Sei sulla pagina 1di 18

Chemicnl Engineering Sciemx,

Printed in Great Britain.

Vol. 41, No. 9, pp. 2197-2214,

1986

REVIEW
POLYETHYLENE
THERMODYNAMICS

OOOS2509/86
$3.00 f0.00
Per&smon Journals Ltd.

ARTICLE

NUMBER

TEREPHTHALATE-I.
AND TRANSPORT

K. RAVINDRANATH

20
CHEMISTRY,
PROPERTIES+

and R. A. MASHELKAR

Polymer Science and Engineering Group, Chemical Engineering Division, National Chemical Laboratory,
Pune 411008, India
(Received 6 July 1984)
Abstract-Polyethylene
terephthalate (PET) is a major polymer with diverse applications. The chemistry of
synthesis and engineering of manufacture of this polymer provides many challenging problems The present
review critically evaluates the work done in this area In the first part of this series the aspects of mechanisms
in PET synthesis and also those of kinetics have been summarized. Total network of main reactions (which
govern productivity) and side reactions (which govern product quality) have been considered. Additionally
thermodynamic and transport data pertaining to PET formation have been reviewed. Future directions of
research have been highlighted.

CONTENTS
1. INTRODUCTION
2. HISTORICAL
3. PET PROCESS
4. CHEMISTRY

2197
PERSPECTIVE

2198

DESCRIPTION

2199

OF PET SYNTHESIS

2200

OF CATALYSIS
5. MECHANISM
5.1. TransesterilIcation catalysis
5.2. Polycondensation catalysis

2201
2201
2202

6. KINETICS
OF MAIN REACTIONS
6.1. Transesterilication kinetics
6.2. Esterilication kinetics
6.3. Polycondensation kinetics
6.4. Equilibrium constants

2202
2202
2202
2203
2204

REACTIONS
7. DEGRADATION
7.1. Mechanism of degradation reactions
7.2. Kinetics of degradation reactions

2205
2205
2207

8. THERMODYNAMIC
8.1.
8.2.
8.3.
8.4.
8.5.

AND

TRANSPORT

DATA

2208
2208
2209
2209
2210
2210

REMARKS

AND

SUGGESTIONS

FOR FURTHER

WORK

2211

NOTATION

2211

REFERENCES

2212

Communication

No. 3570.

and the demand is likely to rise to 10 million


tonnes by 1990 (L&kin, 1981).
Extensive fundamental studies relating to the aspects of synthesis, structure, properties and application
of this major polymer have appeared in the literature
(Heffelfinger and Knox, 1971; Katz, 1977; McIntyre,
1982; Ravindranath and Mashelkar, 1985). However,
aspects of chemistry and engineering associated with
this polymer
have not been reviewed critically. The

world

1. INTRODUCTION

Polyethylene terephthalate (PET) is a major polymer


which is extensively used in the form of fibres, films and
as a moulding material. Forty per cent of the world
synthetic fibre demand is captured by PET. Approximately 5 million tonnes of PET is produced in the

CES41:9-A

TO PET FORMATION

Phase equilibrium
Molecular diffusion in PET
Diffusion in molten polymers
Diffusion experiments in molten PET
Diffusion of volatile products in solid PET

9. CONCLUDING

tNCL

PERTINENT

2197

K. RAVINDRANATH

2198

and R. A. MASHELkAR

primary objective of this review is to fill this important


gap, especially in view of the important advances that
have been made over the last decade.
The present review might appear to be rather
unusual for this jounral since it deals with a specific
product. However, it needs to be emphasized that the
review will not only be of value as a summary of the
current status of the information base, with recommendations for future directions of research, but it
would also show how gaps in the genera1 development
of science and engineering can affect the progress in
our understanding of a specific problem. One sees, for
instance, that inadequacy of general models for
making predictions of diffusion in molten or solid
polymers affects our ability to mode1 specific aspects of
PET reactors. On the other hand, some of the advances
that have been made while undertaking specific modelling efforts on PET will lead to development of genera1
techniques, which can be used by the chemical
engineering profession at large. For instance, the
Effective Flash procedure developed while modelling
the specific transesterification stage in PET can generally be used for handling problems of distillation with
reaction. Indeed, by suitable modifications this procedure can also be used for handling problems of
simultaneous desorption and reaction in a multicomponent system with variable interfacial concentration.
It will be our endeavour to bring out such features in
this critical review. Table 1 shows the immediate
relevance of such work to other areas of chemical and
polymer engineering interest.
We shall be attempting a comprehensive coverage of
chemistry and engineering of PET production. The

reaction networks involved in synthesis of PET, along


with the pertinent kinetic and mechanistic considerations will be discussed first. Special emphasis will be
on the degradation reactions, which are so critical in
controlling the product quality. We shall then review
the currently available thermodynamic and transport
data pertinent to PET formation and bring out the
gaps in information. Rather than being exhaustive, our
emphasis

will

information
directions
design
reactor

be on critical
with

of

of

research.

PET

reactors

modelling

analysis

indication
The
and

of

aspects
the

is considered

the existing

regarding
of

future

engineering

current

status

of

in Part II of this

review.
2. HISTORICAL

PERSPECTIVE

The fibre forming properties of synthetic polyesters


with linear molecular structure became known only in
the 1940s, when Carothers and co-workers obtained
aliphatic polyesters. These polymeric compounds
could not be used for the preparation of textile fibres,
as they possessed low melting points and had poor
hydrolytic resistance. In 1941, English scientists
Whinfield and Dickson proposed that a polyester
obtained by using symmetrical aromatic acids and
aliphatic glycols should be used for the production of
fibres. Following this, ICI undertook extensive research and in 1955 started a commercial plant for the
production of PET. Similar production capacities were
established in the U.S.A. and in the rest of the
European countries.
Synthesis of PET is only one of the steps in the
manufacture of commercial PET fibre. The molten

Table 1. Relevance of the present review to other chemical/polymer


No.

clear

engineering problems

Specific investigation in PET


synthesis/reactors

Related area in chemical/polymer


engineering with relevance

Reaction network analysis in


transesteritication/esteri&ation/
polycondensation

Other polyalkylene terephthalates


(e.g. polybutylene terephthalate)

Pilati er al. (1981)


Rafler et al. (1981)

Thermodynamic

Other polymer-solvent/solute
systems
(e.g. water in nylond, styrene in
polystyrene)

Nagasubramanian and
Reimschuessel (1973)

Reactor models on transesterification and esterification


reactors

Algorithm generally applicable to


esterification reactions and systems
involving distillation with reaction

Block Ulrich (1977)


Kaibel et al. (1978)

Thin film models on multi-component


desorption with reaction

General chemical engineering problems


of multi-component desorption with
reaction

Shah and Sharma (1976)

Engineering models on interfacial


surface generators

Blood oxygenators, waste water


treatment. solvent removal from
molten polymers, etc.

Borchardt (1971)
Denson (1983)

Solid state polycondensation


of PET

Solid state polycondensation of


polyalkylene terephthalates (e.g.
polybutylene terephthalate, polypropylene
terephthalate. etc.) and other solid
state polycondensations (nylon-66)

Buxbaum ( 1979)
Fortunate et ol. (1981)
Griskey and LCC (1966)

Molecular weight distribution


equilibration with interchange
reactions

Melt blending of homopolymers,


preparation of block copolyamides.
copolyesters or copolyester amides

Kotliar (1981)
Devaux et al. (1982)

and transport data

Ref.

2199

Polyethylene terephthalate-I
fibre is spun and subsequently subjected to solid state
processing involving
drawing, texturizing, etc. The
aspects of synthesis-processing-structure-properties
are closely linked. Very extensive work on all these
aspects has been published in the open literature,
which has been classified in Table 2. There have been a
number of studies on rheological
and mechanical
properties and processing, but the studies on other
aspects have not been as extensive. Indeed the studies
related to the basic aspects of catalysis, kinetic mechanisms, reactor modelling,
etc. have appeared
more
extensively only during the past decade or SO.
The level of our understanding of the aspects related
to the science of PET fibre formation has been directly
linked to the developments in other branches of science
and engineering. Just as a specific example, take the
case of PET reactor models. One essentially deals with
the case of reversible equilibrium
reactions, with
simultaneous desorption of several volatile products
and byproducts. Development
of simpler theories of

gas absorption/desorption
with reaction as they appeared in the chemical engineering literature in the
1950% were obviously critical in providing basic frameworks. In 1964, the numerical solution of penetration
theory equations was worked out by Brian et al. (1961)
for a simplified but generalized case of an irreversible
reaction. A simplified model dealing with the case of a
reversible reaction with desorption of volatile species
was presented by Secor (1969). All these developments
were decidedly helpful in the analysis of diffusional
influences in PET polycondensation
reactions. In the
present review, we shall attempt to comment on such
developments, wherever appropriate.
3.

PET PROCESS

DESCRIPTION

Production of PET essentially involves four stages:


(I) transesterification
or direct esterification, (2) prepolymerization,
(3) melt polycondensation
and (4)
solid state polycondensation.
A schematic diagram
(Fig. 1) shows the different stages. In the first stage,

Table 2. A summary of the publications on PET in the open literature


Year

Kinetics/
mechanisms

catalysis/
chemistry

Reactor
modelling

Mechanical/

rheological
properties/

General
references/

Total

review articles

processing

1961-1970
1971-1975
197Sl980
1981-1985

3
8
10
34
3

7
11
17
34
23

1
7
8
11
18

5
8
24
67
16.5

1
2
3
7
5

17
36
62
153
214

Total

58

92

45

269

18

482

195&f

960

ESTERIFICATION
I

STIRRED
250

VESSEL
-

DP

20-30

p-SPOS

280-C

DISC

REACTOR
280so10.000-

2-3
kPa
5Oco-7000 s
I

oP=loo

RING

zsO*c
roo ~9
20.000

-I
s

~1000

DPSi-S-4

TRANSESTERIFICATION
DMT
EG

AND

I
140-220-c
400

k PO

7000-10.000

Fig. 1. Different stages in PET synthesis.

+
Pas

ROTARY

--

I
DRUM

OP2:150

ZOO-240.C
.nn
* D

II

2D,oOD-90.000

MOULDING

-3

2200

K. RAVINDRANATH

bis(hydroxyethyl)terephthalate (BHET) (and some of


its linear oligomers) is prepared by a transesterification
or a direct esterification process. In transesterification
process dimethyl terephthalate (DMT) is reacted with
ethylene glycol (EG) and methanol is removed
continuously. In the direct esterification process
terephthalic acid (TPA) is reacted with EG and the
byproduct, water, is removed. After the development
of fibre grade TPA, most of the plants in the world are
switching over to TPA as a raw material, as it appears
to offer a considerable advantage (Arunpal, 1978). In
the second stage, BHET (and some oligomers) emerging from the first stage is polymerized in a prepolymerization reactor up to a degree of polymerization (DP) of approximately 30. The viscosity of the
reaction mixture in the prepolymerization process rises
to approximately to 5.0 Pas (Ellwood, 1967). The
product is then passed onto the third stage (uiz. final
stages of polycondensation) where it is further polymerized to a DP of about 100 using high vacuum and
a special agitation system. The melt viscosity of the
polymer now builds up to a very high value and mass
transfer limitations become important, necessitating
special considerations for reactor design. Usually
reactors with a facility of mechanical generation of
interfacial area are employed in the continuous process
for this stage. In the batch process both prepolymerization and final stages of polycondensation are carried
out in a single reactor, using conventional helical screw

acNX&

HOCH,CH,OH

and

R. A.

MASHELKAR

or ribbon agitators (Chavan and Mashelkar, 1980).


The solid state polycondensation process (Li-Chen,
1967) is generally employed for obtaining moulding
grade PET having DP > 150. For producing this
product PET melt from the polycondensation stage is
solidified and made in the form of chips. These chips
are subjected to a solid state polycondensation process,
which is carried out above the glass transition temperature, but below the polymer melting temperature.
It may be indicated here that another alternative
route for synthesis of BHET is the direct reaction of
TPA with ethylene oxide. This route offers some
advantages, since the step of manufacture of EG from
ethylene oxide is eliminated. However, there are critical
safety engineering problems due to the explosive
nature of ethyIene oxide. Mares et al. (1969), Bhatia et
al. (1976) and Kamatani (1977, 1978) focussed attention on the kinetics of TPA with ethylene oxide in a
solvent system, whereas Tani and Enoki (1970) and
Lee et al. (1983) studied the kinetics in solvent-free
system. However, since the process has not found
general commercial acceptability in the world, we shall
not review information pertinent to this route.
4.

CHEMISTRY

OF PET SYNTHESIS

In the transesterification process, the main reversible


reactions are as follows (Ravindranath and Mashelkar,
198 1). Ester interchange reaction (l), transesterification reaction (2) and polycondensation reaction (3).

+-CWCH,CH,OH

==z

Em

*COOCH,

E,

CH,OH

EE

HO(.H,CHIOOCe

(1)

(2)

EG=

E,

2~COOCH&H;OH

===z

~COOCH,CH,OOC~

ES-

HOCH,CH20H

(3)

EC;

In addition to this polycondensation reaction, in the


case of esterification process, we have the following
esterification reactions (4) and (5) (Ravindranath and
Mashelkar, 1982a).

*COO

HOCH,CH20H

+COOCH,C,OH

Hz0

(4)

Polyethyleneterephthalate-I
Note that the build-up of polymer chain occurs
essentially due to the polycondensation reactions of
the type given in reactions (2), (3) and (5). Esterification
reactions (4) and (5) also proceed in the transesterification stage and in the polycondensation stage where
side reactions lead to the formation of acid end groups
(see reactions (15) and (20) below]. The equilibrium
constants for the reactions (lk(5) are defined as:
K,

= -,K
e,m,=--2ed

K,=T,K;=2e,g

2zm
egem

,K=%,

2zw
e,eg

es

(6)

where e,, es, e,, m, w, z and g are the concentrations of


methyl end group (E,), hydroxyl end group (E,), acid
end group (E,), methanol (M), water (W), diester group
(Z) and ethylene glycol (EG), respectively. The equilibrium constants are related by
K:, = K,K,

K; = K,K.

(7)

A number of side reactions (Buxbaum, 1968) occur


during the PET synthesis along with the main reactions. It is important to control these side reactions
since the product quality is critically dependent on the
amount of side products formed. We shall discuss this
aspect in detail while considering the degradation
reactions subsequently.
5. MECHANISM OF

CATALYSIS

Selection of a suitable catalyst for PET synthesis is


very critical for production of high grade PET. The
literature suggests numerous catalysts, but few of these
have gained commercial acceptance. A good catalyst
should accelerate both transesterification and polycondensation reaction to an equal degree; it should
have good solubility in the reaction medium and it
should not promote side reactions. Selection of such a
catalyst is an extremely difficult task. Therefore, two
types of catalysts are generally used. These are
transesterification and polycondensation catalysts.
Use of two different types of catalysts in the first
and second stage becomes necessary, since each of the
catalysts has its own typical medium dependence.
Transesterification catalysts are very active in a high
and a low hydroxyl content medium, but they are easily
poisoned by small amounts of acid end groups
(Hovenkamp, 1971; Walker, 1983).
Polycondensation catalysts are insensitive to acid
end groups, but their catalytic activity is known to
increase as the concentration of hydroxyl end groups
decreases. Generally metal catalysts are not employed
in the direct esterification stage since such reactions are
catalysed by acid end groups of TPA (Arunpal, 1978).
However, polycondensation catalysts are the same for
transesterification process as well as for direct esterification process.
5.1. Transesterijication catalysis
Various transesterification catalysts have been cited
in the literature, such as salts or mixed salts of sodium,

220 1

potassium, lithium, calcium, manganese, magnesium,


lead, zinc, aluminium and cadmium. However, the
acetates of zinc, manganese, calcium and sodium have
attained commercial acceptance due to their good
solubility and catalytic effect on transesteritication
reactions. The relative efficiency of these catalysts has
been studied rather elaborately (Yoda, 1971) and
appears to change as zinc acetate > manganese acetate
> calcium acetate > sodium acetate. Yoda (197 1) used
the criterion of electronegativity of metal species for
determining the ordering factor of the catalytic activity
of the metal compounds. Tomita and Ida (1975) found
that the catalytic activity varies with the reactant, but
the electronegativity is the same regardless of the
reactant. Therefore, Tomita and Ida used the stability
constant of the dibenzoyl methane complex for determining the catalytic activity of the metal compounds in the transesterification process. Zinc and lead
acetates having stability constants between 9-11 were
found to be good transesterification catalysts as shown
in Fig. 2. In industrial practice, usually a mixture of
transesterification catalysts is used. The catalyst with
lower efficiency serves as a tracer in many cases. The
catalysts, during the transesterification process, react
with glycol and metal alcoholate and acetic acid are
formed.
M(CH&O&

+ 2ROH = M(OR),

+ 2CH&OOH.
(8)

The acetic acid formed is removed during the transesteriflcation process, thus favouring the formation of
metal alcoholate. However, the presence of any acid
end groups (via TPA appearing as impurity in DMT or
acid end groups formed due to side reactions) reduce
the transesterification rate due to the displacement of
the equilibrium to the left (Walker, 1983). The inhibitory effect of the acids is not related to acidity but it

xO
:
-1

-2

Fig. 2. Rate constantsin the transesterification


of DMT with
EG with various metal acetatesas catalystsagainst stability
constants (log@) of dibenzoal methane complexes of the
correspondingmetal species (Tomita and Ida, 1975).

2202

K.RAVINDRANATH

and

is related to chemistry (i.e. the nature) of the acids. For


example, dicarboxylic
acids in which the carboxyl
groups are present at para positions show extraordinary inhibitory effect compared
to ortho or meta
substitution (Walker, 1983). The mechanism proposed
(Fontana, 1968; Walker, 1983) for transesterification
reaction employing metal ion catalysis is shown in
reaction (9). Here R, R, RZ and R are alkyl groups.

R. A.MASHELKAR
6. KINETICS

OF MAIN

dR1
\\

.*_p
RO-0
R"-0

P-0

/M

(9)

R'

/O\,,

RO--0

R-O

5.2. Polycondensation catalysis


During
the polycondensation
stage, antimony
trioxide or antimony triacetate catalysts are commonly
used in the industry. A few manufacturers use germanium oxide, but reportedly it is not as effective as
antimony compounds.
Rafler et al. (1974), Tomita
(1976a, b), Kamatani and Konagaya (1978), Kamatani
et al. (1980) studied the mechanism of polycondensation reactions and concluded
that the reaction
proceeds by the nucleophilic attack of hydroxyl end
groups, upon the ester carbonyl groups. The polycondensation catalysts facilitate the coordination
of
metal ion to the ester carbonyl bond which increases
the polarity of this bond and facilitates the nucleophilic attack. In the initial stages of polycondensation,
antimony compounds readily form stable compounds
with ligands containing hydroxyl groups and cannot
react with carbonyl groups of esters. Therefore, the
catalytic activity of antimony compounds is less at high
concentration of hydroxyl end groups observed in the
initial stages of polycondensation
(Stevenson, 1969).
As the polycondensation
reaction progresses, the concentration of hydroxyl end groups decreases and the
coordination
of the carbonyl groups to antimony
catalyst becomes possible. Therefore, the activity of the
antimony catalysts increases as the polycondensation
reaction proceeds.
Maerov (1979) studied the effectiveness of antimony
catalyst with varying numbers of hydroxyethoxy
hgands and found the effectiveness to change with the
ligand
number
as
5=3>2>130.
Tomita
(1976a, b) used the stability constant of dibenzoyl
methane complex of each metal species in determining
the activity of the catalysts. Metal compounds with
stability constant of about 12 were found to be more
active as catalysts in the polycondensation
stage.
Tomita (1977b) found that the catalytic activity of
Sb,O,
was very low in the initial stages of polycondensation but it increased with the progress of the
reaction due to the reduction of Sb205 by acetaldehyde formed due to side reactions.

REACIIONS

6.1. Transestertfccation kinetics


Many workers (Griehl and Schnock, 1957; Peebles
and Wagner,
1959; Fontana,
1968; Sorokin
and
Chebotareva, 1969; Tomita and Ida, 1973; Shah et al.,
1984; Datye and Raje, 1985) have studied the transesterification reaction. The progress of the reaction

TI

was followed by determining the quantity of methanol


distilled out. The transesterification data have been
analysed by using a number of models. Invariably, these
models do not take into account the effect of the
variable volume due to the continuous removal of
methanol generated and also the presence of oligomerization reactions [reaction (2)]. This has caused a
considerabIe confusion among workers as regards the
determination of the transesterification reaction order.
In fact, first, fourth or fractional orders have been
reported. However, Fontana (1968), Tomita and Ida
(1973) and Datye and Raje (1985) found that transesterification reaction is overall third order, being first
order with respect to catalyst, EG and DMT. Yamanis
and Adelman
(1976a, b) and Ravindranath
and
Mashelkar (1982b) analysed the experimental data of
different investigators and showed that oligomerization reactions in the temperature range of 175-197C
do not proceed to any significant extent upto about
88 y0 DMT conversion, when the EG to DMT molar
ratio is kept 2 2. At the end of the transesterification
stage, the contribution
of the oligomerization
reactions towards methanol generation and the deactivation of the catalysts by acid end groups cannot be
neglected and these factors should be taken into
account for any realistic models. A summary of the
kinetic data reported by various investigators is given
in Table 3.
6.2. Esterzjicarion kinetics
The kinetics of the esterification process is complicated by the heterogeneity of the reaction system
caused by an extremely low solubility of TPA in EC. At
19OC, for affecting complete dissolution of 1 mole of
TPA, we require at least 77 mole of EG (Krumpolc and
Malek, 1973). In addition, TPA particle size may also
influence the esterification rate (Rod et al., 1976).
Interestingly, however, Kemkes (1969) found that the
rate of reaction is not affected when TPA particle size is
in the range of 20-100 /.LIn addition, the presence of
two bifunctional reactants (TPA and EG) leading to

Polyethylene terephthalate-I

2203

Table 3. A summary of some typical transesterifleation kinetic data reported in the literature
NO.

1
2

3
4
5
6
7

Catalyst

lead oxide
calcium acetate
manganese acetate
antimony acetate
zinc acetate
zinc acetate
zinc acetate

zinc acetate

zinc acetate

Temperature
(C)

Activation
energy
&J/mob

160-220
160-220

52.3
60.7

197
197
160-220
197
160-197
155-l 97
170-l 90

complex intermediate reaction products complicates


the analysis of the esterification process (Krumpolc
and Malek, 1973; Rod et al., 1976). However, Chegolya
et al. (1979),
Reimschuessel
et al. (1979) and
Reimschuessel and Debona (1979) obtained qualitative kinetic data using various model systems, uiz.
benzoic
acid-hydroxyethyl
benzoic
acid-EG,
benzoate,
terephthalic
acid-2-(2-methoxyethoxy)
ethanol. These data are the most reliable to date and
have, therefore, been used in modelling of esterification
reactors by Ravindranath and Mashelkar (1982a).
6.3. Polycondensation
kinetics
The main polycondensation
reaction is a reversible
reaction and the condensation product EG has to be
removed to obtain the desired molecular weight. The
kinetics of polycondensation
reaction was studied by
many investigators (Challa, 196Ob, Cafelin and Malek,
1969; Rafler et al., 1974; Tomita and Ida, 1973) and it
has been generally reported that the main polycondensation
reaction is a second order reversible
reaction. However, the activation energies and the rate
constants obtained by various investigators are not in
agreement. The difficulties in the analysis arise due to
the presence of a number of complications. Firstly,
diffusional limitations on account of desorbing volatile
species (which are products of the reaction) come in
after a DP of about 30 is reached. The melt viscosity
increases with the progress of the polycondensation
reaction. Lastly, a number
of side reactions accompanying the main polycondensation
reaction complicates the analysis and the accurate determination of
rate constants. Many investigators carried out polycondensation experiments in closed systems (sealed
tubes), semibatch reactors (stirred vessels) and thin
stagnant films. There are certain merits and demerits of
each method and these will be examined briefly now.
Closed system.
The analysis of closed systems is
simple since the volume of the reaction mixture
remains constant and the concentration of the reactive
end groups can be easily determined. Challa (196Oa, b)
and Hovenkamp (1971) used such a closed system and
found that the results are sensitive to the experimental
errors. Special difficulties arise at high DP, wherein the

62.8
70.7

0.03-. 1
0.01-0.8
0.19
1.35 x 10-S
0.05-0.3
0.24
0.03-O.175

52.1
60.0

0.08-0.2

Ref.

Rate constant
[(l/mol)2s-]

Fontana (1968)
Fontana (1968)
Tomita and Idi (1975)
Tomita and Ida (1975)
Fontana (1968)
Tomita and Ida (1975)
Ravindranath and
Mashelkar (1982b)
Shah et al. (1984)
Datye and Raje (1985)

concentration of EG becomes
too low for accurate
determination.
This limits the applicability
of the
closed system for analysis only in the range of low DP.
Further, in the absence of EG removal, the degradation reactions leading to side products might dominate, thus complicating the analysis further.
Semibatch
reactor.
Semibatch reactors can be used
for determining the polycondensation
rate constants.
However, the analysis is very complex and the models
developed should take into account the desorption of
EG, the influence of mixing and the presence of
number of side reactions. So far the models reported in
the open literature are too simple and do not take into
account all these complexities. For example, Rafler et
al. (1973a) assumed that the concentration of EG in the
reaction mixture can be negligibly small if high vacuum
was applied or an inert gas was bubbled. The semibatch reactor data were analysed by using second order
irreversible rate equation and the degree of polymerization was simply given as
P, - P,,, = kt.

(10)

Here k is the polycondensation rate constant, P,, is the


initial degree of polymerization
and t is time.
Equation (10) predicts that degree of polymerization
increases linearly with time. However,
the experimental data (Rafler et al., 1973b, Tomita, 1973) show
that DP increases linearly upto a certain time (depending on the temperature and the concentration of the
catalyst) and then decreases due to the presence of
degradation reactions. Tomita (1973) considered the
main degradation reaction [see reaction (20) below] in
the analysis of semibatch reactor data and an expression for DP was derived as
1
-=
p*

1
1+pt

+ zdt
3

(11)

where p and dare parameters for the polycondensation


and degradation reactions. While developing eq. (11)
Tomita (1973) neglected the influence. of backward
reaction and the influence of mixing. The parameters
derived from eqs (10) and (11) are only apparent
values, since these include the effect of mass transfer
and chemical reaction. However, these studies are

K.

2204

and R. A. MASHELKAR

RAVINDRANATH

useful for comparing


the effectiveness of various
catalysts and their influence on the stability of PET
melt. Yu et al. (1980) developed the following semiempirical formula for semibatch reactor
2

1
p,=

(l + 2 C,kr)

C,kPoy

+ 5

0.16K~fiPr

(12)
where C, is the initial concentration of BHET; P,,
vapour pressure of EG; y, mole fraction of EG in the
gaseous phase; E, interfacial area; N, stirrer speed; k,,
the degradation reaction rate constant. Unlike eqs (10)
and (1 l), eq. (12) takes into account the effect of reactor
pressure and the effect of mixing.
Stagnant films. For accurate analysis of any polycondensation data, the influence of mass transfer has
to be taken into account. The information about the
kinetics and mass transport can be obtained from the
experiments conducted in this stagnant films. These are
generally carried out by placing a known amount of
prepolymer in the sample holder and the experiments
are carried out either under vacuum or nitrogen
atmosphere. The analysis of thin film experimental
data has been done, by setting up relatively straightforward mass balance equations of the following form
(Pell and Davis, 1973; Rafler et al., 1979)
as
-=D,~+k(e~--g)
at

(13)

dZ

(14)

-i%=k(e:-g).
at'

Here Do is the diffusivity for EG. Equations (13) and


(14) can be solved with appropriate initial and boundary conditions.
The rate constants and diffusivity
values are obtained by curve fitting. The presence of
side reactions complicate the analysis and meaningful
vahtes cannot be obtained by using thin film experimental data alone. To consider the presence of side
reactions in the analysis, kinetic parameters for side
reactions, esterification
reactions, polycondensation
reaction and the diffusivity values of EG and water are
needed. Such a large number of parameters cannot be

obtained from thin film experimental


data alone.
Therefore the kinetic constants have to be obtained
from the model experiments and the diffusivity data
from diffusion experiments using nonreactive media.
Again the values obtained by thin film experiments
depend on the rigour of the mathematical analysis
since the kinetic and mass transfer data are obtained by
curve fitting. In Part II of this review (Ravindranath
and Mashelkar, 1986) we will be describing such an
analyses in more depth. The pertinent transport and
kinetic data obtained by various investigators are
summarized in Table 4.
6.4. Equilibrium constants
Although many investigators have studied transesterification reactions, information is rather scant on
equilibrium studies pertaining to the reaction of DMT
with EG. An exception is the study by Challa (196Oc),
who determined the individual values of equilibrium
constants Kk and K, by conducting experiments on
reaction of DMT with EG in a closed system. Chalh
found K&, and K, to be 0.3 and 0.15, respectively,
which were found to be independent of temperature
(lSO-195C)
and to the ratio of EG to DMT (1.8 to
2.0). The individual values of equilibrium constants are
required to enable the calculation of DMT conversion.
Chegolya et al. (1979), Reimschuessel et al. (1979) and
Reimschuessel and Debona
(1979) determined
the
esterification equilibrium constants K; and K, using
model compounds
and these were found to be in
et al. (1979) and
the range of 0.75-2.5. Chegolya
Reimschuessel (1980) reported conflicting reports on
the dependence of the esterification equilibrium constants on the composition of reaction mixture. Many
workers (Koepp
and Werner, 1959; Challa, 196Oa;
Fontana, 1968; Hovenkamp,
1971; Chegolya et al.,
1979; Reimschuessel et al., 1979) have undertaken
extensive studies on the polycondensation
reaction
using PET or model compounds. They have estimated
the equilibrium constant (K) to be in the range of 0.4-1,
which was found to be independent of the composition
of the reaction mixture. For model calculations, an
approximate value of K = 0.5 has been used by many
1968; Ravindranath
and
investigators
(Fontana,
Mashelkar. 1981; Kumar et al., 1984).

Table 4. A summary of some typical polycondensation


No.

1
2

Type of
system

Catalyst

Temperature
(C)

Polycondensation
activation
energy (kJ/mol)

Stirred
vessel

Zinc acetate

270-290

99.2

Stirred

Antimony

260-285

77.5

Antimony trioxide
Zinc acetate and
antimony acetate
Manganese
acetate and
antimony acetate

275

58.6
-

vessel

Thin film

Thin film

Thin film

trioxide

kinetic data reported in the literature

270
270

Polycondensation
rate constant
C(/moU s- 1

Ref.

Tomita (1973)
0.5-2.7 x lo- 3
0.01
9.6 x lo-
2.58 x 10-l

Yokoyama

(1978a, b)

er al.

Stevenson (1969)
Pell and Davis (1973)
Rafter et al. (1980)

Polyethylene terephthalate-I

2205

7.1. Mechanism of degradation reactions


Degradation reactions occur during all the stages of
PET synthesis viz. transesterification, esterification,
polycondensation
and also during PET processing.
However, the type and the formation rate of the side
products varies in different stages. For example, most
of the DEG is Formed in the transesterification or
esterification stage and in the beginning of the polycondensation stage, whereas acetaldehyde, acid end
groups and vinyl end groups are formed mainly in the

subsequent identification of various stabilizers and


retarders. The study of kinetics of degradation
is
important also as an aid in building up a rational
mathematical analysis of the synthesis process. We will
now review the studies which have attempted to
elucidate the mechanisms related to such reactions.
Reimschuessel (1980) studied the mechanism of
hydroxyl end group degradation and suggested that
the degradation is likely to proceed via a five membered orthoester type intermediate and acetaldehyde is
Formed via intermediate formation of ethylene oxide
[reaction (1519.

final stages of polycondensation.


Again the rate of
Formation of these products depends on the operating
conditions and also on the type and concentration of

Ethylene oxide can react with EG and hydroxyl end


group forming free DEG or incorporate DEG in the
polymer chain as shown in reactions (16~(18).

7. DEGRADATION

REACTIONS

CHz\
CH,,O
I

HOC,H,OH

c
Hzv:i
HzC

OC~H,OH

HOC:Hs,OCzH.,OH

(16)

+!

0
(17)

(18)

HOGH.OC,H.OH

rH0,2 .

the catalysts used (Kemblowski


and Torzecki, 1982).
The transesterification and polycondensation catalysts
are known to catalyse the degradation reactions. The
important side products which critically control the
product quality are acetaldehyde, diethylene glycol,
acid and vinyl end groups. Even a small extent of
degradation can influence the product quality. For
example, formation and subsequent incorporation of
diethylene glycol (DEG)
in the polymer chain decreases the melting point (Fakirov et al., 1981; HornoF,
1981) and the thermal stability of PET (Buxbaum,
1968). Acetaldehyde formation, even in concentrations
as low as few ppm, causes flavour in the soft drink
bottled in PET bottles (Schaul, 1981). Similarly, acid
end group concentration
in PET determines the
thermal and hydrolytic stability of PET (Zimmermann
and Kim, 1980). It is important to know the mechanism of the Formation of these side products For
minimizing their production by controlling the degradation
rate. This will, hopefully, be useFu1 in

Buxbaum (1968) suggested a seven membered ring


as transition state and the formation of active species
[CH2CHIO]
For hydroxyl end group degradation.
Even though the active species proposed by Buxbaum
(1968) and Reimschuessel (1980) are different, the
reactions leading to the Formation of side products
remain same. DEG is also Formed due to dehydration
of EG.
2HOCH2CH20H

HOCH2CH20CH,CH20H
+ H20.

(19)

Another important side reaction in PET synthesis is


the degradation of the repeating units by a molecular
mechanism with random chain scission at ester links
(Buxbaum, 1968; Bednas et al., 1981). Among thermally weak linkages, i.e. the C-O bonds along the
polymer chain, those located at /I-position from C=O
bonds most likely get subjected to the thermal cleavage

K.

2206

and R. A.

RAWNDRANATH

MASHELKAR

forming acid and vinyl end groups [reaction (201-J.

and mechanism of thermooxidative

Acid and vinyl end groups can react with hydroxyl


end groups of PET and help in reforming the repeating
units [reactions (21) and (22)].

studied by Marshall and Todd (1953) Jenekhe and Lin


(1983), Kamatani and Kuze (1979) Zimmermann et al.
(1979) and Jabarin and Lofgren (1984). The rate of

~COOC
\--/

=CH, + HOCH,CH,CJOC&

=zz.zz

\--/

The net effect is that each broken repeating unit will


reform producing an equivalent amount of acetaldehyde and water. When most of the hydroxyl end
groups have been consumed, the molecular weight will
begin to fall. In such cases, acid and vinyl end groups
will accumulate leading to formation of anhydride
groups and network structures as shown in Fig. 3.
Acetaldehyde formed due to side reactions (15) and
(22) participates in secondary reactions forming
polyenaldehydes
CH,CHO

+ CH3(CH

: CH),-

lCHO

+ (n - l)H*O.
(23)

The complex network of reaction scheme (Goodman


and Nesbitt, 1960; Goodings, 1961; Buxbaum, 1968) in
the degradation of PET is shown in Fig. 3.
The degradation of PET in the presence of oxygen is
also important in the processing stage. The kinetics

~COOC~C,OOC~

degradation was

+ CH,CHO

(22)

degradation increases with an increase in oxygen


content and even small amounts of oxygen greatly
accelerate the degradation rate. Buxbaum (1968) and
Kamatani and Kuze (1979) proposed free radical
mechanism for thermooxidative degradation. The side
products were found to be quite similar to those
obtained
in thermal degradation
In addition,
Kamatani and Kuze (1979) detected the presence of
4,4-biphenyldicarboxylic
acid, 2,4$biphenyltricarboxylic acid and 1,2-bis(4-carboxyphenyl)ethane
as
side products formed during polycondensation and
thermooxidative degradation.
Note that one of the most vital problems in PET
manufacture is the formation of chromophoric substances that lead to colour formation (Buxbaum,
1968). The chromophoric substances are mainly polyenaldehydes from acetaldehyde (Goodings, 1961) and
unsaturated
molecules
from
polyvinyl
esters

COOH

Fig.

3. Reaction

network

in the degradation

of PET.

Polyethyleneterephthalate-I
(Zimmermann and Leibnitz, 1965). These are present
in the range of few ppm and it is extremely dificulr to
isolate them.
In addition to the above side reactions, cyclic
oligomers are formed during the manufacturing and
the processing stages of PET. Even though they are
present in the range of 2-3 % by weight, these compounds cause problems in the processing of PET. The
possible mechanisms for cyclic oligomer formation
could be through the cyclization of linear short chain
molecules or alternatively the polymer hydroxyl end
group can attack ester groups of its own chain as
shown in reaction (24).

these methods gives only apparent rate constants


because of a number of series and parallel reactions
that occur in PET degradation. For example, an
increase in the acid end group concentration due to the
degradation of repeating units is not necessarily accompanied by a decrease in the molecular weight
(Zimmermann and Kim, 1980). Similarly acetaldehyde
formed due to side reactions participates in the
secondary
reactions
(Goodings,
1961).
Again
Zimmermann and Kim (1980) found that the rate of
thermal degradation in the closed system is about three
times higher than that in the open system. However,
the degradation studies are useful in determining the

ol,-,

\-/

n = 1.2.3

Cyclic oligomers can also form due to interchange


reactions taking place within or between
polymer
molecules. Many investigators (Peebles et al., 1969; Ha
and Choun, 1979) studied the possible mechanism of
cyclic oligomer formation and concluded that a backbiting mechanism (cyclodepolymerization) as shown
in reaction (24) is probably operative.

7.2. Kinetics of degradation reactions


The information in the literature on the kinetics of
thermal degradation is somewhat ambiguous. This is
because the role of catalysts in the degradation reactions is not well understood and the method of analysis
employed by each investigator is different. Generally,
the kinetics of degradation reactions was studied either
by measuring the amount of gases (such as acetaldehyde) evolved, or by measuring the change in the
concentration of end groups (acid end groups) or by
measuring the change in the molecular weight (intrinsic

viscosity).

The

information

obtained

through

2207

(24)

... .

effect of various transesterification and polycondensation catalysts on the rate of side products formation.
Pertinent kinetic data are shown in Table 5. Because of
the difficulty in studying the PET system, many
investigators
(Goodings,
1961;
Buxbaum,
1968;
Zimmermann and Kim, 1980; Tomita, 1977a) used
model compounds representing the specific segments
of PET chain for their degradation studies, these
include hydroxyethyl benzoate, ethylene dibenzoate
and diethylene glycol dibenzoate. By using these model
compounds, the number of possible reactions are
minimized and this simplifies the analysis. The kinetic
data obtained from the model compounds are also
shown in Table 5.
Hovenkamp and Munting (1970), Hornof (1981)
and Renwen et al. (1983) have studied the kinetics of
DEG formation and concluded that DEG is formed
mainly during the final stages of transesterification and
in the initial stages of polycondensation. Ha and
Choun (1979) have made some theoretical calculations
for cyclic oligomer formation rate.

Table 5. A summary of some typical kinetic data of diester group degradation reaction reported in the literature
No.

1
2
3
4
5
6

Type of
system

PET
PET
PET
PET
Ethylene
dibenzoate
Ethylene
dibenzoate

catalyst

Antimony trioxide
trioxide
Antimony trioxide
Antimony triacetate
Zinc acetate

Antimony

(s-l)

Activation
energy
(kJ/mol)

275-285
272-292
275-285
270
280

OS-I.0 x 1O-6
1.O-5.0 x lo-
0.34 x lo-
1.2 x 1o-6

158
195
136
123.2

330-350

5-20 X 10-e

Temperature
03

Rate constant

223.6

Ref.

Yokoyama e&al. (1978b)


Tomita (1973)
Yu et al. (1980)
Rafler et 01. (198Ob)
Zimmermann
and Kim (1980)
Tomita (1977a)

K. RAVINDRANATH

2208

Stabilizers are generally added during polycondensation for improving thermal stability and quality of
PET. These are mainly phosphorus compounds such
as trialkyl or triaryl phosphite and the corresponding
thiophosphoric acid esters. The mechanism of the
stabilizer action is not clear from the literature.
Terechowa and Petuchow (1960) tried to improve the
stability of PET by blocking the end groups of PET
with orthophosphoric acid. However, the PET melt
could not be stabilized. Zimmermann (1962) indicated
that the stability can be improved by blocking the
transesterification catalyst which is believed to be
responsible for the thermal instability of PET melt.
Chang et al. (1982) found that the transesterification
catalysts are not blocked by these stabilizers and that
the DMT conversion is independent of the stabilizer
concentration. If the stabilizers are meant for blocking
the transesterification catalysts, then they need not be
used in the TPA process. In fact many PET producers
continue to use stabilizers in the TPA process. In the
presence of phosphoric acid and triphenyl phosphate
(TPP), the polycondensation rate is known to decrease
marginally (Yokoyama et al., 1978c; Kamatani et al.,
1980) and the possible mechanism of this observation
was provided by Kamatani et al. (1980). Chang et al.
(1982) found that TPP is the best stabilizer with
respect to acid end groups and DEG formation rate,
when TPP concentration was kept less than 0.04 o/0by
weight. Arunpal and Vineypal (1982) found that the
presence of water decreased the concentration of DEG
in the esterification process.
8. THERMODYNAMIC
PERTINENT

AND
TO PET

TRANSPORT

and R. A. MASHE~KAR

ant. In the case of TPA process, the first stage involves


dissolution of TPA accompanied by a set of chemical
reactions, and necessitates solid-liquid equilibrium
calculations. The specific systems for which accurate
data are required for PET reactor analysis are shown
in Table 6. For many systems, data do not exist in the
open literature and the data available for some of
the systems cannot be used for any meaningful
calculations.
Griehl and Forster (1956) studied equilibrium polycondensation and presented data in the form of DP as
a function of reactor pressure as shown in Fig. 4. These
data are useful in calculating the reactor operating
pressure that should be maintained for obtaining the
required DP. However, the data should be used with
caution for deducing any equilibrium relationships,
since the contribution of the volatile side products such
as acetaldehyde and water to the vapour pressure is
neglected in the analysis. Acetaldehyde will be under
supercritical conditions under the reaction operating

DATA

FORMATION

8.1. Phase equilibria


In the synthesis of PET, the reaction mixture is
present in a number of different forms. In the first
stage, where DMT is reacted with EG, vapour-liquid
equilibrium considerations for the multicomponent
system BHET-EG-methanol-DMT
become import-

ES

PRLSSUFfE,tOrr

IPOR

Fig. 4. Operating range of pressure for PET reactor indicating molecular weight build-up limitations (Griehl and
Forster, 1956).

Table 6. A summary of current status on phase equilibrium data pertinent to PET synthesis

System

NO.

Type of data
required

Ref.

Remarks

BHET/EG/methanol/
DMT

Vapour-liquid

Baker et al. (1964)

Data on some binary pairs


(EG-methanol) available.
Calculations usually done
by assuming ideality

TPA/EG

Solid-liquid

Kntmpolc and Malek


(1973)
Baranova (1977a, b)

Reasonably good data but dificult interpretation due to interference of chemical


reaction

TPA/BHET

Solid-liquid

TPA/EO

Solid-liquid

Krumpolc and Malek


(1973)
Baranova (1977a, b)
-

BHET/EG/water/TPA

Solid-liquid-vapour

BHET/EG

Vapour-liquid

Flory-Huggins

PET/EG

Vapour-liquid

Very complex analysis, no reliable data or


estimation techniques

n
No reliable data
No reliable data
retation used

Polyethyleneterephthalate--I
conditions and it may exert some influence on EG
removal by the process of supercritical extraction
(Sharma, 1984). No reliable data or estimates in the
range of operations are available to verify this conjecture. If the polymer contains number of acid end
groups, the achievable DP will be limited and there
should be a family of curves in Fig. 4 for various acid
end group concentrations.
For vapour-liquid equilibrium prediction involving
polymeric system Raoults law is known to fail.
Flory-Huggins relationship given below is generally
used.
ln~=ln&+

1-z

>

42+x+$.

(25)

Here p1 is the partial pressure of the solvent, p: is the


solvent vapour pressure, +i is the volume fraction of
solvent (1) or polymer (2), and Vi is the molar volume
of solvent (1) or polymer (2). x is the polymer solvent
interaction parameter and in the absence of actual data
it can be approximately estimated as
x = 0.34 + $T

(6r - S,)l

where 6, is the solubility parameter of the polymer and


8, is the solubility parameter of the solvent, R is gas
constant and T is absolute temperature. Vrentas et al.
(1983) determined the x parameters for different
polymer-solvent systems and found these to be in the
range of0.3-0.5. Although such approximations can be
used for engineering calculations, clearly there is a need
for more rigorous analysis in this area.
8.2. Molecular diflusion in PET
As we have remarked earlier, diffusional phenomena
coupled with chemical reactions play an important role
in the melt polycondensation as well as in the solid
state polycondensation of PET. It is therefore important that we review the current state of our understanding on diffusional phenomena in PET, both in molten
and solid state conditions. Let us first consider the
specific systems for which we will have to understand
the diffusional phenomena. The volatile byproducts
which diffuse out of PET melt are monomeric subacetaldehyde, water, etc.
stances such as EG,
Additionally compounds such as diphenyl carbonate,
diphenylterephthalate, etc. have been added as chain
extenders (Shima et al., 1973) with the objective of
enhancing the polycondensation rate through the
rapid diffusion of low molecular weight/small size
molecules such as phenol. Thus information on diffusion in such specific systems is also required. When
solid state polycondensation occurs the equilibrium
reaction is accompanied also by diffusion of EG
through the solid polymeric matrix which is invariably
semicrystalline in state (Buxbaum, 1979).
In many common physical systems, the diffusing
substrate is inert towards the medium through which
diffusion occurs. However, when reversible chemical
reactions result due to the interaction of the diffusing

2209

species (such as EG in PET), it becomes exceedingly


difficult to undertake any meaningful measurements.
Thus, for instance, studies on diffusivity of styrene
through molten polystyrene are relatively simple in
view of the fact that it is essentially a non-reactive
system, whereas study of diffusion of EG through PET
melt is virtually impossible due to the simultaneously
occurring reversible reactions. This implies that a total
analysis of the reaction-diffusion problem is needed
for deducing the diffusivity values. It is particularly
important to develop a priori models based on model
systems so that diffusivities could be estimated in
the above systems based entirely on structural
considerations.
8.3. Diflusion on molten polymers
Stokes-Einstein equation has been used for describing the diffusivity of a small solute through an ordinary
leading to the famous
solvent; its extension
Wilke-Chang equation (Sherwood et al., 1975) has
also been extensively used. Such equations predict an
inverse dependence of diffusivity on bulk (or macroscopic) viscosity. This equation fails when diffusion in
more viscous systems is considered and therefore
models which are more appropriate for viscous systems have been built. The Gainer-Metzner
model
(Gainer and Metner, 1965) for instance can be cited as
an example. They show that the sensitivity of diffusivity to the bulk viscosity is much lower in higher
viscosity media. However, even such equations fail
when they are used for polymeric systems. One
observes that diffusion of solutes in macromolecular
systems is hardly sensitive to the bulk viscosity of the
system (Mashelkar and Dutta, 1982, Kulkarni and
Mashelkar, 1983a). It is the segmental motion of the
polymer molecule that controls the passage of a small
solute. Therefore, it is not the macroviscosity of the
polymer solution or melt but the microviscosity (related to microbrownian motion of the segment) that
controls the diffusional phenomena. A number of
models developed (Wang, 1954; Clough et al., 1962;
Hoshino
and Sato,
1967) rely on this basic
presumption.
A mode1 that is specifically applicable to diffusion of
small solutes in polymer melts is provided by Duda
et al. (1982) and this will be described in some detail
now. They assume that the chemical potential of the
solvent in the solution can be assumed to follow
Flory-Huggins relationship and incorporate this thermodynamic information into the free volume theory
developed by them to obtain a relationship for the
binary diffusion coefficient D as
D = Do1 (1 x exp

&)(l - 2x41)
wlP:+w,<tir

P&v

1.

(27)

Here wr and w2 are the mass fractions of the solvent


and of the polymer, respectively, p : and P f are the
specific critical hole free volumes of the solvent and the
polymer molecules required for the jump, prFHand y

2210

and R.A.

K.RAVINDRANATH

are the average hole free volume per gram of mixture


and overlap factor for free volume, respectively. The
specific free volume for migration was expressed as

FF-WI
Y

Kll
yEK21

+w,-

Kl,
Y

-Tg,+T1
-[K22-Tg~+T].

(28)

The parameters K,,,


K2r, K,,
and K,,
are free
volume parameters. which can be determined from the
solvent and polymer viscosity data (Duda et al., 1982).
Ts, and Tgz are the glass transition temperatures for the
solvent and the polymer, respectively, which can be
readily calculated. Vrentas and Duda (1979) reconsidered the role of the energy term in eq. (27) giving
Ool = D,, exp[ - E/RT].

(29)

All the parameters in eqs (27 j(29) can be determined


(except DO, E and I;) from the viscosity and the
thermodynamic data, Duda and Vrentas proposed
that the most direct procedure to evaluate the theory
would be to use the limiting diffusivity data to evaluate
D,,, E and C;. They have demonstrated the utility of
eq. (27) by analysing the data on diffusion of toluene in
molten polystyrene, methyl acetate in polymethyl
acrylate, etc. Considering that the diffusivity values
undergo a change by several orders of magnitude in the
range of concentrations covered, the theory appears to
hold a great deal of promise. Although this theory
would appear to be useful in providing us with a
method of a priori prediction for system of our interest
(i.e. EG-PET melt system), the back up data required
for such calculations are not readily available at the
moment.

8.4.

Diflision

experiments

in molten

PET

There have been many efforts to generate data on


diffusion of EG through PET melt by undertaking well
defined experiments, and by interpreting them in terms
of reasonably comprehensive models which take into
account both diffusion and reaction. One such early
effort was by Pell and Davis (1973). They conducted
thin film experiments and deduced a constant value of
diffusivity of EG (D = 1.66 x lop4 cm2/s at 270C) in
the DP range of 50-120. A similar effort was undertaken by Rafler et al. (1979). By fitting the experimental
data with the model, they obtained a variable diffusivity dependent on DP. This observation is not
consistent with other investigators observations that
diffusivity is independent of the molecular weight after
a certain limit (Kulkarni and Mashelkar, 1983a). Later
Rafler et al. (1980) incorporated thermal degradation
and mass transport by convection in their previous
model and deduced a constant value of EG diffusivity
(D = 0.82 x lo- cm2/s) in the DP range of 20-150 at
270C. It needs to be emphasized that due to the
complicated diffusion-reaction problem, the value of
the diffusivity deduced from such experiments is very
much dependent upon the accuracy of the model itself.

MASHELKAR

This problem will be critically examined in Part II of


our contribution.
8.5. Diflision of volatile products in solid PET
The solid state polycondensation of PET is carried
out by heating the polymer below the melting temperature but above the glass transition temperature. The
main reaction product which is volatile diffuses out of
the solid semi-crystalline polymer. It is thus of importance to relate the diffusivity of a small molecule such as
EG to the structural characteristics of the polymer. It
has been experimentally well established that as the
crystallinity of the polymer increases the diffusivity
goes down. There is a need for models, which could
provide quantitative guidelines in such cases. This gap
has been partially filled by the recent work of Kulkarni
and Mashelkar (1983b). Their initial work of Kulkarni
and Mashelkar (1981) was based on modifying
Maxwells treatment by using a pseudo-two phase
concept. However, later on by using a unifying concept
of an altered free volume state description for a variety
of macromolecular systems, they obtained an expression for calculating diffusivity of a small solute in a
semicrystalline polymer. Their simple model is

where DC and D, are the diffusivities in the semicrystalline and purely amorphous polymer, respectively, B, is the jump factor, f, is the fractional free
volume of the amorphous polymer and #, is the
crystalline volume fraction of the polymer.
Experiments have been presented by Chang (1970)
on the kinetics of thermally induced solid state polycondensation process. By assuming that the process
was purely diffusion controlled, he was able to estimate
the diffusivity as a curve fit parameter. Based on his
observation he developed the following equation:
D = 3 x 1O-8 (s)

cm/s.

(31)

Here V is the limiting polymer specific volume. V, and


V, are the polymer specific volumes at time t and at
time t = 0 respectively. This expression can be recast in
terms of the instantaneous crystalline volume fraction
of the polymer and appears qualitatively consistent
with the model developed by Kulkarni and Mashelkar
(1983b) for low a degree of crystallinity.
Due to the importance of water in hydrolytic
degradation reactions, careful drying of PET chips is
normally carried out. Certain amount of data base
therefore exists on diffusion of water through solid
PET. Reference may be made in particular to the
studies by Stockburger and Faulhaber (1969), Yasuda
and Stannett (1962) and Whitehead (1977). Such
studies are useful for the approximate estimation of
drying times, but not perhaps totally adequate for
building up structure-property relationships.
Considering the present state-of-the-aft, it appears
that it is important to develop better predictive

Polyethylene terephthalate-I
methods to enable us to estimate more rigorously the
diffusivity of small molecules in molten and solid
polymeric systems. Such efforts are currently underway and they are likely to lead us to the goal of
successful a priori predictions.

DG

e,, es, e,
E
f.

9. CONCLUDING

REMARKS
FURTHER

AND

SUGGESTIONS

FOR

WORK

Although extensive studies have been undertaken on


the kinetics and mechanisms of reactions leading to
PET synthesis, many critical areas still need to be
explored. For instance, the influence of trace acid
impurities (e.g. benzoic acid or TPA) on transesterification catalysts is still not well understood. Although
ester&cation reactions are conventionally carried out
without a catalyst, the formation of side products is
rather high and it is important to examine the possibility of using certain inhibitors to reduce the side
product formation rate. The isolation and evaluation
of the trace quantities of chromophoric
substances,
causing problems in colour stability is a major issue,
which requires the most sophisticated tools and approach. PET is used for the manufacture of artificial
arteries in current biomedical materials practice. Steam
sterilization at even moderately
high temperatures
(Z 120C) appears to produce side products and oligomers, which cause subsequent problems. It is important to examine degradation kinetics and mechanisms
under such conditions as well. A search for a more
powerful
polycondensation
catalysts which inhibit
side reactions will certainly be useful. The future trends
would be on easily handled soluble catalysts; in view of
the emergence of continuous processes.
Although a fair bit of effort has been spent on
studies related to the mechanism of side products
formation,
reliable
kinetic
studies are still not
available. This limits the scope for accurate modelling.
There is a clear dearth of thermodynamic
and
transport data. Due to the peculiar conditions existing
during the finishing stages of polycondensation
reactions (high temperature, extremely low pressure, high
viscosities, reversible equilibrium reactions and side
products formation), this becomes a formidable task.
More profound advances in our experimental abilities
and theoretical predictions will be needed to bridge
this gap.

NOTATION
B,

c0
d
D

D0
DCll
0,. D,

jump factor
bis(hydroxyethy1)
concentration
of
terephthalate
parameter for degradation reaction
diffusion coefficient
preexponential
factor for eq. (29)
preexponential
factor defined by eq. (29)
diffusivities
in the semicrystalline
and
purely amorphous polymer

kd
K

Kc, Kb
m

N
P

PI
P'I

P
P no
PG

PT
R

t
T
T.1

Ul,

Ts2

02

v,
VO

V
P:,

Py

vFH
W
Wlr W2
X

Y
z

2211
diffusivity of ethylene glycol
concentration of acid, hydroxyl end and
methyl end groups
critical energy per mole needed to overcome attractive forces
fractional free volume of the amorphous
polymer concentration of ethylene glycol
rate constant for polycondensation
reaction
rate constant for degradation reaction
equilibrium
rate
polycondensation
constant
free volume parameters for solvent
free volume parameters for polymer
equilibrium
rate
transesterification
constants
esterification equilibrium rate constants
concentration of methanol
stirrer speed
parameter for polycondensation
reaction
partial pressure of solvent
solvent vapour pressure
degree of polymerization
initial degree of polymerization
vapour pressure of ethylene glycol
reactor pressure
gas constant
time
absolute temperature
glass transition temperature of the solvent and polymer, respectively
molar volume of the solvent and polymer
polymer specific volume at time t
initial polymer specific volume
Iimiting polymer specific volume
critical hole free volumes of the solvent
and the polymer
specific free volume for migration
concentration of water
mass fractions of the solvent and polymer,
respectively
distance from the interface of the film
mole fraction of ethylene glycol in the
gaseous phase
concentration of diester groups

symbols
stability constant of dibenzoal methane
complex
overlap factor for free volume
Y
solubility of the solvent and polymer
a,, $3
respectively
ratio of critical molar volume of solvent
c
jumping unit to critical molar volume of
jumping unit of polymer
volume fraction of solvent and polymer
419 42
respectively
crystalline volume fraction of the polymer
4,
polymer-solvent
interaction parameter
x
interfacial area
E
viscosity
fi

Greek

2212

K. RAVINDRANATH and R. A. MASHELKAR


REFERENCES

Arunpal, A., 1978, Polyethylene terephthalate feed stock


selection and process options. Chem. Age India 29,
733-737.
Arunpal. A. and Vineypal, A., 1982, The effect ofwater and air
contamination on polyethylene terephthalate formation.
Polymer Engng Rev. 2, 123-133.
Baker, T. H., Fisher, G. T. and Roth, J. A., 1964, Vapor-liquid
equilibrium
and
refractive
indices
methof
the
anol+ethylene glycol system. J. Chem. Engng Data 9,1 l-12.
Baranova, T. L. and Kremer, E. B., 1977a, Role of degree of
dispersion of terephthalic acid in the process of its esterification by ethylene glycol during the synthesis of polyethylene terephthalate. Khim Volokna 19(2), 47-58.
Baranova, T. L. and Kremer, E. B., 1977b, Dissolving of
terephthalic acid in the reaction system terephthalic
acid-ethylene glycol esterification products. Khim Yolokna
19(4), 1618.
Bednas, M. E., Day, M., Ho, K.. Sander, R. and Wiles, D. M.,
1981,
Combustion
and
pyrolysis
of
polyethylene
terephthalate--I. The roie of flame retardants on products
of pyrolysis. J. appl. Polymer Sci. 26, 277-289.
Bhatia, S.. Rao, M. G. and Rao, M. S., 1976, A kinetic study of
the reaction between terephthalic acid and ethylene oxide
at atmospheric pressure. J. appl. Chem. Biotechnol. 26,
295304.
Block Ulrich, 1977, Carrying out continuous reactions with
super posed distillation. Chem.-Zng.-Tech. 49, 151.
Borchardt, J. A., 1971, Biological waste water treatment using
rotating discs. Biotechnol. Bioengng Symp. 2, 131-140.
Brian, P. L. T., Hurley, J. F. and Hasseltine, E. H., 1961,
Penetration theory for gas absorption accompanied by
second order chemical reaction. A.I.Ch.E. J. 7, 226231.
Buxbaum, L. H., 1968, The degradation of polyethylene
terephthalate. Angew. Chem. Znt. Ed. 7, 182-190.
Buxbaum, L. H., 1979, Solid state polycondensation
of
polybutylene terephthalate. J. appl. Polymer. Sci: Appl.
Polymer Symp. 35, 5946.
Cafelin, P. and Malek, J., 1969, On the kinetics of polycondensation of bis(2-hydroxyethyl)terephthalate.
Colln.
Czech. them. Commun. 34, 419426.
Challa, G., 196Oa, The formation of polyethylene terephthalate by ester interchange-I.
The polycondensation
equilibrium. Makromol. Chem. 38, 105122.
Challa, G., 196Ob, The formation of polyethylene terephthalate by ester interchange-II.
The kinetics of reversible melt polycondensation.
Makromol. Chem. 38,
123-137.
Challa, G., 196Ck, Ester interchange equilibria from dimethyl
terephthalate and ethylene glycol. Rec. Trau. Chim. 79,
-100.
Chang, T. M., 1970, Kinetics of thermally induced solid state
polycondensation of polyethylene terephthalate. Polymer
&t&g Sci. 10, 364-368. _
Chang, S., Sheu, M. and Chang, N., 1982, Effect of stabilizers
on the preparation of polyethylene terephthalate. J.
Polymer Sci.: Polymer Chem. Ed. 20. 20532061.
Chavan, V. V. and Mashelkar, R. A., 1980, Mixing of viscous
Newtonian and non-Newtonian
fluids. in Advances in
Transport Processes (Edited by Mujundar, A. S. and
Mashelkar, R. A.), Vol. 1, pp.
_ _ 210-252. Wilev Eastern Ltd..
New Delhi.
Chegolya, A. S., Shevchenko, V. V. and Mikhailov, G. D.,
1979, The formation of polyethylene terephthalate in the
presence of dicarboxylic acids. J. Polymer Sci.: Polymer
Chem. Ed. 17, 889-904.
Clough, S. B., Reed, H. E., Metzner, A. B. and Behn, V. C.,
1962, Diffusion in slurries and in non-Newtonian fluids.
A.I.Ch.E. J. 8. 346-350.
Datye, K. V. and Raje, H. M., 1985, Kinetics of transesterification of dimethyl terephthalate with ethylene glycol. J.
appl. Polymer Sci. 30, 205-219.
Devaux, J., Godard, P. and Mercier. J. P., 1982, Bisphenol-a

polycarbonate-polybutylene
terephthalate transesterification-l.
Theoretical study of the structure and of the
degree of randomness in four component polycondensates.
J. Polymer Sci. Polymer Phys. Ed. 20, 18751880.
Denson, C. D., 1983. Stripping operations in polymer processing, in Advances in Chemical Engineering (Edited by Wei,
J., Bischoff, K. B., Thomas, B. D. and John, H. S.), Vol. 12,
pp. 61-105. Academic Press, New York.
Dollimore, D., Gamlen, G. A., Jefferies, M. and Shah, T. H.,
1983, The use of the rising temperature technique to
establish the kinetics of acetaldehyde evolution during
atmospheric
polymerization
of bis(hydroxyethyl)terephthalate. Thermachim. Acta 61, 97-106.
Duda, J. L., Vrentas, J. S., Ju, S. T. and Liu, H. T., 1982,
Prediction of diffusion coefficients for polymer-solvent
systems. A.1.Ch.E. J. 28, 279-285.
Ellwood, P., 1967, Continuous polyester condensation is
obtained with two reactors. Chemical Engineering, 20
November, pp. 98-100.
Fakirov, S., Seganov, I. and Kurdowa, E., 1981, Effect ofchain
composition on polyethylene terephthalate structure and
properties. Mokromol. Chem. 182, 185-197.
Fontana, C. M., 1968, Polycondensation equlibrium and the
kinetics of the catalyzed transesterification in the formation
of polyethylene terephthalate. J. Polymer Sci.: Part A-l 6,
23432358.
Fortunato, B., Pilati, F. and Manaresi, P., 1981, Solid state
polycondensation of polybutylene terephthalate. Polymer
22, 651657.
Gainer, J. L. and Metzner, A. B., 1965, Diffusion in liquids:
theoretical
analysis
and
experimental
verification.
A.Z.Ch.E. Symp. Ser. 6, 7+82.
Goodings, E. P., 1961, Thermal degradation of PET. Sot.
Chem. Znd. (London). Monograph No. 13, pp. 211-228.
Goodman, I. and Nesbitt, B. F., 1960, The structures and
reversible polymerization of cyclic oligomers from PET. J.
Polymer Sci. 48, 423433.
Grlehl, W. and Forster, P., 1956, The determination of mass
reaction constant for polyethylene terephthalate formation. Faserforch. U. Textilteahnik 7, 463468.
Griehl, W. and Schnock, G., 1957, Kinetics of polyesterification by reesterification. J. Polymer Sci. 30, 413-422.
Griskey, R. G. and Lee, B. I., 1966, Thermally induced solid
state polymerization in Nylon 66. J. appl. Polymer Sci. IO,
105-111.
Ha, W. S. and Choun, Y. K., 1979, Kinetic studies on the
formation
of cyclic oligomers
in polyethylene
terephthalate. J. Polymer Sci.: Polymer Chem. Ed. 17,
21032118.
Heffelfinger, C. J. and Knox, K. L., 1971, Polyester films, in
The Science and Technology of Polymer Films (Edited by
Sweeting, 0. J.), Volume II, pp. 587639. Wiley, New York.
Hovenkamp, S. G., 1971, Kinetic aspects of catalyzed reactions in the formation of polyethylene terephthalate. J.
Polymer Sci.: Part A-l 9, 3617-3625.
Hovenkamp, S. G. and Munting, J. P.. 1970, Formation of
diethylene glycol as a side reaction during production of
polyethylene terephthalate. J. Polymer Sci.: Parr A-l 8,
679-682.
Homof, V., 1981, Influence of metal catalysts on the formation of ether links in polyethylene terephthalate. J.
Macromol. ScL-Chem. A15(3), 503-514.
Hoshino, S. and Sato, K., 1967, The diffusion of a small
molecule in a polymer solution. Kagaku Kogaku 31,
961-966.
Jabarin, S. A. and Lofgren, E. A., 1984, Thermal stability of
polyethylene
terephthalate.
Polymer
Engng Sci. 24,
1056-1063.
Jenekhe, S. A. and Lin, J. W., 1983, Kinetics of the thermal
degradation of polyethylene terephthalate. Thermochim.
Acta 61, 287-299.
Kaibel, G., Mayer, H. H. and Seid, Et., 1978, Reactions in
distillation columns. Chem.-Zng.-Tech. 50, 586592.
Kamatani, H., 1977, The reaction of terephthalic acid with

Polyethylene terephthalatel
ethylene oxide in the presence of amines. Nippon Kagaku
Kaishi
10, 1505-1511.
Kamatani, H., 1978, Kinetic studies on the reaction of
terephthalic acid with ethylene oxide to form ether linkage.
Nippon Kagaku Kaishi 9, 1271-1275.
Kamatani, H. and Konagaya, S., 1978, The effect of Mn(II).
Co(H) and Zn(I1) compounds on the polycondensation of
bis(hydroxyethyl)terephthalate. Kobunshi Ronbunshu 35,
787-794.
Kamatani, H. and Kuze, K., 1979, Formation of aromatic
compounds as side reactions in the polycondensation of
bishydroxyethyl terephthalate. Polymer J. 11, 787-793.
Kamatani, H. and Konagaya, S., 1979, Effect of phosphorus
compounds on the activity of catalyst (Zn(II), Co(H),
Mn(I1))
in the polycondensation
of bis(2-hydroxyethyl)terephthalate. Kobunshi Ronbunshu 36. 293298.
Kamatani, H., Konagaya, S. and Nakamura, Y., 1980, Effect
of
phosphoric
acid on
the polycondensation
of
bis(hydroxyethyl)terephthalate catalyzed by Sb(III) compounds. Polymer J. 12, 125130.
Katz, M., 1977, Preparation of linear saturated polyesters, in
Polymerization Processes
(Edited by Schildknecht, C. E.
and Skeist, I.), pp. 468496.
Wiley, New York.
Kemblowski, Z. and Torzccki, J., 1982, The principles of pipeline design for molten polyethylene terephthalate. Polymer
Engng Sci. 22, 141-146.
Kemkes, J. F., 1969, Direct esterification of terephthalic acid
with ethylene glycol under atmospheric pressure. J.
Polymer Sci.: Part C 22, 713720.
Koepp, H.-M. and Werner, H., 1959, Determination of end
groups and molecular weight distribution of polyethylene
terephthalate. Makromol. Chemie 32, 79-89.
Kotliar, A. M., 198 1, Interchange reactions involving condensation polymers. J. Polymer Sci. Macrowwl. Rev.
16,
367-395.
Krumpolc, M. and Malek, .I., 1973, Esterifieation of bengenecarboxylic acids with ethylene glycok 4. Kinetics of the
initial stage of polyesterification of terephthalic acid with
ethylene glywl catalyzed by zinc oxide. Makromol.
Chem.
171,69-81.
Kulkarni, M. G. and Mashelkar, R. A., 1981, On the role of
penetrant structure in diffusion in structured polymers.
Polymer
22, 16651672.
Kulkarni, M. G. and Mashelkar, R. A., 1983a, A unified
approach to transport phenomena in polymeric media-I.
Diffusion in polymeric solutions, gels and melts. Chem.
Engng Sci. 36,925-940.
Kulkarni, M. G. and Mashelkar, R. A., 1983b. A unified
approach to transport phenomena in polymeric media--II.
Diffusion in structured solid polymers. Chem. Engng Sci.
38,941-957.
Kumar, A., Sharma, S. N. and Gupta, S. K., 1984,
Optimization of the polycondensation step of polyethylene
terephthalate formation with continuous removal of condensation products. Polymer
Engng Sci. 24, 1205-1214.
Lee, W. K., Jin. B. Y., Lee, J. S., Kim, J. I. and Yi, S. H., 1983, A
kinetic study of the reaction between terephthalic acid and
ethylene oxide in a nonsolvent system. Chem. Engng Sci. 38.
1021-1029.
Li-Chen, Hsu, 1967, Synthesis of ultra high molecular weight
of polyethylene terephthalate. J. Macronwl. Sci. Phys.
M(4), 801-813.
Lufkin, P., 1981, Synthetic fibres in the eighties. Textile
Month, February, pp. 23-24.
Maerov, S. B., 1979, Influence of antimony catalysts with
hydroxyethoxy ligands on polyester polymerization. J.
Polymer
Sci.: Polymer
Chem. Ed. 17, 403-m.
Mares, F., Hetflezs, J. and Bazant, V., 1969, Reaction of
carboxylic acids with ethylene oxide. I. Kinetics of the
reaction of terephthalic acid with ethylene oxide, Colln.
Czech. them. Commun.
34, 3086.
Marshall, I. and Todd, A., 1953, The thermal degradation of
polyethylene terephthalate. Trans. Faraday Sot. 49,67-78.
Mashelkar, R. A. and Dutta, A., 1982, Convective diffusion

2213

in structured fluids: need for new analysis and design


strategies. Chem. Engng Sci. 37. 969-985.
McIntyre, J. E., 1982, Polyester fibres and films from p-xylene,
in Toluene,
The Xylenes
and their Industrial
Derivatives
(Edited by Hancock, E. G.), Chapter 14, pp. 400444.
Elsevier Scientific, The Netherlands.
Nagasubramanian,
K. and Reimschuessel, H. K., 1973,
Diffusion of water and caprolactam in nylon 6 melts. J.
appl. Polymer
Sci. 17, 16631677.
Peebles, Jr. L. H. and Wagner, W. S., 1959, The kinetic
analysis of a distillation system and its application to
preliminary data on the transesterification of dimethyl
terephthalate by ethylene glycol. J. phys.
Chem.
63,
1206-1212.
Peebles, Jr. L. H., Huffman, M. W. and Ablett, C. T., 1969,
Isolation and identification of the linear and cyclic oligomers of polyethylene terephthalate and the mechanism
of cyclic oligomer formation. J. Polymer
Sci.: Part A-l 7,
479-496.
Pell, Jr. T. M. and Davis, T. G., 1973, Diffusion and reaction in
polyester melts. J. Polymer
Sci.: Polymer
Phys. Ed. 11,
1671-1682.
Pilati, F., Manaresi, P., Fortunato, B., Munari. A. and
Passalacqura. V., 1981, Formation of polybutylene terephthalate: secondary reactions studied by model molecules. Polymer
22, 1566-1570.
Rafler, G., Bonatz, E. and Reinisch, G., 1973a, The kinetics of
the formation of polyethylene terephthalate in an open
Polycondensation
of
bis(B-hydroxyethyl)
systerr+-I.
terephthalate in a stirred polycondensation
system.
Faserforch
U. Textiltechnik
24, 235-239.
Rafler, G., Bonatz, E. and Rein&h, G., 1973b, The kinetics of
the formation of polyethylene terephthalate in an open
system-II.
Effect of thermal side reactions on the polycondensation process. Faserforch
U. Textiltechnik
24,
269-272.
Rafler, G., Rein&h, G. and Bonatz, E., 1974, Kinetics, mass
transport and thoughts about the mechanism of the
formation of oolyethylene terephthalate by metal ion
kl, 253-267:
catalysis. Acta khirn. (Budapest)
Rafler. G., Bonatz E.. Rein&h,
G.. Gaiewski, H. and
Zacharias, K., 1979, Reaction and diffusion ii the melt
polycondensation
of polyethylene terephthalate. Aeta
Polymerica
30, 253-257.
Rafler, G., Bonatz, E., Rein&h,
G.. Gajewski, H. and
Zacharias. K., 1980, Extension of the reaction-diffusion
model of melt polycondensation. Acta Polymerica
31,
732-733.
Rafler, G., Blaeshce, J., Moller, B. and Stromeyer, M., 1981,
The kinetics and mechanism of thermolysis of poly(alkylene terephthalate)s. Acta Polymerica 32, 608-611.
Ravindranath. K. and Mashelkar. R. A., 1981, Modelling of
polyethylene terephthalate reactors. 1. A semibatch ester
interchange reactor. J. appl. Polymer
Sci. 26, 3179-3204.
Ravindranath, K. and Mashelkar. R. A., 1982a, Modellina of
polyethylene terephthalate reactors 4. A continuous-ester&cation process. Polvmer
Ennna Sci. 22. 610-618.
RavindranathfK. and Maihelkar, I?. A., 19826, Reanalysis of
kinetics of transesterification of dimethyl terephthalate
with ethylene glycol. J. Polymer Sci. Polymer Chem. Ed. 20,
3447-34so.
Ravindranath, K. and Mashelkar, R. A., 1985, Recent advances in polyethylene terephthalate manufacture, in
Developments
in Plastic
Technology-2,
Chapter 1, pp.
142. Elsevier Applied Science, London.
Ravindranath, K. and Mashelkar, R. A., 1986, Polyethylene
terephthalate-II.
Engineering analysis. Chem. Engng Sci.
41, in press.
Reimschuessel, H. K., 1980, Polyethylene tereplithalate formation. Mechanistic and kinetic aspects of the direct
esterification process. Ind. Engng Chem. Prod. Res. Dew. 19,
117-125.
Reimschuessel, H. K. and Debona, B. T., 1979, Terephthalic
acid esterification kinetics: 2-(2-methoxy)ethyl terephtha-

2214

K. RAVINDRANATHand R. A. MASHELKAR

lates. J. PoIymer Sci.: Polymer


Chem. Ed. 17, 3241-3254.
Reimschuessel, H. K., Debona, 3. T. and Murthy, A. K. S.,
1979, Kinetics and mechanism of the formation of glycol
esters: benzoic acid-ethylene alvcol svstem. J. Polvmer Sci.:
Polymer
Chem. Ed. 17.321713239.
Renwen. H.. Fena. Y.. Tinzhena, H. and Shimina G., 1983,
The kinetics 07. formation of diethylene gly& in preparation of polyethylene terephthalate and its control in
reactor design and operation. Angew.
Makromoi.
Chem.
119, 159-172.
Rod, V., El Diwani, G., Minarik. M. and Sir, 2.. 1976, Kinetics
of esterification of terenhthalic acid with ethylene alycol.
Colln Czech. &em. Cot&m.
41. 2339-2352.
_
- Schaul. J. S.. 1981. Drvinn and iniection moldina PET for
beverage bottle pref&r&. Polymer
Plast. Tech&l.
Engng
16,209230.
Secoi, R. M., 1969, The kinetics of condensation polymerization. A.I.Ch.E.
J. 15, 861-865.
Shah, Y. T. and Sharma, M. M., 1976, Desorption with or
without chemical reaction. Trans. Inst. Chem. Engrs 54,
141.
Shah, T. H., Bhatty, J. I., Gamlen. G. A. and Dollimore, D..
1984, Aspects of the chemistry of polyethylene terephthalate.
III. Transesterification
of dimethyl
terephthalate with ethylene glycol in the presence of various
catalyst systems. J. Macromol. Sci. Chem. A21 (4), 413443.
Sharma, M. M., 1984, Private communication.
Shima, T., Ursaki, T. and Oka, I., 1973, Improved process for
oolvcondensation of hinh molecular weight PET in the
presence of acid derivatives, in Polymerizai!ion
Kinetics and
Technology
(Edited by Plazer, N. A. J.), Adv. in Chem.
Series No. 128, pp. 183-207. Am. Chem. Sot., Washington.
Sherwood. T. K.. Pieford. R. L. and Wilke. C. R.. 1975. Mass
Transfer, p. 25. McGraw-Hill,
New York.
Sorokin,
M.
F.
and
Chebotareva,
N.
A.,
1969,
Transesterification of the dimethyl esters of terephthalic
Inst. 61,
acid by ethylene glycol. Tr. Mosk. Khim.-Tekhnol.
103106.
Stevenson, R. W., 1969, Polycondensation rate of PET-II.
Antimony trioxide catalyzed polycondensation in static
thin films on metal surfaces. J. Polymer
Sci.: Part A-l, 7.
395-407.
Stockburger, D. and Faulhaber. F. R.. 1969, Drying behaviour
of plastics. Chem.-lug-Tech. 41,4X&461.
Tani, T. and Enoki, K., 1970, Make BHET without solvents.
Hydrocarbon
Processing,
November, pp. 146-150.
Terechowa, G. M. and Petuchow, B. W., 1960, Chemiefbsern
4, 8.
Ton-&a, K., 1973, Studies on the formation of polyethylene
terephthalate: 1. Propagation and degradation reactions in
the polycondensation of bis(2-hydroxyethyl)terephthalate,
Polymer
14, 50-54.
Torn&a, K., 1976a, Studies on the formation of polyethylene
terephthalate: 6. Catalytic activity of metal compounds in
poiycondensation
of
bis(2-hydroxyethyl)terephthalate.
Polymer
17, 221-224.
Tomita, K., 1976b, Some considerations on the cataiysis
mechanism of metal compounds in the polycondensation
process of bis(2-hydroxyethyl)terephthalate.
Kobunshi
Ronbunshu 33.96101.
Tomita, K., 1977a, Studies on the formation of polyethylene
terephthalate: 9. Thermal decomposition ofethylene dibenzoate as a model compound of polyethylene terephthalate.
Polymer
18, 295-297.
Tomita, K., 1977b, Studies on the formation of polyethylene
terephthalate: 10. Elevation of the catalytic activity of
antimony (V) compound in the synthesis of polyethylene
terephthalate. Kobunshi
Ronbunshu
34, 675-678.
Tomita, K. and Ida, H., 1973, Studies on the formation of
polyethylene terephthalate: 2. Rate of transesterification of

dimethyl terephthalate with ethylene glycol. Polymer 14,


5560.
Tom&a, K. and Ida, H., 1975, Studies on the formation of
polyethylene terephthalate: 3. Catalytic activity of metal
comoounds in transesterification of dimethvl tereohthalate
_
with-ethylene glycol. Polymer
16, 185-190:
Tsimmerman. G. and Shaaf, E.. 1973, Chemistry of transesteriftcation catalyzed by me&l ions in the preparation of
Vysokomol.
Soyed.
AI5,
polyethylene
terephthalate.
415422.
Vrentas, J. S. and Duda, J. L., 1979, Molecular diffusion in
polymer solutions. A.1.Ch.E. J. 25, l-24.
Vrentas, J. S., Duda, J. L. and Hsieh, S. T., 1983, Thermodynamic properties of some amorphous polymersolvent systems. Ind. Engng Chem. Prod. Res. Dev. 22,
326-330.
Walker, C. C., 1983. The inhibitory effect of carboxylic
acids on the catalyzed reaction between dimethyl terephthalate and ethylene glycol. J. Polym. Sci.: Polym. Chem.
Ed. 21, 623626.
Wang, J. L., 1954, Theory of the self diffusion of water in
protein solutions. A new method for studying the hydration and shape of protein molecules. J. Am. them. Sot.
76,4755-4763.
Whitehead. B. D., 1977. The crystallization and drying of
oolvethvlene terenhthalate. Ind. Engno Chem. Process Des.
b&. 16; 341-346:
Yamanis. J. and Adelman. M.. 1976a, Significance of oligome&tiou
reactions in the tran&st&ification of dimethyl terephthalate with ethylene glycol. J. Polymer Sci.
Polymer
Chem. Ed. 14, 1945-1959.
Yamanis, J. and Adelman, M., 1976b. Two models for the
kinetics of the transesterification of dimethyl terephthalate
with ethylene glycol. J. Polymer Sci.: Polymer Chem. Ed. 14,
1961-1973.
Yasuda, H. and Stannett, V., 1962, Permeation, solution and
diffusion of water in some high polymers. J. Polymer Sci.
57,907-923.
K., 1971. Catalytic action of metal compounds in
Yoda,
transesterification. Kogyo Kagoku
Zasshi 74, 14761482.
Yokoyama, H., Sano, T., Chijiiwa, T. and Kajiya, R., 1978a,
Polycondensation rate of polyethylene terephthalate at
reduced pressure. J. Japan Petrol. Inst. 21, 58-62.
Yokoyama, H., Sane, T., Chijiiwa, T. and Kajiya, R., 1978b,
An effect of reduced uressure on the formation of nolvethylene terephthalate. J. Japan Petrol. Inst. 21, 77-?9.Yokovama. H.. Sano. T.. Chiiiiwa. T. and Kaiiva. R.. 1978~.
In&et&
of catalyst, stabilizer. and temperature -on the
ethvlene alvcol terenhthalate Dolvcondensation orocess. J.
Jajan.
P%ol.
Inst.21,
208-&O.Yokovama. H.. Sano. T.. Chiiiiwa. T. and Kaiiva. R.. 1978d.
Degradatiod reactions in-ethylene glycoi ierephthalatd
polycondensation
process. J. Japan
Petrol.
Inst.
21,
194198.
Yu, L. Q., Qun, Z. C. and Wen, H. R., 1980, Semiempirical
formula of the ester reaction kinetics. J. Chem. Ind. Engng
(China) 1, 95-101.
Zimmermann, H., 1962, Chemical studies on fibre forming
polyesters-I.
Thermal stabilization of polyethylene terephthalate. Faserforch.
Textiltech.
13, 481490.
Zimmermann, H. and Leibnitz, E., 1965, Chemical studies on
fiber forming polymers-II.
Model experiments on the
thermal decomposition
of polyethylene terephthalate,
Faserforch.
Textiltech. 16, 282-290.
Zimmermann. H. and Kim, N. T., 1980, Investigations on
thermal and hydrolytic degradation of polyethylene terEngng Sci. 20, 680683.
ephthalate. Polymer
Zimmermann, H., Becker, D. and Schaaf, E., 1979, Thermo
analytical and polymer-analytical
investigations
on
thermo-oxidative
degradation of polyesters. J. appt.
Polymer
Sci. Appl. Polymer
Symp. 35, 183-191.

Potrebbero piacerti anche