Content uploaded by Mehrsa Emami
Author content
All content in this area was uploaded by Mehrsa Emami on Jul 31, 2018
Content may be subject to copyright.
Vol.:(0123456789)
1 3
Iran Polym J
DOI 10.1007/s13726-017-0573-6
ORIGINAL PAPER
Gas‑phase polymerization ofpropylene atlow reaction rates:
aprecise look atcatalyst fragmentation
MehrsaEmami1· MahmoudParvazinia1· HosseinAbedini1
Received: 9 March 2017 / Accepted: 14 October 2017
© Iran Polymer and Petrochemical Institute 2017
Introduction
Understanding of the mechanism of polymer growth has
been, and still is, one of the most important fields of hetero-
geneous olefin polymerization studies. Many efforts have
been made by research groups to give insight into the polym-
erization kinetics, catalyst morphology, polymer properties
and mass transfer interdependency [1–3].
Catalyst fragmentation during polymerization affects
polymerization activity of the catalyst as well as proper-
ties of the produced polymer (particle density, porosity,
etc.) [4–6]. The morphology of the final polymer particles
depends on the mechanical and structural properties of both
the catalyst support and the polymer formed at the critical
very early stage of the polymerization [7–9].
The first observation relevant to the mechanism of growth
of polyolefins on heterogeneous catalysts dates back to the
1970s. Bulls and Higgins with the aim of microscopic
images showed fragments of the catalyst particle dispersed
within the expanding polymer [10]. Beside model predic-
tions [11–13], many experimental investigations were car-
ried out by researchers on catalyst fragmentation.
Microreactors have been recently employed for study-
ing catalytic olefin polymerization reactions [14]. The use
of microreactor devices makes it possible for researchers
to perform reactions with good control over mixing, pres-
sure, flow rate, residence time, mass and heat transfer, which
resulted in enhanced reproducibility [15]. Different reactor
and configurations have been designed to study very specific
aspects of the reactions process.
Based on the use of microreactors coupled with visual
and/or infrared (IR) microscopy, many insitu or online study
methods have been developed to investigate particle growth.
With the aim of online microscopy, particle growth kinet-
ics during polymerization had been followed and profitable
Abstract In this work, we reported the development of a
mini-reactor experimental setup for synthesizing of polypro-
pylene with heterogeneous Ziegler–Natta catalysts in gas-
phase. Use of pro-activated 4th generation of Ziegler–Natta
catalyst and preheated monomer feed enabled the polym-
erization reaction to be carried out at constant temperature.
Evaluation of monomer consumption with high precision
(0.01bar pressure drop) allowed the detection of polymeri-
zation yield at low reaction rates. In this regard, polymeri-
zation yield, particle morphology and catalyst fragmenta-
tion were studied, as well. The results of melt microscopy
showed that catalyst fragmentation was developed during
the reaction, and was not restricted to the initial rupture of
catalyst particles. The rate determination showed a peak dur-
ing the polymerization (not necessarily at the initial stage).
The results showed that depending on the reaction condition,
this peak could be either a consequence of a major cata-
lyst fragmentation or overheating. Low reaction yield, large
fragments of catalyst and agglomeration of particles were
considered as evidence of particle overheating and polymer
local melting. As we imposed the results of melt microscopy
for the polymerization conditions, a layer-by-layer fragmen-
tation of the catalyst was found to be the main fragmenta-
tion process, at least at the beginning of the polymerization
reaction.
Keywords Mini-reactor· Propylene· Polymerization·
Gas-phase· Fragmentation
* Mahmoud Parvazinia
M.Parvazinia@ippi.ac.ir
1 Iran Polymer andPetrochemical Institute, P.O.
Box14185/458, Tehran, Iran
Iran Polym J
1 3
results were obtained [16, 17]. Oleshko etal. [18] employed
insitu video-microscopy (combination of a microreactor and
a video camera) to follow the polymer particle growth dur-
ing polymerization of olefins with supported catalysts. Pater
etal. [19] developed a method for online observation of pol-
ymer particle growth, using optical microscopy and infrared
imaging. They measured surface temperatures of growing
particles and estimated particle temperature as a function
of time with infrared imaging method. Abboud etal. also
employed insitu video microscopy connected to a 100-mL
mini-reactor, to follow the particle growth during polym-
erization of olefins with supported catalysts [20–22]. May-
rhofer and Paulik used a microreactor equipped with a video
microscope to follow the growth of polyethyleneparticles
during polymerization. They conducted gas-phase polym-
erizations at conditions close to the industrial set points.
The rate of polymerization was calculated from the obtained
particle growth curves [23]. Video-microscopy methods are
limited by the resolution of microscope and infrared imaging
techniques. The advantage of online techniques is that the
polymerization and analysis are concurrent.
A number of off-line or ex situ techniques have been
developed to study particle growth during polymerization,
as well. In off-line methods, product is recovered and ana-
lysed after reaction. Different specially designed reactors
have been reported in the literature to follow heterogeneous
catalyst fragmentation, particle expansion, and polymeriza-
tion kinetic. Stopped-flow and quenched-flow methods are
two techniques applied for morphological and kinetic studies
in slurry polymerizations [24, 25]. In these methods, two dif-
ferent streams of reactants are contacted inside a mixing ele-
ment, reacted for a defined time and then fed into a quench-
ing vessel [26, 27]. Di Martino etal. [28, 29] conducted
slurry polymerizations of ethylene on a MgCl2-supported
Ziegler–Natta catalyst at very short times (40ms) with a
quenched-flow apparatus. Their investigations provided
detailed information on the kinetic, morphology and poly-
mer properties at the very early stages of polymerization.
Silva etal. [30] developed the short-stop technique in
2005. The major advantage of this technique is its potential
to perform the polymerization reaction at conditions similar
to the industrial conditions in terms of temperature, pressure
andflow rate. In short-stop technique, gas-phase polymeri-
zation was carried out continuously in a fixed bed reactor.
The particles of catalyst were dispersed in an inert seed-
bed and monomer flows over the particles. The reaction is
halted with the flow of carbon dioxide. Using stopped-flow
method, it was possible to study the evolution of the particle
morphology during the first instants of the polymerization
reaction [31, 32]. Some research works were developed by
reactor optimization and calorimetric method to follow heat
transfer, polymerization kinetics and evolution of catalyst
temperature during the first moments of reaction [33, 34].
McKenna etal. [35] have shown that it was not a trivial
task to run experiments that allow one to obtain reliable
data on the kinetic and evolution of material properties and
particle morphology during the early stages of the catalytic
olefin polymerization. The main difficulties come from the
sensitivity of the catalysts to impurities, the small size of
the catalyst particles in comparisonwith very rapid reac-
tions that lead to heat and mass transfer limitations. For all
of these reasons, it is difficult to build an experimental setup
capable of studying all phenomena at once.
In the present work, a mini-reactor experimental setup
was developed to study the catalyst fragmentation and
polymerization yield. Monomer consumption evaluation at
precision of 0.01 bar in gas-phase propylene polymerization
was enabled. The rate of reaction was calculated from early
stages of polymerization using instance monomer consump-
tion approach. The yield of polymerization and catalyst frag-
mentation were studied. Catalyst fragmentation mechanism
and progress were investigated by molten particle micro-
scopic imaging. The results showed that uniform fragmen-
tation of catalyst particles to smaller sub-particles on the
bulk of polymer/catalyst particles played a crucial role in
the yield of reaction.
Experimental
Materials
The Ziegler–Natta type catalyst (industrial grade) supported
on MgCl2 with 2.4wt% of Ti was utilized. Polymerization
grade propylene (99.99%), triethyl-aluminium(TEAL),
cyclohexyl methyl dimethoxysilane (CMMS) and dicyclo-
penthyl dimethoxysilane were provided by Marun Petro-
chemical Co., Iranas external electron donors. Additionally,
distilled n-hexane (polymerization grade, Pars Cylinder Co.,
Iran) dried overnight in the presence of sodium wire and
molecular sieves, were used as the solvent.
Polymerization set‑up
Gas-phase reactions were carried out in a homemade mini-
reactor shown in Fig.1. The reactor was a stainless steel
horizontal cylinder with a volume of 2.8cm3. The reactor
was equipped with two solenoid valves controlled by a logic
controller equipped with homemade software for the feed.
The reactor pressure was monitored with a pressure transmit-
ter. Propylene was purified with three commercial purifiers
and preheated passing through the heating unit, as shown in
Fig.1. The heating unit was a shell and tube heat exchanger
connected to heating bath, with a 4 m length tube to ensure
that the gas temperature reaches the desired temperature.
A heating bath was used to maintain the temperature of the
Iran Polym J
1 3
reaction medium at the desired value. A K-type thermocou-
ple was used to monitor the inlet gas temperature.
The reactor was charged by a pro-activated catalyst in a
dry box to avoid its contamination and then fixed to the gas
line. Afterwards, the reactor was plunged into a heat jacket
(connected to heating bath) to ensure constant reaction tem-
perature during the reaction. Pressure and temperature of the
monomer feed were controlled, temporally. Polymerizations
at different temperatures and pressures with commercial Zie-
gler–Natta catalyst were enabled in this mini-reactor. Mon-
omer consumption evaluation at high precision (0.01bar)
from the beginning of polymerization was implemented. The
rate of reaction was calculated during polymerization using
instance monomer consumption.
Polymerization procedure
Before polymerizations, the catalyst was activated with
triethyl-aluminum (TEAL) and an external electron donor
and dried. At first, solution of TEAL in hexane was allowed
to react with electron donorfor about 5min and then the
mixture was added to the catalyst (Al/Ti molar ratio was 88).
After about 10min, the catalyst was washed with hexane and
dried. After drying, a small amount of catalyst was weighed
into the reactor and it was closed. All these steps were car-
ried out in the dry box. To be able to control the reaction
temperature and pressure, it was necessary to use catalysts
loadings of less than 10mg in the reactor. After weighing,
the reactor was brought out of the glove box and connected
to the propylene line in mini-reactor setup. Polymerizations
with different conditions were carried out using the same
activated batch of the catalyst. Polymerization time ranged
from 1s to 2h, at the end of the pre-set polymerization time,
the reactor was simultaneously vented and fed with CO2.
Polymerization temperature was ranged from 40 to 70°C
and the monomer pressure from 2 to 10bar. Activities of
the catalysts samples were calculated by weighing the final
particles after the polymerization.
Characterization
SEM
Morphology of the produced polymer particles was observed
with scanning electron microscopy (SEM). Particle mor-
phology micrographs were observed using a VEGA II
microscope (Tescan, Czech Republic) operating at acceler-
ating voltages of 10 or 15kV (depending on each individual
case). As the catalyst we used (and especially its morphol-
ogy) is highly sensitive to moisture, it is essential to carry
out SEM observations under inert conditions.
Cross-sections of the catalyst/polymer particles were
studied with high-resolution scanning electron microscope
as well. For preparation of the cross-sections, the samples
were mixed with very light epoxy resin in dry box. The
resin/sample mixtures were cured with the aim of light. The
cross-sections were cut with a microtome (Leica RM2165)
and used for SEM observations as soon as possible to pre-
serve the paticlesstructure. Before SEM measurements the
samples were coated with gold (2nm coating). The coatings
were used to prevent charging of the samples. In the case of
short polymerization times, catalyst/polymer samples were
sealed in an inert vial in the glove box, and just before SEM
observation, they were fixed on a tape and quickly moved
to the chamber of the SEM to prevent their contact with air
and moisture.
Melt microscopy method
Melt microscopy of polymer particles was first employed to
study the fragmentation of catalyst particles during olefin
polymerization by Abboud [22]. In our experiments we used
melt microscopy method to investigate catalyst fragments in
one particle. One particle was placed on a glass plate and
fixed by a metal ring on the microscope platform. Then, the
sample heated to 175°C (above the polypropylene melting
point) and temperature is held constant for 10min. Pictures
were taken from the molten staged sample. The resulting
polymer melt is transparent and allows observation of solid
fragments inside its phase.
Particle size measurement
To measure the size of particles, a thin layer of particles
was spread on a microscopeslide. Image of about 200
particles was taken with the aim of acamera. Diameter
Fig. 1 Polymerization set-up with the mini-reactor: C3 monomer, N2
inert gas, PG pressure gauge, PH gas pre-heat, PR pressure regulator,
R reactor, J jacket, PT pressure transmitter, V storage vessel
Iran Polym J
1 3
of each particle was measured individually, then, mean
particle size was calculated for each sample.
Results anddiscussion
The objectives of this work were to investigate the rate
and yield of polymerization, catalyst fragmentation and
morphology of the catalyst particles at relatively low reac-
tion rates. Most of polymerization experiments were car-
ried out at constant pressure (semi-batch), it means that,
the reactor pressure was kept constant with a tolerance
of 0.2bar. For instance, if the reactor pressure was set
to 6bar, solenoid valve acted after the pressure falls to
5.8bar and fed propylene to the reactor to recover the set
point at 6.0bar. The monomer was fed with the precision
of±0.01bar. Temperature and pressure data acquisition
were performed in the course of reaction. Table1 shows
reaction conditions used in the experiments.
Overall polymer yield was measured by weighing the
catalyst used and final polymer samples. Also for each
monomer injection pulse, time interval and pressure
change were measured and catalyst productivity was cal-
culated. Therefore, a graph of the rate of polymerization
versus time was plotted.
Table 1 Reaction conditions
used in the experiments
* C donor: cyclohexyl methyl dimethoxysilane (CMMS)
** D donor: dicyclopenthyldimethoxysilane
Run Cat (mg) Trec (°C) N2 (bar) P (bar) Ext. donor t (min) Yield (mg
pol/mg cat)
D (mean)
(µm)
E3 6.6 70 1 8 (batch) C*25s 1.5
H5 5 70 1 8 (batch) C 42 10.2
H6 5.1 70 1 6 C*60 24.5
E1 6.9 70 1 8 C 3 5.1 97
G4 2.6 70 1 8 C 24 27
G7 2.2 70 1 8 C 25 36.1 141
G3 2 70 1 8 C 73 84.15 187
F2 4.6 70 1 10 C 60 17.9
NC6-2 5.2 70 1 8 D** 3 11.6 131
NC6-1 4 70 1 8 D 7 21.7 175
NC6-4 4.3 70 1 6 D 3 10.3 162
NC6-3 3.4 70 1 6 D 7 20.5 –
NC4-4 3.3 40 1 8 D 12 30 130
NC4-5 3 40 1 4 D 12 2.8 110
NC5-1 5.5 70 1 2 D 12 4.1 119
NC5-2 4.2 70 1 6 D 12 19.8 181
NC5-3 4.5 70 1 4 D 12 17 200
NC5-4 3.1 70 1 8 D 12 26 215
Fig. 2 Rate profile for experiments carried out at 70°C and different
pressure of: a 8bar and b 6bar
Iran Polym J
1 3
Reproducibility
To ensure that the results were reproducible, experiments
were repeated twice, at least. Figure2 shows the result
obtained under the same polymerization conditions at dif-
ferent reaction times (runs NC6-1 and NC6-2; NC6-3 and
NC6-4 in Table1). As can be seen in Fig.2, rate profiles
match well with each other. Acceptable results were taken
at different polymerization pressures.
Polymerizations atdifferent pressures
A number of experiments were carried out at different mon-
omer pressures (runs NC5-1, NC5-2, NC5-3 and NC5-4 in
Table1). Figure3 shows rate profiles obtained from polym-
erizations performed at 8, 6, 4 and 2bar. It is clear in Fig.3
that the rate of polymerization was strongly affected by
monomer pressure. The higher the monomer pressure, the
higher was the rate of polymerization. Also it can be seen
from Fig.3 that, the rate of polymerization was a decay
type, which is in agreement with the results of other research
works [32, 36, 37]. This can be the result of the intrinsic
instability of the active centres [38–40] and mass transfer
limitations [41–43]. Note that in these experiments, Al/Ti
ratio was 80 (not at the optimum value) thus, productivity
of the experiments shown in Fig.3 was lower than those in
Fig.2.
Catalyst morphology
SEM micrographs of the initial catalyst are presented in
Fig.4 which show that the initial catalyst particles had
fairly rough surface and were consisted of uniform spheri-
cal particles. The catalyst particles had cracks as large as
1μm. These cracks were probably caused by sampling of
the commercial catalyst and the treatment (activation and
drying) we imposed on it to obtain dry activated catalyst.
Mean catalyst diameter (40μm) was calculated from SEM
micrographs.
Effect ofpolymerization time onparticle morphology
To study the effect of polymerization time, polymeriza-
tion reactions were performed at different polymerization
time (runs E3, E1, G7 and G3 in Table1). The morpho-
logical inspection of the surface and bulk of the growing
polymer particles are shown in Fig.5. For samples shown
in Fig.5a–c, polymerization was carried out for 25s and
only one pulse of the monomer was injected into the reac-
tor. The polymer obtained from this polymerization was
3.5mg (less than the weight of the catalyst used: 6.6mg
represented in Table1). Comparing the surface of the
Fig. 3 Polymerization rate profiles obtained from experiments per-
formed at 70°C and different pressure and time
Fig. 4 SEM micrographs of the initial catalyst: a single particle, b its subparticles and c surface cracks
Iran Polym J
1 3
Iran Polym J
1 3
particle after 25s of polymerization (Fig.5a, b) with that
of the catalyst (Fig.4) shows that the surface of the par-
ticle became smoother at initial stages of polymerization,
which means that the produced polymer had encapsulated
the micro grains of the catalyst and caused its surface to
be smooth. This is in agreement with the observations of
Di Martino etal. [28, 29].
The external primarycracks of the catalyst particles were
not filled with the produced polymer. In the cross section
micrograph of the particle (Fig.5c) some non-reacted cata-
lyst moieties can be seen in the centre of the particle. For
polymer sample shown in Fig.5d–f polymerization was car-
ried out for 3min up to the yield of 5.1mg/mg catalyst.
Comparison of the particle in Fig.5a with the one shown in
Fig.5d, e, reveals that the surface of the latter is rougher and
a lot of pores were formed, caused the particle porosity to be
raised. Cross section micrograph of this particle confirmed
the existence of polymer and large pores and deep cracks
in the whole particle (Fig.5f). This type of fragmentation
(layer-by-layer) is in agreement with the studies of Tanike
etal. [25] who conducted slurry propylene polymerization
using a spherical Mg(OEt)2-based Ziegler–Natta catalyst
with stopped-flow technique. Their results showed that poly-
mer species initially filled macrospores mainly located in the
middle layer of the catalysts, then caused the fragmentation
of the middle layer, and finally provoked the fragmentation
of the particle core in a stepwise manner [25]. Further stud-
ies on the mechanism of catalyst fragmentation were carried
out by molten polymer microscopy observations.
Referring to Fig.5g–l (particles obtained at 25 and
73min of polymerization) the surface of particles were
smooth although a lot of holes and cracks were maintained
in the particle. This can be due to the fact that polymer spe-
cies tend to surround macro pores of the catalyst, instead
of cracks.
Catalyst fragmentation mechanism bymelt microscopy
Since the discovery of Ziegler–Natta catalysts for olefin
polymerization, many research groups investigated cata-
lyst fragmentation and particle growth. Two catalyst frag-
mentation patterns were reported as ‘‘layer-by-layer’’ and
“rapid fragmentation”. The ‘‘layer-by-layer’’ fragmenta-
tion means the catalyst particle breakage starts from the
external surface to the centre of the particle until the
whole catalyst particle is fragmented [6, 7, 25]. Pater and
Weickert proposed a different particle growth mechanism.
They showed a rapid fragmentation of the catalyst into a
large number of small sub-particles at the beginning of
the polymerization and the fragments initially were well
distributed throughout the particle [44].
To study the catalyst fragmentation, a number of polym-
erization experiments were carried out at different condi-
tions. Molten polymer microscopy studies were performed
to follow the evolution of catalyst fragmentation during
polymerization. Beside it, the catalyst fragmentation and
reaction yield were compared and studied.
Figure6 shows molten polymer microscopy of two dif-
ferent particles obtained at 25s and 42min of polymeri-
zation with productivity at 1.5 and 10.2mg pol/mg cat,
respectively (runs E3 and H5 in Table1). Polymerization
in these two experiments were carried out batch wise,
means only one pulse of monomer was injected to the
reactor. Pressure drop was due to monomer consumption.
Figure6 shows, as polymerization proceeds (10.2mgpol/
mg cat in comparison with 1.5mgpol/mg cat), the size
of catalyst fragments became smaller as shown in Fig.6b
(about 10μm) confirming development of fragmentation
of the catalyst carrier as polymerization yield increases.
Figure6a clearly shows that the fragmentation was initi-
ated from outer parts of the particle confirming a layer-by-
layer fragmentation at the beginning.
Fig. 5 SEM micrographs of a single particle obtained from polym-
erization at 8bar pressure and 70°C: a, b surface and c its cross sec-
tion at t=25s and yield=1.5mg/mg cat; d, e surface and f its cross
section at t=3 min and yield=5.1 mg/mg cat; g, h surface and i its
cross section at t=25min and yield=37mg/mgcat; j, k surface and
l its cross section at t=73min and yield=85.4mg/mgcat
◂
Fig. 6 Optical microscope images of melted polypropylene par-
ticles at 175 °C with catalyst fragments (solid dots) inside them: a
yield=1.5mg/mg cat and b yield=10.2mg/mg cat, polymerization
performed at 70°C and initiated from 8bar pressure (100× magnifi-
cation)
Iran Polym J
1 3
Surface and cross section SEM micrographs of these par-
ticles are shown in Fig.7. Figure7a shows that the surface of
the particle is smooth. Figure7b shows formation of fibres
between catalyst segments. With progress of the reaction
(Fig.7d, e) fibres of polymer are replaced by the bulk of
polymer, caused the catalyst carrier to be fragmented. In the
cross section image of the particle obtained at 25s of polym-
erization (Fig.7c) pieces of unreacted catalyst exist near the
centre of the particle. In Fig.7f, cross section micrograph
of the particle obtained at 42min of polymerization shows
polymer was formed in the internal part of the particle. Fig-
ure7f shows large pores in the inner parts of the particle.
This may be due to mass transfer limitations into the inner
parts of the particle, which caused higher level of polymeri-
zation at the outer parts of the particle. The outer parts grow
more than inner parts and therefore, the cracks and empty
volume inside the particle were developed.
Yields andcatalyst fragmentation
Evolutions of fragmentation of different particles obtained
from different polymerization yields were investigated.
Figure8 shows polymerization kinetic profile and micro-
graphs of molten particles (runs G4, G7 and G3 in Table1).
In Fig.8a, a secondary peak was appeared at 25min of
polymerization. Micrographs of molten particle are shown
in Fig.8b–d. Figure8b belongs to the sample obtained at
21min of polymerization and Fig.8c shows the sample at
25min of that. By comparing these two figures it is clear
that the number of catalyst fragments are higher. This is an
evidence of catalyst fragmentation at about 25min after the
beginning of the reaction. It can be concluded from Fig.8
that the number of fragments increased as reaction pro-
ceeded and the size of fragments was decreased. In Fig.8b,
appearance of fragments with the size more than 10μm
confirmed incomplete catalyst fragmentation. Gradually,
with the progress of reaction, the fragments were broken
and smaller fragments were appeared.
Fig. 7 SEM micrographs of a single particle obtained from polymerization initiated from8bar pressure and 70°C: a, b surface and c cross sec-
tion at t=25s; d, e surface and f cross section at t=42min
Iran Polym J
1 3
With respect to the melt microscopy images and com-
paring them before and after the peak time, it seems that
the fragmentation is the main reason for enhanced rate and
appearance of the peak. Other works, beside catalyst frag-
mentation, stated overheating and over reduction of Ti+3 to
Ti+2 [40] and formation of cracks [45] as possible reason
for appearance of the peak. SEM micrographs in Fig.6 at
the same experimental conditions do not confirm overheat-
ing. As Fig.6 shows there is no particle agglomeration or
polymer melting inside the particle. Based on the experi-
mental conditions, the peak can be occurred by overheating,
whichis explained in the next section.
Experiments indifferent polymerization pressure
Furthermore, a number of experiments were carried out at
mild conditions at 40°C to study the progress of catalyst
fragmentation. Figure9 illustrates catalyst fragmentation for
particles obtained from experiments carried out at 40°C
and different pressures (runs NC4-4 and NC4-5 in Table1).
Figure9 shows at low reaction yields (2.8mgpol/mg cat)
fragmentation is incomplete and a broad range of fragments
size from 3 to 10μm was existed in the particle (Fig.9a, b).
When polymerization was carried out at 8-bar pressure, the
reaction yield reached to 30mg pol/mg cat and fragment
sizes reduced to less than 2μm (Fig.9c, d).
Also fragmentation of two particles obtained from polym-
erizations at 70°C and different pressures (runs NC5-1 and
NC5-4in Table1) was investigated in Fig.10. Figure10
shows that at low polymerization yields (4.1mg pol/mg cat)
fragmentation was incomplete and fragments with different
sizes up to 10μm with a broad size range can be seen inside
the particle. While, in the particles in Fig.10c, d fragmen-
tation was more uniform and size of fragments reduced to
2μm.
Our results confirmed the observations of Noristi etal.
who performed slurry polymerization of propylene with sup-
ported Ziegler–Natta catalyst at different polymer yields.
They suggested that if the site distribution in the catalyst
was uniform and the polymerization conditions were mild,
the polymer growth started uniformly throughout the cata-
lyst particles, which then undergoes an even and progres-
sive fragmentation into very fine particles homogeneously
dispersed in the polymer matrix [46]. Figures9 and 10 also
show a progressive development of fragmentation inside the
catalyst particle. This shows that the fragmentation is not
limited to the initial stage of the polymerization as stated by
Pater etal. [44], previously.
To confirm our observations, fragmentation of two
particles obtained from different polymerization pres-
suresare compared in Fig.11 (runs H6 and F2 in Table1).
In Fig.11,by increasing the pressure from 6 to 10 bar a
non-uniform fragmentation of the catalyst can be observed.
Table1 showsat higher pressure of 10bar, the yield was
Fig. 8 a Rate profile of polymerization initiated at 8 bar pressure
and 70 °C and optical microscope images of melted particles after:
b 21min and yield=27mg/mg cat; c 25min and yield=36mg/mg
cat; d 73min and yield=84mg/mg cat
Iran Polym J
1 3
decreased in comparison with 6bar pressure. This shows the
vital role of catalyst fragmentation on polymerization yield.
The reason for reduced yield at higher pressure can
be explained by monomer diffusion limitation inside the
particle, causing low reaction progress. At 10bar pressure,
the reaction rate was increased by increasing monomer con-
centration. Polymerization first happens at the surface of
the catalyst particle. The high level of the bulk of polymer
Fig. 9 Optical microscope images of melted polypropylene particles at 175°C with catalyst fragments (solid dots) inside them: polymerization
at 40°C and pressure of a, b 4bar and c, d 8bar after 12min (100× magnification)
Fig. 10 Optical microscope images of melted polypropylene particles at 175°C with catalyst fragments (solid dots) inside them: polymerization
at 70°C and pressure of a, b 2bar and c, d 8bar after 12min (100× magnification)
Iran Polym J
1 3
covers the surface of the catalyst particle, leads to reducing
the monomer transport into the particle. This causes lower
yield and it is the reason of observing the larger fragments of
the catalyst particle in the melt microscopy images. In fact,
the larger fragments are the reason of lower reaction yield.
It is possible that overheating happens if any peak is
observed at this condition. To confirm it, rate profile data
and SEM micrographs are shown in Fig.12 (run F2 in
Table1). As can be seen in Fig.12 a, b, agglomeration of
particles is the evidence of particle overheating and poly-
mer local melting. The polymer layer reduced the particle
heat transfer and overheating can occur that at first increased
the reaction rate, and then by deactivation of the catalyst, it
would be decreased. If Fig.12c is comparedwith Fig.8a,
it can be seen that, at the first moments of the reaction the
primary peak was not appeared, but there is one peak after
about 18min. The SEM micrographs of particles relevent
toFig.8a,have been presented in Fig.5. Theyshow no evi-
dence of agglomeration or polymer melt inside the catalyst
particle. But SEM micrographs of Fig.12 show particles
agglomeration, confirming particle overheating.
Fig. 11 Optical microscope images of melted polypropylene parti-
cles at 175°C with catalyst fragments (solid dots) inside them: polym-
erization at 70°C and pressure of a 6bar andb 10bar after 60 min
(100× magnification)
Fig. 12 a, b SEM micrographs of polypropylene particles and c rate
profile of polymerization at 10bar pressure and 70°C
▸
Iran Polym J
1 3
Conclusion
In this work, gas-phase polymerization of propylene was
performed in a homemademinireactor. The study focuses
on the reaction yield and catalyst fragmentation. Melt
microscopy method was implementedto study the cata-
lyst fragmentation. It was observed that the fragmentation
was started by a layer-by-layer mechanism. The size of the
catalyst fragments inside theparticle showed continues
development of catalyst fragmentation inside the polymer/
catalyst particles and it was not limited to an initial rupture
of the catalyst particle. Comparison of the reaction yield
with melt microscopy showed the crucial role of catalyst
fragmentation in polymerization yield. More fragmenta-
tion and smaller fragment pieces resulted in higher reac-
tion yield.
Acknowledgements The authors would like to express their special
thanks to Mr. F. Hormozi for his assistance in set-up construction and
Mrs. F. Khosravi and Mrs. L.Tolami for microscopic imaging. Fur-
thermore, the authors wish to thank Marun Petrochemical Co. (MPC),
Iran, for supplying catalyst.
References
1. Albizzati E, Cecchin G, Chadwick JC, Collina G, Giannini U,
Morini G, Noristi L (2005) In: Pasquini N (ed) Polypropylene
handbook, 2nd edn. Hanser, Munich
2. Nouri M, Parvazinia M, Arabi H (2015) Effect of fragment size
distribution on reaction rate and molecular weight distribution in
heterogeneous olefin polymerization. Iran Polym J 24:437–448
3. Cecchin G, Marchetti E, Baruzzi G (2001) On the mechanism of
polypropene growth over MgCl2/TiCl4 catalyst systems. Macro-
mol Chem Phys 202:1987–1994
4. McKenna TF, Di Martino A, Weickert G, Soares JB (2010) Par-
ticle growth during the polymerisation of olefins on supported
catalysts, 1–nascent polymer structures. Macromol React Eng
4:40–64
5. Redzic E, Garoff T, Mardare CC, List M, Hesser G, Mayrhofer L
(2016) Heterogeneous Ziegler–Natta catalysts with various sizes
of MgCl2 crystallites: synthesis and characterization. Iran Polym
J 25:321–337
6. Zheng X, Loos J (2006) Morphology evolution in the early stages
of olefin polymerization. Macromolsymp 236:249–258
7. Zheng X, Smit M, Chadwick JC, Loos J (2005) Fragmenta-
tion behaviour of silica-supported metallocene/MAO catalyst
in the early stages of olefin polymerization. Macromolecules
38:4673–4678
8. Salakhov II, Batyrshin AZ, Sergeev SA, Bukatov GD, Barabanov
AA, Mats’ko MA, Sakhabutdinov AG, Zakharov VA (2016) Effect
of titanium–magnesium catalyst morphology on the properties of
polypropylene upon propylene polymerization in a liquid mono-
mer. Catal Ind 8:213–216
9. Taniike T, Funako T, Terano M (2014) Multilateral characteriza-
tion for industrial Ziegler–Natta catalysts toward elucidation of
structure–performance relationship. J Catal 311:33–40
10. Buls VW, Higgins TL (1970) A particle growth theory for het-
erogeneous Ziegler polymerization. J Polym Sci Polym Chem
8:1037–1053
11. Kosek J, Grof Z, Stepanek F, Marek M (2001) Dynamics of parti-
cle growth and overheating in gas-phase polymerization reactors.
Chem Eng Sci 56:3951–3977
12. Fisch AG, Dos Santos JHZ, Secchi AR, Cardozo NSM (2015)
Heterogeneous catalysts for olefin polymerization: mathemati-
cal model for catalyst particle fragmentation. Ind Eng Chem Res
54:11997–12010
13. Najafi M, Parvazinia M, Ghoreishy H (2014) Modelling the cata-
lyst fragmentation pattern in relation to molecular properties and
particle overheating in olefin polymerization. POJ 1:77–91
14. Némethné-Sóvágó J, Bneke M (2014) Microreactors: a new con-
cept for chemical synthesis and technological feasibility. Mater
Sci Eng 39:89–101
15. Zöllner K, Reichert KH (2002) Video microscopy for the exami-
nation of the heterogeneous gas-phase polymerization. Chem Eng
Tech 25:707–710
16. Knoke S, Ferrari D, Tesche B, Fink G (2003) Microkinetic vid-
eomicroscopic analysis of olefin polymerization with a supported
metallocene catalyst. Angew Chem Int Ed 42:5090–5093
17. Ferrari D, Fink G (2005) Video microscopy for the investiga-
tion of gas phase copolymerisation. Macromol Mater Eng
290:1125–1136
18. Oleshko VP, Crozier PA, Cantrell RD, Westwood AD (2001)
In-situ and ex situ microscopic study of gas phase propylene
polymerization over a high activity TiCl4–MgCl2 heterogeneous
Ziegler–Natta catalyst. Macromol Rapid Comm 22:34–40
19. Pater JTM, Weickert G, Van Swaaij WPM (2003) Optical and
infrared imaging of growing polyolefin particles. AIChE J
49:450–464
20. Abboud M, KallioK Reichert KH (2004) Video microscopy for
fast screening of polymerization catalysts. Chem Eng Technol
27:694–698
21. Abboud M, Denifl P, Reichert KH (2005) Study of the morphol-
ogy and kinetics of novel Ziegler–Natta catalysts for propylene
polymerization. J Appl Polym Sci 98:2191–2200
22. Abboud M, Denifl P, Reichert KH (2005) Fragmentation of Zie-
gler–Natta catalyst particles during propylene polymerization.
Macromol Mater Eng 290:558–564
23. Mayrhofer L, Paulik C (2014) Growth kinetics obtained from
single particle gas-phase ethene homopolymerization with a Zie-
gler–Natta catalyst. Macromol React Eng 8:194–200
24. Mori H, Yoshitome M, Terano M (1997) Investigation of a fine-
grain MgCl2-supported Ziegler catalyst by stopped-flow propene
polymerization: model for the formation of active sites induced by
catalyst fragmentation during polymerization. Macromol Chem
Phys 198:3207–3214
25. Taniike T, Thang VQ, Binh NT, Hiraoka Y, Uozumi T, Terano M
(2011) Initial particle morphology development in Ziegler–Natta
propylene polymerization tracked with stopped-flow technique.
Macromol Chem Phys 212:723–729
26. Di Martino A, Broyer JP, Spitz R, Weickert G, McKenna TF
(2005) A rapid quenched-flow device for the characterisation of
the nascent polymerisation of ethylene under industrial conditions.
Macromol Rapid Comm 26:215–220
27. Di Martino A, Broyer JP, Schweich D, de Bellefon C, Weickert
G, McKenna TFL (2007) Design and implementation of a novel
quench flow reactor for the study of nascent olefin polymerisation.
Macromol React Eng 1:284–294
28. Di Martino A, Weicker G, McKenna TFL (2007) Contributions
to the experimental investigation of the nascent polymerisa-
tion of ethylene on supported catalysts, 1. Macromol React Eng
1:165–184
Iran Polym J
1 3
29. Di Martino A, Weickert G, McKenna TFL (2007) Contributions
to the experimental investigation of the nascent polymerisa-
tion of ethylene on supported catalysts, 2. Macromol React Eng
1:229–242
30. Silva FM, Broyer JP, Novat C, Lima EL, Pinto JC, McKenna TFL
(2005) Investigation of catalyst fragmentation in gas-phase ole-
fin polymerisation: a novel short stop reactor. Macromol Rapid
Comm 26:1846–1853
31. Olalla B, Broyer JP, McKenna TFL (2008) Heat transfer and nas-
cent polymerisation of olefins on supported catalysts. Macromol
Symp 271:1–7
32. Machado F, Lima EL, Pinto JC, McKenna TFL (2009) Evalua-
tion of the initial stages of gas-phase ethylene polymerizations
with a sio2 supported Ziegler–Natta catalyst. Macromol React
Eng 3:47–57
33. Browning B, Sheibat-Othman N, Pitault I, McKenna TFL (2014)
A 2-D observer to estimate the reaction rate in a stopped flow
fixed bed reactor for gas phase olefin polymerization. AIChE J
60:3511–3523
34. Martins AR, Cancelas AJ, McKenna TFL (2016) A study of the
gas phase polymerization of propylene: the impact of catalyst
treatment, injection conditions and the presence of alkanes on
polymerization and polymer properties. Macromol React Eng.
doi:10.1002/mren.201600011
35. McKenna TFL, Tioni E, Ranieri MM, Alizadeh A, Boisson
C, Monteil V (2013) Catalytic olefin polymerisation at short
times: studies using specially adapted reactors. Can J Chem Eng
91:669–686
36. Shen X-R, Fu Z-S, Hu J, Wang Q, Fan Z-Q (2013) Mechanism of
propylene polymerization with MgCl2-supported Ziegler–Natta
catalysts based on counting of active centers: the role of external
electron donor. J Physic Chem C 117:15174–15182
37. Severn JR, Chadwick JC (2013) Immobilisation of homogeneous
olefin polymerisation catalysts. Factors influencing activity and
stability. Dalton Trans 42:8979–8987
38. Shimizu F, Pater J, Weickert G (2001) Three-site mechanism and
molecular weight: time dependency in liquid propylene batch
polymerization using a MgCl2-supported Ziegler–Natta catalyst.
J App Polym Sci 81:1035–1047
39. Matsuoka H, Liu B, Nakatani H, NishiyamaI Terano M (2002)
Active sites deterioration of MgCl2-supported catalyst induced
by the electron donor extraction by alkylaluminium. Polym Int
51:781–784
40. Han-Adebekun GC, Hamba M, Ray WH (1997) Kinetic study
of gas phase olefin polymerization with a TiCl4/MgCl2 catalyst
I. Effect of polymerization conditions. J Polym Sci Polym Chem
35:2063–2074
41. Vakili M, Arabi H, Salehi Mobarakeh H (2012) The effect of
SiO2 porosity on activity profiles and comonomer incorporation
in slurry ethylene/butene-1 polymerization by (SiO2/MgCl2/
TEOS/TiCl4) catalyst system. J App Polym Sci 124:5145–5153
42. Zheng X, Pimplapure MS, Weickert G, Loos J (2006) Influence
of porosity on the fragmentation of Ziegler–Natta catalyst in the
early stages of propylene polymerization. e-Polymers 6:356–365
43. McDaniel MP (2011) Influence of catalyst porosity on ethylene
polymerization. ACS Catal 1:1394–1407
44. Pater JTM, Weickert G, Loos J, van Swaaij WPM (2001) High pre-
cision prepolymerization of propylene at extremely low reaction
rates—kinetics and morphology. Chem Eng Sci 56:4107–4120
45. Hassan Nejad M, Ferrari P, Pennini G, Cecchin G (2008) Ethylene
homo-and copolymerization over MgCl2–TiCl4 catalysts: polym-
erization kinetics and polymer particle morphology. J Appl Polym
Sci 108:3388–3402
46. Noristi L, Marchetti E, Baruzzi G, Sgarzi P (1994) Investigation
on the particle growth mechanism in propylene polymerization
with MgCl2-supported Ziegler–Natta catalysts. J Polym Sci Polym
Chem 32:3047–3059