RAMAN APPLICATION IN EMULSION POLYMERIZATION SYSTEMS - Raman Spectroscopy for Soft Matter Applications - page 126

ment costs and lifetime). Sometimes it is better to install a HeNe laser
at 633 nm (which is cheap and long lasting) instead of a solid- state laser
at 785 nm (or 835 nm), even if the 785-nm laser would be better in terms
of reducing the effect of fl uorescence. Ease of maintenance, availability
of spare parts, and 24-hour support by the manufacturer are very impor-
tant issues to be considered for the installation of advanced instruments
in polymerization plants.
5.5.1.3 Requirements of the Probe The following points must be
considered: (1) in turbid mixtures, a confocal probe is a prerequisite for
getting spectra (focus of the lenses should be between 0.5 and 1 mm in
front of the window); (2) the probe material and the housing of the
probe have to withstand the reaction ingredients and the reactor pres-
sure and temperature for a long time; (3) the probe must be free of any
electrical contacts; and (4) the probe (i.e. specially the light path within
the probe) has to withstand the vibrations of an industrial reactor.
5.5.1.4 Position of the Probe in the Reactor Stirred - tank reac-
tors are typically used to produce emulsion polymers. Lab-scale reac-
tors and pilot-scale stirred-tank reactors can be safely assumed to be
perfectly mixed reactors and hence monitoring in any location of the
reactor should be representative of the state of the reactor. However,
in large-scale reactors and with high solids content formulations, this is
not necessarily true, as has been demonstrated by computational fl uid
dynamic simulations of the mixing behavior of viscous polymerization
reactors [90]. Therefore, the location of the Raman probe in a large
reactor is not straightforward. Figure 5.11 shows two potential locations
based on common openings of industrial reactors. The fi rst location is
near the impeller, using an opening on top of the reactor and protecting
tubing. This location will ensure a representative analysis of the whole
reactor content. Potential diffi culties with this setup include (1) diffi cult
maintenance, as accurately adjusting the position of the probe is diffi -
cult; (2) possibility of fouling due the presence of a protruding device;
and (3) high sensitivity to mechanical vibrations due to the length of
the tubing and the proximity to the impeller.
The second location is at the bottom of the reactor. The advantages
of this location are that it is easily accessible and hence maintenance
is possible, and that it is possible to build a device smooth enough to
avoid or minimize clogging. The main disadvantage of this setup is the
uncertainty of whether the latex analyzed at the bottom of the reactor
is representative of the whole content of the reactor. Computational
fl uid dynamic calculations performed in a polymerization reactor