Lloyd L. Handbook of Industrial Catalysts
Подождите немного. Документ загружается.
308 Chapter 7
31. K. Gordon, Progress in the Hydrogenation of Coal and Tar, Institute of Me-
chanical Engineers, Dec. 9, 1946.
32. R. Holroyd, Report on Investigations of Fuels and Lubricant Teams: US
Bureau of Mines Information Circulars, 7370/7375 (1946).
33. K. K. Kearby, Ind. Eng. Chem. 42 (1950) 295; K. K. Kearby, Emmett, US
Patent 2370797-8 (1945).
34. H. E. Drennan, US Patent 2371809 (1945); F. T. Eggertsen and H. H. Voge
(Shell) US Patent 2414585 (1947).
35. F. T. Eggertsen and E. P. Davies, US Patent 2461147 (1949).
36. E. C. Britton, A. J. Dietzler, and C. R. Noddings, Ind. Eng. Chem. 43 (1951)
2871.
37. Massoth and Scarpiello, J. Catal. 21 (1971) 294;. Rennard and Kehl, J.
Catal. 21 (1971) 282.
38. Phillips Petroleum Star Process, Hydrocarbon Processing, June (1976) 133.
39. Linde Process, Chem. Eng. News, March, 1992; Hydrocarbon Processing,
July (2000).
40. Styrene, its polymers, copolymers and derivatives, Ed. by R. H. Boundy,
ACS Monograph 115, Reinhold, New York (1952).
41. I. G. Farben, US Patents 1986241 (1935) and 2110833 (1938).
42. Dow, US Patents 1985844 (1934) and 2036410 (1936).
43. H. W. Ashton and T. W. Flavell, BIOS Report 3057.
44. K. K. Kearby, Catalytic dehydrogenation, in Catalysis, Vol. 3, Ed. by P. H.
Emmett, Reinhold, New York, 1955, p. 481.
45. K. K. Kearby, Catalytic dehydrogenation, in Catalysis, Vol. 3, Ed. by P. H.
Emmett, Reinhold, New York, 1955, p. 480.
46. E. W. Pitzer, US Patent 2866790.
47. Fleming and W. R. Gutmann, US Patent 2990432.
48. W. R. Gutmann, US Patent 3361683.
49. F. J. O’Harra, US Patent 3904552.
50. Sud. Chemie, US Patent 4467046.
51. D. L. Williams, Styrene Catalysts, AICE Meeting, New Orleans, March 6–
10, 1988.
52. (a) Sud Chemie Literature; (b) P. R. Courty and C. Marcilly, A Scientific
Approach to the Preparation of Bulk Mixed Oxide Catalysts, in Preparation
of Catalysts III, Eds. by G. Poncelet, P. Grange and P. A. Jacobs, p. 485 (c)
P. R. Courty and J. F. LePage, Relationship between average pore diameter
and selectivity in iron-chromium-potassium dehydrogenation catalysts in
Preparation of Catalysts II, Ed. by G. Poncelet, P. Grange and P. A. Jacobs,
Elsevier Science Publishers, Amsterdam, 1978, p 293.
53. R. W. Moncrieff, Heywood and Co. Ltd., London, 1963; Chem. Eng. News,
(March 21, 1994) 11; (May 2, 1994) 10.
54. W. H. Carothers, US Patents 2130523, 2130947-8 and 2163636; British
Patents 461236 and 461367 (1937).
Petrochemical Catalysts 309
55. M. McCoy, Chem. Eng. News, (Aug 28, 2000) 10; (Oct 2, 2000) 33.
56. Chem. Eng. News, (April 26, 1971) 30.
57. C. A. Tolman, W. C. Seidel, J. D. Druliner, and P. J. Domaille, J. Organo-
metallic Chem. 3 (1984) 33.
58. E Billig, C B Strow and R L Pruett, Chem. Comm. (1968) 1307.
59. P. J. Hogan and J. R. Jennings, UK Patent 1,546,807 (to ICI plc).
60. R. J. Cozens and J. R. Jennings, Applied Catal. 19 (1995) 297.
61. P. Schlack, I. G. Farben Industrie, German Patent Chem. Eng. News (Oct 2,
2000) 32; US Patent 2241321 (1941); Smith, in The Production of Polymer
and Plastics Intermediates from Petroleum, Ed. by Long, Butterworths,
London (1967).
62. German Patent 748253.
63. O. Wallach, Synthetic Fiber Development in Germany, CIOS Report XXX,
HMSO, London, 1945, p. 13.
64. Wallach, Liebigs Ann. 312 (1900) 171.
65. Hydrocarbon Processing 49 (1970)137; British Patents 1257609 (1971) and
1332211 (1973); US Patents 4092360 (1978) and 4203923 (1980).
66. M. A. Mantegazza, G. Paparatto, G. Petrini, G. Fornasari, and F. Trifiro
(Enichem), Ammoxidation Reaction in Gas and Liquid Phase, p. 353, Ca-
talysis of Organic Reactions, edited by M. G. Scaros and M. L. Prunier,
Dekker, New York (1995).
67. W. Muench, Chimica Ind. Milano 44 (1962) 636; SNIA Viscosa Italian Pa-
tents 604795; Belgian Patents 582793, 603606.
68. Walter, J. Pract. Chem. 45 (1892) 107.
69. P. Woog, Comp. Rend., 145 (1907) 124.
70. P. Woog, French Patent 379715 (1907).
71. Weiss and Downs, US Patent 1321959 (1919).
72. Craver, British Patent 189091 (1920).
73. M. McCoy, Chem. Eng. News, (Oct 2, 2000) 32.
74. R. W. Moncrieff, Man Made Fibers, Wiley, New York, 1971; J. R. Whin-
field and J. T. Dickson, British Patent 578,079 (1941); US Patent 2,465,319
(Assigned to DuPont)
75. P. Short, Chem. Eng. News, (May 15, 2000) 25.
76. C. Masters, Homogeneous Transition Metal Catalysts, Chapman & Hall,
London (1981).
77. A. Tullo, DuPont NG-3 Process, Chem. Eng. News, (March 6, 2000) 25.
78. P. M. Morse, New Polyester, Chem. Eng. News, (Nov 10, 1997) 8.
79. O. Roelen, US Patent 2415102 (1947).
80. Combined Intelligence Sub-Committee Report (CIOS), G 2 RAPO 1.
81. G. U. Ferguson, in The Production of Polymer and Plastics Intermediates
from Petroleum, Ed. R. Long, Butterworth, London (1967), p. 86.
310 Chapter 7
82. R. Wyman, in Selected developments in catalysts, Ed. J. R. Jennings, Criti-
cal Reports on Applied Chemistry, Vol. 12, Blackwell, Oxford (1985), p.
128.
83. L. H. Slaugh and R. D. Mullineaux, Organomet. Chem. 13 (1968) 469; E.
R. Tucci, Ind. Eng. Chem. Prod. Res. Dev. 9 (1970) 516.
84. R. Fowler, H. Connor, and R. A. Baehl, Hydrocarbon Processing, (Sept.
1976) 247; Chem. Eng. News (March 9, 1998).
85. Chem. Eng. News (Oct 10, 1994).
86. K. Weissermel and H-J. Arpe, Industrial Organic Chemistry, VCH, Wein-
heim, 1993.
87. E. Wiebus and B. Cornils, Hydrocarbon Processing, (March 1996) 63.
88. N. von Kutepow,W. Himmele, and H. Hohenschitz, Chem. Ingr. Tech. 37
(1965) 383; Hydrocarbon Processing 45 (1966) 141.
89. F. E. Paulic and J. F. Roth, J. Amer. Chem. Soc. Chem. Comm. (1968) 1578.
90. D. Forster, J. Amer. Chem. Soc. 98 (1976) 846.
91. Catiwa Process, Chem. Eng. News, (July 1, 1996) 8
92. F. C. Phillips, J. Amer. Chem. Soc. 16 (1894) 255.
93. J. Smit, W. Hafner, R. Jira, R. Sieber, J. Sedlmeier, and A. Sabel, Angew.
Chem. (Int Edn) 1 (1962) 80.
94. R. Banks, Discovery and Development of Olefin Disproportionation (Me-
tathesis), American Chemical Society (1983).
95. Hydrocarbon Processing, (March, 1998) 61.
96. Oil Gas J., (Sep. 20 1999) 62.
97. E. R. Freitas and C. R. Gum, Chem. Eng. Prog., (Jan 1979) 73.
8
OLEFIN POLYMERIZATION
CATALYSTS
Catalytic polymerization processes have become increasingly important as the
use of plastics has escalated throughout the world. Many common plastics were
discovered in the 1930s, and after the 1939–1945 war this created a demand for
petrochemical intermediates derived from the refining industry. Ethylene and
propylene are now common building blocks for plastics and by the year 2000,
polyethylene, polypropylene and their copolymers were the most widely used
plastic materials. The gradual development of polyolefin production is shown in
Table 8.1.
Even before Ziegler discovered his catalysts for the polymerization of ole-
fins in 1953,
1
it was well known that high-molecular-weight oils and insoluble
waxes could be formed by the treatment of olefins and synthesis gas with metal
alkyls and transition metal catalysts.
2
However, the first polyethylene process
was the result of research into high-pressure reactions such as those undertaken
by ICI in the 1920s and 1930s.
3
By March 1933, an experiment with high-purity
ethylene and benzaldehyde at 1900 bar pressure and 170°C had produced a sub-
stance like paraffin wax when the pressure vessel had apparently developed a
leak. After several more experiments, in which it was confirmed that polymer
had formed, and that the reaction could be explosive, work stopped until 1935,
at which time it appeared that there might be a commercial demand for the
product. When relatively impure ethylene was heated at 170°C, the pressure
could not be maintained at the target of 2000 bar and polyethylene, in the form
of a white powder, was formed. It had been recognized, of course, that benzal-
L. Lloyd, Handbook of Industrial Catalysts, Fundamental and Applied Catalysis,
DOI 10.1007/978-0-387-49962-8_8, © Springer Science+Business Media, LLC 2011
311
dehyde played no part in the reaction. By 1936, it was found that traces of oxygen
312 Chapter 8
of the oxygen is not to act as a catalyst in the normal sense of the word, but to
act as a source of free radicals. The free radical reacts with a molecule of eth-
ylene, which then undergoes chain growth by a free radical mechanism. The
oxygen is better referred to as an initiator. Details of both molecular weight and
operating conditions had been accidentally disclosed at a conference held in
1935, although experts, including Hermann Staudinger, refused to believe that
the polymerization of ethylene was possible. Fortunately, the minutes of the
meeting did not include details of the disclosure and the ICI patent application
was subsequently granted.
3
A full-scale plant with a capacity of about 100 tons
of polyethylene per year started operation in September 1939.
The ICI patent took precedence over work by Union Carbide in the United
States and by I. G. Farben in Germany. Experts from Germany had visited the
ICI polyethylene plant in 1937, and according to an official from BASF, then
began the work that led to the development of the Lupolen Process during the
war.
5
8.1. LOW-PRESSURE POLYETHYLENE
In 1953 a further step in the development of polyethylene was announced by
Karl Ziegler at the Max Plank Institute in Mulheim.
6
Ziegler began by studying
the reaction of lithium hydride and ethylene. This was not very promising, and
he turned to the reaction of aluminium hydride and ethylene, which gave trieth-
ylaluminium at temperatures in the range 60–80°C. Triethylaluminium is, of
course, spontaneously inflammable in air and reacts explosively with water. The
handling of these dangerous reactive compounds needed the development of a
whole range of new experimental and production techniques. Polymers were
formed by addition of further molecules of ethylene at temperatures around
100–120°C and these were eventually developed as low molecular weight α-
TABLE 8.1. Worldwide Production of Polyolefins.
Year Polyethylene
(million tones year
−1
)
Polypropylene
(million tones year
−1
)
Typical new plant
capacity (tones year
−1
)
1965 3 < 0.05 10,000
1970 6
1975 10 4
1980 15 6
1985 22 8 80,000
1990 30 13 100,000
1995 38 19
2000 50 30 300,000
were acting as a catalyst and process details were filed in a patent.
4
The role
Olefin Polymerization Catalysts 313
olefins. The reaction then continued. The reaction became known as the Aufbau
reaction and produced a linear polymer, in marked contrast to the highly
branched product from the high pressure processes.
The subsequent development of the process for high-density polyethylene
was the result of careful systematic observation by the experimental team. Fur-
ther work on the oligomerization of ethylene to the commercially-valuable C
6
and C
8
α-olefins led to the discovery that the displacement reaction was cata-
lyzed by traces of nickel in the reactor. The displacement reaction became faster
than the growth reaction as the concentration of nickel was increased, and at
120°C, only the dimer was produced. In subsequent investigations of other met-
als as co-catalysts in conjunction with triethylaluminium, it was shown that zir-
conium acetylacetonate inhibited the displacement reaction completely, and that
polyethylene was the only product. Further work led to the development of the
well-known titanium chloride/triethyl aluminum catalyst, which operated at low
temperatures and atmospheric pressure. The reaction became known as the Mul-
heim Atmospheric Polyethylene Process.
6
At about the same time that news of the Ziegler discovery was released, the
Phillips Petroleum Company in the United States announced that it had devel-
oped a medium-pressure, catalytic process (500 psig) to produce a high-density,
crystalline polyethylene.
7
The process was discovered when traces of ethylene in
a flue gas had polymerized over conventional cracking catalysts. The Phillips
catalyst contained chromic oxide supported on silica. The Standard Oil Compa-
ny of Indiana (later Amoco) also introduced a medium pressure process using a
catalyst comprising molybdenum oxide supported on carbon or alumina, but it
did not enjoy the success of the Ziegler or Phillips processes and was only oper-
ated in three full-scale plants.
8
8.1.1. Polyethylene Process Development
The Phillips process, used a different type of catalyst from Ziegler, was availa-
ble for license almost immediately and became a commercial success by produc-
ing a linear, highly crystalline product with higher density than the high-pressure
polymer discovered by ICI. The new polymer became known as high-density
polyethylene (HDPE), whereas the original ICI polyethylene was thereafter
known as low density polyethylene (LDPE). Phillips did not develop a catalyst
for the production of polypropylene.
Ziegler, working in the academic environment of the Max Planck Institut,
did not develop his own process for the polymerization of olefins, but did offer
the rights to his catalysts for license. Licensees were required to develop their
own versions of a polymerization process that was based on Ziegler’s discovery.
It is not surprising that several different versions of the Ziegler processes
evolved to produce HDPE, each based on somewhat different catalyst formula-
tions and operating procedures.
314 Chapter 8
8.1.2. The Development of Polypropylene Catalysts
Ziegler restricted his studies to the polyethylene catalyst, and the next major
discovery was made by one of his licensees. Professor Guilio Natta, who worked
with Montedison in Italy, investigated the polymerization of propylene and other
α-olefins to produce solid crystalline polymers. He found that polymerization
did not proceed when using the brown titanium trichloride used by Ziegler. He
showed that crystalline stereoregular polypropylene
9
could be produced when
using violet titanium trichloride.
8.2. ZIEGLER–NATTA CATALYSTS
The versatile range of low-pressure catalysts, discovered by Ziegler and modi-
fied by Natta in 1953 and 1954 was a tremendous advance in catalysis. It led to
the introduction of a wide range of commercially useful polyolefins, including
polypropylene and many copolymers as well as polyethylene. Since the initial
discoveries, there have been many developments in both the catalysts and the
commercial polymerization processes that have led to increased activity and
productivity.
8.2.1. Early Polyolefin Catalysts
The original Ziegler catalyst comprised a metal alkyl and a transition metal
compound that could be reduced and alkylated in the presence of the metal alkyl
co-catalyst to give active initiating and propagating centers. Since 1954, the
original catalyst has been developed continuously and the active centers modi-
fied with coordinating ligands. Most of the commercial catalysts are heteroge-
neous and the active component is supported on an inert material.
The most widely used transition metal salts have been the relatively cheap
titanium or vanadium halides, reduced with aluminum alkyls. In Ziegler’s work,
titanium tetrachloride was reduced to brown ß-titanium trichloride, which was
able to polymerize ethylene. However, when ß-titanium trichloride catalyst was
used in the polymerization of propylene, the product contained a high proportion
of the gum like atactic polymer, which was not viable for commercial use. In
contrast, Natta, in his work on the polymerization of propylene and other α-
olefins, showed that violet α-titanium trichloride could polymerize propylene to
the useful, crystalline, isotactic polymer. Nevertheless, a relatively large quanti-
ty of atactic polypropylene still had to be separated from the commercial isotac-
tic product.
Most of the early catalyst development was related to the need for a more
active catalyst to improve productivity. With low-activity catalysts, the polymer
must be de-ashed to remove solid catalyst residues so as to meet product speci-
Olefin Polymerization Catalysts 315
fications. To achieve a high productivity, the catalyst particles should fragment
during the process and be diluted with high-molecular-weight polymers. The
most significant difference between olefin polymerization and other catalytic
processes is that the catalyst remains in the product.
In crystalline Ziegler–Natta catalysts, the properties of the active transition
metal centers with their vacant coordination sites are different, depending on
their position on the fractured crystal surface. This leads to variations in both
activity and reducibility and consequently to the wide variations in polymer mo-
lecular weight observed in commercial products. The continuing development of
catalysts has therefore been related to the need for the formation of active more
regular sites, which can then produce polymer chains with a more uniform mo-
lecular weight and, in the case of polypropylene, higher stereospecificity. This
subsequently led to the use of a range of supported titanium catalysts for both
polyethylene and polypropylene production and, more recently, to the use of
single-site metallocene catalysts. These improvements have led to increases in
the range of possible products even further.
8.2.2. Ziegler’s Brown Titanium Trichloride
The early heterogeneous Ziegler catalysts were prepared in the reactor by the
reaction in solution from a mixture with a molar ratio of titanium tetrachloride to
triethyl aluminium of less than 2.5:1 to prevent excessive reduction of the titani-
um species. The organotitanium complexes that formed decomposed spontane-
ously at ambient temperature to a precipitate of the brown metastable ß-titanium
trichloride.
10
This is a very superficial description because the nature of the tita-
nium product depends on many factors, such as the relative concentration of the
components, the solvent used, the order of addition, and the temperature of the
reaction. The aluminum alkyl acts as an alkylating agent and the resulting alkyl
titanium species decomposes to a lower valence state. Triethyl aluminum is a
stronger alkylating agent than diethyl aluminum chloride and was preferred for
polyethylene catalyst production.
A controlled molar excess of the aluminum alkyl cocatalyst was added to
the reduced titanium (~2:1) before the catalyst entered the reaction zone. All the
components were carefully purified and handled in a dry, inert atmosphere. De-
spite these precautions, the catalyst had a low surface area and was not very ac-
tive. Most of the titanium was inaccessible due to the formation of relatively
large crystals and relatively few active sites were formed at the crystal edges.
Furthermore, ß-titanium trichloride was not stereospecific for the polymerization
of higher olefins.
Activity and stereospecificity depended, therefore, not only on the catalyst
preparation but also on the size and crystallinity of the particles. The careful
control of conditions and the optimum concentration of reactants were necessary
to form particles that fragmented during use to expose fresh active sites and to
316 Chapter 8
increase productivity. Higher productivity, that would allow the use of the pol-
ymer without de-ashing to remove catalyst particles, remained an objective.
8.2.3. Natta’s Violet Titanium Trichloride
The important contribution made by Natta was the modification of the Ziegler
catalyst for the polymerization of propylene. Pasquon, one of Natta’s colleagues,
reported that,
“he assumed stereospecificity to be associated with the regularity of the cat-
alyst surface and used the crystalline violet ß-titanium trichloride instead of
liquid titanium tetrachloride to prepare his catalysts.”
11
In this way, Natta was able to develop a stereospecific catalyst based on
α
-
titanium trichloride, produced by reducing titanium tetrachloride in hydrogen at
500°–800°C, before combining it with the triethyl aluminum co-catalyst.
The catalyst was further improved by producing a stereospecific form of α-
titanium trichloride. The metastable ß-titanium trichloride which also contained
co-crystallized aluminum trichloride was heated to temperatures of 300°–400°C.
The transformation proceeded via intermediate γ-titanium trichloride, which
could also produce stereoregular polymer at temperatures of 100°–200°C. The
α- and γ-crystalline modifications of titanium trichloride have layered hexagonal
and cubic crystal structures, respectively, and are both active catalysts.
Stauffer produced titanium trichloride commercially by reducing titanium
tetrachloride with aluminum metal in an aromatic solvent at about 250°C.
12
The
aluminum could be activated by ball milling, with the addition of a small
amount of aluminum trichloride. This procedure formed high-activity γ-titanium
trichloride containing co-crystallized aluminum trichloride. It was soon realized,
from experience during commercial production of polypropylene catalysts, that
both the α and γ forms of titanium trichloride were more than ten times as active
after the crystals had been ground or milled to produce smaller particles.
13
Mill-
ing converts α- and γ-titanium trichloride to the more active δ-titanium trichlo-
ride and increases the surface area from about 5 m
2
g
−1
to 20 m
2
g
−1
. The crystal
structure of a range of titanium trichloride catalysts is described in Table 8.2.
Early commercial catalysts were supplied to a strict specification. The aver-
age particle size was usually in the range 10–100 μm with an activity exceeding
1000 g polypropylene per gram of catalyst and an isotactic index exceeding
88%. Continuous, rather than batch, operation was possible with the new, ex-
ternally formed catalyst but the efficiency was still low and the polypropyl-
ene still contained unacceptably high levels of both catalyst residues and atactic
polymer.
Olefin Polymerization Catalysts 317
TABLE 8.2. First-Generation Ziegler–Natta Catalysts (1957–1960).
Modification Crystal form
a
%Polypropylene
stereoregularity
(AlEt
2
Cl co-catalyst)
b
Preparation
β
-TiCl
3
(brown)
Globular; fibrous chainlike
structure
40–50 In situ
α
-TiCl
3
(violet)
Platelets; hexagonal close
packed
80–92 External: H
2
reduction
of TiCl
4
at 300
o
–400
o
C
γ
-TiCl
3
(violet)
Platelets; cubic close
packed
80–92 External: H
2
reduction
of TiCl
4
at >150
o
C
TiCl
3
.
0.33AlCl
3
Hexagonal globules (10–
100 μm)
>88 External: Al reduction
δ
-TiCl
3
(violet)
Larger secondary particles;
random alternation of
α
-
and
γ
- packed particles
80–92 External: milling of the
α
-and
γ
-forms, from
1960 onwards
a
Charges in the close-packed chlorine layers of
α
-and
γ
-TiCl
3
balanced by hexagonal clusters of
smaller titanium atoms in spaces between chlorine layers. To maintain electroneutrality there are
rows of empty spaces in the titanium layer with lower charge density, which simplifies disintegra-
tion when milled. The small crystals agglomerate and are more active.
b
Catalyst more active with AlEt
3
but less stereospecific.
Although polypropylene was only produced on a small scale from 1957,
and despite the relatively greater importance of high-density polyethylene, work
on improving polypropylene catalysts led to a better overall understanding of
both catalysts and process operation.
8.2.4. Second-Generation Propylene Polymerization Catalysts
By 1960, the original preparation of brown ß-titanium trichloride in situ was no
longer used by most producers. It was replaced by commercial γ-titanium tri-
chloride, which contained an isomorphous form of aluminum chloride in the
molar ratio (3TiCl
3
.AlCl
3
). This was ball milled to form δ-titanium trichloride,
which increased activity and led to the formation of a higher proportion of iso-
tactic polymer. However, even after these improvements in catalyst perfor-
mance, there was still a limitation to the catalyst activity because the aluminum
trichloride in the catalyst reacted with the triethyl aluminum co-catalyst to form
quantities of ethyl aluminum dichloride, which is detrimental to the polymeriza-
tion reactions. Titanium trichloride crystals were also relatively large despite the
milling treatment to reduce particle size.
A further improvement in performance was achieved by the addition of
electron donors (Lewis bases), such as esters, ethers, or amines, to the catalyst,
which can form complexes with aluminum alkyls. Electron donors could be
added either with the co-catalyst, or during the γ-titanium trichloride milling
stage, depending on the compound used or the result required. In addition to