Polyethylene: High-pressure
classes of di-alkyl peroxides, peroxy alkyl esters,
Low-density polyethylene (LDPE) was discovered in
1933 and attained commercial importance because of
its favorable properties and manifold application
potential, e.g., for electrical insulation or film material.
More than 60 years after its discovery there was still,
around the end of the twentieth century, a growing
market for LDPE, about 1.5% p.a. in western econ-
omies, and up to 8% in emerging markets like south-
east Asia. Plant capacities have risen starting from
5000 t yr−" in 1940 through 1.0i10& t yr−" in 1980
up to 3.0i10& t yr−" in 2000 and further capacity
increases are on the way.
LDPE is characterized by its molecular weight and
branching structure. It has a statistical distribution of
short and long chain branches, formed intrinsically via
intra- and intermolecular radical transfer reactions.
Polymer properties are correlated to the temperature
and pressure during polymerization and can be tuned
by controlling these values.
Two generally different continuously operated high-
pressure processes, the autoclave and tubular reactors,
have to be distinguished (Gropper 1992). Oxygen
and\or organic peroxides are used as initiators for the
free radical polymerization. Molecular weight can be
controlled by adding chain transfer agents such as
saturated hydrocarbons, α-olefins, ketones, or alde-
hydes to the reaction mixture (Gropper 1992, Doak
1992, Ehrlich and Mortimer 1970, Luft 1980).
1. Polyethylene High Pressure Processes
Both the autoclave process and the tubular reactor
process are operated at high pressures up to 2400 bar
(autoclave) and maximum 3500 bar (tubular), respect-
ively. A rough general scheme of LDPE processes is
given in Fig. 1.
Ethylene is compressed in several stages up to the
desired polymerization pressure. Behind a preheating
section the reactor part follows, where the free radical
polymerization occurs. Monomer conversion can
reach values up to 40% in the tubular process
compared with up to 25% in autoclaves. Under
polymerization conditions the polymer is dissolved in
the supercritical ethylene resulting in a one-phase
reaction mixture. The separation of polymer and
unreacted monomer occurs typically in two stages at
pressures of about 300 bar and 2 bar, respectively.
Separation in two stages has the advantage that the
total energy stored in the compressed ethylene is not
lost. The unreacted ethylene is then recycled to the
appropriate compression stages. The polymer is ex-
truded, pelletized, degassed, and transferred to storage
silos.
For initiation and molecular weight control in the
free radical polymerization of ethylene a combination
of several substances is usually applied.
Typically used free radical initiators belong to the
peroxy carbonates, or di-acyl peroxides. The choice of
the initiator mainly depends on its half-life time at
application temperature. To generate a more or less
constant radical concentration level over the range of
polymerization temperatures (e.g., T l 150–300 mC) a
combination of different initiators is commonly ap-
plied. A typical mixture consists of a low T decom-
posing peroxide (e.g., tert-butyl peroxypivalate) and
a high T decomposing peroxide (e.g., di-tert-butyl
peroxide) dissolved in hydrocarbons.
Molecular weight is controlled by chain transfer
agents (modifiers). Their effectiveness depends on their
chemical structure and concentration, temperature,
and polymerization pressure. Usually the effectiveness
of chain transfer agents is increased with rising
temperature and reduced pressure.
One reactor type, the high pressure tubular reactor
consists of a large number of jacketed tubes with
a total length to diameter ratio between 1.0i10%
and 4.0i10%. Polymerization temperatures can reach
340 mC. Radical initiators are introduced at several
injection points starting the polymerization to form
polyethylene. Monomer per pass conversion depends
on the product grade and is higher for products with
low density and molecular weight.
Tubular reactors are offered in two generally differ-
ent forms. With the single-feed tubular reactor the
total ethylene flow from the hypercompressor is
preheated to 150–180 mC and fed to the inlet of the first
reaction zone. Reaction is started by injection of
organic peroxides. As the reaction mixture cools after
the first reaction peak, additional peroxide initiator is
added to start a second reaction zone. There can be
further peroxide injection points giving a total of three
to five reaction zones. A typical axial profile for
temperature and monomer conversion along a tubular
reactor is shown in Fig. 2.
With the multifeed tubular reactor the compressed
ethylene is split into several streams and fed into the
reactor. 30–70% of the total ethylene quantity is
preheated to 150–180 mC before the polymerization is
started in the first reaction zone. The other 70–30% of
the ethylene is cooled and further split into more or
less equal quantities, which are injected at different
locations along the reactor. Together with addition of
initiator this results in the desired number of reaction
zones. Additional reaction zones can be created by
dosing of initiator directly to the reactor. In addition
to the cooling by pressurized water, as typical for the
single-feed tubular-type, the reaction heat in this
reactor type can be absorbed by the cold gas feeds,
having the disadvantage that this part of the gas feed
has only a reduced residence time in the reactor. The
balance of these two effects can result in higher
conversion rates, depending on the product grade. The
multifeed tubular reactor can have, in contrast to the
single-feed tubular one, a stepwise increase in reactor