(32c) Lewis Acid Catalyzed Pyrolysis of Polystyrene in Low-Temperature Molten Salts: Experiment and Simulation

Brown, A., Worcester Polytechnic Institute
Timko, M. T., Worcester Poly Institute
Datta, R., Worcester Polytechnic Institute
Deskins, N. A., Worcester Polytechnic Institute

Waste polystyrene is a potential energy and chemicals
resource that is a major component of the plastics that are abundant in
municipal solid waste.  To valorize polystyrene waste, a major challenge is
obtaining high yield and selectivity of the styrene monomer.  Various reactor
designs and catalysts have been reported previously.  In our work, we have
studied pyrolysis of polystyrene in ternary salt mixtures at or near their
eutectic temperatures.  Specifically, we have focused our attention on ternary
mixtures of chlorides, with at least one component being ZnCl2, as
these mixtures have melting temperatures comparable to those used for
polystyrene unzipping, i.e., 300 °C and greater.  Molten salt mixtures have
several important advantages for polystyrene de-polymerization, notably
excellent heat transfer as well as potential for dispersion of acid and/or base
co-catalysts.   To improve performance further, we used an inert sweep gas that
removed volatile decomposition products as they were formed. 

First, we examined the benefits of polystyrene pyrolysis in
a 60/20/20 ZnCl2/NaCl/KCl molten salt mixture compared to thermal
pyrolysis, finding that styrene yield and selectivity improved in the presence
of the salt.  The benefit was most pronounced at lower temperatures, with
appreciable liquid yields (>60%) being achieved even at modest temperatures
of 350 °C where thermal pyrolysis was much less effective.  We then sought to
examine the effects of polystyrene loading, ZnCl2 addition, and
composition of the supporting electrolytes (e.g., ratio of NaCl and KCl and/or
addition of LiCl).  This work revealed that polystyrene pyrolysis in a 40/30/30
ZnCl2/NaCl/LiCl mixture could achieve an overall liquid yield of
over 90% and a styrene yield of 75%.  The molten salt pyrolysis performance
compares favorably to yields reported in the literature, for example in the
presence of an Fe-based catalyst or with microwave enhancement.  Brønsted acids
and bases were examined as catalysts.  ZSM-5 and phosphotungstic acid greatly
reduced the liquid yield, instead producing char.  Thermal conditions converted
Ni(OH)2 into NiO which was removed by the inert sweep gas as green
particulate material.

We then sought to understand the chemical role of the molten
salt using density functional theory (DFT) simulations.  In particular, we
investigated the potential for the metal cations present in the molten salt
acting as Lewis acids via aromatic-cationic interactions.  We focused on Zn+2,
as this is the strongest Lewis acid in our systems.  Using DFT, we calculated
bonding energies of Zn+2 to a styrene trimer for a series of
geometries.  These calculations showed that the most stable arrangement was for
the Zn+2 to intercalate between adjacent aromatic rings.  The Figure
is the stable geometry calculated for the intercalated configuration.  However,
the intercalated geometry does not result in substantial changes in
carbon-carbon bonding, with only a modest changes (1-2%) in the bond lengths of
several carbon-carbon bonds being observed.  A second geometry, in which the Zn+2
positions directly on the aliphatic backbone, results in a substantial increase
in the length of the carbon-carbon bond connecting a-methyl styrene to the remainder of the trimer.  The backbone
position is less thermodynamically stable than the intercalated configuration
and we hypothesize that the intercalated configuration is first fully populated
in the salt solvated polymer, followed by population of the backbone
configuration.  This hypothesis also may explain the benefits of the Li salt. 
Li+ has a higher charge-to-surface-area ratio than either Na+
or K+, which allows Li+ to participate more strongly in
cation-aromatic interactions.  Moreover, optimal Lewis acid loading is required
to maintain high styrene yields without co-producing a-methyl styrene that reduce selectivity to the monomer.  Brønsted
acids, when combined with Lewis acids, catalyze carbon-carbon bond breaking and
bond formation with low selectivity, leading to char production.  These results
form the basis for designing industrial recycling processes for obtaining high
monomer yields from polystyrene pyrolysis.

Figure 1. Optimized geometry of the styrene trimer
participating in a Lewis acid/base interaction with Zn+2.  Carbon
atoms (yellow), hydrogen (white), zinc (blue) are color labeled for clarity.