Bayat 2012

Download as pdf or txt
Download as pdf or txt
You are on page 1of 5

Quantification of Polyethylene Wax in Semibatch Laboratory Reactor

and a Study of Different Parameters on Wax Production as a


By-Product in the Slurry Ethylene Polymerization
Mohammad Hossein Bayat, Majid Abdouss, Mehran Javanbakht
Department of Applied Chemistry, Amir Kabir University of Technology, Tehran, Iran 15914

Correspondence to: M. H. Bayat (E-mail: mhbayat@gmail.com)

ABSTRACT: In coordination polymerization of alpha-olefins utilizing Ziegler-Natta catalysts via slurry, polyolefin is produced with
other by-products such as polyethylene wax. Wax is defined as a linear short chain saturated by hydrocarbons in the relevant indus-
tries. Hexane and heptane are two of the most common suspending agents for the slurry process and wax molecules are soluble or
swollen in these kinds of liquids. After termination of polymerization, wax separates from the main product by physical processes. A
procedure for wax quantification in a laboratory reactor is suggested. Also, impacts of various factors such as reaction duration, pressure,
and cocatalyst on wax generation have been investigated. VC 2012 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 127: 1027–1031, 2013

KEYWORDS: polyethylene wax; ethylene; Ziegler-Natta catalysts; coordination polymerization; slurry phase; semibatch reactor

Received 19 June 2011; accepted 16 April 2012; published online 9 May 2012
DOI: 10.1002/app.37916

INTRODUCTION The knowledge of wax content in the mother liquor is an essen-


tial criterion for the system to control in industry. Mother liq-
The application of Ziegler-Natta catalysts, based on transition
uor is used as a suspending agent which has been recovered
metals, for the polymerization of olefins has been developed
from polyethylene by a physical separation and contains wax,
over the last 50 years. Various aspects of these catalysts and
cocatalyst, and mainly a suspending agent. The suspending
their polymerization have been reported in several reports and
agent is returned to the production line by distillation and the
articles.1–6
wax remains.
Wax is low molecular weight tail in the molecular weight distri-
Many different factors are able to change the wax amount in
bution of polyolefin, produced as the by-product of ethylene
the slurry phase such as temperature, pressure, concentration of
polymerization by Ziegler-Natta catalysts in the slurry phase,
molecular weight controlling agents (hydrogen), cocatalyst, type
and because of the separation costs and decrease in main prod-
of suspending agent, existence and concentration of electron do-
uct yield it is not desirable. Wax generation during polyethylene
nor compounds, presence of water, impurities and … .
production using slurry phase is not avoidable though no wax
generation happens when different production methods are The kinetic scheme of coordinative polymerization of a-olefins
applied (for example in gas phase/fluidized bed reactor). Waxes describes other reactions known as chain transfer reactions
have molecular weight in the range of 200–1000 g/mol but the other than the main reaction.10,11 Four types of chain transfer
desired polymer is in the range of 10,000–6,000,000 g/mol.7,8 reaction include:
Upon the initiation of polymerization, oligomer, wax, and long 1. Transfer to ethylene by transferring hydrogen from the
chain polymers generate. Polyethylene is insoluble in the sus- b-carbon of the propagating chain to the coordinated
pending agent and can easily be separated by filtration. Oligom- monomer:
ers separate from the suspending agent simply because of their
low molecular weight and exit out along with the unreacted
monomer when the reactor depressurizes. Together with the Cl Ti  ðCH2  CH2 Þn  R þ CH2 ¼ CH2 ! Cl 
increase in chain length, the melting point increases. Ti  CH2 CH3 þ CH2 ¼ CH  ðCH2  CH2 Þn1  R

C 2012 Wiley Periodicals, Inc.


V

WWW.MATERIALSVIEWS.COM WILEYONLINELIBRARY.COM/APP J. APPL. POLYM. SCI. 2013, DOI: 10.1002/APP.37916 1027


ARTICLE

2. Transferring a hydrogen atom from the b-carbon of the (>99.9%), TEAL (Triethylaluminum) solution 1.0M in hexanes
propagating chain to the transition metal of the active site (Sigma Aldrich) and TIBAL (Triisobutylaluminum) solution
(b-hydrogen elimination): 1.0M in hexanes (Sigma Aldrich)].
Catalyst
Cl Ti  ðCH2  CH2 Þn  R ! Cl Ti  H A Mg(OEt)2-Supported TiCl4 catalyst was used for this study.
þCH2 ¼ CH  ðCH2  CH2 Þn1  R The catalyst used contains 2.7% Ti (w/w). The procedure of
preparing this type of catalyst has been described in the litera-
ture.12–15 In order to increase the dosage accuracy, the prepared
3. Transfer to organoaluminum compound (co-catalyst): catalyst was diluted 10 times with dried hexane.

Cl Ti  ðCH2  CH2 Þn  R þ AlR03 ! Cl Ti  R0 EXPERIMENTAL


þR02 Al  ðCH2  CH2 Þn  R Polymerization
The polymerization was carried out in a 5 L steel jacket Büchi
autoclave reactor fitted with a thermocouple equipped with a
4. Transfer by molecular hydrogen: mechanical seal stirrer (variable speed, anchor form) in the
slurry phase.
Cl Ti  ðCH2  CH2 Þn  R þ H2 ! Cl Ti  H
After running out of all moisture and air by nitrogen, 2700 mL
þCH3  CH2  ðCH2  CH2 Þn  1  R of hexane and 2.0 mL of TEAL (1.0M in hexane) were added
and the mixer was turned on to annihilate any probable con-
As it can be seen the scheme also anticipates the generation of tamination. After 5 min, 1.0 mL of the diluted catalyst (contain-
lower molecular weight components such as waxes. ing 0.042 mmolTi) was added by using dried syringe. The reac-
With this in mind, a common suspending agent in the slurry tor was warmed up to 80 C and ethylene was supplied
process for the production of polyethylene is often Hexane or continuously at 6 bars for 1 h. The polymerization was termi-
Heptane. Wax molecules are liquid or solid (against oligomers) nated by injecting ethanol while the mixer was working after
and the density is about 0.74–0.85 g/cm3 that is very near to switch off the ethylene supply and depressurizing the reactor to
polyethylene (0.88–0.94 g/cm3); however, it can be removed by atmosphere pressure. It has been observed in some literatures
physical operation. Wax is soluble (low molecular weights) or that hydrochloric acid was applied to stop the reaction. This
partially soluble (higher molecular weights) in the suspending procedure is not suitable because hydrochloric acid may solve
agent. Higher molecular weight waxes swell in the suspending some titanium in the final product and wash it out so it would
agent (especially at elevated temperatures); therefore, the density interfere in Ti residua determination in the final product.
decreases to near the suspending agent and this makes it feasible
Wax Determination Procedure
to separate physically from the polymer along with the mother
After termination of the polymerization, the reactor cooled
liquor. The generated wax can be measured after drying by
down to below 40 C and untreated and generated gases were
gravimetery and this is the exact procedure that is employed in
slowly released by the upper valve and then this valve closed.
industry and our base for estimation in a laboratory reactor.
Hexane was gradually transferred out by the force of nitrogen
Since polyethylene is the main product of Ziegler-Natta catalyst to a preweighted seal vessel by means of a stainless steel pipe
systems, polyethylene wax is barely applied and mentioned in connected to the lower valve (Lower valve should not open too
literatures. In this article, a method for the determination of the much to avoid outflow of any polyethylene particles along with
total wax is described and then the effect of some of the most hexane). After that the lower valve closed and 2700 mL of hex-
important variants (those which are common to consider in ane was ejected and then after closing the valves, the mixer was
Ziegler-Natta type systems and have the most effects on effi- turned on for 5 min. Hexane was added to the rest and the
ciency and the way that reaction continues) on the amount of washing stage was repeated to ensure that the entire wax is
wax is studied. Although there are several different definitions extracted from the reactor. The vessel was reweighed to get the
for wax and in some cases low molecular weight polyethylene is net weight of the gathered mother liquor.
regarded as wax, the by-product of ethylene coordinative poly-
Considering that the mother liquor contains cocatalyst (wax
merization applying the slurry method is considered in this
carries a high concentration of aluminum, unlike polyethylene)
article.
and it partially converts to aluminum oxide (nonvolatile) during
the drying stage, and also evaporates along with other volatiles,
MATERIALS
it should be totally converted to get a repeatable and reliable
Ethylene and Nitrogen (purity > 99.99%) were passed through result. TEAL reacts with water and releases hydrocarbons and
columns packed with moisture and oxygen scavenger, respec- short chain alcohols and ultrafine aluminum oxide. After trans-
tively. Hexane (Merck) was dehydrated by refluxing through a ferring 400 mL of mother liquor and 50 mL of hydrochloric
distillation column using sodium as a dehydrating agent under acid (1N) to a screw cap Erlenmeyer, the mixture is exposed to
dry nitrogen. Then hexane was dried over a molecular sieve of vigorous shaking for 10 min. The suspension was settled down
0.4 and 0.5 nm (same portion) for 24 h to reach less than 5 for a night to ensure complete separation of the phases and
ppm. Other reagents were used without purification [Hydrogen deposition of aluminum oxide. The upper layer is the modified

1028 J. APPL. POLYM. SCI. 2013, DOI: 10.1002/APP.37916 WILEYONLINELIBRARY.COM/APP


ARTICLE

Table I. Standard Reaction Condition (Exp No1)

Duration Cocatalyst H2O ppm in


Exp no (min) (mmol) Hexane
1 60 2 (TEAL) 5
Total Ti Residue in Al Residue in Al/Ti
P (Bar) PE (ppm) PE (ppm)
6 3 10 48

The reference state which applied for this research and just the targeted
item is changed and traced.

mother liquor. For a double determination, two 250-mL conical


flasks with ground stopper were preweighted (m1). To avoid any
evaporation during weighting, a 100-mL glass syringe was used.
The syringe was filled by the modified mother liquor and was Figure 1. The effect of polymerization duration on yield and PE/Wax ratio.
discharged two times. The syringe was filled by the modified
mother liquor and then weighted (ms1) and the mother liquor
was introduced to a conical flask. The syringe was reweighed
½g ¼ K  Mav
(ms2) so the difference is the exact weight of the sample (m0).
where K and a are 6.77  104 and 0.67, respectively.16
m0 ¼ ms2  ms1
The amount of titanium in the prepared catalyst and other
The hexane phase was separated off by evaporating and the resi- minerals (Al and Ti) in the final polymer were determined by
due was dried in a vacuum oven in two steps (First stage for inductively coupled plasma atomic emission spectroscopy (ICP-
about 35 min at 500 mBar at 60 C, the second stage for 20 min AES). Microwave digesting was applied to be decomposed.17
more at 20 mbar at 80 C). After cooling to room temperature The number of double bond per chain was calculated by FTIR
the remainder was determined by weighing the flasks (m2) and spectroscopy.18 Moisture content in applied hexane was quanti-
then the wax percentage was calculated according to the equa- fied by the Karl-Fischer method.19
tion below:
RESULTS AND DISCUSSION
Wax content % ¼ ððm2  m1 Þ=m0 Þ  100 The reference condition for polymerization (standard state) has
been described before. Other experiments were done in the
And finally the total wax in the batch laboratory reactor: same conditions and only the targeted item has been changed
[see Table I (experiment No.1)].
Total wax ðgÞ ¼ ðWax content %
Total gathered mother liquor ðgÞÞ=100 The Effect of Polymerization Duration
As shown in Table II and Figure 1, the polymerization yield
increases and then reaches a constant level but the wax to poly-
CHARACTERIZATION
ethylene ratio decreases by increasing the reaction time. This
The total obtained polymer was fully gathered and then washed indicates that most of the wax was produced in the early stages
with ethanol and dried in vacuum at 50 C to a constant weight. of the polymerization, also with regard to the total wax that
The viscosity average molecular weight (Mv) of the polymer was reached a steady level at the end of the reaction. Another inter-
obtained using Ubbelohde suspended level dilution viscometer. esting fact deduced from Table II is that the double bonds per
The temperature of the viscometry was maintained at 133 6 each polymer chain increase when reaction time increases. This
1 C by boiling chlorobenzene in the vapor jacket of the viscom- attests that near the end of the polymerization (when less
eter. The Mv was calculated from the intrinsic viscosity using amount of wax is generated) transfer to monomer and b-hydro-
the Mark-Howink Sakurada equation. gen elimination are the main cause of the chain transfer. These

Table II. Polymerization Duration and Polymer Specifications (Exp No2-6)

Exp no Minutes Yield (g) MW Wax (g) Wax/PE *10000 [C¼


¼C]
2 5 239 450,000 0.42 17.6 0.18
3 15 305 560,000 0.45 14.8 0.21
4 30 402 790,000 0.52 12.9 0.24
1 60 545 960,000 0.62 11.4 0.50
5 90 546 950,000 0.60 11.0 0.55
6 120 550 970,000 0.63 11.5 0.63

WWW.MATERIALSVIEWS.COM WILEYONLINELIBRARY.COM/APP J. APPL. POLYM. SCI. 2013, DOI: 10.1002/APP.37916 1029


ARTICLE

Table III. Ethylene Pressure Effect on Wax Content and Polymer Table V. Cocatalyst Concentration and Type (Effect on Wax Content and
Specifications Polymer Specifications)

Total Exp Cocatalyst Yield Wax Wax/PE


Exp Pressure Yield Wax Wax/PE no (mmol) (g) MW (g) *10,000 [C¼
¼C]
no (Bar) (g) MW (g) *10,000 [C¼
¼C]
1 2 (TEAL) 545 960,000 0.62 11.4 0.50
7 2 84 480,000 0.76 90.5 0.29 13 4 (TEAL) 511 800,000 0.94 18.4 0.37
8 4 291 890,000 0.62 21.3 0.44 14 6 (TEAL) 453 720,000 1.40 30.9 0.25
1 6 545 960,000 0.62 11.4 0.50 15 2 (TIBAL) 714 1,180,000 0.57 8.0 0.41
9 8 577 1,000,000 0.60 10.4 0.52

this means when hydrogen is applied, the influence of other


series of experiments (Exp1-6) also imply that there is more transferring agents are negligible. Experiments 10, 11, and 12
probability to form wax when shorter polymers are being gener- demonstrate that the wax generation intensifies with the pres-
ated. Another interesting fact that may be inferred form these ence of any molecular weight reducing agent.
data is that the used catalyst contains several types of active
centers (this is why its PDI is higher than 2) therefore the cen- The Effect of Cocatalyst
ters that form relatively low molecular weight PE fractions TEAL is one of the most common cocatalysts in these kinds of
(which contain the largest wax fraction) are unstable and may polymerization systems. Also, it is reported in many papers that
die after 10–15 min. The centers that produce the high molecu- the utilization of other types of cocatalyst such as TIBAL are
lar weight polymer are more stable (although they also die after able to change the polymerization path. According to data from
about 60 mininutes). This is why the wax fraction decreases Table V, the wax content enhances to the higher level when the
with time. cocatalyst concentration (TEAL) increases but the double bonds
decrease. Thus, in the absence of hydrogen, the organometallic
The Effect of Ethylene Pressure plays the main role in chain transferring. Regarding the reduc-
As it can be seen from Table III, there is a predictable ratio tion of molecular weight, the polymer yield does not reduce as
between the ethylene pressure and the final yield. Seeing that expected and this means that there is a broadening in molecular
the polymerization is supposed to be carried out in the slurry weight distribution (higher value for PDI). Therefore, it could
phase, based on Henry’s law, ethylene concentration increases in be perceived that any chemical agent that is able to broaden the
hexane by increasing the pressure. If the polymerization reaction MWD causes more wax generation. In experiment No. 15,
is running at increasing ethylene pressure, the average molecular another kind of cocatalyst was used. TIBAL had different effects
weight of the polymer increases. Consequently, the fraction of on the polymerization (Table V). Despite it being added in the
the low molecular weight component in the molecular weight same molar portion to the reaction medium, a different ratio of
distribution (wax) strongly decreases. In lower pressure, the the wax to polymer is gained and clarifies the important role of
lessened polymer is yielded with the lower molecular weight. In the organometallic compound on the wax generation.
this case, more wax and a higher ratio of the wax to polymer is
obtained. The Effect of External Electron Donor
External donors have been found to play an important role in
The Effect of Hydrogen the coordination polymerization of a-olefins.20,21 It is well
In these series of experiments, hydrogen was applied as a molecu- known that external donors affect catalyst activity, MW and
lar weight modifying agent loaded at 25 C before ethylene to the MWD. The Ziegler-Natta type catalysts are sensitive to the com-
defined pressure. After closing the hydrogen valve, the reactor pounds which have oxygen and nitrogen in their molecular
was warmed up to 80 C, and then ethylene was supplied continu- structure.22,23 THF (Tetrahydrofuran) was used as an external
ously to the reactor to keep the total pressure constant at 6 bars. electron donor. By dosing 2 mmol THF, some changes occurred
to the reaction route. The expected MW increased in spite of
As it is shown in Table IV, wax increases significantly when
the decrease in the yield of polymerization and meanwhile, the
hydrogen is used as the chain-transfer agent. Hydrogen also
wax content reduced. According to this experiment, an external
reduces the double bonds number in each polymeric chain and

Table VI. External Donor Effect (THF - Exp 16) and Impurities Effect
Table IV. Hydrogen Pressure Effect on Wax Content and Polymer
(H2O - Exp 17) on Wax Content and Polymer Specificationsa
Specifications

Hydrogen Exp Yield Wax Wax/PE


Exp PRESSURE Yield Wax Wax/PE no THF/H2O (g) MW (g) *10,000 [C¼
¼C]
no (Bar) (g) MW (g) *10,000 [C¼
¼C]
16 50 mmol 422 1,050,000 0.35 8.3 0.48
1 0 545 960,000 0.62 11.4 0.50 THF
10 1 207 230,000 1.24 59.9 0.27 17 45 ppm 403 590,000 1.05 26.1 0.48
11 2 127 140,000 1.86 146.5 0.16 H2Oa
12 4 43 90,000 2.23 518.6 0.08 a
When Fresh Hexane Applied Without Any Pretreatment.

1030 J. APPL. POLYM. SCI. 2013, DOI: 10.1002/APP.37916 WILEYONLINELIBRARY.COM/APP


ARTICLE

donor can also significantly affect the wax quantity (see Table REFERENCES
VI). As the external donor changes, the mechanism of the 1. Natta, G.; Pino, P.; Corradini, P.; Corradini, F.; Danusso, F.;
monomer insertion to the active sites of the transition metal Mantica, E.; Mazzanti, G.; Moraglio, G. J. Am. Chem. Soc.
complex even the kinetic of the polymerization changes, there- 1955, 77, 1708.
fore the wax content varies.
2. Bohm, L. L. Polymer 1978, 19, 553.
3. Kim, I.; Choi, H. K.; Kim, J. H.; Woo, S. I. J. Appl. Polym.
The Effect of Impurities Sci. 1994, 52, 1739.
Regarding the high sensitivity of Ziegler-Natta catalysts to mois- 4. Marques, M. M. V.; Nunes, C. P.; Tait, P. J. T.; Dias, A. R. J.
ture, hexane and other components are dehydrated before use. Polym. Sci. A Polym. Chem. 1993, 31, 209.
In this experiment (Exp. No 17), fresh hexane is used without 5. Abedi, S.; Sharifi-Sanjani, N. J. Appl. Polym. Sci. 2000, 78,
any pretreatment or dehydration. Other parameters are retained 2533.
similar to Experiment No. 1 (the standard reaction condition).
6. Alt, F.; Schoneborn, P.; Bohm, L, inventors; Basell Polyole-
The initial water content of the fresh hexane was determined
fine GmbH, assignee. U.S. Pat.7,008,898 (2006).
(45 ppm). Experimental data shows that less yield and more
wax has been obtained (Table VI). 7. Finlayson, M. F.; Garrison, C. C.; Guerra, R. E.; Guest, M.
J.; Kolthammer, B. W. S.; Parikh, D. R.; Ueligger, S. M,
inventors; The Dow Chemical Company, assignee. U.S.
CONCLUSIONS Pat.6,054,544 (2000).
8. Troughton, M. J. Handbook of Plastics Joining (A Practical
In the coordination polymerization of ethylene in the slurry
Guide), 2nd ed.; Norwich, New York: William Andrew, 2008.
phase, wax is an inevitable by-product. This is the integral part
of any olefin polymerization and could be defined as the low 9. Lyons, W. C.; Plisga, G. J. Standard Handbook of Petroleum
and Natural Gas Engineering, 2nd ed.; Burlington, Massa-
molecular weight fraction of the combined polymer material.
chusetts: Elsevier, 2005.
The wax fraction depends on two principal parameters, the av-
erage molecular weight of the polymer and the presence of low 10. Natta, G.; Pasquon, I. Adv. Catal. 1959, 11, 1.
molecular weight components in the molecular weight distribu- 11. Bohm, L. L. Polymer 1978, 19, 545.
tion. The factors that affect any of these two properties affect 12. Virendra, K. G.; Marayil, R. Macromol. Chem. Phys. 1996,
the relative content of the wax. It is feasible that the quantifica- 197, 1937.
tion of wax in a laboratory reactor by gravimetery be carried 13. Franke, R.; inventor; Hoechst Aktiengesellschaft, assignee.
out in order to study different variants in conditions that the U.S. Pat.4,859,749 (1989).
total wax is separated and purified. In the semibatch reactor,
14. Bohm, L. L.; Herrmann, H. F.; Berthold, J.; Hohner, G.;
wax is generated in the early stage of reaction but continuous
Lecht, R.; Vetter, H. J., inventor; Clariant GmbH, assignee.
reactors generate steadily. Several various chemical or physical
EP Pat0,607,773 (1994).
factors are able to change the wax quantity although in view of
the fact that there is a big difference between the molecular 15. Dashti, A.; Ramazani, S. A. A.; Hiraoka, Y.; Kim, S. Y.;
Taniike, T.; Terano, M. Polym. Int. 2009, 285, 52.
weight of wax and the main product, it could be concluded that
wax is the result of chain transfer reactions to other reaction 16. Brandrup, J.; Immergut, E. H.; Grulke, E. A.; Abe, A.; Bloch,
components such as cocatalyst and particularly hydrogen. Also, D. R. Polymer Handbook, 4th ed.; New York: Wiley, 2005.
as result of deactivation of the catalyst (catalyst poisons and any 17. Sakurai, H.; Noro, J.; Kawase, A.; Fujinami, M.; Oguma, K.
impurities) the wax content increases and these kinds of reac- Anal. Sci. 2006, 22, 225.
tions have the most likely portion and were observed to occur 18. Goldenberg, A. L.; Lubetskii, S. G. Vysokomol. Soedin. 1963,
at the beginning of the reaction, especially. 5, 905.
On the basis of the results, since physical or chemical parame- 19. Standard Test Method for Water Using Volumetric Karl Fi-
ters are able to vary wax quantity, the lower and constant wax scher Titration; ASTM E 0203–08; ASTM International:
content as a by-product in continuous industrial production is West Conshohocken, PA, 2008.
an indication of desirable purity in raw materials. Also, the 20. Sacchi, M. C.; Forlini, F.; Tritto, I.; Locatelli, P.; Morini, G.;
nonexistence of catalyst poisons enhances minimum fluctua- Noristi, L.; Albizzati, E. Macromolecules 1996, 29, 3341.
tions in physical variants such as temperature, pressure, and et 21. Sacchi, M. C.; Forlini, F.; Tritto, I.; Mendichi, R.; Zannoni,
cetera. G.; Noristi, L. Macromolecules 1992, 25, 5914.
The authors express their gratitude to the department of 22. Boor, J. Ziegler-Natta Catalysts and Polymerizations; Aca-
Applied Chemistry of Amir Kabir University of Technology and demic Press: New York, 1979.
Arak Petrochemical Research Centre for their help and assis- 23. Moore, E. P. Polypropylene Handbook; Munich: Hanser,
tance in carrying out this project. 1996.

WWW.MATERIALSVIEWS.COM WILEYONLINELIBRARY.COM/APP J. APPL. POLYM. SCI. 2013, DOI: 10.1002/APP.37916 1031

You might also like