by CHDF does not give accurate results with increasing conversion,
the number of particles (Np) of emulsion Run Number 2 with whereas the calculated diameter of particles based on the seed
increasing time. The number of particles (Np, dm3 water) was particle size measured by CHDF gives good agreement with the
calculated using the following equation: experimental values. This might be attributed to the existence of
the electrical double layer on the particle surface where its
6rw M x Mx thickness could be increased by adding the hydrophilic monomer
Np ¼ ¼ 1:81 1021 3 (1)
prPD3n Dn (NMA) in the feed stage.
Fig. 7. Pattern formed by the coagulum on the blade of the agitator (a), simulated
flow pattern with a Rushton turbine blade (b), schematic three-dimensional views
of the trailing vortex pair (c) [10].
Table 5
Evolution of particle diameter and number of particles at 300 rpm, 400 rpm, and 500 rpm using the 8 cm diameter fluid-foil impeller.
Seed 60 202 1.004 1.1 198 1.006 1.2 215 1.005 0.9
Feed 30 284 1.005 1.3 295 1.004 1.2 327 1.004 0.9
Feed 60 341 1.004 1.4 356 1.004 1.2 382 1.006 1.0
Feed 90 378 1.005 1.5 398 1.005 1.3 432 1.006 1.0
Feed 120 412 1.008 1.4 439 1.005 1.2 463 1.006 1.0
Feed 150 459 1.006 1.3 465 1.005 1.3 500 1.005 1.0
Table 6
Evolution of particle diameter and number of particles at 300 rpm, 400 rpm, and 500 rpm using the 8 cm diameter Rushton turbine impeller.
Seed 60 201 1.006 1.0 201 1.006 1.1 216 1.004 0.9
Feed 30 299 1.003 1.0 301 1.004 1.2 332 1.004 0.9
Feed 60 352 1.003 1.2 361 1.004 1.2 381 1.006 1.0
Feed 90 390 1.005 1.3 395 1.005 1.3 439 1.006 0.9
Feed 120 431 1.005 1.3 441 1.005 1.2 481 1.005 0.9
Feed 150 470 1.005 1.2 474 1.005 1.2 524 1.004 0.9
developed mixing is 5 s for the pitched blade. Some dead zones are lowest turbulence. Therefore, a higher collision frequency between
observed near the blade in the case of the Rushton turbine particles using the Rushton impeller led to larger particle size.
impeller. The zones may be caused by trailing vortices, which can It was observed that the seed particle sizes for both impeller
have a major effect on coagulum formation. types were very similar at 300 rpm and 400 rpm, whereas at
500 rpm, the size was much larger than at the other speeds. The
3.4. Effect of extent of agitation on the particle size and particle seed particle size also differed with impeller diameter. The number
number of particles decreased with increasing impeller diameter and
speed. There are two possible explanations for this. One involves
Tables 5–7 show the volume-average diameter (Dv), the the shear stress and its effect on the homogeneous-coagulative
polydispersity index (PDI = Dw/Dn), and the number of particles nucleation mechanism. The latter involves the precipitation of
(Np) of the emulsions synthesized under different agitation growing radicals (primary precursors) to form nuclei as well as the
conditions. aggregation of nuclei that were already generated unstably
From the CHDF results, mono-disperse particles are observed (precursors). More nuclei (precursors) can aggregate in the
for each sample with increasing time. The particle size of the beginning of the seed stage by the induced high shear stress
emulsion increased with the degree of agitation for both impeller (500 rpm), which is proportional to the square of the impeller
types and smaller numbers of particles were produced at higher speed and, eventually, the number of particles is decreased. The
impeller speeds. The particle sizes of the emulsions synthesized second explanation involves the distribution of the surfactant. The
using the Rushton impeller, are generally larger than those higher agitation intensity produces smaller droplet sizes and thus
prepared using the fluid-foil impeller, most notably for the final more droplets. As a result, the surface area of monomer droplets is
emulsions. This is because the Rushton turbine impeller is more increased, resulting in more surfactant adsorption. Fewer surfac-
dominant in shear than the fluid-foil impeller. Velocity-time tant molecules are thus available to stabilize the nucleated
profiles for three different impellers (Rushton turbine, pitched polymer particles. Hence, higher agitation produces a smaller
blade (458), and fluid-foil impeller), which indicate the degree of number of particles. However, to fully explain the mechanism
turbulence [13]. The Rushton turbine impeller produced the during the entire polymerization, a combination of the two
highest turbulence, while the fluid-foil impeller produced the explanations might be in order, namely, fluid mechanics (i.e.,
hydrodynamic and colloidal interactions) and surface science (i.e.,
adsorbed surfactant on particles).
With increasing impeller speed, especially for the Rushton
Table 7
turbine impeller at 500 rpm, the coagulation rate is seen to change
Evolution of particle diameter and number of particles at 300 rpm using the 6 cm
diameter Ruton turbine impeller. with time. The several rate stages in the stirred tank coagulation
phenomena. The primary rate stage was associated with collisions
Time (min) 300 rpm
among small particles [14]. The secondary stage displayed an
Size (nm) Np (1015 dm3 H2O) accelerated coagulation rate with large, porous particles resulting.
Dv PDI The collision efficiency is enhanced by penetration of fluid flow
into the aggregates which attain large porosities [15]. Also, the
Seed 60 199 1.006 1.2
Feed 30 298 1.004 1.2 capture efficiency in hete-coagulation (collisions between dissim-
Feed 60 352 1.004 1.3 ilar size particles), is much higher than in homo-coagulation [16].
Feed 90 382 1.006 1.5 Therefore, coagulum formation is affected not only by static
Feed 120 425 1.006 1.4
factors, but also by dominantly dynamic factors such as large
Feed 150 458 1.005 1.3
particle size.
436 Y.-H. Choi, W.-K. Lee / Journal of Industrial and Engineering Chemistry 16 (2010) 431–436
4. Conclusions Acknowledgment
The number of particles was found to decrease with increasing This work was supported by the Pukyong National University
degree of agitation, and was correlated to impeller diameter, Research Abroad Fund in 2008 (PS-2008-052).
speed, and type. Two explanations are being considered. First, the
References
higher agitation intensities produce an increased surface area of
monomer droplets, affecting surfactant adsorption. Thus, higher [1] V. Lowry, M.S. El-Aasser, J.W. Vanderhoff, A. Klein, C.A. Silebi, J. Colloid Interface
agitation can produce smaller numbers of particles because fewer Sci. 112 (1986) 521.
surfactant molecules are available to stabilize the nucleated [2] H.W. Park, J.-D Kim, J. Ind. Eng. Chem. 15 (2009) 578.
[3] C. Bonardi, Ph. Christou, M.F. Llauro-Darricades, J. Guillot, A. Guyot, C. Pichot, New
polymer particles. Second, these results might be explained by Polym. Mater. 2 (1991) 295.
shear stress and the homogeneous-coagulative nucleation [4] S.D. Tobing, A. Klelin, J. Appl. Polym. Sci. 79 (2001) 2230.
mechanism. This mechanism is not confined only to the [5] J. Garrett, P.A. Lovell, A.J. Shea, D.R. Viney, Macromol. Symp. 151 (2000) 487.
[6] C. Laureau, M. Vicente, M.J. Barandiaran, J.R. Leiza, J.M. Asua, J. Appl. Polym. Sci. 81
precipitation of growing radicals to form nuclei, but also the
(2001) 1258.
aggregation of nuclei that were previously generated. More nuclei [7] A. Klein, E.S. Daniels, in: P.A. Lovell, M.S. El-Aasser (Eds.), Emulsion Polymerization
will be aggregated in the beginning of the seed stage by induced and Emulsion Polymers, John Wiley and Sons, Chichester, 1997, p. 207.
high shear stress and, subsequently, the number of particles is [8] S. Krishnan, A. Klein, M.S. El-Aasser, E.D. Sudol, Polym. React. Eng. 11 (3) (2003) 335,
1532.
decreased. [9] S.H. Cha, J.C. Kim, J.S. Lim, C.S. Yoon, J.H. Koh, H.I. Chang, S.W. Kim, J. Ind. Eng.
It was found that coagulum formation increased with time and Chem. 12 (2006) 115.
impeller speed. Most of the coagulum is formed on the impeller, [10] R.G. Gilbert, Emulsion Polymerization: A Mechanistic Approach, Academic Press
Inc., San Diego, 1995.
shaft, and baffles. With increasing rotational speed of the impeller, [11] K. Van’t Riet, J.M. Smith, Chem. Eng. Sci. 30 (1975) 1093.
the differences in the amounts of coagulum in the emulsions [12] K. Van’t Riet, W. Bruijin, J.M. Smith, Chem. Eng. Sci. 31 (1976) 407.
between the different RPM were seen to increase. This is because a [13] J.Y. Oldshue, J. Chem. Eng. Prog. 85 (1989) 33.
[14] G.B.J. De Boer, Chem. Eng. Res. Des. 67 (1989) 301.
trailing vortex is created by the high velocity gradient for the [15] K.A. Kusters, J.G. Wijers, D. Thoenes, Chem. Eng. Sci. 52 (1977) 107.
Rushton turbine impeller. [16] S.G. Mason, T.G.M. van de Ven, Colloid Polym. Sci. 255 (1977) 468.