Propylene
oxide (PO) homopolymer and PO/ethylene oxide (EO) copolymers,
initiated by the starter glycerol (GLY), are the most popular polyols
employed in the production of elastic polyurethanes. The molecular
weight (MW) of these polymers falls in the range 2000–6500
Da. The simulation of polymerization and copolymerization of EO and
PO to obtain high-molecular-weight polymers can hardly be made by
using a deterministic model that requires the solution of a system
of many differential and algebraic equations (many thousands in some
cases) with a high calculation time. A new application of a stochastic
model has recently been proposed by the authors that can simulate
EO and PO homo- and copolymerization in a few minutes of calculation
time that results much efficient especially in the case of copolymer
simulation. In other words, while deterministic and stochastic approaches
can be considered almost equivalent for homopolymerization, we found
that for complex alkoxylated copolymers the Monte Carlo (MC) method
is preferred. The method, already described in our previous work,
was employed to simulate the behavior of a bifunctional starter as
ethylene glycol (EG), giving not only the description of the time
profile of all the occurring reactions but also the microstructure
of the polymers obtained. In this work, the developed MC kinetic model
has been extended to the simulation of EO and PO homo- and copolymerization
using the trifunctional starter GLY. Different systems will be considered
as examples of simulation, such as the GLY-PO homopolymerization,
the GLY-EO-PO random copolymerization, and the GLY-EO-PO block copolymerization.
Results will be presented for the mentioned systems by considering
the effect of the different EO/PO molar ratios, the time evolution
of the alkoxide/substrate molar ratio, the occurrence of the PO to
allyl group side reaction with related consequences, and the molecular
weight distribution. Moreover, the model is also able to give detailed
information on the single-chain microstructure in terms of sequence
length distribution. A detailed description will be given, in this
work, about the selection of the most reliable kinetics parameters.