Ethylene dimerization and oligomerization to butene-1 and linear α-olefins

Ethylene dimerization and oligomerization to butene-1 and linear α-olefins

Cut&& Today,14 (1992) 1-124 Elsevier Science Publishers B.V., Amsterdam PART 1 1.1 DIMEBIZATION OFETHYLENE TO BUTENE-1 INTRODUCTION The dimeriz...

7MB Sizes 0 Downloads 0 Views

Cut&& Today,14 (1992) 1-124 Elsevier Science Publishers B.V., Amsterdam

PART 1

1.1

DIMEBIZATION

OFETHYLENE

TO BUTENE-1

INTRODUCTION

The dimerization of ethylene is an important route for the selective production of butene-1. The considerable interest in this field by academic institutions and industry is reflected by the large number of patents and publications covering this subject during the last few years. However, the stimulus in this direction was provided by the pioneering studies of Karl He explored the use of organoaluminum Ziegler in the early 1950’s. compounds in the selective dimerization of alkenes. Ever since, workers in this area have investigated many novel catalytic systems based on transition metal complexes. These complexes afford a new attractive mean of achieving high selectivity to butene-1 by ethylene dimerization.

1.2

INDUSTRY STATUS

Butene-1 can be produced by a variety of methods; including the following L21: 1. Refinery operations (isolation from fraction of C4 of crude oil and petroleum products), 2. Steam cracking of C4 hydrocarbons, 3. Butane dehydrogeration, 4.

Co-product

from a-olefin

manufacturing,

5.

Catalytic ethylene dimerization,

6.

Butyl alcohol dehydration,

7.

Pyrolysis of butylacetate

and

and butylchloride.

Only the first five methods are of direct interest to the industry. Over 60% of butene-1 is produced from refinery and steam cracking operations. The remainder is obtained as a-olefin co-product [3-141. The estimated production of butene-1

from a-olefins

is 120,000 metric ton/year

(1988).

By 1993, the

production is expected to reach 240,000 metric ton/year as co-product from aolefins [l]. Although the United States and the European countries are the only major sources of butene-1 in the current world market, countries such as Saudi Arabia (currently producing 50,000 metric ton/year), Argentina, Brazil, and China will start production of butene-1 in the coming years. Since these countries do not have enough domestic source of butene-1 (due to primarily cracking natural gas liquids for ethylene production), selective dimerization or ohgomerization of ethylene seems to be an economical method to produce butene-1 or other a-olefins [3,4]. Butene-1 is industrially generated at Exxon, Texas Petrochemicals USA, Du Pont Canada, Huels Germany, and Shell UK by separation from various

2

butylene streams. It is also produced as a co-product in the a-olefins processes operational at Chevron, Ethyl, Shell, and Mitsubishi [8-141. Saudi Arabia became a major butene-1 producer when Sabic’s affiliate Petrokemya operated its world-largest plant that employs the IFP-Alphabutol technology. This process utilizes promoted titanium catalytic system that selectively dimerizes ethylene to butene-1 and precludes its isomerization to butene-2 which otherwise necessitates costly products superfractionation 143. Moreover, this process is characterized by its low investment, widely available feedstock, and its polymerization grade butene-1. This makes the process economics very favorable, particularly when ethylene is obtainable at low marginal cost [4-61. Two more plants implementing the Alphabutol process have started production of butene-1 in Thailand and China in 1989 and 1990, respectively. The main use of butene-1 is in polyethylene as comonomer. Although butene-1 is less expensive than its main competitors in the linear low density polyethylene (LLDPE) comonomer market, hexene-1, and to a lesser extent, octene-1 continue to capture a greater share because of their enhanced performance properties including greater tear strength and stress crack resistance. The large number of patents in this field assigned to more than 35 companies and institutes world-wide, shows the importance of this process from the viewpoint of practical industrial applications. Companies such as IFP, ICI, Dow, Phillips, Shell, Mitsubishi and Mitsui and several research institutes in the world performed experimental work using Ziegler-type catalysts to dimerize ethylene to butenes. A series of homogeneous and heterogeneous catalytic systems have been used at extremely high dimerization rates and butene-1 selectivity. The major catalytic systems are those based on titanium or nickel catalytic systems. Table 1.1 presents a list of companies having patents in ethylene dimerization to butene. 1.3

REACTION CHEMISTRY AND MECHANLSM

The dimerization of ethylene is governed by three essential features that influence the formation of butene-1. These are: the degree of polymerization, the mode of linking ethylene molecules, and the position of the double bond. The degree of polymerization or the number of ethylene molecules forming the hydrocarbon chain depends on the frequency of chain transfer reactions with respect to the propagation reaction. However, the predominance of chain transfer reactions depends on the conditions of shifting hydrogen atoms in the hydrocarbon compounds. This dominance is frequently achieved by increasing the reaction temperature in order to lower the formation of high molecular weight polymers during the dimerization reaction. Another factor that limits the degree of polymerization is the use of high concentration of ethylene during the reaction (limiting conversion) in order to prevent any codimerizaThis codimerization is due to the tion of the formed butene-1 and ethylene. high affmity of butene-1 to the catalyst [15-171.

3

Table 1.1 Companies

having patents in ethylene dimerization

Titanium

Ref.

Nickel

Ref.

Allied Chemical Chemische Werk Chemopetrol Erdoel Chemie IFP Indian Petrochem. Mitsubishi Mitsui Phillips Snamprogetti Tokuyama Soda Toyo Soda Union Carbide Others

P16* P17 Plo-12 P15 Pl-5 P6 P25,P26 P13,P14 P7 I% P18 P19-23 P27 P29-42

BASF CNRS CommonwealthSC. ICI Intl. Synthetic Nest.e Oy Phillips Tokyo Inst. Toyo Soda Others -

P48 P49 P50 P46,P47 P51 P54 P56-60 P52,P53 P55 P62-70

-

to butene Other catalytic systems Dow (AI) du Pont (Ta) Goodyear (W) Grace (Pd) ICI (Ru) MIT (Ta) Mitsubishi (V) Mitsui (Mn, Fe) Mobil (Ru) Petro-tex (Pd) Phillips (Co, Cr) Shell (Pd, Ru) Sun Oil (Co) Toray (Co)

Ref. P99,Ploo P87,Pss P81 P75 P74 P89,P90 P93 P94 P73 P76 P86,P91, P92 P72, P77-80 P84,P85 P82

*A reference number preceded by the letter “P” refers to the patent reference presented at the end of the review.

Another feature that influences the formation of butene-1 is the mode of linking the ethylene molecules during dimerization, which depends on two fundamentally different steps, initiation and propagation. The contribution of the initiation step to the microstructure is essential to the formation of butene-1[171. The position of the double bond in the butenes product is reflected by the ability of the catalyst to mobilize the hydrogen atoms during the dimerization reaction. It is defined either in the initiation or transfer step or it results from a subsequent isomerization after the formation of butene-1 1171. Actually, the insertion step is often the limiting step in defining the position of the double bond C151. This is due to the fact that the active center of the catalyst is incapable of accommodating two ethylene molecules simultaneously. However, if the isomerization occurs, it can be kept at a minimum by increasing the concentration of ethylene in the reaction medium, consequently limiting its conversion [Xl. Due to the large number of catalytic systems used to dimerize ethylene, it is quite possible that each catalytic system (based on the metal) operates according to a different mechanism. It is even possible that the same metal reacts via different mechanisms depending on the reaction conditions and cocatalysts applied [El. However, two different and representative reaction mechanisms that are widely accepted for ethylene dimerization are the

4

hydride and cyclic intermediate mechanisms [X-18]. Simplified schemes for the two mechanisms are shown in Figures 1.1 and 1.2, respectively [15]. The hydride mechanism, which is generally the most accepted one, consists of three steps: insertion, propagation, and elimination. The first step is the insertion of ethylene into a metal-hydrogen bond. Further insertion of ethylene into the metal-alkyl bond leads to growth products (propagation step). The elimination of the formed products (butenes) occurs by P-hydrogen elimination that allows the catalyst (L&H) to re-enter the catalytic cycle. The types of reactions that occur during the hydride scheme are the following (ko and ke are rate constants for the progagation and elimination reactions, respectively):

Polymerization: Oligomerization: Dimerization:

A considerable number of insertion steps will occur and polyolefins are formed. Oligomers with geometric molecular weight distribution (Schulz-Flory type> will result. Only dimers are partially formed.

kP * ke kP - b k, a ke

However, which one of these reactions prevails, is dependent on the nature of the metal and its oxidation state, the nature of ligand, and the reaction conditions [l&18]. Transition metals of Groups IV and VIII favor P-hydrogen transfer (elimination) that is important in dimerization and oligomerization reactions. In contrast, Groups V and VI transition metals promote propagation reaction that makes them active polymerization catalysts. Upon using the same metal, it has been reported that the chain propagation decreases as the metal’s oxidation state increases [Xl. Moreover, the nature of the coordinated ligands has been found to exert a similar influence. While the donor ligands tend to promote chain propagation, acceptor ligands exhibit the opposite trend [15,171. The cyclic intermediate mechanism is based on the formation of metallacyclopetane intermediates that are converted to the olefin complex via 1, 3 hydrogen shift. Elimination of the product olefin and coordination of ethylene lead to the formation of the bis (ethylene) complex [2,15,191. The course of the reaction can be significantly influenced by the nature of the metal, the ligand supplied, and the reaction conditions. The reaction cycle is complete when the This reaction does not distinguish metallacyclopentane is regenerated. between the two possible modes of addition of a metal-carbon or a metalhydrogen bond to an unsymmetrically substituted olefin. The direction of this addition is very important for obtaining linear or branched products. Reaction mechanisms, based on the above two schemes, are presented under each catalytic system as discussed in Section 1.4. However, some reaction mechanisms are not fully established yet. 1.4

CATALYTIC

SYSTEMS

A large variety of catalytic systems has been used to dimerize ethylene to butene-1. Different compounds of nickel, palladium, titanium, rhodium,

5

L,M-C-C-R

L,M-R’

Figure1.1. Reaction mechanism by hydride intermediates [151.

Q R

M =b

H

H

R \

2R -CH

-Cy

Cy-C-CH-C&

Figure 1.2. Reaction mechanism by cyclic intermediates Cl51

6

zirconium, tantalum and cobalt together with (or without) various aluminum alkyls, containing or not containing halogen atoms, and supported (or not supported) on carriers, were used as catalysts. The homogeneous catalytic systems used in ethylene dimerization consist of active metal, Lewis acid or co-catalyst, additive, and solvent. Catalysts used in the presence of a reducing agent or Lewis acid are referred to as Ziegler-Natta catalysts 1171. However, the commonly used active catalysts are titanium and those of Group VIII transition elements, especially the first two series of the Group, Fe, Co, Ni and Ru, Rh, and Pd f173. These elements are generally active when bonded with less than two anions or reduced to a lower oxidation state by a variety of alkyl derivatives of aluminum and other reducing agents. The co-catalyst is usually a Lewis acid that enhances the dimerization rate and has a profound effect on the course of the reaction 116,171. It plays the double role as a reducing agent and as an acceptor for halide ions that are responsible for the dimerization reaction f173. Catalyst modifiers or additives such as phosphines are usually added to the catalyst system to provide better selectivity for the reaction [16,17]. The modifiers influence the mode of linkage of ethylene molecules and inhibit the formation of high molecular weight polymers. Solvents often play an important role in enhancing the dimerization rate and promoting the catalytic activity because of their reducing power [ZOI. ParaEnie, cyclopara~nic, olefinic, aromatic, and chlorinated hydrocarbons are used as solvents for the homogeneous catalytic systems [2,16,20,211. In general, the reported different approaches for the activation of the homogeneous catalytic systems ethylene dimerization can be summarized as follows r151: 1. Catalyst formation by reaction of transition metal complexes with org~ome~lic compounds, such as AlR3, 2. Complexation of transition metal complexes with a Lewis acid (X,AlR3.,, BF3, where X is a halide and n=l-31, 3. Oxidation addition of inorganic or organic acids to transition metal complexes forming hydride species, and 4. Use of alkyl or aryl transition metal complexes stabilized by ligands with electron donor or acceptor properties. Several papers on ethylene dimerization used heterogeneous catalytic systems with high activity [22,23], By anchoring the active catalysts complex to polymeric, zeolitic or ordinary supports, the high selectivity of homogeneous catalytic systems can be combined with the easy product separation of heterogeneous systems [151. There is a close relationship between homogeneous and heterogeneous catalysts systems used in ethylene dimerization; the support of the active site in the heterogeneous system plays the role of the ligand on a coordinated metal atom in the homogeneous system [17]. The changes in the nature of the support drastically alter the behavior of an active site in the same way that the ligands in the homogeneous catalysts modify the electronegativity and the polarizability of a metal atom, and its reactivity 1223. The catalytic systems (both homogeneous and heterogeneous) used in ethylene dimerization are classified and presented according to the type of

7

transition conditions,

metal used. They are discussed in terms of activity, selectivity to b&me-l, and reaction mechanisms.

reaction

1.4.1 TitaniumBasedCatalysts !WOR)4 AlRs Systems Titanium-based catalysts have been developed and reached the stage of The Alphabutol process of the Institut industrial commercialization. Francais du Petrole UFP) uses a homogeneous titanium based catalyst that selectively dimerizes ethylene to butene-1 14, Pl-P53. The catalytic system based on Ti(OR)4-AlR3, where R is an alkyl having 1 to 6 carbon atoms or an aryl, is the most selective for butene-1 formation among other available titanium catalytic systems [4,15-l&21,24-371. Ethylene dimerization studies on this type of catalyst have been conducted at the IFP, Indian Petrochemical Owing to the low isomerization Corporation, and other research centers. activity of titanium, butene-1 is formed exclusively along with small amounts The aluminum alkyl of the of cis/trans butenes, hexenes and polymer. catalytic system exhibits high Lewis acidity that reduces the electron density on the active titanium centers. The influence of reaction conditions on the conversion of ethylene, selectivity to butene-1 and formation of by-products (oligomer and polymer) was studied for the Ti(OR)4-AIR3 catalyst system 124-331. The parameters studied were: nature of alkyl aluminum, nature of alkoxy group around titanium, Al/Ti molar ratio, solvents, temperature, pressure, and catalyst modifiers (additives).

Alkyl Group The effect was studied solvent, and compounds. establish the to butene-1. order:

of the nature of alkyl group in titanium and organic aluminum in a glass reactor at 3O”C, atmospheric pressure in hexane 2 h reaction time 1241. Ti(OBu)4 was used with various AlR3 On the other hand, AlEt was used with various Ti(OR)4 to influence of the type of alkyl group on the selectivity of the system The activity of Ti(OBu)4-AlR3 systems decreased in the following

Al (C4Hg)3 > Al (C3H7)3 > MEt3 > Al (C4HgMI The activity of Ti(OR)4-AlEt

(1.1)

systems decreased also in the following order:

Ti (OBu)4 > Ti (OEt)4 > Ti (OC5H11)4 = Ti 0%iCsHll)4

> Ti (OCsH5)4

(1.2)

However, the selectivity for both of the systems decreased in the reverse order. Therefore, from the point of view of maximum catalyst selectivity, it is better to use a system of Ti(OCsH5)4-Al(C4Hg)2H, while from the point of view of maximum catalyst activity the best system is Ti(OC4Hg)4-Al(C4Hg)3. These results agree with studies performed by other investigators [161. The alkyl group attached to the aluminum remains intact during dimerization and the reaction proceeds exclusively through titanium alkyl complex 1251. The effect of the type of alkyl group in titanium and aluminum for the system Ti(OR)4-

8

AlR3 on butene-1 selectivity is presented in Table 1.2. The results indicate that systems consisting of AlR3, where R is C2H5 or CH3 radical, and Ti(ORY4, where R’ is a straight chained aliphatic radical containing 2-6 carbon atoms, have the highest catalytic activity and selectivity in dimerizing ethylene to butene- 1 I21. Table 1.2 Effect of the type of alkyl group in organic aluminum AlR3 and titanium alkoxide Ti(OR)4 catalytic systems in ethylene dimerization at 3O”G 1 atm, and 2 h in hexane C241 C&&&&&Jn

Pl-0~

. .

ht.%~

Conversion Methyl- EthylButene-1 pentene Butene-1 Polymer (%)

TilOR)4

AIR3

Ti(OBu)4 Ti(OBu14 Ti(OBu14 Ti(OBu1.4

AlEts AlPr3 AlBu3 Al(Bu)2H

46.6 53.5 63.1 81.9

66.0 56.3 59.5 95.5

-

Ti(OPr14 Ti(OBuf4

AlEts AlEt

37.6 42.4

62.5 79.8

Ti(O~5Hl1)4 Ti(O-iCgH& Ti(OPhI4

AlEt AlEt AlEt

36.7 36.4 17.2

63.0 62.7 99.7

10.7 13.1 13.0

22.2 28.4 26.2

2.1 2.2 1.3 4.5

5.0 6.0

11.6 13.4

0.9 0.8

5.7 5.5

10.3 10.8

1.0 1.0 0.3

-

This parameter is widely recognized to have a significant control on the selectivity of the ethylene dimerization to butene-1. It has been reported that when the Al/Ti molar is ~10, the dimerization reaction will be favorable. On the other hand, an AUTi molar ratio of >lO will lead to the domination of the pol~e~za~on reaction [16,26-291. This observation has been confirmed by a study conducted in heptane at 26’C and 1 atm, which achieved high conversion of ethylene (70%) associated with 75% selectivity to butene-1 when the AlpTi was 7.6 [16]. The level of the C6 oligomer and the polymer formed during the reaction at lower AlLI? molar ratio has been noticeably minimized. Whereas at AlfFi molar ratio of 18, higher conversion was achieved at the expense of the selectivity of the system to butene-1. Within this line, another study observed maximum selectivity to butene-1 at AliTi molar ratio of 3 in pentane C291. A relevant observation has been reported by Belov who observed an optimum Al/‘I’i molar ratio of 8-10 when the temperature range is lo-3O’C; and 3-4 at temperatures greater than 50°C [30]. A positive aspect of dimerizing ethylene at low Al/Ti molar ratio is related to the decline in the rate of deactivation of titanium active center caused by the presence of free AlEt3.

9

Solvent

The nature of solvent used in the catalytic system, Ti(OR)4-AlR3, is critical for the dimerization reaction [P&3-P121. Selection of the solvent is made in such a way that the solvent will not compete with ethylene to coordinate on the metal center or neutralize the needed Lewis acidity of the co-catalyst [171. Although various solvents have been used in the dimerization of ethylene, aliphatic hydrocarbons are often preferred 1161. The effect of the use of ether solvents, such as diethyl ether or dibutyl ether was studied at 40-60°C, 4-16 atm and lo-50 Al/Ti ratio [20]; the principal reaction product being butene-1. However, when the reaction was conducted in hydrocarbon solvents (butane, heptane, iso-octane, benzene, and toluene), the products next to butene-1 were: 0.5-5.0 vol.% butene-2, up to 5 wt.% higher olefins, and 1.5-8.0 wt.% polyethy lene. The use of low-boiling solvents, such as ethyl chloride and diethyl ether is preferable because they facilitate heat removal, isolation of butene-1, and rectification of the solvent [203. Butene-1 selectivity reached 98% when ethyl chloride was used as a solvent 1311. The activity of the dimerization reaction in various solvents decreased in the following order: n-heptane > Cll-Cl4 hydrocarbons > petroleum ether (40-6O’C) > p-xylene > benzene > CC14 [16]. The nature of the solvent used in the reaction significantly influences the rate of deactivation in the catalytic system Ti(OR)4-AlR3 [23. In toluene or in other aromatic solvents the deactivation of Ti active centers proceeds at a somewhat higher rate than in aliphatic media, such as heptane. In diethyl ether solvent and in the presence of hydrogen, the rate of reduction of titanium, which favors the formation of active centers, increased [211. The presence of hydrogen in the reaction media explains the role of hydride complexes of titanium during dimerization and the nature of active centers in the catalytic system of Ti(OR)4-AlR3 1211. Butene-1 can also be used as a solvent for the Ti(OBu)d-AlEts catalyst system (Al/Ti=lO) [321. The selectivity to butene-1 as product increased to 93% compared to 92% and 89% when heptane and toluene were used as solvents, respectively. However, the initial dimerization rate was independent of the solvent type, and the catalyst deactivation behavior varied with the solvent nature 1321.

Temperature The effect of temperature on ethylene dimerization and butene-1 selectivity was the subject of several studies [16,301. Increasing the reaction temperature from 25 to 50°C was accompanied by an increase in ethylene conversion and an increase in butene-1 selectivity 1161. Other reaction conditions were 12 atm, Al/l? ratio of 8 and C&11 hydrocarbons cut as solvent. The formation of hexenes, presumably by a codimerization of butene-1 and ethylene, was observed to increase with the increasing temperature. The effect of temperature on the kinetics of ethylene dimerization was studied in the range 30-90°C in heptane solution at Al/Ti ratio of 4 1303. It was found that the maximum butene-1 yield can be obtained at an optimum temperature of lo-3O’C. The activation energy of ethylene dimerization calculated from initial rates was 8.73 kcal/mol. As indicated earlier, an increase in the reaction temperature accompanied by a decrease in the Al/l? molar ratio reduces the rate of polymer formation [30].

10

Pressure The conversion of ethylene during dimerization reaction in the presence of Ti(OR)4-AIR3 catalyst increases with increasing pressure [16,27,301. Over a

range of 4-14.5 atm and 60°C in Cll-Cl4 as solvent, the conversion increased from 88.7% to 97.6%. The initial ethylene pressure has no effect on the selectivity of butene-1. The formation of higher olefins decreases with increasing pressure. However, the polymer formation increased from 0.9% at 4 atm to 3% at 14.5 atm [16]. Catalyst Modifiers

Polymer formation during ethylene dimerization is inhibited by the use of catalyst modifiers such as organic esters of orthophosphoric acid, diphenylamine, phenothiazine, and phenylacetate [16,33]. The inclusion of these compounds in the amounts of 0.1-l mole ratio of AlR3 leads to certain reduction in polymer formation associated with a decrease in catalyst activity [P351. The use of catalyst modifiers suppresses active centers responsible for polymerization [4,16,33]. Polar solvents and electron donor additives seem to stabilize the titanium IV complex that is responsible for dimerization. Various additives, such as tert-phosphines, phosphite, amine, and cyclic ether were used in the catalytic system Ti(OR)4-AIR3 at a temperature of 35’C, 12 atm, AYTi of 8 and Cll-Cl4 hydrocarbon cut as solvent. The amount of additive used was 0.1 mol/mol AlEt [16]. The optimum results were obtained when the sequence of addition to the reactor was: catalyst, Ti(OR$ followed by additive and then the co-catalyst AIRS. The results are presented m Table 1.3. The beneficial advantage of these additives was the suppression of higher olefins and polymer formation [161. The selectivity to butene-1 in the presence of various additives decreased in the following order: P(OPh)3 >THF > No additive > PP~Q> Et3N > PBu3. Table 1.3 Effect of catalyst modifiers on ethylene conversion and butene-1 selectivity for Ti(OBu)&lEt3* Cl63 Additive

Conversion (%)

PBu3 PPhs EtaN THF PCOPh>a No Additive

93.9 90.2 86.2 56.9 38.8 97.5

* Temperature = 35°C Pressure = 12 atm All Ti molar ratio = 8

Butene- 1 61.5 67.3 67.2 78.3 84.3 75.3

selectlvltv . . (Q ) Hexenes OcDtenes 37.0 28.0 29.2 17.1 15.7 21.2

Solvent = [email protected] hydrocarbon cut AdditiveIAIEt3 = 0.1 (molar ratio) Reaction time = 1 h.

Polymer

1.4 1.7 3.6 4.6

0.1 3.0

2.2

1.3

11

The modification of Ti(OR)4-AlR3 catalyst system by the addition of oxygen allowed a W-20% increase in the yield of butene-1 and simultaneous enhancement of the process selectivity [Ml. Mechanism of Ethylene Dimerization Using C?‘i(OR)4-ALRs A number of mechanistic hypotheses has been proposed to explain the selective dime~zation of ethylene using Ti{OR)4-AlR3 catalyst system [2,4,19,33-371. The mechanism of the dimerization reaction proceeds according to the following scheme for Ti(OBu)&lMes system C341: 1. Reduction of Ti compound by AlR3, 2. Oxidation of the reduced Ti compounds by ethylene with the formation of an active center con~i~ng 2 Ti atoms, 3. Complexation of the active center with 2 moles of ethylene, and 4. Hydrogen-atom transfer from one activated ethylene molecule to the other with subsequent formation of butene-1. The reaction is characterized by the formation of a titanium (IV) cyclopentane species which then decomposes to butene-1 by intramolecular hydrogen transfer and ~-elimination [4] as shown in Figure 1.3. The cyclic character of the intermediate explains the high selectivity to butene-1. The oxidation number of titanium is important in the dimerization reaction. Ti(II1) is known to be involved in polymerization while TiUV) favors dimerization [41. The formation of intermediate complexes containing titanacycle and homoallylic species was investigated by the decomposition pathway of titanacyclopentane [191. The intermediate was synthesized by the reaction of 1,4 ~li~iobutane and Ti(OIW4. This intermediate gave hydrocarbons (51.7% butane and 34.4% butene-1) upon warming from -WC to room temperature.

C”2 \ / Ti

CH2\ CH2

-Ti

kH

t

CH/ 3 Ti (IV)

Figure 1.3. Titanacyclopentane formation dimerization by titanium based catalysts [43.

CH2

Butene-l

as intermediate

in ethylene

The formation of intermediate titanium complexes during ethylene dimerization was proposed [2,35]. The reaction mechanism of Ti(OBu)4-AIR3 catalytic system is shown in Figure 1.4. The intermediate groups behave as binuclear active centers through the formation of bititanium ethylene-bridged

12

l-Active

Center

Formation

/H--A

{Hz-CH2----H2$CH2 )Ti-CH2 -CH2 -T’(

5sH9

-

)Ti-CH2-

CH2-Ti(

CH=CHz Z-Insertion ethylene

of toordinated in the C-H bond H2C=CH2 H2C=CH2 \J ,Ti-CH2-CHZ!i( -

3-Rapid

kH=CH2

CH-CH2

CH2- CH, \I ,Ti-CH+&-ii<

Chain Transfer 39

ji-CHz-CHz-Tt

./

* C2H4-

,v

,Ti-CHt-,H,-Ii<

1.4.

Proposed mechanism for Ti(OBu)-AlMeg catalytic

C4H8

CH CH2

:H’ CH2

Figure

l

system

[351.

complexes in the presence of AlMe [35]. The main features of the proposed mechanism are the following: l The participation of ethylene in the oxidation reaction of low valent titanium compound; l The formation of bititanium centers, which are active centers in ethylene dimerization to butene-1; and l The p~i~pation of free AlR3 in the ‘deactivation’ reaction of active centers. Other mechanisms for titanium-based catalysts were proposed by Cossee C361and other researchers [18,371. The proposed mechanism for bimetallic titanium-aluminum complex is presented in Figure 1.5 [18]. In Cossee’s mechanism, the propagation and intramolecular hydrogen transfer, responsible for butene-1 formation are favored decisively by titanium The stabilization of titanium d-orbital was suggested to be d-orbitals. important for catalytic activity [361. The coordination of two ethylene molecules and their arrangement to butene-1 on a Ti-coordination sphere via T&Et and Ti-Bu groups were proposed and discussed by Belov et al. [331. Ethylene dimerization was studied on a supported catalyst system of Ti(OEt~4-AlEt3 [23, P7]. Ti(OEt)4 was supported on porous carriers, such as

13

5

b”9

CzHI -/

‘~.~i,o’x.A,/

+ l- butene

‘\.

CH2- CH3

Figure 1.5. Proposed mechanism for bimetallic titanium-aluminum complex catalytic system [183. SiO2, AlgO3, and alumino-phosphate (P/A1=0.9). The reaction was performed at 95”C, 37.4 atm in isobutane solvent. The study compared the results of supported catalytic systems to the conventional unsupported catalyst system of Ti(OR)&lR3. Different approaches were used in supporting the catalyst; the best approach was to treat the carrier with Ti(OEt)4 and then reduce it. AlEt present in the reactor was used to reduce the catalyst in situ. The activity of catalysts supported on SiO2 and aluminophosphate was quite good when compared with the homogeneous catalysts. The best activity achieved was at an AVl’i ratio of 1.4 for the silica-supported catalyst. The kinetics of ethylene dimerization using supported-titanium catalytic systems were different from the unsupported systems. The dimerization rate was high and the catalyst productivity of butene-1 was more than the unsupported catalyst as shown in Figure 1.6. Supporting the Ti(OEt.14catalyst gives an improvement in overall productivity of the system because the active sites seem to be stabilized on the support. Another benefit of this system is that it minimizes the reactor fouling associated with liquid systems. Much of the polymer produced (1.5 wt.%) was recovered as tiny beads on the catalyst surface instead of the reactor walls as experienced in the use of the unsupported catalysts [23]. A novel catalytic system based on Ti(OBu)4-AlEts-MgC12 showed high activity for concurrent dimerization and polymerization of ethylene 1381. The experiments were carried out in Cll-Cl4 solvent at 27°C and 1 atm. Comparative tests were performed on a conventional Ti(OBu)4--AlEts system and a Ti(OBu)&lEt+MgC12 system for ethylene dimerization over a range of AhTi molar ratios of 5-100 (MgITi molar ratio was 0.23). Mg-Ti catalyst was characterized by high initial rates but low conversion. The selectivity to butene-1 showed an initial decrease up to AlIl’i ratio of 50 followed by a sharp increase to 95% at AVIS of 100 and at a conversion of 20%. The Mg-free catalyst did not show any change in butene-1 selectivity over the Al/Ti ratios of 5 to 100. The results are presented in Table 1.4. The advantage of the Mg-Ti catalyst is that no solid polymer was formed in any of the experiments. The results indicate that by the use of one titanium compound, Ti(OBuj4, it is possible to perform dimerization/polymerization

14

10

15

20

25

35

Reaction Time (min 1 Figure 1.6. Kinetics of ethylene dimerization for supported and unsupported Ti(OEt)4 catalytic system [233. Table 1.4 Comparative results of ethylene dimerization using Mg-Ti and Ti (OBukAlEt systems carried out at 27”C, 1 atm and for 45 min in C+C!14 solvent [383 Complex complex Mg-Ti Mg-Ti Mg-Ti Mg-Ti Ti(OBu&AlE t3 Ti(OBu)4-AlEt Ti(OBu)&lEt3 Ti(OBu)&lEta

AlITi ratio

Conversion (%Jo)

5 $ 106 5 10 50 100

Butene-1

Hexenes

a

iii 45 ii 88

41 39

13 44 18” 20 25

12

Octenes 3 8 I2 4 2 1

2 9

15

reactions. A similar study was conducted using dual-function catalysts of Ti(OPrL-AEt3 and TX&-MgC12 systems [26]. The activity of the dimerization component of the dual catalyst was sufIicient to provide the necessary amount of butene-l needed for the polymerization reaction. The dimerization reaction was conducted in heptane solvent at 90°C yielding butene-1 at a selectivity of more than 90% with a small amount of hexenes and polymer (2-4 wt.%>. The presence of the dimerization component in the dual function catalytic system resulted in a two-fold increase in its reactivity [26]. The use of dialkylaminoalanes, HzAlNR2 as co-catalysts instead of alkyaluminum in Ti(OR)4 systems, showed that high butene-1 yields can be obtained using alane derivatives [40]. Selectivity to butene-1 was 95% in toluene solvent at 65”C, 21 atm and for 3 h. Cyclopentadienyl titanium trichloride, C,TiC13, and sodium amalgam dimerized ethylene to butene-1 at room temperature and atmospheric pressure in toluene solvent [41]. The selectivity of the product was 80% butene-l and 20% hexene-1 with traces of butane, hexane and hexene-2. The mechanism proposed for this reaction was based on the stepwise loss of Cl from CsH5TiC13 by reduction with Na amalgam. Another titanium-based catalyst used in ethylene dimerization to butene was (C4H& Ti(dmpe1, where dmpe = 1,2 biscdimethyl phosphino ethane) 1421. The mechanism of this reaction is similar to the Ti(OR)h-AlEt system through the formation of titanocyclopentane complex. The kinetics of the reaction showed the firstorder dependence on the catalyst and the olefin. The reaction of Cp2Ti(PMe&, in excess ethylene under 50 atm pressure and room temperature in pentane solvent produced a bis-ethylene complex, CpzTiCgH4 [433. The ethylene complex was converted to titanacyclopentane species that releases butene-1 and trans-butene as well as Cp2Ti fraction upon fragmentation. The resulting CpnTi fraction regenerates the starting material (bis-ethylene-complex) by uptake of ethylene, thus completing the catalytic cycle. The reaction cycle was repeated four times until the pressure dropped to 15 atm. 1.4.2 Nickel J3asedCatalysts Among the large number of nickel complexes commonly used to catalyze the formation of butenes from ethylene, nickel halides stand as the favorite for reasons related to its low cost. In addition, other sources of active nickel catalysts have been introduced in the form of an organic acid salts, an acac, an olefin, a diene or an ally1 complexes. The coordination of some donor ligands, such as phosphines, to the Ni metal center is known to adjust the metal-to-ligand ratio so as to have a better control on the ratio of dimerization vs. oligomerization reaction [171. Since the discovery of their specific activity to catalyze the dimerization reactions, the nickel based catalysts have been the subject of intensive mechanistic and kinetic studies directed toward the determination of the active sites responsible for this criteria. Earlier nickel based catalysts were homogeneous; however, recent attempts have aimed toward supporting these catalysts to improve their performance.

ilrwid CadcJyats The majority of these catalytic systems, which have exhibited activity for the ethylene dimerization, are composed of a nickel precursor, a phosphine ligand, and an organoaluminum compo~d as a co-catalyst. These catalytic systems derive their dimerization activity from the remarkable ability of nickel to control the mode of linking of olefins, and its high specific activity [44-571. Tris (ethylene) nickel (0) forms an active catalyst with Lewis acids for ethylene dimerization. For example, maximum activity was observed with a molar ratio for BF3:Ni of 3 and AlCl3:Ni of 4 [453. Bisttriphenyl phosphine)aryl nickel (II) bromide and SF3 etherate in benzene or methylene chloride showed a high activity for ethylene dimerization to butenes through a nickel hydride complex [46]. The reaction was carried out at mild conditions (O”C, atmospheric pressure) with 95% butene selectivity. The activity of nickel (ID complex was highly increased due to the substitution of the halogen ligand by an aryl ligand, such as naphthyl or tolyl. The system of bis(t~pheny1 phosp~ne) pentachloro phenyl (chloro) nickel (II) activated with silver perchlorate catalyzed ethylene dimerization in bromobenzene at OOC,and atmospheric pressure [47]. The products (mainly butene-2) were improved by the addition of catalytic amounts of PPh3. Nickel palmitate-AlEt. (0,Ol:l wt ratio) was also found to dime&e ethylene at 40 atm and llO-130°C to butene-1 (510%) [48]. The life of the catalyst was extended to 200 h by successively increasing the temperature to 125’C. Pisman et al. 1491,using a similar catalyst system, proposed a carbonium intermediate on the nickel surface with activation energy of 17.4 kcal/mol. Dimerization of ethylene on a PPh3-Ni (I) catalyst system was carried out at room temperature and atmospheric pressure in propylene carbonate solvent to produce mainly 2-butenes [50]. Brunet et al. [Sll investiga~d the introduction of tris (3-s~fophenyl~ phosphine salt into I(?‘$methallyl)Ni(COD)]PFs to increase the selectivity of dimerization to 1-butene. Selectivities in butenes up to 96% could be obtained. The authors proposed a nickel hydride mechanism as shown in Figure 1.7. Aryl nickel complexes such as [NiB~~6F5)~PPh3)21and [Nicl(c6cl5)(PPh3~2] were used for dimerization of ethylene in the presence of AgC104. The role of Age104 was to abstract the halide ion and/or triphenyl phosphine. The amount of [Ni(ClG~)(o-CsX~)(PPh3)1 was correlated with the dimerization activity [52]. A similar trend has been observed in metallacycles that was reported to play a major role in olefin dimerization to produce linear and cyclobutenes in toluene solvent at 25°C and 5.7 atm 1531. Among a number of nickel complexes tried in a systematic study aimed at finding the best candidate system to catalyze the dimerization of ethylene, the x-ally1 nickel halide complexes have exhibited an outstanding activity upon their combination with aluminum halide [54]. Ghymes et al. investigated the influence of each of the reaction parameters on the course of the reaction that produces a mixture of butenes and hexenes. Optimum yield of butene-1 was achieved at lOo”C, >40 atm, and at the AlEt concentration of ~4 mokliter 1551.

~~~

17

[HNil+

b f

Csl + [HNi]+ [HNi]+

[HNi]+

Figure 1.7. Mechanism of ethylene complexes (Ni+ = NiL,+) [51].

dimerization

with

cationic

nickel

This implies that through the control of the reaction parameters, the reaction could be switched toward the selective production of hexene-1. The nickelocene (dicyclopentadienyl nickel) complexes have shown a remarkable activity associated with a good selectivity to produce butene-1 when it was utilized as a catalyst for the ethylene dimerization 1561. The reaction was conducted at 2OO’C and 41 atm whereupon the nickelocene undergoes a homolytic decomposition to produce excited nickel atoms thought to be responsible for the ethylene molecules dimerization in a fashion similar to that observed in heterogeneous catalysis. Tsutsui proposed a three-step mechanism for this reaction as shown in Figure 1.8. A series of (qs-arenes) NiR2 complexes (where R=SiF3, SiCl3, CsF5, and arene = benzene, toluene, and mesitylene) were tested as homogeneous catalysts for ethylene dimerization 1571. The highest activity was reported for (qs arene)_Ni(SiCl$z. The kinetic analysis showed approximately first order dependence on ethylene and nickel catalyst. A mechanism based on Ni-H species as intermediate has been proposed for the dimerization reaction catalyzed by this class of Ni complexes is shown in Figure 1.9.

CH2

H2C\\

H,c’/\

J

CHz=CHzb

CH2

.

-

iN’\ I

H2C

Hidr t-P.

*

.CH~-CH~-CH~--CHP

I

CHs=CHa

C&=CHy

Low temp.

I

higherdigomcr

,CHs ?I42

&H2 H2C

I

.

H2C\ ‘CH2 J

//CH2 H2C

H2C\\

\

/

CH2

CHt=CHs jN’\

/“; III

+

CH2=CHCH,CH,

Figure 1.8. Mechanism system 1561.

of ethylene

dimerization

with nickelocene

catalyst

Hetemgeneous Nickel Catalysts Heterogeneous nickel dimerization catalysts consist of active nickel compounds deposited on a solid support. Nickel oxides and nickel complexes, supported on silica, silica-alumina, Y-zeolite, X-zeolite, and polymeric material have been reported to be active species for dimerization of ethylene 158-76, P46P49J A series of supported nickel oxide catalysts have been prepared by Matsuda et al. [58]. Those supported on an acidic silica-alumina demonstrated an ethylene dimerization activity coupled with an isomerization tendency of the formed butene-1 to butene-2. This was attributed to the presence of both Lewis

19

R’

ici,

R

Solvent

in high excess

tl X @ R’

+ Ni

@

‘R

CH2 = CH, CH2 = CHR

f/X

e

e

Ni

R ’

0

‘CH2CH2CH2CH,

Ni,

X

CH2 - CH3

* . CH2 = CH2 R A



NI(

\;Ht CH2 CH,

Figure 1.9. Proposed catalytic catalytic complex 1571

cycle

for ethylene

dimerization

by nickel

and Brijnsted acid sites with the former being responsible for the generation of butene-1 whereas the latter is thought to be the reason behind the formation of butene-2 via a direct route from ethylene. The role of the acid sites of the NiOSiO2 catalysts was the subject of an earlier deuterium tracer study 1591. Unlike Matsuda’s interpretation, it was concluded that the deuterium involved in the catalyst did appear in the product butene by isomerization and not by direct dimerization. Yashima et al. [60], reported that ethylene dimerization can be catalyzed The study, conducted at selectively using NiY-RhY and RuY zeolites.

20

temperatures ~100°C and 0.26 atm, showed that for NiY and RhY, the active sites consist of a zero valent Ni and Rh highly dispersed over the zeolite framework. The activity of dehydroxylated nickel-chrysotiles (NixMg3JOH14Si205) was studied for ethylene dimerization [61]. The effect of activation conditions on catalytic dimerization and hydrogenation of ethylene has been reported. The active sites for dimerization were plausibly nickel cations exposed on the surface. The conversion of ethylene was 2.5, 5.5 and 9.3% for chrysotiles of Ni/Mg=l and Ni/Mg of l/4 and Ni chrysotile, respectively, after 1 h reaction at 50°C and initial ethylene pressure of 0.4 atm. The catalytic activities of nickel silicates (montmorillonite and antigorite) for ethylene dimerization and butene isomerization were studied at 20°C, 0.3 atm for one hour 1621. The variation in catalytic activity was closely correlated to the surface acidity of the catalyst. The active sites were protonic in the montmorillonite and nonprotonic on the antigorite. It was shown that the higher the final pH in precipitation from acidic solution of nickel salt-sodium silicate mixture or the higher the Ni/SiOa ratio, the more formation of antigorite was obtained. Two optimum temperatures of activation observed were 100 and 600°C. The first activation was ascribed to the montmorillonitetype site, while the second activation was ascribed to an antigorite-type site. Solid NiO-TiO2 treated with H2SO4 was used to catalyze ethylene dimerization to butenes at 20°C and 0.37 atm. The activity was studied as a function of catalyst composition, preparation methods and activation procedures 1631. The catalytic activity of NiO-ZrOs modified with SO4-2, PO43 and B03-3 anions were studied for the dimerization of ethylene [64]. The dimerization activity was correlated with the acid strengths of the catalysts and activation conditions. Ligands such as CO, C2H4, or C3Hs can be selectively introduced to the coordination sphere of X-zeolite or silica supported Ni(1) ionic complexes leading to the formation of a variety of cationic species of the formula [Ni(L>,J+, the structure of which islargely determined by the applied pressure [65]. At lower pressures (<40 torr), the cationic Ni complexes favor either the tetrahedral or the square planar structure, whereas the higher pressure enforces these complexes to adopt the trigonal bipyramid symmetry. All of these supported Ni(1) complexes are active dimerization catalysts. On a parallel line, the selective introduction of phosphinated ligands into the silica supported Ni(1) ion, afforded the square planar complexes [Ni(PEts)z(Oz)l, where 0s is the surface oxygen ions acting as a ligand. , An EPR study, conducted for this system, confirmed the presence of a metallocyclopentane intermediate as illustrated in the reaction mechanism showed in Figure 1.10 1661. This tailored-made catalytic system has shown a remarkable selectivity for the dimerization of ethylene to butene-1 (95%) at a conversion of 23% [661. This selectivity was attributed to the electron density donated by the basic phosphine ligands to the supported Ni+ ions that allows an easier desorption of With the ionic compounds containing basic and bulky the products [67]. phosphines, very high activity and selectivity to butene-1 were achieved.

Trimers

F’igure 1.10. Proposed reaction mechanism for ethylene dimerization on supported Ni(I) complexes 1661. Grafting nickel complexes onto polymeric carrier and employing it to catalyze the dimerization of ethylene have been the focus of intensive research activities 169-73, P62-P641. It has been suggested that the activated stability of these catalysts was enhanced substantially upon chelating the Ni(I1) onto the carrier in the fashion shown in Figures 1.11 and 1.12. Echmaev et al. [703 prepared and tested supported NW) complexes and their interaction with alkylaluminum chloride over a wide range of temperatures less than 27OC. Polyethylene modified with radiation grafted polyacrylic acid, poly-4vinylpyridine and polymethacrylic acid was used as a support. The reduction of Ni(I1) by alkylaluminum chloride, during the dimerization of ethylene to finely dispersed metallic nickel, led to the deactivation of the catalytic system. The polymer support retarded the aggregation of Ni(0) complexes, thus preventing the deactivation of the catalyst. Similarly, the polymer supported NiC12 systems have demonstrated a good turnover rate associated with high selectivity to butenes [711. Its best achieved selectivity to butanes was 908, of which 80% was butene-1. The principle of synthesis and activation of gel immobilized-metal containing catalytic system was discussed 1721. The catalyst activation was accelerated by the use of chlorohydrocarbons. Borovkov

22

Ni (II) + AIR, Cl3_, -

Ni (1) + Ni (OlI ;IR, C13_x lfi (0) +Ni

Ni (0) colloidal

(111

AIR, Cl,,, I

Ni III + AIR,Cl3,,-Ni

(II) + Ni (0)

Figure 1.11. Formation of active centers for Ni(I1) catalytic ethylene dimerization [691.

[-KHz--CH-I+-

+

1 Ho ‘CH3

NH?-yH

-

]-KH,-Cr-,+-

+

CH3 -C=N\

?=”

HO,

system used for

0-C 0\

rH I

HO-/ I +Ni(lll

-

]--(-CHz-fH

0

-It-

i 0

-OH

Figure 1.12. carrier [691.

Scheme

of Ni(I1)

anchoring

on the

surface

of a polymeric

et al. [73] investigated olefin dimerization and isomerization on tetrahedral bivalent nickel ions supported on aerosil. The conversion at temperatures 203V’C was 4-10% while selectivity to 1-butene was 40-70%. Dimerization active sites were reported to be nickel hydride. A mechanism for dimerization and isomerization was proposed as shown in Figure 1.13. The kinetics of ethylene dimerization over a Ni-NaY zeolite catalyst were studied [74]. The rate of ethylene sorption and diffusion of ethylene molecules in the solid was found to be a rapid step. The influence of intracrystalline mass transport on the rate of dimerization was studied between 70 and 210°C.

23

CH2fW2

H... Ni+ZCzbeH...

Nifll-

s~j~W-W

-Ni

-

CH2

H...Ni+CH,-CH2-CH=CH2 CH,-CHfCH2-{Hz CHS- CH2-CH jCH2 H...Ni H _ .CHI-CH2-II/-CH, Ni Cb-CHI CH-CH, H...Ni CHS-CH+CH-CHa+Ni...H Figure 1.13. Scheme of ethylene catalytic systems [733.

dimerization

reaction

by bivalent

nickel

The dimerization was second order with respect to ethylene and the rate was represented by the following equation: -d”CzHd dt where

=

k [PC2H212

(1.3)

k = 1014 exp.(-2l/RT)

The deposition of ternary complexes prepared by the interaction of Ni(CsH70&-EtzAlCl-PPha over Al203 or Al203-SiO improved the stability and the activity of these catalytic systems for the selective dimerization of ethylene to butene-1 in comparison to their homogeneous counterparts. The obersed kinetic parameters were found to be comparable with those of the homogeneous systems 1761. 1.4.3

other Catalytic Systems

The dimerization of ethylene can be selectively catalyzed by CoX(PPh313, in the presence of a Lewis acid in halobenzene solvent 1771. The catalyst was prepared from Co(acac)a and PPb in ether followed by the addition of halobenzene. The reaction was carried out at 0°C and atmospheric pressure. The effects of various solvents and Lewis acids on catalytic activity were studied. Ethylene was readily absorbed in halobenzene solvents whereas no absorption was observed in other solvents including benzene, THF, and dichloromethane. Bromobenzene was the most effective solvent among halobenzenes, the selectivity to 1-butene being as high as 90%. Among a set of Lewis acids tried, BF* OEta was chosen to be the best candidate for reasons related to its capability to activate the Co(I) complex and its higher stability 1773. The cobalt hydride complex [(H_(N21Co(PPh$al, catalyzes ethylene dimerization without the presence of Lewis acids and at room temperature [783. A mechanism was proposed in which ethylene was inserted into the cobalthydride bond and then a second molecule of ethylene was inserted into the

24

generated cobalt-alkyl bond. The displacement of the dimer by ethylene regenerated the cyclic process. Cobalt (II) or cobalt (III) complexes together with organometallic compound as a co-catalyst were used as catalysts for ethylene dimerization [18, PSZ-P86]. The product mixture consisted of mainly 2-butenes. The optimum molar ratio of Co-catalyst/Co-complex was between 2 and 5. The exposure of ethylene to the tantalum based complex [CpTa(CHCMeg) Cl21resulted in its rapid and selective dimerization to butene-1 by the plausible mechanism given in Figure 1.14 [79-83, P87-P901. The deuterium labelling studies have furnished evidence that the dimer does not form by insertion of an ethylene molecule into a tantalum-ethyl bond but via a tantalacyclopentane intermediate which then rearranges to the dimeric olefin 181-831. The formation of these Ta5+ metallocycles is thought to be due to the coupling of two olefins around the Ta3+ in a [TaCpC&l fragment [SU. The high selectivity of these tantalum based catalysts to butene-1 was attributed to the fact that isomerization of the dimer is negligible. The reduction of the polymerattached [CpTaCle] with an alkyl lithium resulted in the formation of dimerization catalyst with a modest turnover from butene-1 of 18 mmoflmmof Ta-hr at 100°C and 5.4 atm [853. L

L 1 04 /-

‘I”Ul L

Czh

IL

-

I

L f

Id

1 Te

3

-

)YTle\ L

Figure 1.14. Proposed mechanism for ethylene dimerization to butene-1 by tantalum catalytic complex [831.

Hydrated rhodium chloride (RhCls-H20) in alcoholic solution converted ethylene to a mixture of butenes [86]. The yield and composition of the butene mixture were found to be dependent on the temperature, reaction time, and catalyst concentration. The highest butene-l yield (38%) was obtained at 30°C and 1000 atm with 3% RhCls-Hz0 in methanol. RhCls-H20, having a Cl-IRh ratio higher than 3, in the presence of HCl, dimerized ethylene at a lower pressure (100 atm) and 50°C E87,881;99% of the initial product was 1-butene but isomerized rapidly. Isomerization was avoided by lowering ethylene conversion. A mechanism was proposed involv-

25

ing: (i) coordination of the olefin with the metal to form a rhodium (I) complex, (ii) oxidative addition of HCl to give ethyl complex, (iii) insertion of coordinated ethylene to give a butyl complex, (iv) conversion of complex formed in step (iii> to butenyl complex, and (v) replacement of butene molecule from the butenyl complex by ethylene, thus, regenerating the starting complex. It was found that the rate of dimerization varies directly with hydrogen ion, rhodium and ethylene concentrations. An activation energy of 17.2 kcal/mole was reported. Rhodium chloride supported on silica gel, showed a higher activity to ethylene dimerization compared with the homogeneous rhodium chloride catalyst system 1891. The observed activation energy of the reaction was about one-half (7 k&/mole) of that obtained in homogeneous catalysis. The reaction was carried out in a fixed bed glass reactor at 2&X! and 0.6 atm using different types of supports. The amount of rhodium supported on the carriers depended on the type of carrier and decreased in the order: silica gel > silica-alumina > alumina. However, the dimerization activity per unit weight of the catalyst increased in the reverse order. The presence of hydrogen chloride during the reaction enhanced the catalytic activity as in a homogeneous catalyst system. Butene-l was formed predominantly at the initial stage but decreased with time as it was isomerized to butene-2. Rhodium complexes with tin chloride ligands [Rh2C12(SnC13)& in acidified alcoholic solution catalyzed ethylene dimerization [go]. The complex was more active and stable than RhCleH20 having a maximum activity at Sn/Rh ratio of 2 and hydrochloric acid concentration higher than 0.5 mol/l. The dimerization reaction was first order with respect to the catalyst and ethylene concentrations; it was selective to normal butenes (3% butene-1 and 97% cis/trans butenes-2). Palladium complex, in particular PdC12, readily dimerizes ethylene to butenes. This reaction was found to be influenced by the type of solvent used. No dimerization reaction occurred in a series of polar and non-polar solvents, and it proceeded smoothly only upon using an oxygen containing solvents. Thus, it was concluded that the rate of ethylene dimerization decreases in the order: weakly dissociative > dissociative > non-dissociative solvents 1911. Kusunoki and coworkers observed a similar influence of the reaction’s solvent and its selectivity to butenes [92]. The selectivity was found to decline in the order: acetic acid > ethylene dichloride > benzene > chloroform > ethanol. Dichloro bis(benzonitrile) palladium (II) complex in benzene solution was found to catalyze dimerization exclusively to a mixture of butenes (4% 1-butene, 36% c-2-butene, and 60% t-2-butene) 1931. The reaction rate was first order with respect to ethylene. There was an induction period before the start of the reaction. The extent of this period was reduced by the presence of hydrogen releasing olefin, such as 3 methyl-1-butene; which also enhanced the reaction rate without the formation of codimer. Arene Pd(D complexes dissolved in benzene were reported to catalyze ethylene dimerization [94]. Conversion of 95% to butenes was achieved but the product mixture mainly consisted of butene-2.

26

The kinetics of ethylene dimerization catalyzed by PdC12 dissolved in chloroform were established 1951. The reaction was carried out in a stirred batch reactor at 27-47’C temperature and 1-4 atm pressure. The rate of reaction followed 1.5 order and first order kinetics with respect to ethylene and catalyst concentration, respectively. An activation energy of 27 kcal/mole was reported. Few studies on ethylene dimerization using heterogeneous palladium catalysts have been published 196-102, P76]. Dimethylsulfoxide-palladium chloride (DMSOjxPdC12 deposited on silica gel catalyzed ethylene dimerization [961. The butene yield and the selectivity were maximum at 90°C using 0.1% of Pd complex deposited on silica gel in acetone; 0.3% of K+[(DMSO)PdC&$ in phenyl chloride deposited on silica gel achieved the highest selectivity at 100°C. In a similar study [971, the effect of temperature (40-EO”C), the amount of complex deposited and the nature of the solvent on the degree of dimerization, were investigated. Various solvents including hydrochloric acid, acetic acid, acetone and acetonitrile were used. Butene yield (61%) was obtained at 90°C with 0.1% of the complex was deposited on silica gel in acetone. PdCl2 supported, on silica gel in acetic acid, effected ethylene dimerization with a high activity and selectivity, while palladium complexes deposited on 1” alumina showed maximum stability [981. Palladium loaded by ion-exchange on Ca-X zeolite showed an activity towards ethylene dimerization [99]. Selectivity to 1-butene was low even at low ethylene conversion. The effects of temperature and palladium ion concentration, on ethylene dimerization using pretreated palladium exchanged Na-X, Ca-X, Na-Y and Ca-Y zeolites were studied [loo-1021. The reaction was carried out in static and flow systems, over a temperature range of 2585°C. Reaction started after an induction period which was shorter for CaPd-X. This period decreased with an increase in temperature and the palladium content, which was due to the migration of Pd species towards sites accessible to ethylene. The activity increased and the induction period shortened with NaPd-Y when compared to NaPd-X at similar Pd concentrations. No significant difference between CaPd-X and CaPd-Y zeolites was observed. High conversions were observed with some of the catalysts, but the selectivity to 1-butene was low. It was found that Pd+ ions were the active intermediates for the reaction which occured by the addition of two ethylene molecules. to the same Pd+ ion as shown in Figure 1.15 I1021. Lanthanides&rsed Catalysts

Lanthanides (samarium and ytterbium) exhibited an activity for ethylene dimerization I1031. The catalysts were prepared by metal vapor deposition using tetrahydrofuran. The product mixture consisted mainly of cis and trans 2-butenes. The initial rate was high at elevated temperatures and pressures. Based Catalysts Niobium oxide supported on porous silica and activated by W irradiation in the presence of ethylene catalyzed the dimerization and isomerization of

Niobium

21

Pd*

+

Cp H4

-

Pd’-11

CH2

QH4

-

CH2

//CH2

CH

2

\ geometric

Pd+ -

isomeritotion)

CH2\\

/ CH2

I

CH2

Pdf

-

CH iI--Pd’

-

C4Hs

+

Pd+

CH

I CH3

Figure 1.15. Proposed reaction mechanism palladium catalytic complex [ 1021.

for ethylene

olefins at room temperature. However, conversion and the selectivity to 1-butene was 66% 11041.

dimerization

by

was not higher than 5%

Chnnnium andZirconium Based Catalysts The organophosphine chromium (III) complex and ethylaluminum dichloride were found to dimerize ethylene in good conversion and selectivity [105, P91-P94]. The dimerization reactions were conducted in a 300-ml stirred autoclave in which the catalysts were prepared by the interaction of chromium complex and organoaluminum compound in a solvent (chlorobenzene or cyclohexane). The results indicated that the dimerization activity was a function of the organoaluminum type and the reaction conditions. A catalyst system comprising CrC13(Py)3 and EtAlC12 dimerized ethylene to a mixture of butenes at a selectivity of 83% compared to trace amount of butenes formed when Me3Al2C13 was used as a cocatalyst. A productivity of 4700 g-butenes per g-Cr complex was obtained using a catalyst system consisting of CrC13(4EtPy)a and EtAlC12. The selectivity to butenes was 83% out of which butene-1 was 50% and the balance was cis/trans butene-2. Stirring toluene solutions of [Zr(nCsHs)2(dmpe)l2 at 70°C under 1 atm of ethylene yielded 13% 1-butene at 94% ethylene conversion in 50 h 1421. The rate of reaction followed first order kinetics. A metallocyclopentane complex

mechanism wa8 proposed. A similar catalyst, [CpZrMe(dmpe)zl, was reported as a catalyst precursor for the selective dimerization of ethylene to lbutene ClOSl. It8 8yntheBi8 and preliminary study on it8 reactivity were reported. !l’ungstenandiUolybdenumBaaedCatalysts The dimerization of ethylene to butene-1 in high selectivity and yield was accomplished in the presence of a reduced tungsten complex and Et2AlCl in benzene [107, PSll. Ethylene was converted at a rate of 184,000 mole per mole of tungsten in 1 h at 40°C, and 34 atm of ethylene. Butene-1 selectivity reached 98%. A molybdenum complex [(x-C6H6)-MO(x-allyl)cl]2, with ethyl aluminum dichloride, catalyzed the dimerization of ethylene in benzene medium, at 20°C [1081. Aluminum Alkyls Catalysts Trialkylaluminum (AlR3) ha8 been extensively used in several processes involving olefins. Aluminum compound8 are usually used a8 co-catalysts in many dimerization catalytic systems. The kinetics of homogeneous dimerization of ethylene to 1-butene catalyzed by triethylaluminum (AlEts) were studied in gas phase at temperatures ranging between 160 and 230°C [1091. The catalyst was recovered quantitatively from the product mixture. The only side reaction was the isomerization of butene-1. Conversion of 99% was achieved and butene-1 was the primary product. The proposed reaction mechanism is summarized by the following reactions: LWt312

M

2AlEt3

(1.4)

AlEt + CzH4

*

Et2AlBu

(1.5)

Et2AlH + butene-1

(1.6)

AlEt

(1.7)

)

EtnAlBu Et&H+

C2H4

*

The rate determining step was the addition of AlEt to ethylene to form n-butyldiethyl aluminum, (EtzAlBu) (equation 1.5). The reaction between triethylaluminum and ethylene was studied over a wide range of operating condition8 (80-22O”C, 1-9 atm, l-10 wt.% catalyst) CllOl. The reaction rate was first order with respect to the concentration of the dissolved ethylene in the liquid and to the triethylaluminum compound. The reaction rate increased with increasing temperature although it was Slow below 12OOC.Butene-1 was the major gaseous product. 1.6

DIMERJZATIONPRQCESSES

Four dimerization processes based on different catalytic systems are assigned to IFP, Phillips, MIT, and Dow. Only the IFP process (Alphabutol) ha8 reached the stage of commercialization and its description is based on the

29

information obtained from IFP and some related publications 14-61. The other processes, which are conceptual, are discussed based on information obtained from various patents. 1.6.1 IFP-Alphabutol process The Alphabutol process, developed by the IFP, is currently the only available

commercial process for ethylene dimerization to butene-l 141. In the early 19808, IFP developed, patented and commercialized the Alphabutol process which selectively dimerizes ethylene to butene-1 in the presence of a catalyst comprising Ti(OBu)4 and AlEt [Pl-P31. The process uses the technology similar to the IFP-Dimersol process. There are eight licensed Alphabutol plants with four in operation [4-61. Pmcess Description The Alphabutol process consists of four sections [51: 1. Catalyst preparation step, 2. Dimerization reactor, 3. Catalyst removal section, and 4. Distillation product recovery. The flow scheme of the Alphabutol process showing the four sections is presented in Figure 1.16. The catalyst used is a Ziegler-Natta type, believed to be’s titanium based catalyst, named LC2253 (IFP proprietary) and triethylaluminum, AlEts, as cocatalyst. The catalyst components are combined in the plant itself and pumped from storage tanks to the reactor after dilution with butene-1. Fresh and recycled ethylene are fed to the liquid part of the reactor through a gas distributor. The homogeneous catalyst components are continuously fed to the reactor section. The dimerization reaction is carried out at about 50-60°C and 25 atm and a reaction time of about 5-6 h. The homogeneous catalytic reaction proceeds at an ethylene conversion of about 80-85% per pass and a selectivity to butene-l approaching 93%. The exothermic heat of reaction (-25 kcal/gmole) is removed via an external classical pump-around the device equipped with a cooler [41. The vapor flow rate is maintained very low and the main part of the eMuent is withdrawn from the bottom of the reactor as a liquid containing the spent catalyst 153. In the catalyst removal section, the liquid withdrawn from the reactor, is vaporized in two steps: the first vaporization is made in a classical exchanger and the second is achieved in a thin film evaporator. The residue containing the spent catalyst and a small amount of hydrocarbons are incinerated [51. At the distillation section, the catalyst-free hydrocarbon portion of the reactor effluent proceeds to the ethylene recycle column where unconverted ethylene is distilled overhead and recycled to the dimerization reactor. The pressure of the recycled ethylene is chosen on the basis that ethylene can be directly recycled to the dimerization reactor [4,53. A small Cl/C2 purge stream removes inert methane/ethane from the recycle gas. The bottoms from the ethylene recycle column are fed to the butene-1 purification column where comonomer grade butene-1 (99.7%) is distilled overhead and sent to storage facilities. The purification column bottoms are mainly oligomers of C6+ which

Figure 1.16.

ETHYLENE

Flow

CH

J I

CATALYST REMOVAL

1,

INCINERATOR

4

‘.:.:.:.:.:.:.~:.:.:.:.:.:.:.:.:.: .................

........... .::.:.:.:.: 0 C6 CUT

b

l-BUTENE

b

-c

I-BUTENE COLUMN

scheme of IFP-Alphabutol process for butene-1 production via ethylene dimerization [5].

I

r T-

T

4-

RECYCLE COLUMN

31

are sent to gasoline blending facilities. The overall yield of butene-l based on ethylene fed to the plant approaches 95% [3,41. Typical analysis of the polymerization grade butene-1 produced by the Alphabutol process is presented in Table 1.5. Table 1.5 Typical analysis of IFP Alphabutol butene-1 El Composition

Concentration limits

Butene-1 Other CgS (butenes and butanes) Ethane Ethylene

99.5 0.3 0.15 0.05

Impurities

wt ppm max

c15 Oh?finS Ethers (as DME) Sulfur, Chlorine Dienes, Acetylenics CO, CO2, 02, HzO, Methanol

50 2 1 5 each 5 each

wt wt wt wt

% % % %

min max max max

1.5.2 Phillips Ptwcess A novel method for preconditioning an olefm dimerization reactor has improved the selectivity and the yield of dimers significantly [P57,P58]. The method can be applied to homogeneous dimerization processes using at least two component hydrocarbon-soluble catalyst systems. The catalyst used in the Phillips process consists of ethyl aluminum dichloride and bis (tri-n-butyl phosphine) nickel dichloride prepared in dry n-pentane. Ethylene dimerization takes place in a circulating loop reactor. The reactor preconditioning method is carried out according to the following steps: 1. The reaction system is filled with a suitable cleaning solvent and circulated for 20-30 minutes; 2. The cleaning solvent is dicharged and a fresh batch of reaction solvent is charged to the reactor; 3. Nickel catalyst addition is started at a rate sufficient for the anticipated initial flow of ethylene into the reactor; 4. Aluminum compound addition is started at a rate to give a controlled Al/Ni molar ratio less than 2:l; 5. Ethylene flow is started; 6. The Al/Ni ratio of less than 2 is maintained for about 25 h during the preconditioning or startup period; 7. The catalyst and the feed rates are adjusted to get the desired product rate.

32

The flow rates of nickel complex and aluminum compound are adjusted to obtain AliNi molar ratio higher than 2, by the end of precondition/startup period. The reaction can be carried out as long as it is desired or until fouling of reactor surfaces requires a shutdown for cleaning. The system should be water free before the start of the above procedure of preconditioning [P57]. The AvNi molar ratio being less than the stoichiometric ratio of 2 is found to be the key feature in reducing the initial preconditioning period. The stoichiometric ratio of 2 was found based on the proposed reaction mechanism, as shown in Figure 1.17. Ethyl aluminum dichloride reacts with the nickel complex to generate the cationic species I, which then coordinates with ethylene to give compound II. This undergoes an insertion reaction followed by an ethylene coordination to produce the butyl-ethylene complex III. By a simple hydrogen transfer, compound III is converted to ethyl-butene complex IV, which in turn undergoes an ethylene displacement to regenerate compound II and gives 1-butene that isomerizes to butene-2. Complex V is displaced by ethylene to regenerate compound II and give butene-2. Thus, two moles of aluminum compound are required for one mole of nickel complex [P571.

8.

L\,,,/Cl L’

EtAICI2> ‘Cl

L\,,,/Et LI

EtAICI2 ‘Cl

w

[

Lb/Et L’

]

:,A,,,

3

I

Figure 1.17. Reaction mechanism of ethylene dimerization by nickel-based catalytic system CP573.

33

When the system operates with Al/N ratios lower than the stoichiometric molar ratios, part of the excess nickel catalyst adheres to the walls of the reactor. The aluminum alkyl halide comes in contact with the adhered nickel complex and activates it rather than forming an aluminum halide coating on the wall of the reactor. The coated halide catalyzes polymerization reaction and leads to a low dimer yield and fouling of the reactor. The effect of Al/N molar ratio on butenes selectivity is shown in Figure 1.18.

X CONVERSIONOF ETHYLENE . PER PASS TO 6UTENES

0

2

Effect of Al/X butene-1 yield [P57].

4 Almi

l

X SELECTIVITYTO'DIWR

A

XYIELD PERPASS

6 6 MCI_RATIO CF INITIAL CONDITIDNINSW REKTOA

10

molar ratio on the conversion,

1:2

selectivity,

and

The catalytic Phillips dimerization process converts ethylene to butene utilizing a nickel based catalyst system according to the preconditioning method described above. The process flow scheme is presented in Figure 1.19. The process consists of three steps: a reaction and two quench steps. In the reactor section, ethylene is fed to the reaction vessel where it comes in contact with a mixture of diluent butenes and the two catalyst components. The catalyst is circulated through the reactor by an external pump. The circulating mixture is passed through a cooler before it enters the reactor. Fresh catalyst components are pumped in by two pumps that are also used to control the molar feed ratio of the catalyst components. The Al/N molar ratio is controlled at a value (0.7-1.0) below that of the stoichiometric ratio (2.01, during the initial startup period and at a higher value (3-5) (at least 10% above the stoichiometric) during the operating phase. Typical reaction conditions

34

Butene-1

I

Ethylene

Catalyst

1

I

kfyd

-p--J Catalyst

kystic Purge

Purge

2

LEGEND RVF PCFigure 1.19. Flow process [P571.

scheme

of Phillips

ethylene

Reactor External Vessel Sand Filter Pump Cooler

dimerization

to butene-1

are 48“C and 13.7 atm and average residence time of the reactants in the circulating loop being 30 minutes. The internal surface area to volume ratio of the reactor system is 51 ft-1. High velocities are maintained in the reactor to minimize fouling and temperature rise across the reactor. The liquid product from the reaction section, containing unreacted ethylene, product butenes, and catalyst, is sent to the catalyst quench section. The catalyst is deactivated by treatment with 2 wt.% acetic acid and separated from The catalyst-free butene stream of the the butene product in an extracter. extracter effluent proceeds to a neutralization vessel where it is contacted with

35

dilute caustic soda solution. The butene stream leaving the neutralization vessel is filtered, distilled, and recovered as product [P571. Table 1.6 and Figure 1.18 show the results of pilot plant scale ethylene Runs 1 and 2, in which high initial dimerization using the above process. AWNi ratios were used, gave a low butene yield. A higher yield and selectivity to dimers is observed in Runs 3 and 4 in which lower Al/Ni ratios were used. 1.5.3 Maasachuaetts Institute of Technology 0 Pnwess A conceptual process for ethylene dimerization in the presence of tantalum or niobium based catalysts has been developed by MIT researchers [3,82,125,P61]. The technology is based on a metal hydride-based homogeneous catalyst that selectively dime&es ethylene to butene-1. The particular catalyst is a neopentylidene complex of tantalum or niobium [3]. A speculative flow scheme for the production of butene-1 via ethylene dimerization is presented in Figure 1.20. The preparation of the homogeneous catalyst is rather a complex process; the tantalum complex is prepared by reacting trineopentyl tantalum dichloride, Ta(CH2CMe3)&!12 and neopentyl lithium LiCH2CMe3 in octane solvent to yield thermally stable neopentylidene tantalum catalyst in quantitative yield [821. Process Description Fresh and recycled ethylene plus an octane solution of the tantalum catalyst are fed into the dimerization reactor operating at 100 atm and 80°C. The dimerization takes place in a homogeneous liquid phase and proceeds rapidly at the rate of one mole of butene-l formed per min per mole of the catalyst [3]. The heat of reaction is removed by excess ethylene leaving the reactor as an overhead stream. Ethylene is cooled and recycled to the reactor along with The reaction is conducted under an oxygen-free, fresh ethylene feed. anhydrous environment to prevent deactivation of the catalyst. The reaction is maintained at 80°C in order to minimize the potential isomerization of butene1 to butene-2 [3, P61]. Ethylene conversion is about 20% per pass and the assumed product mixture from this process is 95 wt.% butene-1 and 5 wt.% butene-2 [P611. Effluents leaving the dimerization reactor are sent to a liquid-vapor separator, where ethylene is separated from butenes mixture. The separator bottoms proceed to the solvent recovery column which produces butenes overhead stream and bottoms solvent stream (containing the catalyst in solution). The solvent stream is recycled back to the reactor while the butenes are sent to an extractive distillation column using the Nippon Zeon extractive distillation process. A high purity butene-1 (99.9%) is produced from the extractive distillation column 133. 1.5.4 Dow Proa?ss A mixture of ethylene and butene-1 is prepared by the dimerization of ethylene in the presence of organic aluminum compound AIR3 in a boiling solvent reaction zone 1P99,PlOOl.

10.6

8.5

2.3

0.8

3.0

1

2

3

4

5

Run No.

Average initial catalyst molar ratio AvNi

56

1115

497

304

494

Total proeess time th)

43

29

31

46

41

21

11

21

21

22

conmt,lons Pressure Temp. (atm) (“C)

..

1.5

3.3

5.1

5.6

9.2

83.6

94.2

83.9

93.9

85.7

Average C2= conversion (%I

with ethylaluminum

Avg. operating catalyst molar ratio AVNi

Table 1.6 Results of pilot plant studies on ethylene dimerization butylphosphineMichloronicke1 [P571

87.1

93.5

89.4

64.3

66.0

m c4+ (%I

12.3

6.5

10.6

35.7

34.0

73.3

88.1

75.0

60.4

56.5

Average g. C4= yield (%I

and bis(tri-n-

select. C6+ (%I

dichloride

Figure 1.20.

SEPARATOR

‘i_tl

-

;+T”,

-

SOLVENT RECOVERY COLU?lN

Conceptual flow scheme of the MIT ethylene dimerization

DltlERlZATlON REACTOR

-_--_ _ _ ----

_-__

)r

I

process [3].

z

Z

C

BUTENE - 2

STH.

EXTRACTWE DISTILLATION

BUTENE - 1

38

FmcessDescription The flow scheme for the production of butene-1 and ethylene mixture via ethylene dimerization is presented in Figure 1.21. High purity ethylene is fed into the dimerization reactor operating at 27 atm. The dimerization takes place in a homogeneous liquid phase of EtA13 and tetradecane solvent. Ethylene flows through a gaseous diffuser or a sparger to disperse or break up the ethylene gas for better contact. Ethylene to Et&l weight ratio is in the range of 4,000 to 8,000. A prepared solution of 0.4 wt.% EtaAl in tetradecane is added to the reactor and maintained at a specific level by liquid level controller [P991. The homogeneous liquid is recycled through a feed reservoir. The catalyst concentration in the dimerization reactor was determined by measuring the amount of ethane liberated by careful hydrolysis of the solvent . aliquots [P991. A conventional heating device is used to heat the liquid-gas mixture to 277°C. At the upper end of the vertical dimerization reactor, a conventional contact device such as mesh packing or ceramic rings is used. The reactor eflluent proceeds to a cooler where dissolved ethylene is separated from the butenes stream. A reflux drum is provided for the condensation of solvent vapor and the liquid is recycled back to the reactor. The pressure in the dimerization reactor is maintained within predetermined limits (27 atm) by a pressure controller valve. The mixture of ethylene and butene-1 proceeds from the reflux drum through the pressure controller valve into the outlet line. After 5 h of reaction time, ethylene conversion reaches 25.7% and the product distribution is mainly butene-1 at 95.5% selectivity, and small amounts of hexenes and other oligomers but without any polymer formation. The unit ratio for the grams of butene-1 produced per gram of Et&l was 159.

1.5.6Comparison of DimerizationProcems The titanium based system used in the commercial IFP-Alphabutol process shows a high butene-1 selectivity associated with minimum formation of byproducts such as cis/trans butenes-2, higher olefins and polymers. The process is best suited for the production of butene-1 because of the simple process configuration needed in the dimerization of ethylene and distillation of the butene-1 product. ,The Phillips process uses a loop reactor and is characterized by the formation of cis/trans butenes-2 due to the isomerization activity of the nickel based catalyst used in the process. Low ethylene conversion and high butene-1 selectivity are obtained in the processes assigned to MIT and Dow. The MIT process uses a neopentylidene tantalum complex at a high operating pressure, while Dow uses a triethyl-aluminum catalyst at high displacement temperatures. The MIT and Dow processes are suited as a first step for the production of linear low density polyethylene by the copolymerization of ethylene with the butene-1 product. Table 1.7 presents the

39

Products

{ethylene and butene- 1)

1 Pf eheater 2 Gaseous diffuser 3 Heating device 4 Level of AIRa in solvent 5 Level ~ontro~lef 6 Control valve

7 8 9 10 11

Control mesh packing tooler Reflux drum Pressure controller Control valve

Figure 1.21. Ethylene dimerization to a mixture of butene-l and ethylene Dow Process I;p991.

277

Tantalumbased

Triethylaluminum

MIT

Dow

*mixture of butenes.

80

Nickelbased

Phillips

48

50&O

Titaniumbased

IFP

T;{*. *

Catalyst system

Process assignee

27

100

13.7

25

PressZe (atm)

25.7

Tank growth reactor is used, Low ethylene conversion

Catalyst preparation is a complex method. Recovery of catalyst is required.

95

95.5

Loop reactor is used. High formation of cis/trans butene-2. Product super-fractionation is needed.

Low isomerization and polymerization activity of the catalyst

Remarks

50-&Y

93

8045 85-W

Butene-1 selectivity (%I

conversion (%I

.. DDe atinp cd t one Ethylene

Table 1.7 Comparison of ethylene dimerization processes

41

operating conditions, ethylene conversion, butene-1 selectivity, and other features of the above mentioned dimerization processes. 1.6

CONCLUSIONS

The analysis of progress made in the field of catalytic ethylene dimerization during recent years shows a trend towards the utilization of titanium complexes associated with aluminum alkyls. The review of publications and patents in the field of ethylene dimerization has revealed that homogeneous titanium based catalysts and various metal complexes of nickel, tantalum, palladium, rhodium and other metals, have been used. However, ethylene dimerization in the presence of these metal catalytic systems, other than titanium, gives a complex mixture of products consisting of butene-1, cis/trans butenes-2, hexenes, octenes, and polymeric by-products. The possible industrial extension of these complexes is still dependent on the improvement of their dimerization rate and butene-1 selectivity. Three essential factors are usually considered in the evaluation of a suitable catalytic system for the dimerization of ethylene: selectivity to butene-1, yield of butene-1 per unit weight of the catalyst, and the required process equipment. This, to a considerable extent, is associated with the fact that the formation of by-products, even in small amounts (2-5 wt.% of the reacted ethylene) requires substantial modification of the entire process. The formation of by-products not only lowers the yield of butene-1 and its purity, but also increases the ultimate cost of butene-1 and reduces the process efficiency. In the case of butenes-2 formation, a costly product super fractionation is needed; periodic reactor cleaning is required in the case of polymeric by-product formation. Among other catalytic systems used in the dimerization of ethylene, only titanium based catalysts and especially titanium alkoxides associated with aluminum alkyls, have reached the stage of commercialization due to the low isomerization activity of these catalysts. Titanium tetrabutoxide, Ti(OBu)b, and triethylaluminum, AlEts, have been found to be ideally suited for the dimerization of ethylene to butene-1, giving high yields of butene-1 at ambient temperature and pressure. Based on the above factors and with a view of enhanced catalyst activity and selectivity, analysis of results in the literature, indicates that the homogeneous system based on Ti(OBu)&lEtg is the most selective one and is readily available among other catalytic systems used for ethylene dimerization. The published literature on this catalyst include the investigation of several parameters such as Al/Ti molar ratio, operating conditions, solvent type, pretreatment in the presence of hydrogen and addition of catalyst modifiers to suppress polymer formation.

42

PART2 2.1

OLIGrOMElUZA~ON

OF ETHYLENE TO cc-OLEFINS

INTRQDUCTION

The oligomerization of ethylene is of considerable academic and industrial interest for the synthesis of linear a-olefins. Numerous publications and patents in this field reveal the scientific and commercial significance of this process [15,1253. The intensive research on organoaluminum chemistry by Ziegler and his coworkers in the 1950’s made it possible to convert ethylene to C4-C2o a-olefins having an even number of carbon atoms. It provided the technical basis for the synthesis of high purity a-olefins in relatively unlimited quantities. However, since the inaugural discovery of trialkylaluminum by Ziegler, there has been considerable sustained interest in the identification of new oligomerization active catalyst complexes. Various transition metal catalysts like nickel, titanium, zirconium, and others have proven to be powerful oligomerization catalysts [15]. Present commercial processes for making a-olefins are based on trialkylaluminum (Chevron and Ethyl) and nickel complexes (Shell). In the past twenty years there has been a rapid increase in the number of patents concerning the catalytic oligomerization of ethylene to linear a-olefins. This increase has been parallel to the extensive research and development of new organometallic complexes as selective catalysts for the oligomerization reaction. Patents in this field were originally devoted to the modification of Ziegler’s trialkylaluminum route to os-olefins in order to produce a selected range Of &OlefinS mainly in the C12-Cls detergents range. Since 1975, a number of patents have appeared, emphasizing the evergrowing industrial interest in the field. Many different types of catalytic complexes have been used, but only a few have been commercialized. Catalytic systems, such as aluminum alkyls have been used in patents assigned to Chevron (Gulf), Ethyl, Dow and Conoco (now Vista); nickel based catalysts have been used by Shell, Chevron (since 19811, Asland and Union Carbide patents; titanium based catalysts have been used in patents assigned to Exxon (Esso), Mitsui and others, while the zirconium based catalysts have been used recently in patents assigned to Exxon, Idemitsu, and Shell. This section provides a comprehensive review of ethylene oligomerization by organometallic catalysts. Related publications and patents to date have been reviewed and evaluated in terms of catalytic systems used, reaction mechanism, product selectivity, and the available process technology. 2.2

INDUSTRY STATUS

Linear a-olefins are currently produced commercially by the following processes 111: 1. Ethylene oligomerization,

43

2. Paraffin

wax cracking,

3. Paraffin

dehydrogenation,

4. Alcohol

and

dehydration.

The ethylene oligomerization and alcohol’ dehydration processes produce high purity even-numbered chains a-olefins while the other two processes Companies produce both odd and even numbered olefin compounds. producing a-olefins via ethylene oligomerization are Ethyl, Chevron, Shell, and Mitsubishi [1,3,141. Chevron shut down its wax cracking olefin plant in 1984 after it acquired Gulf with its ethylene based a-olefin plant [l]. Shell had three wax cracking olefin plants in Europe and at least two of then are no longer operational. The paraffin dehydrogenation process is used mainly by Shell to produce internal olefins for their linear alkyl benzene plant [l]. It is likely that this plant has been shut down after the new Shell plant in Geismar, Table 2.1 presents a list of Louisiana, USA started operation in late 1989. companies having patents in ethylene oligomerization to a-olefins using various catalytic systems. Table 2.1 Companies

having patents in ethylene oligomerization

to a-olefins

Aluminum

Reference

Nickel

Reference

Titanium

Reference

Zirconium

Reference

Chevron (GulD Conoco (vista) Dow Chem. Ethyl Sumitomo

PlOl-113

Amoco

P164

Exxon

P175-178

Exxon

P185-190

P126-128

Ashland

P161,P162

IFP

P180

Idemitsu

P191-195

P121,P122 P114-120 P123-125

Chevron Exxon Phillips

P148-160 P171 P165,P166

Miteui Mobil Societa Italian Union Carbide

P183,P184 P179 P182

Mitsubishi Montecatini Shell

P196 P197,P198 P130-147

P163

Oligomerization of ethylene is currently the primary source of linear aolefins and is likely to be the basis for such production for many decades [ll. The trend away from wax and paraffins processes has occurred for reasons of economics and product quality. Table 2.2 lists the present and future plant capacities for a-olefin production using ethylene oligomerization process. 2.2.1 Chevron Chemical Company In 1966, Gulf (acquired by Chevron in 1983) became the first manufacturer of commercial high purity a-olefins from ethylene oligomerization. Its plant at Cedar Bayou, Texas, produces

even carbon-number

a-olefins

in the range of

44

Table 2.2 Plant capacities of linear a-olefins produced by ethylene oligomerization processes [1,3,141 Producer and location

Capacity*

Initial Operation/yr

Expansion*

Completion date

North America Chevron/Cedar Bayou, Texas EthyVPasadena, Texas ShelVGeismar, Lousiana Shell/Canada

125 400 340 200

1966 1971 1977

125 55 ?A0 Planned

SheIVStanlow, UK EthyVFeIuy, Belgium ChemopetroV Czechoslovakia Nizhnekamsk/Russia

220 200 100 75

1979

50

30 50

1970 1989

Asia Mitsubishi/Japan Idemitsu/Japan *Thousand

1990 1989 1989

1989

1991 1992 1990

metric ton per year.

C4-C2o+ [91. The plant is currently rated at 125,000 ton/year. Chevron further added 125,000 ton/year to the existing capacity in early 1990. In the Chevron process, a-olefins are synthesized from ethylene using the high temperature oligomerization process in the presence of a Ziegler catalyst (triethylaluminum). 2.2.2 EthylCorporation Ethyl Corporation is currently the second largest producer of a-olefins in the United States. Its plant in Pasadena, Texas, is rated at about 430,000 tons/year of a-olefms and 115,000 ton/year of linear primary alcohols [61. Almost 60% of the plant output (Cs-Cl0 olefin range) is aimed for the plasticizer production 181. Ethyl-Belgium has its first European cr-olefin plant in Feluy, Belgium with an initial capacity of 200,000 ton/year. The plant started production in the second half of 1991. Most of the a-olefins produced are in the plasticizer range (Cs-Clo) and polyolefins (C4-Cs). Ethyl’s initial a-olefin production in 1977 (68,000 ton/year) was timed to coincide with the start up of Monsanto’s C7-Cl1 plasticizer alcohol plant. Ethyl’s a-olefin production capacity has increased several times since 1977 (22,000 ton/year>. The process used by Ethyl is a modified Ziegler process which has a recycling scheme that allows short-chain monomers to continue to grow to longer chain lengths as required [1,31.

The development of the Shell Higher Olefin Process (SHOP) by Shell’s research divisions in Houston, Texas and Amsterdam, Holland and the first commercialization of the process at Geismar, Louisiana in 1977, were directed by the requirements of the a-olefins and fatty alcohols markets. The Geismar plant, built mainly to supply 0x0 alcohol market [3,10], had an original capacity rating of 190,000 ton/year which was expanded to 300,000 ton/year in mid 1986 111. Another SHOP plant is located in Stanlow, UK with a design capacity ~220~000 t&year 1111. The SHOP is a combination of various processes, all developed or modified (but not invented) by Shell, which produces a mixture of a-olefins by stepwise chain growth of ethylene using a non-Ziegler catalyst system [ill. SHOP is a an excellent example of the type of creativity that recombines and develops previously available knowledge in ethylene oligome~za~on using complexes of nickel chelates catalytic systems. Both lower and higher a-olefins can be converted to the desired carbon number by isomerization and disproportionation processes C33. 22.4 Mitsubishi Chemid Company ~tsubis~ Chemical Company, Japan has a linear ol-olefins plant based on Chevron/Gulf technology. The company started a-olefin production in 1970 and its current production is rated at 30,000 ton/year 1141. 222.5 0therA.nn0dPlan~

An a-olefin plant, based on Ethyl technology with a capacity of 200,000 ton/year, started operation in Russia in 1990. Another plant, being built based on Chevron technology for Chemopetrol, Czechoslovakia with a planned rate of 120,000 ton/year, is scheduled to start operation in 1992. Idemitsu, Japan has started a-olefin production at a rate of 50,000 ton/year in early 1991. SABIC, a major ethylene producer, has announced its interest in building a plant for the produc~on of a-olefins 1121. At various times, a-olefin plants reportedly have been considered for Australia, India, Japan, Canada, Mexico, New Zealand, and other countries, citing among other things, the low-cost ethylene and captive markets as incentives [ll. To date, many more plants have been considered and even announced than have been built.

23

APPLICATIONS OF ALPHA-OLEPINS

Alpha olefins are versatile chemical intermediates for a wide variety of industrial and consumer products and their use has been advanced for many applications. Tables 2.3 and 2.4 show different a-olefin markets and the end uses of each olefin cut 111. The major use of a-olefins is either in su~a~~ts (detergents) or polyethylene comonomers. The development of biodegradable

46

Table 2.3 Applications of a-olefins [ 101 I (X-Olefin Carbon Number

I

I Reactant

I

I Product

Applications Polybutylene HDPE,LLDPE Esters PVC plasticizers Synthetic lube oils Diols. hydroxyethers and amines

Herbicides and plastics Acids, detergents, alcohols Surfactants Surfactants Surfactants Surfactants. lube oil additives Amines

Cationic surfactmts Amines, amides

Paper sizing, leather, preservatives

Antistats

Table 2.4 Alpha-olefin markets* 1 .,141 Application

Market share [%)

Detergents LLDPE Plasticizer HDPE Polyalpha olefin Lube oil additives Fatty acids Tert amines Polybutene * Estimated total a-olefin production European countries).

38 30 8

7 7 4 2 2 1 was 1.2 million

or-Olefin Cut

h-h3 c4--c8 CS-cl0 c4-c8 Cl0 c12-Cl8 hZ--cl6 c12-cl6 c4 metric ton in 1990 (excluding

East

47

surfactants for laundry detergents has been facilitated by the availability of aolefins. These cost-efficient derivatives mean effective cleaning in their enduse applications with minimal environmental impact on waterways in highly populated areas [1,8,11]. Tougher linear low density polyethylene plastic films containing a-olefins comonomers mean less damage to goods in shipments [l,lO]. Other applications of a-olefins are in the fields of plastics, synthetic lubricants, polymer additives, and many other industrially useful chemicals. Alpha olefins used in detergent applications have higher market value than those used for plasticizer alcohols. Hence, research has continuously sought to produce selectively the a-olefins of the most preferable carbon numbers [l&131. 2.3.1 De-u? The use of a-olefins in detergent manufacture

is mainly in the production

0x0 alcohols, linear alkyl benzene (LAB), and cc-olefin sulfonates

(AOS).

of

The

growth in this application is occurring as ol-olefin derivatives replace other less biodegradable surfactants [ 11. Alpha olefins are hydroformylated to primarily give straight-chain alcohols with odd numbers of carbon (Cl1 or more). The detergent alkylates are another product arising from the higher olefins. The reaction with aromatic compounds gives secondary alkylates. Alpha olefins also provide the detergent market with AOS mostly derived from C14-Cl8 feeds by treatment with sulfur trioxide and subsequent hydrolysis. AOS detergents provide superior performance in cold hard water, and are mainly used in light duty detergent applications. A total of 480,000 ton/year of ol-olefins have been used to manufacture detergents during 1989 [l].

Alpha olefins (butene-1, hexene-1 and octene-1) are used as comonomers with ethylene and other monomers in relatively small quantities (615%). They improve the stress cracking resistance, melt strength, and rigidity properties that are particularly desirable in blow molding grades of polyethylene 111. An addition of 3% hexene-1 to ethylene produces a high density polyethylene (HDPE) with a better stress/crack resistance than the homopolymer 1111. Butene-1 and hexene-1 are preferred comonomers for gasphase LLDPE processes while octene-1 is mostly used in liquid-phase LLDPE processes. Alpha-olefins depend, to a large degree, on the rate of LLDPE growth, the amount of comonomer used, and which one of the three comonomers, butene-1, hexene-1 or octene-1, are most useful. The LLDPE share in the combined LLDPE and HDPE market increased from 18% in 1985 to 30% in 1990. This increase required 400,000 ton of C4-C!8 a-olefin at 8% of LLDPE-comonomer 111. 2.3.3 Other Applications Alpha olefins (Cs-C1o) can be hydroformylated to give primarily straight chain alcohols with odd carbon numbers (Cl1 or less), which are then reacted

with phthalic anhydride or other acid to produce esters [ll]. The ester is used as plasticizing agent in flexible poly(viny1 chloride) (PVC). Alpha olefins also react with carbon monoxide to give odd numbers carboxylic acids which can be converted to esters, useful as lubricant additives, or used as the free acids in lubrication [ill. Mercaptans and sulphides can be produced from cr-olefins, mainly C4-CU. These low sulfur compounds are useful intermediates in the production of some specialty chemicals and as chain-transfer agents in emulsion polymerization and the production of rubber and detergents [ll]. Similarly, a number of nitrogen containing compounds, often with surfactant properties, can also be made from a-olefins. Oligomerization of some a-olefins and especially Cl0 (decene-1) produces a lubricating oil base stock. Although more expensive than mineral oils, such synlubes have some advantageous properties, particularly at very low temperatures Elll. Other miscellaneous uses of a-olefins include paper, fabric, leather, enhanced oil recovery, waxes, and fuel additives [ll. 2.4

REACTION CIBXISTRY

AND KINETICS

The principles of ethylene oligomerization and its chemistry are common to all processes producing a-olefins. The dimerization mechanistic aspects, presented in Part 1, apply to ethylene oligomerization reactions. However, catalytic oligomerization can be represented as a combination of chain growth and displacement reactions that take place at an org~ome~llic site (M-R). The chain transfer to monomer occurs randomly and at a sufficient rate to produce low molecular weight olefms [15]. The conventional growth and displacement reactions are represented by the following reactions 11261: Chain Growth (2.1)

M-R + nCHz = CH2 + M~CH2CH2)~R Displacement [email protected]&,

R + CH2 = CH;? -+ M-CH2CH3 + CH2 = CH(CH$JH$,_l R (2.2)

The displacement reaction may also take place in two steps: M--tCHxCH& R -+ M-H + CH2 = CH(CH&Hz),.l M-H + CH2 = CH2 + M-CH2CH3

R

(2.3) (2.4)

The “hydride insertion-alkyl migration” mechanism seems to be the generally accepted one for ethylene oligomerization. Figure 2.1 presents the geometric distribution of products by the metal-hydride mechanism 1151. The

49

LnM-H

-

JI c=c

b”-

C2H5

%P--c&3

A

LnM-H

+ butene-1

t A

II

L,M-H

+ hexene-1

nC=C

LnM-R

T A

L,M-H

A

L,M-H

+ higher cc-olefins

nC=C t

II LnM-R

Figure 2.1.

Geometric

distribution

+ polyethylene

of products

by metal-hydride

mechanism

Ml. direction of ethylene addition to a metal-carbon or a metal-hydrogen bond is very important for obtaining linear or branched products. Types of additions are known as Markownikov and anti-Markownikov for linear and branched products formations, respectively [Xl. The course of reaction can be significantly influenced by the nature of the metal, the ligands applied, and the reaction conditions. ReadZioIlKinetic9 alldProductDiatritnltion Kinetic studies on ethylene oligomerization are conducted to calculate the rate constants of the propagation and chain termination reactions [1271. Experimental data from different catalytic systems are used in the kinetic evaluation.

50

Ethylene polymerization, in general, is characterized by the absence of chain termination reactions (the length of the molecules are determined by chain transfer). The molecular weight determining reaction in polymerizaHowever, in ethylene oligomerization, the tion is the P-hydrogen abstraction. relative frequency of the P-hydrogen abstraction has been pushed far enough to obtain oligomers instead of the high molecular weight polymers. Chain propagation and chain termination reactions have been considered in the following kinetic evaluation of ethylene oligomerization 11271. The rate of chain propagation is given by: (2.5)

-rp = kp [C*l [Ml

where kp is rate constant, [C*l is concentration of active catalyst sites, and [IVII is concentration of monomer. The rate of chain termination may be either: rt = kt [C*][M] or

(2.6a)

rt = kt [C*l

(2.6b)

depending on whether the monomer is involved in the rate determining step of the chain transfer (equation 2.6a) or not (equation 2.6b). Evidently, the overall kinetics depend upon this mechanistic feature. Where equation (2-6b) is valid, the monomer is used up by chain propagation and termination steps. Therefore, the overall rate of monomer consumption is: (2.7a)

-d[MYdt = (kp + kt) [C!*l[Ml If equation (2.6b) is valid, then the rate of monomer consumption -d[MYdt = k, [C*l[Ml In Flory given takes

is: (2.7b)

either case, the most probable distribution of molecular weights (Schulzdistribution) is to be expected. The probability of chain growth (K) is by equations (2.8a) and (2.8b) depending on whether or not the monomer part in the determining step of the chain termination 11271.

K = rp/(rp+ rt) = [l + (kt/k,)l-1

(2.8a)

K = r&r,+

(2.8a)

rt) = [l + (ktflr, [M1)l-1

The weight fraction (Mp) of the olefin products as a function of the degree of oligomerization (N) is given by the following equation: Mp = (InzK)(N)(KP) A plot of product weight distributions presented in Figure 2.2.

(2.9) (Mp) versus the growth factor K is

51

0.6 Growth factor, K Figure 2.2.

Alpha-olefin

product distribution

0.6

versus growth factor (K) 11281.

Thus, an evaluation of the molecular weight distribution of ethylene oligomerization products provides an independent way of distinguishing between the two proposed reaction schemes (equations 2.6a and 2.6b) The individual rate constants kp and kt can be calculated from data on overall rate and weight fraction (equations 2.8 and 2.91, provided that the concentrations of monomer [Ml and active sites [C*l are known [127]. Table 2.5 presents the estimated rate constants of ethylene oligomerization with different catalyst systems. The rate constants were determined from experimental data using the above mentioned approach C1271. On evaluating the results presented in Table 2.5, it seems that the temperature was the only variable for the adjustment of the growth factor (K) and hence for the optimization of the molecular weight range (equation 2.8 and Figure 2.2). The results indicate that the titanium catalysts require temperatures less than -20°C in order to become active as oligomerization systems, whereas zirconium catalysts work satisfactorily at 40-80°C. The great advantage of the zirconium based catalysts was that the reaction rate was two orders of magnitude higher than titanium systems. The apparent rate constants for the nickel systems, on the other hand, are extremely low even at high temperatures. This might be due to catalyst depletion by decomposition of the complex and precipitation of Ni(0) sponge [1271. Alpha olefin product distribution in catalytic ethylene oligomerization is usually governed by the Schulz-Flory type distribution obtained by the following equation [31:

52

Table 2.5 Estimated rate constants systems Cl271 Transition metal

Catalyst system*

Ti Ti Ti ZT ZT Zr Ni Ni

A A. B C D E F G

*A B C D E F G

= = = = = = =

of ethylene

T

P

(‘0

(atm)

-45 -20 -20 80 55 40 75 169

12 12 11 l0 35 ll 39 42

oligomerization K 0.74 0.82 0.75 0.78 0.68 0.69 0.70 0.85

k+t 2.8 4.5 2.9 3.5 2.1 2.2 2.3 5.6

with different

catalyst

k, (mol-see-l)

kt

8x10-3 6~10-~ 7x10-2 2 2 4 2x104 2x10-2

3x103 1.3x10-2 2x10-2 0.6 0.8 2 lx104 3x103

(mol-sec-l)

(OEt)3TiCl/EtAlC12/toluene TiClq/t-BuOtt/Et3A12Cl3/chlorobenzene Zr(OPr)4/Et3Al2C13/toluene Zr(OPr)4/Et3A12Cb$heptane Bz4Zr/lQAl2C13/toluene (Bz = benzyl) Bis -1,5+yclooctadiene nickel (O)/diphenyl carboxy methylphosphin&enzene. Bis-1,5-cyclooctadiene nickel (O)/diphenyl (N,N-diphenyl-carbamyl methyl) phosphinetbenzene.

where K = growth displacement, Kcl),

factor (ratio of the rate chain growth to the rate of Cn = number of molecules of a-olefins with chain length

of n, and C&+2) = number of molecules of a-olefins with chain length (n+2), the next member of the series. The product distributions attained in industrial ethylene oligomerization processes are quite close to the theoretical distributions. A calculated weight distribution

of product a-olefins is presented in Table 2.6 and Figure 2.2 as a

function of K (K=0.4-0.85). A typical industrial ethylene oligomerization aolefin plant selects conditions to give K in the range 0.7-0.8. The K factor can be varied by adjusting catalyst composition and reaction conditions [1281. Alpha-olefins obtained by large quantities of catalyst (stoichiometric reaction) follow a different product distribution governed by the Poisson pattern of distribution [33: XiNi a(-N/i!)

(2.11)

where Xi = mole fraction of product containing i units of monomer (ethylene), with carbon chain length of 2i, and N = arithmetic average number of monomer (ethylene) units attached to each atom of the organometallic catalyst after completion of growth (degree of oligomerization). The Poisson type distribution of cc-olefins is obtained in a two-step ethylene oligomerization method, separate growth, and displacement steps. Displacement can be greatly reduced during the growth step by limiting ethylene

53

Table 2.6 Theoretical weight distribution of alpha-olefins produced by single-stage ethylene oligomerization (Schulz-Flory type) [31 Chain length c4

C6 G3 Cl0 Cl2 Cl4

Cl6 Cl8 C20 C22 C24 C26 C2a C30 C30+ Total

0.4

0.5

45.00 27.00 14.40 7.20 3.46 1.61 0.74 0.33 0.15 0.06 0.03 0.01 0.01

33.33 25.00 16.67 10.42 6.25 3.65 2.08 1.17 0.65 0.36 0.20 0.11 0.06 0.03 0.02 100.00

100.00

0.6

(M 0.65 0.7 (wt.%)

22.86 18.15 20.57 17.69 16.46 15.34 12.34 12.46 8.89 9.72 6.22 7.37 4.27 5.47 2.88 4.00 1.92 2.89 1.27 2.07 0.83 1.47 0.54 1.03 0.35 0.72 0.22 0.50 0.38 1.12 100.00 100.00

13.85 14.54 13.57 11.87 9.97 8.14 6.52 5.13 3.99 3.07 2.35 1.78 1.34 1.01 2.87 100.00

0.75

0.8

0.85

10.00 11.25 11.25 10.55 9.49 8.31 7.12 6.01 5.01 4.13 3.38 2.75 2.22 1.78 6.75 100.00

6.67 8.00 8.53 8.53 8.19 7.65 6.99 6.29 5.59 4.92 4.29 3.72 3.21 2.75 14.67 100.00

3.91 4.99 5.65 6.01 6.13 6.08 5.90 5.65 5.33 4.98 4.62 4.26 3.90 3.55 29.04 100.00

concentration and by operating between 100 and 12O’C. The Poisson distributions of a-olefins can be calculated for any chosen value of the average degree of oligomerization, N. Table 2.7 presents a calculated weight distribution of product a-olefins as a function of N (N=4-7). A typical industrial two stage process is adjusted to give an approximate average degree of oligomerization, N=5 131. Figure 2.3 compares the theoretical weight distributions of a-olefin products calculated for the Poisson distribution and geometric (Schulz-Flory) distribution types. A generalization of the Schulz-Flory statistics was proposed by a mathematical model of ethylene oligomerization [1291. The model was able to predict the distribution of all the products from the adjustment of a limited number of kinetic parameters. 25

CATALYTIC SYSTEMS

A variety of catalytic systems oligomerize ethylene to higher molecular weight olefins. The formation of linear a-olefins is possible when the catalyst is selective for ethylene only. Cooligomerization with product olefins can lead to branching, depending on the mode of addition [VI]. Three general classes of ethylene oligomerization catalysts have been reported in the literature, those based on Ziegler types such as trialkylaluminum, Ziegler-Natta types such as transition metals with reducing agents or

54

Table 2.7 Theoretical weight distribution of alpha-olefins produced by two-stage ethylene oligomerization (Poisson distribution) [31 . . . Avem of ol-on CN) Chain length

4

4.5

5

6

6.5

7

1.49 4.47 8.95 13.42 16.10 16.10 13.80 10.35 6.90 4.14 2.26 1.13 0.52 0.23 0.14 100.00

0.98

0.64

bv”t::, C4 C6 C8 Cl0 Cl2 Cl4 Cl6 Cl8 020 C22 C24 C26 C23 C30 Above C30 Total

7.46 14.93 19.90 19.90 15.92 10.61 6.06 3.03 1.35 0.54 0.20 0.07 0.02 0.01 100.00

5.06 11.37 17.06 19.19 17.27 12.96 8.33 4.69 2.34 1.05 0.43 0.16 0.06 0.02 0.01 100.00

3.39

8.48 14.13 17.67 17.67 14.72 10.51 6.57 3.65 1.83 0.83 0.35 0.13 0.05 0.02 100.00

2.26 6.21 11.38 15.65 17.21 15.78 12.39 8.52 5.21 2.86 1.43 0.65 0.28 0.11 0.06 100.00

3.18 6.89 11.20 14.56 15.77 14.65 11.90 8.59 5.59 3.30 1.79 0.89 0.41 0.30 100.06

2.24 5.22 9.13 12.78 14.91 14.91 13.05 10.15 7.11 4.52 2.64 1.42 0.71 0.57 100.00

Lewis acids, and those that can be deduced from the metal complexes (one component catalysts). The Ziegler-Natta types utilize compounds of three transition metals - titanium, zirconium or nickel [125,1301. Among Ti derivatives the most frequently used is TX14 in combination with alkyl aluminum halides. Zirconium compounds, such as Zr(OR)4 in combination with ethylaluminum sesquichloride (Et3AlzC13), oligomerize ethylene to a mixture of pure linear a-olefins. Nickel derivatives used for ethylene oligomerization include nickelocene, nickel acetylacetonates, and ylides. The use of complex nickel catalysts as a non-Ziegler-Natta type, has by far attracted the greatest interest among other transition metal complexes [15]. The main deficiency of the other catalytic systems is polyethylene and branched oligomers formation as side products. The catalytic systems used in ethylene oligomerization are classified in this section according to the type of active organometallic species used in the catalytic system. The catalysts are discussed and reviewed in terms of their methods of preparation, activity, reaction conditions, catalyst-product separation, product distribution, proposed reaction kinetics, and mechanisms. 2.5.1 Aluminum Alkyl Catalysts The discovery by Ziegler that aluminum alkyls could be synthesized directly from aluminum, hydrogen, and olefins, increased interest in aluminum

55

Typical Weight Distribution of Alpha-Olcfins from Combined Growth/Displacement Reaction, with Growth Rate/Displacement Rate Ratio = 0.7

_____ -I-

Typical Weight Distribution of Alpha-Olcfiis From Two-Stage Reaction. Poisson Distribution with Average Monomer Units per Chain N 8 5.0

I,

I-

/

I-7

4

I

I

6

8

I

I

I

I

10 12 14 16 NO. OF CARBON ATOMS IN CHAIN

Figure 2.3. ‘I+heoretical weight distribution Poisson distribution types> 131

,

18

I

20

of alpha olefins (Schultz-Glory

and

56 alkyls as intermediates in large volume industrial processes E131,1321. T~ethyl~u~num (Et&), which is used in the commercial processes for the synthesis of a-olefins, is readily prepared on commercial scale by a two-step reaction process as follows: Hydrogenation:

2Et& + Al + 3f2 Hz + 3EtzA1H

(2.12)

Ethylation~

3EtzAlH + 3C2& -_) 3Et&l

(2.13)

Aluminum alkyls react with ethylene in catalytic reaction (Schulz-Flory type distribution) and stoichiometric (Poisson type distribution) yielding growth products, The rate of addition of ethylene molecules is first order in both ethylene and R3Al monomer. The reaction proceeds by the formation of a coordinate bond between the ethylene double bond and the unfilled octet of the aluminum atom in the monomer with subsequent electron rearrangement leading to the insertion of ethylene between the aluminum-carbon bond of the aluminum alkyl. In catalytic reactions, the EtaAl concentration is made low enough so that there is no economic need for its recycle. It can be removed from the product olefin stream before distillation by hydrolysis with aqueous NaOH or aqueous H&J04 11311. Typical reaction conditions are 176-287% and 140-275 atm which favor both displacement and chain growth reactions El331 as shown in Figure 2.4, The disadvantage in this approach is that it piovides small amounts of the more valuable Cl2 a-olefin and most of the product is C4-Cl0 olefins with C4-Cs predominating [X311. In stoichiomet~c reactions, the amount of aluminum alkyd used is large and there is a need for its recovery. However, the recovery is extremely dif& cult since the boiling points of Et3Al and dodecene-1 (Cl21 are quite close together 11321. This problem is usually referred to as the Ziegler dilemma since Ziegler’s group worked for many years without finding a commercially practical solution for this problem. A laboratory scale solution was carried out by Ziegler through the complexation of aluminum alkyls with metal salts (MX) like NaCl, KC1 or NaF [132,1341. Such metal salts form two different complexes such as 1:l complex, MX:RaAl and 1:2 complex, MX:BRsAl. Thus, the addition of 1:l complex to the EtsAl-olefin mixture, leads to the formation of 1:2 complex and the olefin mixture can be readily separated leaving a residue of MX*2Et& The 1:2 complex is then distilled under quite low pressures (10-S bar) to remove one mole of Et3Al for recycling to the growth step leaving the 1:l complex to use in the next separation cycle. The complexation approach is represented by the following reactions t1311: Chain Growth (2.14)

Et&l + nC2H4 -+ (RCH2CH&U ~~s~~ace~ent 3RCH = CH2 + Et3Al -I-MX. Et&l +

3RCH = CH2 + MXe2Et3Al

(2.15)

57

Ethylene recycle I

Olefins + Catalyst

catliilyst I

Olefin product

Catalyst’residues

Figure 2.4. Schematic ffowsheet of one-step ethylene oligomerization method

l3m. E&Al Regeneration MXq2Et3Al +

lEtsA + MX*Etfl.

(2.16)

Another approach to the Ziegler dilemma was to carry out the ol-olefin synthesis in two stages as shown in Figure 2.5 11331. In the first stage, Cs-Cl0 a-oh&ins are produced by a conventional chain-growth/displacement loop with the separation of EtaAl from the C&lo olefins after displacement. Typical reaction conditions for this stage are 93120°C and 100-200 atm. The Cs-Clo aolefins enter a second chain-growth/displacement loop in which Cc-C 10 aluminum alkyls are converted to C&-Cls aluminum alkyls at a temperature range of 260-315*C and 7-20 atm. The Cs-Cl0 a-olefins are used to displace C12-Cla a-olefins from the chain growth product. These C&-C!ls olefins can be readily distilled from the Cc-Clo aluminum alkyls produced by displacement, while the C&lo alkyls are recycled back to the chain growth [131X The advantage of this approach is that it provides a much greater control of the chain length distribution of the a-olefin product compared to the single chain-growth/displacement loop. The use of glycols and ethanolamines were investigated in the separation of alkyl aluminum (Bu&l) and ethylene oligomers. The transalkylation of the a-olefin product with methylstyrene and treatment with glycols (propylene, triethylene or ethylene glycol) produced a solid, precipitate that was removed by filtration. Treatment with ethanolamines (mono-, di-, or trie-than01 amine) produced a viscous oil that was insoluble in organic solvents and was separated by decantation,

58

Ethylene feed

A A

Displacement reactor Displacing olefin make-up

Figure 2.5. [133].

Schematic

flowsheet

of two-step ethylene oligomerization

method

The kinetics of the chain growth and displacement reactions of ethylene oligomerization by AIR3 systems were studied using a reactor model to simulate a bench scale plug flow reactor 11331. The model was used to predict the reactor’s output composition on the basis of 64 possible simultaneous reactions (8 growth and 56 displacement). The kinetics of the chain growth reaction were taken from literature [1101, while the kinetics of displacement were determined by the displacement of tributylaluminum by octene-1 in 1 cm3 plug flow reactor and the .displacement of trioctylaluminum by hexene-1 in a 150 cm3 continuous stirred tank reactor (CSTR). The Arrhenius expressions for the growth and displacement constants were found to be [110,1331: known

=

(8.61005) exp L-6.279 x 107/RTl

(2.17)

kdisplacement

=

(1.667)(1012) exp I-l.2988 x lO*/RTl

(2.18)

Ethylene oligomerization experiments were conducted in a bench-scale plug flow reactor in the presence of AIR3 as shown in Figure 2.6. The reactor (175 cm31 consisted of 0.95 cm coil carbon steel tubing, 6.2 m in length and 0.6 cm internal diameter. The coil was immersed in a constant temperature

59

~

Pelargonic acid

To G.C. and receiver Catalyst feed

Heating oil (Essotherm N69) 0.6 cm ID 622 cm carbon steel coil

I Figure 2.6. Diagram ization 11331.

Electric heaters 2

of a bench scale reactor

system for ethylene

oligomer-

oil bath and maintained at llO-13O”C, 100 atm and 5.7-6.5 min. reactor holdup. Ethylene was heated to 100°C and mixed with AIR3 at the inlet of the reactor at a molar ratio of 0.16-1.83. The reactor eflluent was treated with pelargonic acid to liberate the alkyl group from the aluminum alkyls in the form of alkanes before gas analysis by GC. The measurements of the product composition (Cs-C1s) from the bench scale reactor were found to compare favorably with the simulation results. However, differences between the experimental and theoretical results were attributed to the presence of nickel impurities in the reactor tubing and fittings which increased the displacement rate in the experimental reactor. The predicted results for the two types of catalysts used, trihexylaluminum and tributylaluminum, were satisfactory from an industrial point of view and could be used as the basis for a preliminary a-olefin plant design [1331.

60

Although all the ethylene is essentially oligomerized to C4-Czo+ a-olefins in the presence of AlEt (200°C and 34 atm), a very small amount of solid polymer is unavoidably formed during the reaction. This polymer is undesirable because it tends to deposit on the reactor tube surface, interferes with heat transfer and requires periodic reactor shutdowns for cleaning [P1021. A test procedure was devised to accelerate the production of solid polymer during ethylene oligomerization and to provide an indjcation of the effectiveness of inhibitors on polymer formation [PlOl-P105,P1121. A commercial colloidal alumina was added to the AlEta solvent mixture (6.0 wt.%) to increase the polymerization. The reaction between AlEt and alumina on metal surfaces formed a co-catalyst which reacted with ethylene to produce the solid polymer. The reactions were carried out in a one-gallon autoclave equipped with a stirrer, a bottom outlet line, inlet lines for adding ethylene, a thermowell for a thermocouple to measure temperature and a pressure gauge outlet. The temperature was maintained at 200°C, pressure of 34 atm, and 8 h reaction time in cyclohexane solvent. The amount of ethylene reacted was determined by the difference in the amount of ethylene metered into the autoclave and the amount of ethylene discharged at the end of the test period [P1121. The amount of polymer formed during the reaction was collected and separated from the product olefins and solvent. The effectiveness of various polymer inhibitors was determined by comparing the obtained results with a similar test where no additive was used. The results are presented in Table 2.8. The data presented in Table 2.8 show that the lubricating oil (containing 57 ppm nitrogen and 218 ppm sulfur) reduced polymer formation more than any other inhibitor tested, and elevated catalyst efficiency. The lubricating oil provided the best combination effect of all materials tested, considering both reduced polymer formation and increased catalyst efficiency. Table 2.8 Effect of additives on polymer inhibition and catalyst efficiency during ethylene oligomerization in the presence of AlEt and colloidal Al203 (200 “C, 34 atm and 8 h in cyclohexane)

Run #

Inhibitor None Lubricating Oil * Lubricating Oil * Phenothiazine Phenothiazine 2-Mercaptobenzothiozole Diphenyl amine Dodecyl sulfide

Inhibitor Amount (g)

20 ml 40 ml 0.25 0.5 0.1 0.2 10

* Based on ethylene reacted. t Grams of ethylene per gram of AIEt3. t Contains 57 ppm nitrogen and 218 ppm sulfur.

Polymer cata1ystt produced (ppm)* efficiency [email protected] 325 49 231 232 91 526 239

110-120 126 x33 89 109 109 162 129

Reference 11 P112 P112 P112 P104 P104 PlOl P103 P102

61

A series of ethylene oligomerixation tests were performed to determine the effect of sulfur concentration in a lubricating oil upon polymer formation and catalysts efficiency [P112]. The tests were conducted in the presence of 6.0% Et&l (based on solvent) in a tubular reactor immersed in a bath of boiling water maintained at 202°C and 230 atm. Various lubricating oils derived from natural crude oils, which had been hydrotreated to various sulfur levels, were used in the oligomerization tests A blank test was performed in a sulfur-free solvent such as a c22_c36 recycle stream of a-olefins. The results, presented in Table 2.9, indicate that the lubricating oil containing 4,800 ppm sulfur reduced polymer formation compared with the sulfur-free solvent; however, it disadvantageously reduced the catalyst efficiency. The other lubricating oils containing 50-1700 ppm sulfur did not show any reduction of polymer formation compared with the sulfur-free solvent [P1121. Table 2.9 Polymer inhibition in ethylene oligomerization by EtsAl in the presence of sulfur-containing lubricating oils (202”C, 230 atm) [P1121 t Run #

Type of lubricating oil

1 2 3 4 5 6

C22-C36 U-Okfin Heavy neutral Bright stock # 150 Double bydrotreated neutral Extract of heavy neutral Bright stock # 200

Feed (wt/%I

Sulfur (ppm)

4.3 4.7 4.3 4.7 6.0 4.1

250 500 50 4,300 1,700

Polymer formation (ppm) 1$20 495 622 552 m, 439

Catalyst efficiency* Xl0 196 231 213 153 230

*Grams of ethylene per grams of Et3Al.

The effects of unsaturated hydrocarbons (butadiene, dicyclopentadiene, isopentene and isobutylene) on solid polymer formation and catalyst efficiency, during ethylene oligomerization to ol-olefins, were studied [P1051. The tests were conducted in a 15-liter tubular reactor of 166 m length and 1 cm inside diameter. Small amounts of unsaturated hydrocarbons (0.15 wt% based on total charge) were added to Et&l-cyclohexane solvent maintained at 202°C and 265 atm. The results are presented in Table 2.10. An appreciable reduction in polymer formation was achieved in the presence of a controlled amount of dicyclopentadiene. The degree of ethylene conversion to a-oleflns was determined using two different methods [P106,P107]. In one method, ethane was injected at a ratio of 1:306 based on ethylene in the feed charges and the ratio of inert gas to unreacted charge in the gaseous phase of the product was taken as the degree of ethylene conversion. It was not necessary to separate the gaseous and liquid products to measure the extent of ethylene conversion. Thus, a feed charge containing 1:300 ethane-ethylene mixture was converted to a-olefins. The product contained 1:lOO of ethane-ethylene mixture indicating that 66.7%

62

Table 2.10 Effect of unsaturated hydrocarbons on polymer formation during ethylene oligomerization in the presence of Et3Al and cyclohexane (202°C and 265 atm)

I?1051

Polymer Run #

inhibitor

1 2 3 4 5 6

No additive Butadiene Butadiene Dicyclopentadiene Isopentene Isobutylene

Inhibitor amount (%I

Ethylene conversion (%I

Reaction time (hr)

Polymer formation (mm)

Catalysts efficiency*

0.97 0.45 0.62 2.21 4.48

63 61 66 66 64 66

21 xl SS4 32 23 24

281 87 198 108 235 189

112 98 112.8 84.7 97 109

*Grams of ethylene per gram of Et3AI.

ethylene was converted to a-olefins [P106]. In another method [P1073, the extent of ethylene conversion was determined by measuring the amount of liquid dodecene-1 (Cl21 formed. The amount of Cl2 in the total olefin product is constant (10 wt.%) over the temperature range 198.9-205°C. The total weight of liquid and gaseous a-olefins was calculated by multiplying the weight of the product-sample by the percentage of Cl2 in the liquid product-sample and then multiplying the result by 10. The ratio of total cc-olefin products to Et&l in the liquid product divided by the ratio of ethylene feed to Et3Al f ?ed, multiplied by 100, gave the extent of conversion. The advantage of this method is that it does not require the analysis of the total a-olefin products lP1071. Following the oligomerization reaction, the Et3Al catalyst was destroyed without recovery [P102,P108,P1091. The catalyst was deactivated by contact with sufIlcient acid, base, water or alcohol to react stoichiometrically with the catalyst. When water was added for deactivation, the alkyls were converted by a hydrolysis reaction to paraffin which remained as impurities in the product olefins CpllOl. The paraffin produced during catalyst deactivation remained as impurities in the range of 1.4-1.7 wt.% in each product olefin. Another disadvantage of using water for deactivation was that the recycled ethylene had to be dried before its contact with Et3Al in the oligomerization reactor [P102,PlOSl. The paraffin content of the produced a-olefins was reduced by subjecting the olefin products to a temperature greater than 260°C several seconds before deactivating of the catalyst [p109]. The ethane content in the aolefins dropped from 1.81 wt.% (without heating) to 0.36 wt.% when the crude a-olefin products were heated for 1.17 set at 315.6”C and 40.8 atm prior to deactivation of the catalyst. 2.5.2 Nickel Baaed Catalysts The criterion of activity in the nickel catalyzed oligomerization of ethylene appears to be the generation of nickel hydride intermediate complex.

63

Therefore, it is not surprising that, under suitable conditions, particularly every type of nickel complex is active and as a result, a large number of publications have appeared. In general, the nickel component is often a nickel salt modified by Lewis acid and tertiary phosphine. The oligome~zation reaction can be carried out in solution or in heterogeneous phase. The homogeneous catalyst has full activity only in the presence of a Lewis acid and hyd~~rbon solvents or alcohols_ Polar solvents are preferable since they are able to dissolve ionic intermediates in the absence of a Lewis acid, but at a reduced rate 11361. Supported nickel catalysts may be prepared in a conventional way by depositing nickel salts on the support, such as SiOs, AlzQ and others, and treating it with the Lewis acid or hydrolysis of an organonickel with the support. The kinetics of ethylene oligomerization reaction involving various nickel catalysts have been studied. In general, the reaction rate is first order with respect to catalyst concentration and second order with respect to ethylene concentration. The presence of phosphines leads to an acceleration of the oligome~zation rate. The enhancement was found dependent on the phosphine nature and ~on~entra~on. Shell commercial process, SHOP, utilizes a non-Ziegler homogeneous nickel ligand catalyst. The process/catalyst is selective to mono olefins with traces of only ~desirable products, such as dienes and paraffins. The boot of linear a-olefins is in the range of 96-97% for all carbon mimbers Cl371, K&m et al. E135,13?-1403 investigated a rmmber of unicomponent catalysts chosen from among square planar nickel that possess a five-member chelate ring containing at least one phosphorous atom. Nickel complexes were readily prepared by the reaction of benzoylmethylene, t~phenyl-phospho~e, and PPh3, as presented in Figure 2.7. The phenyl transfer from phosphorus to nickel is a novel reaction and so far the detailed mechanism is unknown [1371, The high oligome~~ation activity of complex I (Figure 2.7) was demonstrated between 30”12O*C,and 9.9-98.7 atm, giving 399% linear defins having >98% CIolefins and conversion between 10 and 100% [137,1391. The convers~en of ethylene and product distribution was controlled by adjusting reaction temperature and pressure as well as the concentration of the phosphine. The activity of this complex was ~05~ mol ethylene per mol of complex. The experimental data showed that the use of n-hexane as suspended medium resulted in a high molecular weight linear polyethylene formation, The s~thesi$ of ENi~~~C~H~~~l[Ph~PC~~C001Ni as a catalyst for the SHOP oligomerization process was reported [140,139]. Two isomeric forms were identified in teluene solution, (CsH13) ligand being 4-enyl (1-a) and 40% q3 ally1 (I-W as shown in Figure 2.7, The catalyst was tested under a wide range of operating conditions C75-9!Y’Cand 9.9-98.7 atm) yielding 99+% linear oleflns ha&g a-alefins in the range of 93-99% at activities of 0.6 mol of ethylene per mol of Ni-see. It was proposed that ethylene oligome~z~tion proceeds via a

( COD I2 Ni +

PhzP -CH-

[

/O ‘OH

Ph

60 %

Complex

Figure 2.7.

Ph

40%

la

Complex

lb

Synthesis of PO-chelate nickel complexes I(a) and I(b) [139].

Michaelis-Menton type mechanism, the active site being that of the Ni-H complex center as shown in Figure 2.8. The rate equation for ethylene oligomerization was given by: rolig =

[Ni*l k K[C2H41 I+ mC2H41

(2.19)

where [Ni*l is the catalyst concentration, and k and K were calculated to be 0.18 s-1 and 0.52 Vmol, respectively. A series of one-component catalysts, shown in Figure 2.9, were prepared and their catalytic performance for ethylene oligomerization was evaluated [141]. These complexes were found selective for a-olefins synthesis (>99% linear olefins in the C4-Czo range) 11371. The three steps for oligomerization reactions are shown in Figure 2.10. Internal and branched olefins are formed when a-olefins coordinate themselves on the nickel, followed by the isomerization or further insertion steps. The high linearity and the excellent a-olefins selectivity of the SHOP catalysts were attributed to the fact that ethylene is a much better nickel coordinating ligand than the cc-olefin products [137]. The data showed that the addition of phosphines to these complexes increased the electron density on the nickel atom, thus weakening the bonding to other ligands. Phosphines not only occupied a free coordination site on the metal, but also supported the elimination step by decreasing the strength of the nickel-alkyl bond. The nickel complexes, shown in Figure 2.9, are closely related to complex I in terms of structure and catalytic activity for ethylene oligomerization. The arsenic homologous compounds IV and complexes V and VI which contain a

65

Figure 2.8. Simplified mechanism for ethylene oligomerization with PO stabilized nickel hydride intermediate (p = propagation; e = elimination) 11371. phosphorus oxygen ylide ligand were also found to be good oligomerization catalysts for ethylene [1371. Complex IV was prepared by reaction of benzoylmethylene-triphenylarsane with Ni(COD)z in toluene [1421. At 40°C and 14.8 atm, complex IV oligomerized ethylene to linear olefins (>95%) at 70% conversion with a-olefins selectivity being 70%. The catalytic activity was reported to be 1600 mol ethylene per mol nickel per hour. A nickel based complex was prepared by the reaction of 3diphenylphosphino-l,l,l,trifluoro-2-triflu-2-propanol (complex VII) and Ni(COD)z in toluene to produce a catalyst for ethylene oligomerization. The complex was tested at 50°C and 49.4 atm and gave 98% linear olefins of which a-olefins predominates at 99% and the activity was more than 5000 mol of ethylene per mol of Ni(COD)z. The reaction proceeded through the nickel hydride complex (VIII) formed as shown in Figure 2.11 [143]. A novel complex (4-cyclooctene-1-yl)-bis(trifluoro acetamido) nickel oligomerized ethylene to 85% linear olefins at 11500 mol ethylene/m01 nickel 0441. The catalyst precursor was prepared by reaction of Ni(COD)z and 1,3 diketone according to the equation shown in Figure 2.12. This catalyst was found selective to neohexene that can be produced at 40% conversion at 16OT. However, a better conversion was obtained at lower temperature upon addition of boron trifluoride. The maximum conversion at 120°C using the Ni/BFs catalyst was 66%. The ratio of BFdnickel complex was between 0.5 and 2.

66

Ph ph\ ph3P ’ Complex I

Ni ’ -Y” ‘0%

Complex II

Ph

Complex 111

ph Ph

fh\ ><

Ph ‘zp/ \ I -‘N

$”

Ni Ph3P /

f’h ‘P’

‘oljc\

,,h3p/Nf,~t\Ph

Ph

Complex IV

Complex V

Complex Vi

Figure 2.9. Homogeneous one-component model nickel complexes [1371.

a) assoclat

Ion /R’ LnNi-R

+ H2C=CHR’

KI +

LnNi,

/‘I

R = H or alkyl

R

b) Insertion

n - mode L,NI iso - mode

cl elimination L,Ni-CHZ-CHR’R

Figure 2.10. Mechanism catalysts 11411.

k,lim,

L,Ni-H

of ethylene

+ CH2=CF(‘R

oligomerization

by nickel-ligand

67

+ PCY3 Ph2

Complex VII

Figure 2.11. [1433.

Formation

Complex VIII

of the nickel hydride complex VIII (Cy = cyclohexyl)

0 =

c,/R

0 =

c’

Ni (CODI

Y-0

-COD

'R

N

Y=]CH Ligand

1

1

Complex

_3_

G

R = R’

q

Figure 2.12. Preparation of active catalyst precursors bis(l,5-cyclooctadiene) nickel [Ni(COD)21 and l,&diketones

CF3

by the reaction [1441.

of

Various organometallic compounds activate the sulfonated nickel ylide complex IX shown in Figure 2.13 11451. Aluminum alkoxides, such as AlEtzOEt, AlEt(OEt)z and Al(OEtj3, significantly increased the oligomerization activity by a factor of 20-100. The distribution of a-olefins (C4-C40+ range) produced by the sulfonated nickel ylide-aluminum alkoxide catalyst followed the Schulz-Flory molecular weight distribution law [ 1451. The kinetic analysis of ethylene oligomerization reaction with the binary catalytic system showed that the number of active centers is proportional to the nickel complex concentration. The activation energy for ethylene oligomerization was 27 kJ/mol at 50-90°C and 7.1-35 atm. The kinetic steps followed The initial rate of initiation, chain propagation, and chain termination. ethylene consumption under steady state was expressed by the following equation:

ComplexIX Figure 2.13.

I$=

Sulfonated nickel ylide organometallic

kiCmC* (keC, + kt) kiC, + kt

complex 11451.

(2.20)

where Ri = initial rate; ki, k,, kt = initiation, propagation, and termination rate constants, respectively; Cm = monomer concentration; and C* = concentration of active centers. Knudsen [1461 investigated the effect of various acid components of homogeneous nickel catalysts prepared from Ni(C!OD)g, dicyclohexylphosphine system for ethylene oligomerization to alpha-olefins. The fluorinated acids, which had the best activity, produced good purity &-olefins. The effect of solvent type on catalyst selectivity was also studied showing that 2-pentanol gave the best purity of a-olefins although its reaction rate was lower than toluene. The effect of acid-to-nickel ratio on catalyst activity showed that O-l.25 to 1 is the optimum ratio for dicarboxylic acid and trifluoroacetic acid. The activity enhancement, by increasing the acid concentration, was consistent with the nickel hydride mechanism [146]. Brown and Masters [147] tested a series of catalysts derived from Ni (SacSac) anion). The activity and PR3Cl (&d&lC= C5H7f$= pentane-2,4-dithionate product distribution from ethylene oligomerization were studied as a function of phosphine ligand and reaction temperature (-20 to 35°C). Activities of the order of 4 x 106 mol of ethylene converted per mol of nickel per h were reported at quite a long catalyst life. The results showed that the product distribution (C4-C12) was dependent on reaction temperature. The formation of higher molecular weight oligomers was favored at higher temperatures. The butenes formed were mainly 2-butenes indicating the high olefin isomerization of these catalysts. The cationic ally1 nickel (II) complexes [Cs&NiL2]PPs, where L = P (OPh)a, P(OThym)s, SbPb, or l/2(COD), are efficient catalysts for ethylene oligomerization [148]. The catalytic tests, carried out in dichloromethane at 10 atm, showed that the activity increased sharply in the following order according to the type of ligand used: P(OPh)s
69

The product distribution included dimers (mainly 2-butene), trimers, and traces of Cs_Clo olefins. Sulfonated nickel ylides were found to be potent single component catalysts for oligomerization of ethylene between 50-12O’C and lo-15 atm. Both aromatic solvents (benzene and toluene) and polar solvents (alcohols, dioxane and THF) were used. The catalysts showed higher activity as a result of sulfonation when compared with corresponding nickel ylide catalysts [P148P1603. The advantage of the sulfonated nickel ylides is the presence of the sulfonate group which induces solubility in polar solvents, such as water and methanol. This allows easy catalyst separation by extractive techniques, such as aqueous ammonium hydroxide, which is not possible with corresponding nickel ylides. The effects of catalyst preparation procedures, activation conditions, process conditions, poisons, and additives on the activity and selectivity of the sulfonated nickel ylides, were investigated. The catalyst activity decreased (five fold) as the temperature increased from 50 to 12O’C. The presence of water in the reaction medium did not ruin the activity up to 10.5 wt.%, but caused a drop in activity at higher concentrations. A shift towards lighter oligomers was caused by an increase in water concentration [P1491. The activity of the sulfonated nickel ylides was ruined by the presence of oxygen even at low ppm levels. The data showed that catalyst activity, life productivity, and product distribution could be correlated to reaction operating conditions. Operating conditions were derived for each of the series of sulfonated ylides CP156-P1601. The effect of solvent on catalyst performance and ease of separation from the oligomers were studied. Triethylene glycol, dimethyl ether, alcohols, THF, toluene, among others, were compared with methanol, being the only solvent to give two well defined phases. The top phase contained ethylene oligomers, while the bottom contained methanol/catalyst [P155]. Bicomponent catalyst consisting of sulfonated nickel ylide and alkyl aluminum alkoxide in toluene solution between 30 and 12O’C and 6.6-14.6 atm was used to oligomerize ethylene to a-olefins. The data showed oligomerization activities with aluminum ethoxides to be more productive (20-100 times), giving C4-Czo olefins in a Schulz-Flory molecular weight distribution [145,P155]. Hete~geneous Nickel Based Catalysts Nickel complexes can be heterogenized in order to combine the outstanding selectivity of homogeneous catalyst systems with the easy recovery of heterogeneous catalysts 1149-1591. The impregnation of complex I, shown in Figure 2.7, on silica and silica-alumina yielded an active supported solid phase catalyst (SSP). Nickel was chemisorbed on the surface of the solid phase 11531. However, when compared with similar homogeneous systems, the turnover number decreased remarkably and the acidic silica-alumina increased the activity towards co-oligomerization of the primary products and double bond isomerization. This resulted in decreasing the linearity and the a-olefin content of the products. The main disadvantage of the SSP-catalysts

70

the diffusion control of the oligomerization reactions that resulted in an enrichment of polymeric waxes in the pores of the catalyst support. The absence of solvents caused hindrance in dynamic ligand action and a change in the catalyst selectivity. To eliminate the support effects caused by oxidic carriers, complex I was grafted onto polystyrene by the mode shown in Figure 2.14. The heterogenized catalyst I yielded oligomers containing about 99% linear a-olefins. The selectivity of the heterogeneous catalyst was higher than the homogeneous complex in the presence of excess triphenyl phosphine. Without the addition of phosphine, catalyst X shown in Figure 2.14, was active for several runs only. was

0

0

P - lCW2 I,., PPh2=CH-&Ph

+ Ni Icod . PPhg

Ph’

where

‘PPhs

Complex

X

@ =polystyrene chain.

Figure 2.14. Grafting nickel complex II onto polystyrene support 11371. Impregnation of silica and silica-alumina with the nickel complexes II and III (Figure 2.9) gave an active supported solid phase catalyst [1541. The support alone or impregnated with a Ni(I1) salt showed no activity under reaction conditions. Supported complex I on silica gave highly linear products, while on silica-alumina it yielded branched products. Beach and Kobylinski [155] reported the use of di-carbonyl-tris (cyclopentadienyl nickel) on silica-alumina for oligomerization of ethylene with turnover frequency greater than 105 mol ethylene per mol nickel complex. The catalyst, when tested at 150°C and 34 atm, produced C4-Cl4 olefins at 95% selectivity. The effect of nickel concentration, acidity of support, and reaction conditions on oligomerization of ethylene over a nickel exchanged amorphous silicaalumina catalyst were investigated [151,152,156-1581. The acid strength of the support was found to have strong influence on the activity and selectivity of a partially exchanged nickel/silica-alumina catalyst [ 1561. Ethylene activity and molecular size of the product, measured at 3OO”C, 11 atm, and 8-60 h, in a fixed-bed continuous flow microreactor, correlated quite well with the acidity of the support. The product distribution showed an increase in dimers and a decrease in trimers as the conversion increased. The cracking activity of the catalyst was also enhanced by increased acidity of the support, while

71

deactivation, although moderately high, was not strongly affected by the Results indicated that the activity was approximately support type. proportional to the nickel content but with a significant positive bias at lower nickel concentrations [15’71. The selectivity to light olefins (dimers/trimers ratio) increased at higher nickel concentration. This was accompanied by an increase in deactivation rate and a decrease in the recovery of catalytic activity recovery on subsequent regeneration cycles. The effects on product distribution and catalyst deactivation were due to the selective exchange of nickel onto the sites with highest acid strength [X71. The effect of operating conditions on oligomerization of ethylene in the presence of Ni/silica-alumina was investigated between 120-38O’C and l-20 atm [158]. The data showed that the conversion improved by increasing the temperature or pressure as well as by decreasing the space velocity. Catalyst deactivation increased by increasing temperature and pressure, but was independent of space velocity. The rate of ethylene conversion was approximately first order with respect to ethylene. The product spectrum shifted to higher molecular weight olefms when the temperature and pressure increased or space velocity decreased. Elev and coworkers [159,160] reported the formation of coordinatively unsaturated NiU) ions on silica and alumina surfaces by the photoreduction of Ni(I1) in hydrogen or thermal reduction at 297’C. EPR results from coordination with CO, Hz, H20, NH3, 02,or N2 molecules and oligomerization reaction rate showed that Ni(1) could be the precursors of active complex. Ethylene oligomerization was investigated at an initial ethylene pressure of 0.002-0.2 atm and reaction time of l-20 min. After comparing various activation procedures, it was concluded that thermal reduction was the best activation method. 2.55 Titanium Based Catalysts Linear cc-olefins, Cd-Czo+, having an average molecular weight 70-300 and 90-100% purity were prepared by ethylene oligomerization in the presence of homogeneous TiC14-RAlC12 catalyst systems [126,161-164, P175-P1841. The effects of catalyst composition and pretreatment, reaction temperature, type of solvent and pressure on the oligomerization rate and average molecular weight of the products were investigated [126,1611. Ca&&st Compsition andpretreatmat Catalyst composition is the most critical variable, especially at high oligomerization temperatures, because of its effects on the stability of the soluble catalytic mixture [126]. The soluble titanium-based catalyst is usually prepared by mixing an alkyl aluminum chloride with titanium tetrachloride under conditions that prevent decomposition of the alkylated titanium product to a heterogeneous Ziegler type catalyst. At temperatures above 25V, dialkyl aluminum chloride rapidly reduces Tic14 to insoluble TX13 complexes that produce high molecular weight polyethylene [1611. On the other hand, monoalkylaluminum dichlorides and mixtures with aluminum chloride react with Tic14 to yield high active soluble catalysts up to about 25’C. The

72

monoalkyltitanium trichloride produced is stabilized by complexation aluminum chloride resulting from the following alkylation reaction:

6 RAlc13 + TiCl*

=

c1 \JR\ c1 /

\Cl/mcS

with

6+ c1 \ti/c<

=

c1 /

\dmcJR

* (2.21)

[RAlCl3I’[RTiC131+ = RAlC13’ + RTiC13+.

The active soluble catalyst is a polarized complex or ion pair with some possible dissociation to the free ions under favorable solvation conditions [1261. The effects of Al/Ti molar ratio on catalyst activity for various catalyst compositions are presented in Table 2.11. The oligomerization reaction was carried out at 4O”C, 54.4 atm and 1 h reaction time in heptane and chlorobenzene solvents 11261. Alpha olefins produced by the EtAlClz-AlC13-Tic14 catalyst system were in the range of C4-C3o at high purity (>99%). The addition of one mole of either AlC13, EtAlC13 or Tic14 to a catalyst system comprising of EtAlC13-Tic14 and having an Al/Ti molar ratio of 1.0, increased the catalyst activity. This effect is due to the production of complex anions that increased the concentration of catalytically active ion pairs 11261. An increase in the concentration of the EtAIC13-Tic14 catalyst in a benzene-heptane mixture, increased the steady-state rate of ethylene oligomerization and the degree of oligomers polymerization 11621. However, it decreased the content of butenes in the reaction products, effectiveness of the catalyst and the degree of electrolytic dissociation of the RTiC13*AlC13 complex. Table 2.11 Effect of Al/Ti ratio of various catalyst compositions on oligomerization activity and product molecular weight at 4O“C, 54.4 atm and 1 h reaction time [1261 Catalyt* A/B/C (mmoles) 0.5/o/0.5 0.5/0.5/0.5 0.5/o/1.0 0.5/o/0.5 1.0/o/0.5

AlITi (molar ratio) 1.0 2.0 0.5 1.0 2.0

Solvent CsH5CI CgH5C1 heptane heptane heptane

Activity [g(C4-C30)/fliC14-hrl 964 1155 100 110 165

Average Mol. wt. I.% lz 178 168 182

* A/B I C: EtA1Cl.3 IAlCl3/TiCl4.

Titanium based catalyst pretreatment is carried out to allow the catalyst components to react under mild conditions which lead to the alkylation of the titanium component without reduction to heterogeneous Ziegler catalyst. The pretreatment of the EtAIC12-Tic14 complex at 15°C for 30 min in ethylene medium, increased the catalyst activity through the reaction time and allowed better control of product molecular weight [126]. However, excessive catalyst

73

pretreatment severity produced a Ziegler catalyst end a large amount of polyethylene. Pretreatment severity is usually increased by increasing temperature, pretreatment time, catalyst concentration, or Al/Ti molar ratio [126,1611. Table 2.12 presents the weight percent distribution of linear a-olefins (CdC!30+fcorresponding to the average product molecular weight that is used as a criterion for oligomerization catalyst selectivity El261. Tbree representative distributions are presented which emphasize on low, intermediate and high molecular weight fractions. At an average molecular weight of 113, the products are 91.2% C4-Czo, whereas at 261, the products are 62.2% C22+ (mainly waxes). Table 2.12 Alpha olefins distribution molecular weights [ 1261 Carbon number C4 C&l0 ClL420

C22c28 C3o+

corresponding

to different

average

product

Average nroduct molecular weight (wt.%1 136 261 113 10.7 2.6 16.4 34.6 11.8 43.5 35.7 23.7 31.3 6.5 11.6 16.8 2.3 7.4 45.4

The type of solvent used in the titanium-based oligomerization catalytic systems depends on the co-catalyst available in the reaction mixture and other conditions like temperature and pressure [126,161,1621. In the presence of ~~kyl-aluminum chloride in the reaction medium, aromatic solvents are used at temperatures below 25°C. However, in the presence of monoethylahuninum dichloride, saturated hydrocarbon solvents are used in the temperature range of 2570°C. The oligomerization activity of Tic&EtAlC12, operating at 5O”C, 54.4 atm, 1 h reaction time and AkTi ratio of 8.0, decreased in the following order, according to the type of solvent used El261: Chlorobenzene > Benzene > Xylene > Heptane. The linearity of the a-olefm products was high for all solvents used. However, the average product molecular weight obtained in heptane solvent was very high due to the formation of C3o+ wax. In the presence of a mixture of benzene-heptane, an increase in the benzene concentration increased both the electrolytic dissociation of the EtTiC13*AlC13 complex and the rate of oligomerization associated with a decrease in the average molecular weight of a-olefins [1621. Low molecular weight a-olefins can be obtained in any

74

aromatic or saturated solvent by increasing catalyst acidity (AMY) or by decreasing the operating temperature [lSl]. Various compounds are usually added, in the range of 0.1-2.0 g/100 g of reacted ethylene, to the titanium catalytic system as modifiers, to enhance the activity of the catalyst and to increase the linearity of the product olefins. Some of these compounds claimed in the patents are tertiary phosphines or phosphites [P179,P181], Cl-C4 alcohols [P175,P1781, dienes [P1841, Lewis acids lP1801, and methylchlorocyclopentane or ferrocene [P176,P177].

Tempemturw The effect of temperature

on ethylene oligomerization rate and average product molecular weight was studied over a range of -20 to 50% for a TiC14EtAlC12 catalyst system in heptane solvent [126]. The results are presented in Table 2.13. The average product molecular weight of a-olefins showed an increase with increasing temperature. However, at higher temperatures MO “C) linear high melting a-olefin wax was a major part of the total product. The same trend was also observed in xylene and chlorobenzene solvents. This unusual feature of titanium based systems is directly opposite to the reported trend for the effect of temperature on ethylene oligomerization observed on Ziegler type catalysts 11261. It is believed that the chain transfer to monomer becomes more favorable to propagation only at low temperatures. This is due to the fact that P-hydrogen abstraction becomes more facile in the alkyl titanium species complexed with ethylene by the increase of positive charge on the titanium [126]. The temperature effect on the product a-olefin’s molecular weight is, thus, taken as a good evidence that the titanium catalyst species are present in the ionic form [ 1611. Table 2.13 Effect of temperature on ethylene oligomerization rate and product molecular weight for TiC14-EtAlC12* catalyst system in heptane solvent and 2 h reaction time [126] Temperature (‘C)

-26 0 :

AlITi (molar ratio)

1 1 1 0.25

Pressure

(atm)

Activity

gC4-Cs&l’iC14-hr)

;:: :::

: 105 26

Average product

molecular weight 113 143 163 200

*The catalyst systemwas pretreated at 50°C for 30 min.

High ethylene concentration is required to obtain pure linear a-olefins in the presence of titanium-based catalysts 11611. The effect of ethylene pressure on oligomerization rate and a-olefins linearity was studied over the range from

75

atmospheric pressure to 34 atm. Olefin purity increased sharply with increasing ethylene pressure above 3.4 atm and reached 90-100% above 6.3 atm D613. High ethylene pressures are needed to suppress copolymerization and to maintain tbe same olefin purity when cx-olefins concentration increases with the reaction time. Table 2.13 presents the effect of increasing pressure and temperature on the oligomerization activity of Ti~l4-~t~~lz catalyst Cl261, In the presence of EtTiCls*AIC13 catalyst system and conducting the oligomerization reaction at [email protected]*C, 25-30 atm, AVl’i molar ratio of 2-16 VTiCl4 concentration of 0.5-1.5 g/l) in benzene or toluene solvent, linear a-olefins, co&o, were obtained with more than 95% selectivity 11631. The narrow range of the reaction conditions is due to the possibility of the occurrence of side reactions at the oligomerization active sites and the formation of different types of active centers. Tembe et al. 11641found the titanium (IV1 [email protected] system, a selective catalyst for the synthesis of linear a-olefins having carbon numbers from C4 to C2s. The cr-olefins produced with this system generally follow the Schulz-Flory dist~bution. Both Al/Ti and temperature affect the linearity of a-olefins produced in the oligome~zation reaction [X64]. The P-hydride mechanism suggested for the titanium-based oligomerization catalysts was based on the following experimental observations [161]: 1. The rate of chain transfer is high relative to propagation under mild conditions, 2. Catalyst activity increases with increasing solvation or solvent polarity, and 3. Molecular weight decreases with decreasing temperature and with increasing solvent polarity. Comparative ethylene oligomerization experiments were carried out in the presence of a homogeneous catalyst (~Cl4-EtAlCl2) and a heterogeneous catalyst system ~Cp2TiCl2-BzAlCl) supported on polybutadiene E1653. The activity of TiCId-EtAlClz system was evaluated in toluene solvent at 2O”C, 1 atm and APTi ratio of 8 (Tic14 concentration was 1.1 g/l). The pofymersupported catalytic reaction was carried out in heptane solvent, 3O”C, 4 atm and Al/Ti ratio of 20 (catalyst concentration was 20 g/l>. The kinetic measurements were made in a batch reactor at constant ethylene pressure and a stirrer speed that assured the absence of external diffusion constr~n~ on the reaction rate. The polymer supported catalyst was in the form of discrete particles that did not agglomerate or adhere to the walls of the reactor 11651. Tbe effective activation energy of the polymer supported catalyst system, in the temperature range of 20-7O”C,, was found to be 27 kJ/mole compared with 44-46 kJ/mole for a similar homogeneous system. Ethylene oligomers, obtained by the two catalyst systems, were linear a-olefins; however, the molecular weight distribution of the polymersupported system was significantly different from the Schulz-Flory distribution. The proportion of the middle fraction a-olefins was increased. Rgure 2.15 presents the molecular weight distribution of ethylene oligomers

76

I

0 TiCI& - EtAICL2 * (C5H42 TiQ- RzAlCl

Alpha Olefin Carbon Number Figure 2.15. Alpha-olefins distribution for the titanium homogeneous TiCld-EtAlClz and (C5H5)2TiClz-R2AlCl polybutadiene [165].

based catalysts: supported on

obtained by the two types of titanium based-catalyst systems. The polymersupported catalyst system was characterized by the high stability of the catalyst sites and insensitivity to the impurities of the solvent and reactant compared to the homogeneous system. Deactivation and regeneration cycles were repeated many times with the initial process rate completely restored. This behavior is attributed to the protection of the catalyst active sites immobilized in the support, the water repellency of the rubber support (polybutadiene) and the isolation, of the catalyst sites from each other [1651. 2.5.4 Zix~nium Baaed Catalysts Zirconium alkoxides or carboxylates in combination with an organoaluminum compounds catalyzed ethylene oligomerization reaction to produce higher a-olefins [166-169, P186-P1901. The performance of a catalyst system comprising Zr(OCO-iPr)4 and Et&&II!13 in organic solvents was evaluated over a temperature range of 60-100% and a pressure range of 5-20 atm in different solvents, such as heptane, naptha, cycle-hexane, toluene and decene, and Al/Zr molar ratio of 9.5-25.6 [166]. The reaction product consisted of a mixture of higher linear a-olefins. The reaction proceeded in three main steps Since four moles of Et3A12C13 per initiation, propagation, and termination. one mole of zirconium compound were required in the initiation step, the

77

catalyst system showed activity only at Al/Zr molar ratios higher than eight. The average degree of oligomerization, which is a measure of the chain length of the oligomer product, decreased as AllZr molar ratio was decreased. The type of solvent used had no influence on the product distribution [1661. A kinetic model was formulated and the rate constants of the three main steps of the reaction were calculated 11661. Cationic type active centers of the catalyst system comprising Zr(OCO-iPr)r and Et3AlzC13,were formed in the initiation step without any induction period. The rate of catalyst deactivation during the course of the reaction followed a second order kinetics. The maximum concentration of the active centers was 20% of the initial concentration of Zr(OCO-iPr)4 in a catalyst system having an Al/Zr molar ratio of 12 [1671. The zirconium carboxylates were prepared from crude ZrC14 and C5-Cs fatty acid fraction. When a longer chain fraction was used to prepare the catalyst, initial activity and catalyst efficiency decreased during the reaction. A selectivity of 9598% to linear oligomers was achieved; however, the nature of the aluminum compound influenced the selectivity. Highest selectivity of 97.4% was achieved when Etl.sAlC11.5 was used. The presence of excess acid or water in the catalyst system decreased the selectivity for linear oligomers to 60% [X81. Zr(OPr)r in combination with Et3A12C13oligomerized ethylene to a-olefins, mainly C!1&!14, at an optimum Al/Zr molar ratio of lo-12 [1691. The cc-olefins yield increased as the temperature was increased and as the catalyst concentration was decreased. The rate of reaction was faster at a higher ethylene pressure. However, higher pressure effected the product distribution adversely. A combination of Zr(PhCH214 or Zr(CHz:CHCH2)4 with alkyl aluminum chloride was used as a catalyst to oligomerize ethylene to linear a-olefins in the C4-C2o range. The activity of these catalyst systems was significantly higher than other zirconium based oligomerization catalysts [170]. Experiments were performed in toluene solution at 40-80°C, and 10 atm ethylene pressure. The a-olefins product selectivities were: 11.8% C4, 45.1% C6-Cl0, 38.2% C12-C2o and 4.6% Czz+. A higher activity (62 g product/mmol catalyst/atm C2H4) was achieved when Et3A12C13was used as a cocatalyst. The activity decreased according to the type of alkyl aluminum chloride, in the order of: Et3Al2C13>Me#U2C13>Et~AlCl. It was suggested that the aluminum component enhanced activity by increasing the positive charge of zirconium through the successive replacement of ally1 or benzyl groups in zirconium compound by chlorine and by the formation of polar binuclear species. The increased positive charge of zirconium increased the rate of termination, which resulted in very little polymer formation 11701. The reaction of an organic-ester with zirconium tetrachloride produced a complex which oligomerized ethylene in the presence of EtzAlCl[171]. C&40 a-&fins were produced in a Schulz-Flory distribution with 90% linear olefins. Exp&iments were conducted in heptane solvent at 130-150 “C and 60 atm. The catalyst activity at 150°C was (4.1-8.0) x 104 mol ethylene/m01 Zr. Average

78

molecular weight of the product decreased from 140 to 96 as Al/& molar was increased from 7.5 to 15.3. The presence of water in concentrations greater than about 260 ppb resulted in polymer formation in significant amounts. Similar activity and selectivity were observed when an ester complex of Zr(OPr)r was used; however, a higher quantity of Et&ICI was required to activate the catalyst system [1711. ZrC14 and Et2AlCl oligomerized ethylene at Al&r molar ratios higher than 10 with a maximum Al/& ratio of 30. The activity was rednced sharply at higher temperatures. The catalyst activity was low when Et3A1&13 was used; however it improved in the presence of BuONa [172J.

Zeolites, such as ZSM-5 and Y-aeolite, were used to catalyze the oligomerization of light olefins 11731. Ethylene oligome~zation over hydrogen forms of ZSM-5 and dehydroxylated Y-zeolite was studied by Kustov et al. E1741.HZSM-5 contained Bronsted acid sites while the dehydroxylated Y-zeolite contained Lewis acid sites. The oligomerization reaction was carried out at 27°C in the infrared spectrometer cell (IR-cells) under static condition. The infrared spectroscopic data showed that the cationic ethylene oligomerization was initiated by Lewis acid sites of the dehydroxylated Y-zeolite. The products were mainly branched oligomers. Scheme 1 of Figure 2.16 shows the reaction mechanism of ethylene oligomerization on dehydroxylated Y-zeolite C1741. On the other hand, the spectroscopic results of ethylene oligomerization over HZSM-5 showed that the products were linear oligomers. Two possible reaction mechanisms of ethylene oligomerization in the presence of Bronsted acid sites were proposed as shown in Figure 2,16 &homes 2 and 3). A mechanism similar to Scheme 2 was also suggested by Derouane et al. on H-ZSM-5 [1753. However, a carbonium ion mechanism, which leads to the formation of C&-C7 branched oligomers, was suggested by Kofke et al. 11761, Oligomerization reactions carried on H-ZSM-48, which has a different pore structure from H-ZSM-5, yielded products containing shorter oligomers and exhibited lower ethylene conversion 11773. MoZyhdmum Bcxsei-i Cdysts A slightly reduced molybdenum oxide supported on y-alumina exhibited activity for ethylene oligome~zation [lSl]. Linear and branched C&-C16 oligomers were formed at 5O”C, with a maximum distribution at C14. An oxidized form of the catalyst produced highly branched oligomers. Lithium Based Ca&dysts

n-Butyl lithium (BuLi) in the presence of TMEDA catalyzed ethylene o~gome~zation to a product consisting of up to Cl5o olefins 11823. The reaction was carried out over a temperature range of O-20% and a pressure range of O2 atm. Initial BuLi concentration was varied from 0.1-l molk The molar ratio of ~D~uLi was varied from 0.08 to 3.0. A first order rate of reaction

79

Scheme - 1

Cationic mechanism

Scheme- 2

Carbonium fon Mechanism

Scheme - 3

Ethoxy Croup Mechanism

I?igure 2.16. Reaction mechanism of ethylene oligomerization using zeolite catalysts c1741. with respect to ethylene and BuLi concentrations was observed for all T~ED~uLi molar ratios used. However, the rate was independent of TMEDA concentration. It was proposed that a noncomplexed BuLi was responsible for the initiation and that the rate of elimination step was relatively slow C1821. Rodriguez et al. El833 proposed that a 1:l complex of t-3u~i-TMEDA was responsible for initiation. The proposition was suggested that the rate of reaction increased as the ratio of T~D~t-B~ was increased from 0 to 1, at a fixed BuLi concentration; and that the rate was independent of TmDAltBuLi at a ratio greater than 1. The reaction with t-BuLi-TMEDA was conducted in hexane at PC and 1 atm, A similar dependency of the rate on amine/n-BuLi ratio was observed by Crassous and Schue working with the

80

catalyst system, having different amines such as dimethyl-diethylene diamine and per&methyl-diethylene triamine [1841. Magnin et al. 11851, found that the dependency of the rate of reaction on TMBDA/t-BriLi molar ratio was of the order of 0.5, in a molar ratio range of O-l at a fixed t-BuLi concentration. It was proposed that an associated complex of t-BuLi-TMEDA, with a ratio of 1:l and a degree of association of 2, was responsible for the initiation step. !2s6

OLIGOMERZATIONPlVXESSES

Ethylene oligomerization has become the most important source of linear aolefins for new plants because it can be controlled to give the desired product distribution. At present, ethylene oligomerization or chain growth of ethylene is the primary source of linear a-olefins and is likely to be the basis for such production for many decades to come. Moreover, ethylene is a pure commodity raw material that can be purchased under long-term contracts [186,187]. All catalytic routes to a-olefins synthesis use organometallic catalysts of the Ziegler or related type. The processes are similar in principle, but the approach, catalytic system, reaction conditions and product distributions are different [ll. Commercial a-olefins processes and other significant processes are described in this section in terms of reactor configuration, reaction conditions, catalytic system and product distributions. Basic block flow diagrams are presented with each process to give an overall scheme of the process. However, the comparison and selection of an optimum ethylene oligomerization process is complicated because it must be based on an evaluation of a-olefin product quality, the ability to sell the full range of olefin product and the inherent economics of the process [1,31. 2.6.1 Commercial Process There are two approaches - the so-called catalytic or a one-step process and the stoichiometric or two-step process - for the production of a-olefins from ethylene. At present there are three commercial processes, originated by Chevron (Gulf), Ethyl, and Shell 133. Both Chevron and Ethyl have the original trialkylaluminum technology licenced by Ziegler of the Max Planck Institute for Coal Research, Mulheim, Germany; however, Ethyl has extensively modified the Ziegler olefin approach by using separate growth and olefin elimination steps with stoichiometric amounts of triethylaluminum. Shell has developed a unique nickel complex catalyst for the ethylene oligomerization section of the SHOP process [1,3,186-1901. The production of selected a-olefins or peaking to certain olefin range, product distribution flexibility, polymer inhibition and the ability to avoid pluggage or to handle it when it occurs, represent the key proprietary areas for the commercial producers [l]. It is believed that the current producers (Chevron, Ethyl and Shell) will continue to study new processes or

81

modifications to their present processes and that the development of new routes or processes will be the subject of future investigations 111. Cbron Pmcess The Chevron a-olefin process is based on Ziegler aluminum alkyl chemistry and related Gulf patents, since Chevron acquired Gulfs ol-olefins technology in 1983 [PlOl-P1131. Gulf was the first manufacturer of commercial high purity a-olefins from ethylene using Ziegler’s one-step ~owt~~splacement approach, the Chevron plant at Cedar Bayou, Texas, USA, which started production in 1966, is currently rated at 125,000 tons& of cc-olefins. Chevron’s total a-olefins capacity reached 250,000 ton& at the end of 1990 [3,141. The Chevron process uses a one-step catalytic approach to the a-olefins synthesis: simultaneous chain growth and displacement. The catalyst system is believed to be t~ethyl~~in~. The approach is based on carrying out both chain growth and displacement simultaneously under conditions favoring displacement (high temperature and large excess of ethylene). Only even-carbon number a-olefins are produced with product distribution skewed toward the lower carbon numbers, C4-Cl0 a-olefins [3,1871.

The block flow diagram of Chevron’s cx-olefins process is shown in Figure 2.17. Fresh and recycled ethylene are compressed to 230 atm and preheated to 180°C and then manifolded to each reactor in a five-reactor chain [3,PllOl. Ethylene is contacted with sufficient quantities of metered reaction solvent (n-heptane) and t~ethylaluminum catalyst (Et&l) solution. The E&Al-heptane solution is prepared batch-wise (once per shift). Each vessel is agitated and blanketed with high purity nitrogen (low water) to prevent EtsAl/water decomposition. The amount of Et3Al needed during the reaction is about 0.4 wt% of the ethylene reacted [PllOl. The use of solvent/catalyst solution increases the amount of ethylene in the liquid phase particularly during the initial stages, when no product olefin is available to act as an ethylene solvent. The solvent comprises 2-15 wt% of the total charge, which is enough to improve catalyst and reactor efficiency {PllO]. The use of tubular reactors reduces back mixing and minimizes the possibility of dimerization and the formation of branched olefin products [3]. Each of the reactor is simply a long, coiled pipe or alternatively, a group of Upipes connected in series, immersed in water or hydrocarbons to remove the large reaction exotherm. This system provides a very effective means of heat transfer in which the oligomerization temperature can be controlled within narrow limits. In the reactors, the Et3Al adds ethylene molecules in sequence forming linear alkyl groups. Simultaneously, the alkyl groups are displaced by ethylene to form a-olefins. The reactor temperature is maintained at 180°C and the residence time is 15 minutes and about 60% of the exotherm is absorbed by a 20°C temperature rise of the reactant fluids [3,PllOl. The residual exotherm

Ethylene

Make-up (EtsAI + Heptanel

Ethylene @

I

Figure 2.17.

Recycle

Block flow diagram of Chevron/Gulfs alpha-olefin process [S].

E

a-Olefins Products

To Fuel *

83

is removed by generating process steam on the shell side of the reactors. One or two spare reactors may be needed if fouling by solid polymer formation is a A small amount of dodecyl sulfide, benzothiazole, serious problem. diphenylamine or other compounds [PlOl-P1051 is added to inhibit the deposition of solid polymer on the inside of the reactor tubes without greatly reducing the catalyst efficiency. The selectivity t-o cc-olefins is in the range of 94-98% at an ethylene conversion of 3460% ISI. The effluent from the reaction system consists of a homogeneous mixture of normal a-olefins, unconverted ethylene, heptane, complexed Et3Al and small amounts of vinylidenes as well as branched and internal olefins. The effluent is fed to a gas-liquid separator step, where unconverted ethylene is flashed off and returned to the reactor feed compressor. Liquid product from the flash separator is passed to hydrolysis and phase separation steps. A solution of 5 wt.% NaOH is added to react with the complexed catalyst in an agitated pressure vessel to form sodium aluminate and paraffins [3,P108,PllOl: AlR3 + Hz0 + NaOH + NaAlOz + 3RH

(2.22)

The aqueous phase containing the sodium aluminate is separated and passed to waste water treatment facilities, while the paraffins remain in the organic phase with the a-olefin products. The organic phase is cooled and dried using molecular sieves before being sent to an a-olefins splitter column in the distillation section. The bottoms from the splitter, containing C&C18 and heavier a-olefins, are sent to a vacuum distillation unit for recovery of individual carbon number olefins (Clo,Clz, and up to Cl& All distillation towers utilize valve trays and are designed for various vacuum conditions. The final olefin product has a total a-olefin linearity content of 86-97% [PllOl. Approximately 71% of the product distribution is in the C&J18 main a-oletin product range with about 15% each of C4 and Czo-C40 constituting the remainder. Single-carbon number olefins up to Clg, as well as various blends such as Ce-Cs, Clz-Cl4 and Cle-C18 are produced; these blends are produced by back-blending pure single carbon-number olefins. The quality and distribution of Chevron’s a-olefins are discussed in section 2.6.2 along with the properties of a-olefins produced by Ethyl and Shell. Ethyl Pmcess

The information on Ethyl’s a-olefins process, a modified Ziegler two-step process, is based on patents covering a wide range of possible process options [P114-P1201. Various process developments and modifications have been commercialized at Ethyl’s a-olefin plant in Pasadena Complex, Texas, USA, which started production in 1971 and is currently rated at 430,000 metric ton&r of a-olefins and 135,000 metric ton/yr of linear primary alcohols, produced simultaneously. The Pasadena complex includes recycle operations, many chain growth reactors and various transalkylation (olefin

displacement) reactors. These many steps allow Ethyl to tailor its a-olefin and primary alcohol carbon number distributions to meet the changing demands of the market [1,8,186]. The process operated by Ethyl is basically a two-step process - separate chain growth and displacement steps. In the first step, long chain trialkyl aluminum, produced by the reaction of ethylene and triethylaluminum, is grown at low temperature to a desired average length and then sent to a high temperature displacement reactor. The displacement (transalkylationl is carried out in the second step operating at high temperatures and in the presence of excess ethylene or C&lo a-olefins that are used to displace the alkyl chains of the growth product [1,3]. By separating the growth and displacement steps, a more favorable product distribution pattern with a sharp peak in the desired carbon number range can be achieved; the distribution is more peaked toward the C6-Clo region. However, this procedure requires relatively large quantities of triethylaluminum in the growth step and introduces an increase of branched-chain products, particularly 2-alkyl a-olefins [33. The large catalyst quantities associated with the two-step process require that the catalyst be recovered rather than destroyed. The recovery of aluminum alkyl from the olefin product appears to be an economic necessity in the Ethyl process because of the high cost of aluminum alkyl. Straight distillation of aluminum alkyls from the olefln product is npt feasible because of the close boiling points of triethylaluminum (bp=194”C) and one of the most valuable a-olefins, 1-dodecene (bp=213.4W [3]. Several aluminum alkyl recovery systems have been patented by Ethyl [PllS-P1201; however, the system employed commercially comprises the utilization of C&lo cx-olefins instead of ethylene in the displacement step. This system requires a source of supplementary C4-Cl0 aolefins that are produced by Ethyl in an additional one-step conventional aluminum alkyl process [1,3]. Process

Description

The block flow diagram of Ethyl’s a-olefins process is shown in Figure 2.18. The process consists of four stages - triethylaluminum synthesis, one-step catalytic oligomerization of ethylene (combined growth and displacement), two-step stoichiometric chain growth/displacement, and a-olefin product separation 131. Triethylaluminum is produced by a two-step hydrogenation and ethylation process. In the first step, active aluminum metal, hydrogen make up, solvent and recycled triethylaluminum are reacted at 12 atm to yield diethylaluminum hydride. The diethylaluminum hydride is then contacted with ethylene under pressure in a second step to yield triethylaluminum. The net make of one mole of triethylaluminum is used in the following oligomerization step of the process, while two moles are recycled to the hydrogenation reactor of the triethylaluminum synthesis step [3,1871. Triethylaluminum is used in two parallel ethylene oligomerization reactors,

4

‘I

‘I

*

4

I #1



4

+

1

I

c

Figure 2.18.

Chain Growth +2

t-

Chain Growth

+

1

Catalytic Oligomerzation

r Et3AL

J Et,Ai Ah!%!!$ Synthesis +-, Ethylation (Two steps)

Ethylene

l

Separation

&

-

Combined

:,

oL y,_c,__

H20

NaALO2

Phase Separation

9

-

.

.

Fractlonatlon 01

t

n -rIlefIn

J

Cb -Cl2 a _ Ole

+ To Fuel

fr..

-r..

1

C&S

J

Growth

transalkylatlon

AuFr~ct~~t ,[email protected]

r-4

Orying

Block flow diagram of Ethyl’s alpha-olefin process [3J.

Transb Alkylation +2

-D

a - Olefins

Hydrolysis

C4 -Cl2

-

c Alkylation

Trans-

Gas Separation

t

NaOH Hz0

86

the combined growth/displacement and the chain growth of the two-separate stoichiometric step. The second step of Ethyl’s process comprises the preparation of low molecular weight olefins (mostly C4-Clo) in a conventional one-step Ziegler growth process (similar to Chevron’s process). The one-step process operates at a temperature of 160-275’C and a pressure of 135270 atm. Only a catalytic amount of alkylaluminum is used and is destroyed after the reaction by hydrolysis with sodium hydroxide. The large proportion of low molecular weight a-olefins (C4-Cs) produced at this step, is used in the displacement step. The two-step stoichiometric chain growth and displacement operate at 80-100°C and 100-200 atm for the growth step; the separate displacement step proceeds at 2453OO’C and 7-20 atm. The displacement (transalkylation) is carried out in the presence of excess lower molecular weight olefins (C4-Clo) rather than ethylene. The alkyl aluminums having chain lengths from 12 to 18 are displaced by the lower molecular weight oleflns, giving alkyl aluminums with chain lengths of 4 to 10 and free Cl&J18 a-olefins. The lower aluminum alkyls from the transalkylation step are recycled to the chain growth section or to a second stoichiometric chain growth step for further growth. The free a-olefins (mostly C12-C18) are sent to a separate oletins recovery section. This stream, which contains some branched a-olefins, introduced during the transalkylation step, is treated with triisobutylaluminum (iBu&l) in the purification step to form trialklylaluminum and isobutylene. In this step, internal olefins react very slowly during this exchange and thus pass through unreacted. The internal olefins are then easily separated together with the displaced isobutylene from the resulting linear alkyl aluminum [3]. A stream of Cl2 a-olefins fraction is sent to a second transalkylation step, where it is reacted with the growth product from the second stoichiometric growth step as shown in Figure 2.18. This produces an easy-to separate equilibrium mixture of a-olefins (mostly C4-Clo) and growth product (mostly C12). The light a-olefins are sent to olefins recovery and combined with the C4-Cl2 a-olefins from the one-step catalytic oligomerization section 131. The growth product (mostly Cl2 trialkyaluminum) is combined with the C12--C18 product from the purification/transalkylation step and sent to a conventional alcohol processing section where the combined product is oxidized and hydrolyzed to alcohols. Internal olefins, recovered from the oxidation reactor, are combined with the internal olefins produced during the purification treatment 131. Approximately 82% of the olefin production is in the C~-C~FJcx-olefin range, with about 15% as C4 and under 2% as C2o-C4o. The exact distribution of Ethyl’s a-olefins varies according to market demands for olefins and alcohols change. A consequence of Ethyl’s peaking technology is that a degree of branching is introduced into the a-olefin product, the amount of linear aolefln in the product decreases as the molecular weight of the product increases. The specifications of a-olefins produced by Ethyl are presented in

8’7

section 2.6.2 along with other a-olefins processes.

produced

by the Chevron

and Shell

The commercial application of the SHOP began at Shell’s Chemical complex in Geismar, Louisiana, USA, in 1977 after almost 10 years of development [3,11,128,186-19O,P140-Pl471. The second SHOP plant is located in Stanlow, UK and the third in Geismar, which started production in late 1989. The combined a-olefins production of the three SHOP plants is one million tons& [1,14]. In the first stage of the SHOP multistage process, ethylene is oligomerized in the presence of a Shell proprietary nickel-ligand catalyst to a broad range of linear a-olefins having a geometric distribution. Alpha olefins which can be sold (in general Cl,-&20 range) are distilled off, while olefins outside the market range (less than Clo and greater than C2o) are isomerized and disproportionated over heterogeneous catalyst systems in a second stage. The third stage converts olefins of less than Clo and greater than C2o into products in the Clo-C2o detergent range in a metathesis step. The combination of the three stages of the SHOP process offers a unique technology to tailor the carbon number distribution and the amount desired of a-oletins [p141]. The novel feature of this process is the way in which the relatively low range, are valued lighter and heavier cr-olefins, outside the detergent upgraded to detergent range internal olefins by a combination of isomerization and disproportionation steps.

Process Description Certain aspects of the SHOP process shown in Figure 2.19, are covered by patents assigned to Shell Oil Co., USA and Shell International, Holland. The ethylene oligomerization section of the SHOP process is discussed in detail since a-olefins are produced at this step. The oligomerization section consists of a combined growth/displacement reaction in the presence of excess ethylene catalyzed by nickel-based catalyst [128,P1411. The reaction system is essentially a series of time tanks to which compressed ethylene and catalyst/solvent are fed. The reaction is carried out at 80-120X! and 70-130 atm in 1,Cbutanediol solvent. A three-phase mixture (solvent containing catalyst, oligomer product and ethylene gas) is recirculated through the reactors to ensure intimate mixing of the three phases. The heat of reaction is removed by water-cooled exchangers and the rate of reaction is controlled by the rate of catalyst addition to the reactors. Rapid mass transfer of ethylene from gas to liquid phase is ensured by high recirculation velocities. A high degree of ethylene saturation in the catalyst solution is needed for maximum reaction rate and product linearity [128]. The nickel catalyst complex used in the oligomerization section is prepared by dissolving two moles of nickel chloride per mole of bidentate ligand (diphenyl phosphino benzoic acid) in 1,4-butanediol solution [3,15,128,P147]. The mixture is pressurized with ethylene to 87 atm at about 40°C. Boron hydride is added at a molar ratio of two borohydrides per one atom of nickel

Catalyst and Solvent Make-up

Ethylene

l

Figure 2.19.

I’ 71

Recycle

t b Flash

-

I

Phase Separation

To Alpha-Alcohol Production

Cll’CH

Block flow diagram of Shell’s higher olefins (SHOP) process [1411.

Cc/ Cl0 Recycle

Solvent and Catalyst Recycle

Chain Growth Reactor

Ethylene

l

8

89

when the system has equilibrated [P1471. The catalyst preparation reaction is then completed under ethylene pressure of 100 atm. Nickel concentration in the catalyst system is in the range of 0.001-0.005 mol %. Catalyst make-up in solvent (1,4-butanediol) is pumped, as required, into the inlet stream of the reactor [3,P1403. The solvent used in the oligomerization reaction is an oxygen-containing polar organic compound in which the product a-olefins are insoluble. Aliphatic diols of 2-7 carbon atoms per molecule including vicinal alkanediols and alpha-omega alkane diols are the preferred solvents, such as 1,4-butandiol. An essential requirement is that the nickel catalyst complex is soluble in the reaction solvent [P147]. Process Discussion Fresh and recycled ethylene are mixed with a large quantity of recycled solvent before being fed to the reactor system. As hydrocarbons are formed in the reactor, they separate from the solvent-catalyst phase and enter the oligomer (hydrocarbon> phase. The reaction product is first separated into gaseous and liquid phase in a gas liquid separator. The surplus (undissolved) ethylene, which separates, is let down through a pressure control valve to the suction of the final stage of ethylene compression. The mixed liquid phases are subjected to a fairly crude separation procedure in a high pressure liquid cyclone, being the primary liquid/liquid separator. The solvent and entrained hydrocarbons are removed in this separator and recycled back to the inlet of the reactor via the reactor circulating pump [3,P141]. The reactor pressure and flow rates are controlled by the circulating pump and the ethylene pressure control valve. A second high efficiency separation step is used to separate the hydrocarbons leaving the primary liquid/liquid separator. Almost all the remaining solvent phase is removed and sent to the solvent recovery section. The solvent stream is distilled under high vacuum in a recovery column to remove water, light organics and remaining quantities of catalyst. The hydrocarbons leaving the secondary liquid/liquid separator are washed with the recovered pure solvent to remove any dissolved catalyst in a high-pressure solvent wash tower [3,186]. This recovers some of the catalyst and eliminates uncontrolled polymerization and blockage problems. The recovered solvent is then sent back to the catalyst preparation area for mixing with catalyst make-up and solvent. The washed hydrocarbons are then directed to a product flash vessel to separate the gases (mainly ethylene) and liquid hydrocarbons. The gas stream is sent to the debuteniser column of the ol-olefins separation section. C4+ olefins are recovered at this stage before recycling the ethylene. Liquid hydrocarbons from the product flash vessel are washed with pure water to remove remaining traces of solvent and inorganics. The products are then directed to the deetheniser column of the a-olefins separation section for ethylene recovery and recycle [3,186,P141]. The hydrocarbon products are separated in a series of distillation columns for recovery of individual a-olefins. Olefins in the Cc-C&o range are separated into various single carbon cuts and blends for sale while the remaining are

90

sent to the isomerization/disproportionation sections. The a-olefin products are characterized by a geometric distribution with a molar growth factor (K) of 0.78. The value of K is usually selected on the basis of the required product spectrum and whether supplementary low molecular weight olefins are needed for the isomerizationldisproportionation sections. The specifications of a-olefins produced by Shell along with other a-olefins produced by the Chevron and Ethyl processes are presented in section 2.6.2. The rest of the olefin products, for which no ready market exists, are recombined, isomerized, and disproportionated to Q-C 18 linear internal olefins which are hydroformylated by the Shell Hydroformylation (SHF) process to linear detergent range alcohols. The isomerization is carried out over a heterogeneous magnesium oxide catalyst at a temperature of 80-14O’C and pressure of 4-21 atm. The disporportionation is carried out in the liquid phase over a catalyst bed of rhenium oxide supported on alumina at a pressure of 3-20 atm and 80-140 atm. The reaction cleaves the olefin molecules at the double bond position and recombines the fragments. This system yields only lo-15 wt.% per pass of the desired detergent range internal olefins, CllCl4

l.31.

2.6.2 Comparisonof CommercialprocesseS The production

of a-olefins by ethylene oligomerization

is the first step in all

commercial a-olefin processes operated by Chevron, Ethyl and Shell. The processes of Chevron and Ethyl differ in some details but are both based on the catalytic influence of trialkyl aluminum [13]. The Ethyl process includes some recycling to give a better control over the product distribution in order to achieve a higher yield of the most desirable C&I!14 a-olefins. Shell’s process, which employs a nickel-based catalytic system in the oligomerization section, gives the most desirable product in terms of highest a-olefins and lowest paraffin contents [188-1901. However, the ligand that is coordinated to the Ni metal center is too expensive to prepare, and the pressure requirement of the process is high. Chevron’s process appears to be the simplest and may have the lowest capital cost for a given volume of cr-olefins [1,3,187]. Ethyl allows increased production of specific carbon numbers mainly in the plasticizer range olefins, while suppressing others. A consequence of Ethyl’s peaking technology is that a certain degree of branching is introduced into the cr-olefins product. Plant size appears to be a significant asset for Shell and Ethyl. Different operating conditions used in the three commercial processes have resulted in various reactor configurations. The reactors used by Ethyl and Chevron are heat exchangers which remove the exothermic heat of the oligomerization reaction, while, heat exchangers are used separately in the system of Shell’s process 111. The a-olefin product distribution is quite different for the three processes. Chevron has a Schulz-Flory distribution at a growth factor of 0.7, Ethyl’s CXolefins have a Poisson distribution with a degree of oligomerization of 5, while

91

Table 2.14 Actual a-olefins product distribution of various commercial ethylene oligomerization processes 131 a-olefin 1 C4 CS C8 Cl0 Cl2 Cl4 Cl6 Cl8 C20+

14.0 13.0 15.0 12.5 10.0 8.0 6.5 5.5 15.5 100.0

Total

Shell’s

a-olefms

W

.

0

11.5 19.1 22.1 18.4 14.4 7.5 3.5 3.5

100.0

have a Schulz-Flory

11.0 13.0 13.7 13.0 12.4 11.1 9.8 8.6 7.4 I100.0

distribution

at a growth

factor

of 0.8

[3,128]. The a-olefins product distributions of the three commercial processes are presented in Table 2.14 and Figure 2.20. Shell’s process gives a high linear a-olefins content at higher carbon numbers, but its butene is of lower quality. The Chevron and Shell processes produce a higher cr-olefin content at the higher carbon numbers (Cl4+) than does the Ethyl process. Only Chevron distills and stores a linear olefin above C2o [l]. An overall comparison of Chevron, Ethyl and Shell a-olefins specifications is presented in Table 2.15. Ethyl’s a-olefins tend to contain rather more branched components than olefins from the Chevron and Shell processes. The extent of formation of Ethyl’s branched oligomers increases quite markedly with carbon number. The secondary reactions leading to impurities such as internal and vinylidene olefins, dienes and paraffins, are the characteristics of an ethylene oligomerization reaction [131. 2.6.3 Other Ethylene Oligomerization F%oeses Considerable research has been recently carried out to develop a new catalytic cc-olefin process by ethylene oligomerization using a catalyst system of high activity and stability. A number of processes have been reported in the literature which appear to be variations of the current commercial processes. The announced processes use different types of Ziegler-Natta catalysts modified with additives and dissolved in various solvents so that they selectively oligomerize ethylene to the most useful a-olefin compounds. Vista (Conoco), Exxon (Esso), and Dow piloted a-olefins processes in the early 1970s; patents assigned to Chevron in the 1980s used sulfonated nickel

-

I

-

1 ,-,

4

6

8

CATALYTlC

SHELL CATALYTIC (SHOP)

-

i

! 10

12

14

I

ETHYL STOICHIOMETRIC

-CHEVRON

-

t

-

-

-

-

t

\

a -

16

18 20

22 24

26

28 30

32

34

36

38

40

ALPHA-OLEFIN CARBON NUMBER

Figure 2.20. Comparison of alpha-olefins distribution produced by Chevron, Ethyl, and Shell processes 111. ylide catalytic systems other than their current aluminum alkyl based commercial process; Idemitsu and Exxon patented processes using zirconium based catalytic systems; and Union Carbide announced a new catalytic route for the production of commercially useful a-olefins using an organophosphorous sulfonated nickel catalytic systems. Lummus CrestiVista took Conoco’s patents to at least a precommercial stage and are offering their aluminum alkyl based technology for license. Idemitsu has started a-olefin production in early 1990 at a rate of 50,000 ton&r. Processes that have a high yield of Cs-(218 a-olefms and low yield of less desirable butene and Czo+ olefins, have the advantage in today’s market.

93

Table 2.15 Overall comparison of typical olefins produced by various oligomerization processes [13]

qualities of Cs-Cl8 acommercial ethylene

Property : a-olefins Branched olefins Internal olefins Paraffins Total monoolefins

91-94 1.6-7.8 1.6-7.8 1.6-7.8 98.6

W.

63-97.5 1.9-29.1 0.6-8.2 0.1-0.8 C99.0

96.97.5 1.0-3.0 1.0-2.4 co.1 99.9

Prospective a-olefins processes assigned to Lummus Crest/Vista, Idemitsu, Exxon and Union Carbide are discussed with special emphasis on the catalytic system used, process reaction condition, a-olefins distribution and other significant features. Lummrca CrestiVista Process The Lummus CrestMsta a-olefin process is based on Ziegler’s aluminum alkyl chemistry and on the experience gained by the operation of a semicommercial a-olefin unit during 1960s by Conoco Chemicals. Vista acquired Conoco and its a-olefins/alcohols technologies at Lake Charles, Louisiana, USA in 1984 [1911. The first stage of Conoco’s process for the production of primary alcohols is similar to the first stage of Ethyl’s process, i.e., aluminum alkyl catalyst chain growth proceeds almost to the exclusion of displacement, leading to a similar Poisson distribution of carbon chain lengths. Process Description The description of the Lummus CrestNista process is based on Conoco’s patent [P1281 and brochures obtained from Lummus Crest Inc., New Jersey, USA [191,1921. The stoichiometric process is characterized by the high yield of Cs-Cl0 a-olefins, low butene-1 yield, and the production of aluminum sulfate as a by-product. A block flow diagram of the LummusiVista a-olefins two-step process (separate chain growth and displacement) is shown in Figure 2.21. The first section of the process comprises the continuous preparation of triethyl aluminum from aluminum, recycle ethylene and hydrogen. The resultant purified Et3Al is combined with recycled solvent and recycled Et3Al and fed into the growth reactor operating at 115120°C and 95-115 atm. The product distribution of linear a-olefins is controlled by the concentration of aluminum alkyl in the feed and the rate of ethylene addition and distribution

Figure 2.21.

EtgAl Preperation * & F%arilication

Hydrogen

w

-

Album

e Ethylene

EW _overy

A

A

-)

Hydrolysis

Alpha olefin distillation

High purity ehlm

-w

A

Block flow diagram of the Lummus Crest/Vista alpha-olefin process 11911.

Nickel nephthenate

Recycle complex

Recycle solvent + Et&l

product

Alpha &fin

b

P

95

along the length of the tubular reactor. A small amount of carbon monoxide is added to the ethylene feed to act as a polymerization inhibitor [1921. The grown long chain alkyls are displaced by excess ethylene in the presence of a soluble nickle naphthanate at 455O‘X and 85 atm. The Et&l in the displacement effluent is complexed with lean tetramethyl ammonium chloride complex at 60-65°C and 30 atm. The resultant mixture is flashed to 50-60°C and 1.5 atm to recover unreacted ethylene for recycle. The mixture is decanted to separate rich tetramethyl ammonium chloride complex from product olefins and solvent. The dissolved olefins and uncomplexed Et3Al in the rich tetramethyl ammonium chloride complex are extracted by recycle solvent. The rich complex is vacuum pyrolyzed to recover pure Et3Al for recycle to the growth reactor and lean tetramethyl ammonium chloride complex for recycle to complexation step. The decanted product olefins are washed with sulfuric acid to remove residual aluminum alkyl as aluminum sulfate, neutralized with NaOH and separated into the desired olefin products by distillation 11921. Since the process does not depend on the recycle of olefins or aluminum alkyls to the growth or displacement reactors to achieve the desired product distribution, only the growth reactors’ operating conditions are adjusted to shift the olefin product distribution. The aluminum alkyl to solvent ratio and ethylene injection distribution of the growth reactor are adjusted to obtain the desired olefins range. For any proposed plant that maximizes the production of Clo olefin, there is a significant amount of latitude in the distribution of the other olefins. Table 2.16 shows the extremes of the olefin product distribution with a minimum of 20 wt.% Cm a-olefin. Analysis of the crude olefins fed to a-olefins distillation section, allows feed forward control to adjust the reflux ratios of the affected product columns [1921. Table 2.16 Lummus Crest/Vista a-olefin product distribution range with minimum 20% of Cl0 ar-olefin[192] .

a-olefin c4 c6 CS Cl0 Cl2 Cl4 cl6 cl8 c2ch

Total *M =

.

Product &stnbution (wt.%) Lower M* Medium M Higher M 10.9 6.7 5.0 21.7 15.1 11.7 24.9 20.9 17.6 20.0 21.0 20.0 12.4 16.4 17.8 6.2 10.4 12.9 2.6 5.5 7.8 0.9 2.5 4.1 1.5 0.4 3.1 100.0 100.0 100.0

Degreeof oligomerization.

Idemitsu’s a-olefins process was developed in the late 1980s and is based on several patents that describe in detail the various aspects of the catalytic onestep ethylene oligomerization process [P191-P1951. The catalyst used in the Idemitsu process is believed to be a complex compound prepared from zirconium tetrachloride, ethyl aluminum sesquichloride, and triethyl aluminum in the presence of a ligand compound such as sulfur, phosphorous or nitrogen. The catalyst system has a high activity and excellent stability for the production of high purity a-olefins in the range C4-Cla. The unique feature of the process is the method used in the separation of polymeric byproducts during the course of the recovery of unreacted ethylene from the reaction mixture. The precipitated polymer is crushed into tiny particles and is then heated with the a-olefin fraction to dissolve the crushed polymer particles. This method reduces polymeric clogging in the pipings, valves, heat exchangers, pumps and the recovery line of the unreacted ethylene which are experienced by the conventional filter-separation [P1951. The catalyst system, ZrC14, Et3A12C13 and EtsAl, is prepared at a temperature range of 50-70°C in an aromatic solvent or at 60-80% in alicyclic solvent. The concentration of ZrC14 is selected in the range of 60-120 mmolesfliter of solvent [P1911. The performance of the catalyst system is greatly influenced by the order of the successive introduction of the three catalyst components at the time of preparing the catalyst. It has been found that satisfactory results are obtained only when the catalyst components are introduced in the following order - triethylaluminum is introduced first followed by the successive introduction of ZrC14 and ethyl aluminum sesquichloride or by the successive introduction of triethylaluminum and ethyl aluminum sesquichloride followed by the introduction of ZrC14. The activity of the catalyst system is greatly decreased when the catalyst system is prepared by introducing the three components in a different order [P191]. The molar ratio of aluminum to zirconium (Al/Zr) is in the range l-15 and the molar ratio of Et3A12C13/Et&l is in the range 2-10. The molar ratio of ligand compound to ZrC4 is usually in the range l-20 for sulfur compounds and 0.5-10 for phosphorous or nitrogen compounds. Examples of ligand compounds, used in combination with the catalyst system, include dimethyl disulfide, thiophene, trioctyl phosphine and aniline. Process Description The flow chart of the Idemitsu a-olefin process is shown in Figure 2.22. The process comprises several steps including a step for the catalytic oligomerization of ethylene to form a-olefins, a step for the recovery of unreacted ethylene, another step for the deactivation of the catalyst and deashing of the product, and a final step for the fractional distillation of the solvent and the respective compounds of the a-olefin products. Fresh and recycled ethylene are fed along with the catalyst/solvent mixture to the reaction vessel maintained at 12O”C, 60 atm and 40 min reaction time. The Al/Zr molar ratio is 7 and the molar ratio of EtaA12Cla:EtaAl is 3.5:l IP1951.

Figure 2.22. Idemitsu’s conceptual separation IP1953.

1

Pump

alpha-olefins

Shear Cutter #

Ethylene Recycle

1 spent Catatyst

SOiUtlOil

I wastef

process with an improved method for polymeric by-product

Shear cutter #Z

I Deashing Tank

H20 0

Phase Separator

98

Cyclohexane is used as the reaction solvent and thiophene is used as the catalyst ligand. The reaction mixture, after completion of the reaction, contained about 0.05 wt.% polymer as a by-product having a viscosity-average molecular weight of about one million. The reaction mixture is first subjected to an adiabatic flashing treatment to evaporate the unreacted ethylene and to precipitate the polymeric by-product in the second flashing tank operated at 80-90°C. The precipitated polymer particles, having a particle diameter distribution in the range of submicron order to several tens of mm, are then transferred into the first shearing cutter wherein the polymer particles are crushed to a particle diameter not exceeding 1000 pm. The reaction mixture is then transferred into the thirdstage flash tank in which it is freed completely from the unreacted ethylene. The reaction mixture is admixed with an amine compound as a deactivating agent in a volume of 20-100 ml of ammonia/water per liter of the reaction mixture. The mixture is transferred into the second shearing cutter where it is agitated to deactivate the catalyst and the polymer particles are again crushed to have particle diameters not exceeding 1000 pm. Thereafter, the reaction mixture is admixed with water for washing and transferred into the deashing tank [P195]. The effluent from the deashing tank is transferred to a phase separator where it is separated into organic and aqueous phases. The aqueous phase is discarded as waste while the organic phase, containing the ol-olefins compounds and the polymer particles, is mixed with the reaction mixture from the reservoir tank. The temperature of the organic phase is increased from 40-50°C to 100°C so as to rapidly dissolve the polymer particles before being fed to a heat exchanger where it is further heated to a temperature of 120-130°C. The reaction mixture, maintained at a high temperature in the reservoir tank, is transferred to the distillation process for the recovery of solvent, separation polymer.

of a-olefin

compounds,

and removal

of the by-product

The a-olefin products have the following composition: 14.9 wt.% C4, 15.0% C6, 14.1% Cs, 40.2% Clo-Cls, and 15.8% C2o+ fraction. The olefin compounds are useful as monomers of polyolefins, comonomers of various kinds of copolymer+ resins, starting material of various kinds of plasticizers or surface active agents. The present process can be used with other Ziegler-type catalysts without disturbing the continuous operation of the production facilities and even without removing the by-product polymer out of the production line during the course of the reaction [P195]. LZxmmPmcess Patents assigned to Exxon (Esso) in the early 1970s describe a catalytic process for the production of linear a-olefins by ethylene oligomerization in the presence of a titanium based catalytic system [P175-P1781. The process closely resembles the high. temperature aluminum alkyl process practiced by Chevron. Patents assigned to Exxon in the 1980s used a zirconium based catalytic system similar to the catalyst used by Idemitsu process [P185-P1901.

99

The ethylene oligomerization reaction is carried out at 130°C, 68 atm and 0.5 h reaction time in heptane solvent. The catalytic system is ZrCl4, EtAlC12 or Et&W1 and the a-olefins content of the products reaches 98% [P1861. The titanium based process is operated at -30 to +30°C, 14-140 atm., and 1-5 h reaction time [P175-P1771. The catalyst system comprises TiC14, AlEts, Et&Cl or EtAlCl2 modified with various compounds, such as phosphine, methylchlorocyclopentane or ferrocene. The modifiers are added in the range of 0.1-2.0 g/100 g of reacted ethylene, to increase both the linearity and molecular weight of the olefin product and the catalyst activity [P1773. The oligomerization of ethylene is normally carried out in the presence of inert diluent such as aromatic diluents (benzene or xylene). The choice of diluent is critical because of its effect on the reaction system, ethylene adsorbability, molecular weight, and type of olefin product. The diluents, such as paraffin and cycle-paraffin solvents, exhibit low ethylene adsorbability, and yield reaction products predominantly composed of high molecular weight polymers Cp1783. The oligomerization reaction can be carried out in a batch or continuous reactor operating under essentially plug-flow conditions. Stirred tank reactors may be operated in series to approach plug flow [126]. Pilot plant studies were conducted at the Exxon Research Laboratories at Baton Rouge, Louisiana, USA, in the early 1970s to determine the feasibility of continuous operation of a five-gallon unit using a titanium based catalyst [126]. The results confirmed the expectations, based on batch operation; however, single stage continuous operation was not optimum [1261. The control of product distribution is accomplished by adjusting catalyst and oligomerization variables to yield the desired molecular weight over a range of 70-300 of the total product. The C2o+ products are separated by distillation or crystallization to obtain white, high melting waxes. In general, high molecular weight is obtained by increasing the alkyl content of the alkyl aluminum chloride, increasing oligomerization temperature or decreasing solvent polarity.

Process Description The solvent and catalyst components of titanium based or zirconium based systems are mixed in a continuous catalyst preparation reactor in which pretreatment severity is controlled by concentration, temperature, and residence time. The solution containing catalyst is fed into one or more reactors depending upon the size of the plant. Oligomers having the desired product concentration are passed into a catalyst quench tank where soluble catalyst residues are removed. The catalyst is apparently not recycled, but is deactivated after the reaction. The critical point in preventing copolymerization is to terminate the polymerization activity of the catalyst within about one minute after the unreacted ethylene is flashed off from the reactor effluent. The conventional agents used to terminate the catalyst activity include water, methanol, butanol, and ethylene glycol [P1751. Another agent, sodium hydroxide, is often added to the reaction mixture to neutralize the catalyst activity. Unreacted ethylene is flashed and dried for recycle. The solution is heated to dissolve wax product and filtered to remove any polymeric by-products. Distillation is

100

carried out in a number of towers, depending on the number of a-olefin, cuts. The recovered solvent is dried and recycled to the catalyst preparation step [1261. The a-olefins products obtained by the Exxon process using TiC14-Et3Al catalytic system are highly linear and have the following composition: 7.9 wt.% c4, 39.4% c6-cl0, 33.4% Clz-Clg and 16.7% C2o+. A major advantage of Exxon’s a-olefin process over triethyl aluminum growth and displacement process is the ability to obtain higher purity a-olefins at all carbon numbers, particularly above C2o [ 1261. Union CdideProcess Union Carbide Corporation has developed a new catalytic route for the production of commercially useful a-olefins by ethylene oligomerization [193,194, P1631. The heart of the nickel catalyst system is a novel ligand promoter having phosphorus and sulfonate moieties capable of bidentate coordination with various transition metals. Catalytic bidentate ligands include phosphines, containing coordinating functionalities, such as carboxylates, ketonates, amides and phenoxides. Some of these ligands have been used by Shell in its nickel complexes utilized as selective oligomerization catalysts. The research conducted by Union Carbide in the late 19808, was targeted at identifying new and novel catalytic active phosphorous ligands possessing other unusual coordinating functionalities. These efforts resulted in the discovery of high active catalysts based on a new class of ligand promoters, organophosphorus sulfonates [P163, 1931. The catalyst system is the reaction product of three components comprising a transition metal compound (usually nickel (II)), a catalyst activator, and an organophosphorus sulfonate ligand. Nickel salts, in particular, those of sulfonate, tretraflouroborate, and chloride hexahydrate give the most active catalysts followed by chromium, copper and cobalt [P1631. Specific examples of useful transition metal compounds are NiC12 (anhydrous), Ni(COD12, NiCls-6H20, Ni(BF&-6H20, Ni(I1) acac, CrCls-6H20, CuClz-2H20, and Co(OOCCH+4H20. The catalyst activators are reagents capable of transferring a hydride, alkyl or aryl group to the nickel complex. A wide variety of reagents including lithium and sodium borohydride, triethyl boron, borane-THF, triethylaluminum are suitable catalyst activators. Many of these reagents induce significant catalyst activity at low temperatures, thus permitting the oligomerization reaction to be conducted even at ambient temperature [P163, 1931. The organophosphorus-sulfonate ligand is usually in the alkali metal form (Li or Na) and contains at least one benzene ring having a trivalent phosphorus atom and SO3M group located in an ortho position in the benzene ring (M is selected from the group consisting of hydrogen, alkali metals or NR4/PR4). The ligand is prepared by the ortho-lithiation of aromatic sulfonates with n-BuLi and the reaction of the intermediate with The ligand cone angle and phosphorus organophosphorus chlorides.

101

basicities have strong influence on the catalyst activity and the Schulz-Flory’s a-olefin product distribution [193]. Ligands possessing cone angles from about 150 to 180°C produce very active catalysts and offer a broad choice of SchulzFlory a-olefin product distributions. Ligand basicities also influence the product distribution to a lesser extent. More basic phosphorus tend to lower the growth factor of the Schulz-Flory distribution. The typical catalyst concentrations are in the range of about 10 ppm to about 1000 ppm of the transition metal. The molar ratio of catalyst activator to metal salt is typically in the range O.l-3:l. The optimum molar ratio of metal salt to organophosphorus sulfonate is considered to be about 2 to 1 [P1631. The most preferred solvent for ethylene oligomerization is sulfolane in which the catalyst is soluble but not the oligomer. Sulfolane has the following advantages - good partitioning of the organophosphorus sulfonate in the sulfonate phase; very high catalyst activity/productivity using appropriate generation ratios; increased a-olefin selectivity by minimizing olefin isomerization caused by catalystioligomer contact; lower catalyst usage and concentrations; and reduction of the solvent degradation problem [P163]. Optimum reaction conditions are quite different for various ligand structures. For example, the more alkylphosphorus substituents in the phosphinosulfonates, the more stable is the catalyst at higher reaction temperatures. Optimum oligomerization temperatures are in the range of 30140°C with a pressure range of 27-136 atm. However, a commercial unit is run in the range of 60-130°C [P1631. High process productivity is achieved by optimum selection of the organophosphorus sulfonate ligand, transition metal compound, catalyst activator, concentration, solvent, and reaction parameters. For example, a sodium salt of ortho-diphenyl phosphino paratoluenesulfonic acid/nickel chloride hexahydrate/sodium borohydride based catalyst gives good activity at 3O”C, 65 atm, 1000 ppm nickel salt, and a reaction rate of 0.5 gram-mole per liter per hour; the cc-olefin products are mainly in the C4-C12 range [P1631The catalyst turnover frequency achieved in sulfolane solvent is greater than 10 ethylene molecules/nickel atom per second compared to a turnover frequency of 0.6 for the nickel diphenylphosphino acetic acid based catalyst used in the oligomerization step of the SHOP process [1931. Typical a-olefins distribution produced using Ni(BF4)2*6H20 in sulfolane at 80°C and 65 atm, contains 17.6 wt.% C4, 42.7% Cf#lo, 31.6% C!l&I!ls, and 8.1% C2o+ olefins. High catalyst activity, productivity, and 96+% linear cc-olefins purities can be achieved under appropriate reaction conditions and low catalyst concentrations. Alpha-olefins yield benefits from reduced catalyst/oligomer contact which otherwise can lead to secondary reactions (isomerization and homologation) between olefin products and catalyst components [ 1931. Dour Process Dow patents discussed the preparation of C4-Cl0 a-olefins, having high content of hexene-1 and octene-1, in a two-step process comprising a growth step

102

in a conventional coil reactor [PI211 or tank reactor CP1221,followed by a displacement step in a coil reactor [Pl21] or a static mixer [P1221. The growth reaction is carried out in the presence of a mixture of Et3Al and Bu3Al in tetradecane solvent, at llO-12O’C and 20-50 atm. The displacement reaction is carried out in the presence of recycled ethylene and/or butene-1 at 250-270X and lo-20 atm. The growth reaction zone contents of the tank reactor are recirculated in 0.5-3 min. through an external heat transfer zone to remove heat of reaction. A typical a-olefin product distribution obtained by a tankgrowth reactor type consists of 22.2% C4, 60.1% Cs, 15% Cs, 2.3% Clo, and 0.1% C12+olefins Cp1221. zB.4 ~~nof~~~ Et3Al is currently used in the commercial production of linear a-olefins by processes assigned to Chevron and Ethyl. Almost all the information on oligomerization of ethylene using Et3Al is patented by Chevron and Ethyl except a few patents assigned to Dow, Vista, and Sumitomo [P121-P1281. Tables 2.17 and 2.18 present typical operating conditions, c+olefins selectivity, and si~ificant remarks of ethylene oligome~zation processes using Et&l catalyst. Table 2.17 Operating conditions and ol-olefins selectivities t~ethylaluminum in the oligome~zation of ethylene Process licenser

m Temp.

Pressure

B Temp.

Pressure

(“C)

(atm)

(‘C)

mm)

of processes

using

a-olefins selectivity (wt.%) C4 c12-Cl8 c20+ c6410

Chevron1 f Gulfl

SO-220

140

-

-

13.5

40.5

30.0

15.5

Ethyl2

100-130

99106

280-320

lo-20

11.5

59.6

26.9

-

Dow233

llO-l20

20-50

250-270

10-20

22.2

77.4

0.1

*

Vista2

115120

95-115

60-100

2.3-6.9

6.7

57.0

34.8

l One-step Ziegler process. 2 Two-step modified Ziegler process. 3 A mixture of’Et&l and BudI

was used.

1.5

103

Table 2.18 Comparison of ethylene oligomerization processes that use triethylaluminum in the production of a-olefins Olefins used in displacement step

Butene-1 yield

C&lo yield

Commercial one-step

__

High

Low

- Catalytic amount of EtgAl is used - High temperature oligomerization - Formation of polymeric byproducts - Limited product distribution flexibility - High yield of Cgo+ waxy olefins.

Commercial two-step

c6-cl0

High

High

- Et3Al preparation and

Process licenser

Process type

Chevron

Ethyl

(Gum

Remarks

recycle

- High temperature displacement step

- Product distribution flexibility - Formation of polymeric byproducts - Alcohol production. Dow

Two-step

Cg and/or

High

High

- Mixture of EtgAl and BugAl is used - High yield of hexene-1 and octane-l - Tank growth reactor with external recirculation - No ethylene recycle to the growth step

Low

High

- Preparation of [email protected] - Low temperature displace-

c4

Vista (Conoco)

Two step

c2

ment step using nickel catalyst - Aluminum alkyls conversion by complexation - Product distribution flexibility - Production of alumina as co-product.

104

L7

CONCLUSIONS

The evaluation of results obtained in this review shows that ethylene oligomerization is of considerable academic and industrial interest for the synthesis of linear a-olefins. These olefins are used in the manufacture of plasticizer alcohols, special detergents, synthetic lubricating oils, and as ethylene comonomers. The catalytic systems reported in the literature include aluminum alkyls and a variety of transition metal complexes of nickel, zirconium, and titanium. The activity and selectivity of these complexes depend on the electronic and steric factors of the central metal ion. Currently, the commercial a-olefin processes are assigned to Chevron, Ethyl, Mistubishi, and Shell. Both Chevron and Ethyl processes are based on the triethylaluminum (Et3Al) technology licensed by Ziegler, while Mitsubishi is using the Chevron’s technology. Shell has developed its own oligomerization process based on a non-Ziegler nickel-ligand catalytic complex. The olefin product distribution of these processes is highly dependent on the type of catalysts used, reaction conditions, and process design. Chevron produces a wide range of a-olefins in a single-step catalytic Et&l process. in a two-step

Ethyl maximizes the production of plasticizer range a-olefins stoichiometric Et3Al process. Shell produces a high-purity

detergent range olefins by upgrading the lower and higher a-olefins using isomerization and disproportionation steps. Process research by a number of companies has resulted in various. approaches to the production of a-olefins by ethylene oligomerization and development of new catalytic systems. Vista’s process is similar to the Ethyl technology with some modifications in the separation and recovery of the Et&l. The main a-olefin products are in the plasticizer range associated with the production of alumina as a co-product. Dow uses aluminum alkyls in a two-step ethylene oligomerization process for the production of C4-Cs olefins The Exxon and Idemitsu a-olefin processes used as ethylene comonomers. use zirconium based catalytic systems to produce a wide range of olefins. Chevron, Union Carbide, and Phillips have developed nickel based catalytic systems for the selective production of linear a-olefms. The current research and development efforts in the field of ethylene oligomerization are directed towards the development of new-improved highly selective catalytic systems, as well as the improvement of existing process technology. The objective of these efforts are directed towards the development of active, stable and selective catalyst in producing narrow range a-olefins. In evaluating these catalytic systems, ease of preparation, catalyst performance, and potential for industrial application are used as guidelines. The trend away from aluminum alkyls and titanium based oligomerization catalysts to nickel and zirconium based systems is due to several factors. These include severe reaction conditions, difficulty in catalyst separation from olefin products and the pyrophoric nature of aluminum alkyls as well as the low product quality and low stability of titanium tetrachloride catalytic systems.

105 The performance of sulfonated nickel ylides is superior to other nickel based catalysts. Sulfonation has improved catalyst activity, productivity, catalyst solubility in polar solvents and ease of separation from oligomer products. The sulfonated nickel ylides are operated at low temperature and pressure as well as at low catalyst to ethylene ratio. Metal oxides and zeolites are generally less selective catalysts than transition metal complex catalysts. High branched oligomers are mainly formed in the presence of these catalysts, particularly at high temperatures. Zirconium based catalysts are stable and active for oligomerizing ethylene to linear cr-olefins. The catalysts used are zirconium tetrachloride, zirconium alkoxides and ester complex of zirconium tetrachloride in combination with chloroalkylaluminums, diethyl zinc and diethyl aluminum alkoxides or The additives, such as hexamethylbenzene, diamides co-catalysts. phosphines or organic disulfides, are used to reduce the formation of polymeric by-products. The product distribution is highly dependent on the molar ratio of AVZr used. Zirconium tetrachloride with cocatalysts other than chloroalkylaluminums produce higher molecular weight olefins. Zirconium alkoxides require higher amounts of cocatalysts compared to zirconium tetrachloride. The catalyst system, ester complex of zirconium tetrachloride with diethylaluminiumchloride, changes to a polymerization catalyst in the presence of even trace levels of water.

ABH3EVIATIONS

AC acac Ar atm Bu Bz COD COT CP

CY Dmpe EPA F i min M MTBE Me :-AC OPh R?

Acetyl group acetylacetonato ligand (2,4-pentanedionato) Aryl group Atmospheric pressure Butyl group Benzyl group Cyclooctadiene Cyclooctatetraene Cyclopentadienyl group cyclohexyl group 1,2-bis (dimethylphosphino) ethane Electron paramagnetic resonance Ethyl group hour is0 minute The central metal in a complex or a metal atom on a surface Methyl tertiary butyl ether Methyl group normal Acetoxyl group, acetato anion Phenoxy group Bifunctional phosphine Phenyl group

106 pph3

Pr

? SPl &F TMEDA X

Triphenyl phosphine Propyl group Pyridine An alkyl or aryl group Square planar tertiary Tetrahydrofuran N,N,N,N-tetramethyl ethylene diamine Halogen or halide anion

ACKNOWLEDGMENTS The authors acknowledge the support of the Research Institute, King Fahd University of Petroleum and Minerals and the Saudi Basic Industries Corporation, SABIC (Research Project PN 210881 in conducting this work. They also thank Syed A. Ali for preparing the references database and Nihal Ahmad for preparing the final manuscript.

1 2 3 4 5 6 ;: 9 10 Z 13 14 15 16

G. Lappin and J. Sauer teds.), Alpha-Olefins Applications Handbook, Marcel Dekker, New York, 1989. F. Marta and D. Kallo teds.), Mechanisms of Hydrocarbon Reactions, Elsevier, Amsterdam, 1975. Chem. System Inc., Butene-1 and Other LLDPE Comonomers, Report by Chem. Systems Inc., New York, 1986. D. Commereuc, Y. Chauvin, and J. Gaillard, Hydrocarbon Proc., 63, 11 (1984) 118. IFP, Alphabutol Process for Butene-1 Manufacturing, IFP Publications, Paris, 1988. A. Hennico, J. Leonard, A. Forestiere, and Y. Glaize, Hydrocarbon Proc., 69,3 (1990) 73. W. Keim, Angew. Chem. Int. Ed. Engl., 29 (1990) 235. Ethyl Technical Bulletin, Alpha Olefins, Ethyl Chemicals Group, 1981 Chevron Technical Bulletin, Gulfene Olefins, Chevron Chemical Company, 1988. Shell Technical Bulletin Neodene Alpha and Internal Olefins, Shell Chemical Company, 1989. M. Sherwood, Chemistry and Industry, Dec. (1982) 994. N. Alperowicz, Chemical Week, Nov. 29 (1989) 11. A. H. Turner, J. Am. Oil Chem. Sot., 60 (1983) 623. P. Layman, Chem. & Eng. News, 66,22 (1988) 9. W. Keim, A. Behr, and M. Roper, Comprehensive Organometallic Chemistry, G. Wilkinson et al. (eds.), Vol. 8, Pergamon Press, Oxford, pp. 371-384, 1982. S. M. Pillai, G. L. Tembe, M. Ravindranathan, and S. Sivaram, Ind. Eng. Chem. Res., 27 (1988) 1971.

107 17

23 24 25

27 23 23 30

31 32 33

39 40 41 42

43 44 45

G. Lefebvre and Y. Chauvin, Aspects of Homogeneous Catalysis, R. Ugo, ted.), vol. 1, Carlo Manfredi, Milan, pp. 107-201, 1970. S. Pillai, M. Ravindranathan, and S. Sivaram, Chem. Rev., 86 (1986) 353. A. Bre, Y. Chauvin, and D. Commereuc, New J. of Chem., 10 (1986) 535. P. E. Mathovskii et al., Kinet. & Catal., 19 (1978) 209. P. E. Matkovskii et al., Petroleum Chemistry,. 22 (1982) 76. D. Ballard, Catalysis-Homogenous and Heterogenous, Proc. Intl. Symp. B. Delmon et al. teds.), Elsevier, Amsterdam, pp 253-262,1975. P. D. Smith, D. D. Klendworth, and M. P. McDaniel, J. Catal, 105 (1987) 187. V. Zhukov and G. Belov, Petroleum Chemistry, 22 (1982) 150. M. Khan and A. Martell, Oligomerization of Ethylene and Alpha Olefins, Homogeneous Catalysis by Metal Complexes, Academic Press, New York, pp. 135-141, 1974. D. L. Beach and Y. V. Kissin, J. Polym. Sci., Polym. Chem. Ed., 22 (1984) 3027. P. Vernov and G. Belov, Neftepererab. Neftekhim., 8 (1977) 25; Chem. Abstracts 87 (1977) 167481. V. Zhukov and N. Shestak, Izv. Sev. Kauk. Nauchn. Tsentra Vyssh. Shk., 3 (1975) 93; Chem. Abstracts, 83 (1975) 193844. J. Vybihal, Chem. Prum., 31 (1981) 64; Chem. Abstracts, 96 (1982) 69463. G. Belov and V. Smirnov, Petroleum Chemistry, 18 (1978) 223-227. S. Yamada and I. Ono, Bull. Japan Petrol. Inst., 12 (1970) 160. G. Belov and A. Brikenshtein, Net&p. Nefiekhim., 6 (1989) 14. G. Belov and Z. Dzhabieva, Tezisy Dokl., 12th, 2 (1975) 201; Chem. Abstracts, 85 (1976) 159101. T. Dzhabiev, Nefiekhimiya, 14 (1974) 700. G. Belov, T. S. Dzhabiev, and I. M. Kolesnikov, J. Mol. Catal., 14 (1982) 105. P. Cossee, J. Catal. 3 (1964) 80. P. Cassoux, F. Crasnier, and J. Labarre, J. Organometallic Chem. 165 (1979) 303. S. S Sivaram and V. K. Gupta, ACS Pet. Chem. (Preprints), 24 (1989) 595. Y. V. Kissin and D. L. Beach, J. Polym. Sci., Polym Chem. Ed., 24 (1984) 1069. S. Cesca, W. Marconi, and M. L. Santostasi, J. Polym. Sci., Part B, 7 (1969) 550. K. Shikata et al., Kogy Kagaku Zasshi (Japan), 68 (1965) 2266; Chem. Abstracts, 64 (1966) 15721. S. Datta, M. B. Fischer, and S. S. Wreford, J. Organometallic Chem., 188 (1980) 353. H. Ah, K. Schwind, and M. Rausch, J. Org. Chem., 349 (1988) C7. P. W. Jolly and G. Wilke, The Organic Chemistry of Nickel, vol. 2, Academic Press, New York, pp. l-33, 1975. P. E. Mathovskii et al., Neftekhimiya, 22 (1982) 317; Chem. Abstracts, 97, (1982) 144267.

108

46

47 48 49

E 52

55 56

57 58

67

70 71 g

76

K. Maruyama, T. Kuroki, and T. Mizoroki, Bull. Chem. Sot. Japan, 44 (1971) 2002. T. Koike, K. Kawakami, K. Maruya, and T. Mizoroki, Chem. Lett., (1977) 551. I. Chernikovaand I. Pisman, Zerb. Khim. Zh. 6 (1965) 35; Chem. Abstracts, 64 (1966) 19385. I. Pisman and I. Chernikova, Azerb. Khim, 2 (1968) 73; Chem. Abstracts, 69 (1968) 105782. S. SibiRe, J. Coulombeix, and J. Perichon, J. Mol. Catal., 32 (1985) 239. J. Brunet, A Sivade, and I. Tkatchenko, J. Mol. Catal., 50 (1989) 291. Y. Ishimura, K. Maruya, and Y. Nakamura, Bull. Chem. Sot. Japan, 56 (1983) 818. R. Grubbs and A. Miyashita, J. Am. Chem. Sot., 100 (1978) 7416. U. Birkenstock, H. Bonnemann, and B. Bogdanovic, Adv. Chem. Ser. No. 70 (1968) 250. G. Ghymes, A. Grobler, and A. Simon, Kern. Lapja., 20 (1965) 570; Chem. Abstracts, 64 (1966) 6471. M. Tsutsui and T. Koyano, J. Polym. Sci., Polymer Chemistry Ed., 5 (1967) 681. S. B. Choe, H. Kanai, and K. KIabunde, J. Amer. Chem. Sot., 111 (1989) 2875. T. Matsuda, H. Miura, and K. Sugiyama, J. Chem. Sot., Farad. Trans. I, 75 (1979) 1513. K. Kimura, A. Hideo, and A. Ozaki, J. Catal., 18 (1970) 271. T. Yashima, Y. Ushida, and M. Ebisawa, J. Catal., 36 (1975) 320. Y. Ono, N. Kikuchi, and H. Watanabe, Preparation of Catalysts IV, Delmon et al. teds.), Elsevier, Amsterdam, 1987,519. J. R. Sohn and A. Ozaki, J. Catal., 610980) 29. J. Sohn and H. Kim, Bull. Korean Chem. Sot., 6 (1985) 248. J. R. Sohn, H. W. Kim, and J. T. Kim, J. Mol. Catal., 410987) 375. L. Bonneviot, D. Oliver, and M. Che, J. Mol. Catal., 21 (1983) 415. F. X. Cai, C. Lepetit, M. Kermarec, and D. Olivier, J. Mol. Catal., 43 (1987) 93. F. X. Cai. C. Lepetit, M. Kermarec, and D. Olivier, Dimerization of Ethylene and Propylene with Tailor made Nickel Complexes Supported on Silica, Proc. 5th Int. Sym. on Homog. & Heterog. Catal., Y. Ermakov et al. (eds.), VNU Sci. Press, Utrecht, 1986,419. R. Ceder, G. Muller, J. Sales, and J. Vidal, J. Mol. CataI., 68 (1991) 23. A. Pomogailo, F. Khrisotomov, and F. Dyachkovskii, Kin. Catal., 26 (1985) 953. S. B. Echmaev and I. N. Ivleva, Kinet. Catal., 21(1980) 1099. J. C. Carlu and C. Caze, Reactive Polymers, 13 (1990) 153. V. A. Kabanov and G. Grishin, Kin. Catal., 26 (1984) 1230. V. A. Borokov, V. Kaverinsky, and V. Kazansky, Homogeneous and Heterogeneous Catalysis, Elsevier, Amsterdam, 1975, 253. L. Riekert, , J. Catal., 19 (1970) 8. I. V. Nicolescu, L. Botez, and E. Angelescu, Revue Roumaine de Chemie, 35,2 (1990) 199. E. Angelescu and M. Udrea, Proc. VIth Intl. Symp., Sofia, Heterogeneous Catalysis, 1987,55.

109

77 78 79 80

81 : : 86

K. Kawakami, T. Mizoroki, and A. Ozaki, Bull. Chem. Soe. Jap., 51(1978) 21. P. Sun, L. A. Yomanoto, and S. Ikeda, J. Am. Chem. Sot., 90 (1968) 7170. S. Mclain, J. Sancho, and R. Schrock, J. Am. Chem. Sot., 101 (1979) ~~c~~n, J. Sancho, and R. Schrock, J. Am. Chem. Sot., 102 (1980) 5610. R. S&rock, S. Mclain, and J. Sancho, Pure and Appl. Chem., 52 (1980) 729. R. Schrock and J. Fellmann, J. Am. Chem. Sot., 100 (1978) 3359. J. D. Fellmann, G. A. Rupprecht, and R. R. Schrock, J. Amer. Chem. So& lOl(1979) 5099. S. Mclain and R. Schrock, J. Am. Chem. Sot., 100 (19’78) 1315. B. Chang, C. Lau, R. H. Grubbs, and C. Brubaker, J. Organometafhc Chem., 2810985) 213. T. Alderson, E. Jenner, and R. V. Lindsey, J. Am. Chem. Sot., 87 (1965)

105 106

Earner Ace Chem Res l(1968) 186 R: Cramer: J. Am. Chem. Sb’c., 87 (1965) 4717. N. Takahashi, I. Okura, and T. Keii, J. Am. Chem. Sot., 97 (1975) 7489. V. Igantov and N. Borunova, Nef’tekhimiya, 22 (1982) 749. K. Kawamoto, T. Imanaka, and S. Teranishi, Nippon Kagaku Zasshi, 91(1970) 39; Chem. Abstracts, 72 (1972) 99767. Y. Kusunoki, R. Katsuno, and S. Tsutsumi, Bull. Chem. Sot. Japan, 39 (1966) 2921. T. Kitamura, K. Maruya, M. Yoshihiko, and Ozaki, Bull. Chem. Sot. Japan, 45 (1972) 1457. P. Pertici and G. Vitulh, Tetrahedron Lett., 21(1979) 1897. S. Delgado, R. Arvelo, D. Trujillo, and J. Dominguez, Indian Chem. Eng., 29 (1987) 3,11. Y. Usov, N. Kuvshinova, and E. Chekourovskaya, Deposited Document, SPSTL 898, (1980) 12; Chem. Abstracts, 97 (1982) 91661. Y. Ysov, N. Kuvshinova, and E. Chekourovskaya, Kinet. Catal., 19 (1978) 1309. Y. Usov, N. Kuvshinova, and E. Chekourovskaya, Nefiekhimiya, 18 (1978) 369. J. Michalik, H. Lee, and L. Kevan, J. Phys. Chem., 89 (1985) 4282. A. K. Ghosh and L. Kevan, J. Phys. Chem., 93 (1989) 3747. A. K Ghosh and L. Kevan, J. Phys. Chem,, 92 (1988) 4439. A. K. Ghosh and L. Kevan, J. Amer. Chem. Sot., 110 (1988) 8044. H. Imamura et al., J. Mol. Catal., 52 (1989) 25. A. Morikawa, T. Nakashima, I. Nishiyama, and K. Otsuka, Proc. 8th Int. Catal. Congress, vol. 4, Verlag Chemie, Weir&am, 1985, 815. E. A. Zuech, J. Polym. Sci., Polym. Chem. Ed., 10 (1972) 3665. Y. Wiefstra, S. Gambarotta, and M. Chiang, Organometallics, 7 (1988)

107 108 109

:8% Wideman H Menapace and N Maly J Catal 43 (1976) 371 M. Green and J. Knight, J. C&em. So;., Da&on Trani., 70 (1974) 311. K. Egger, Trans. Farad. Sot., 67 (1971) 2638.

87 : : 92 93 g 96 97 93 99 100 101 102 103 104

110

110 111 112 1X3 114 :: 117 118 119 l26 Z l23 124 Z 127 Z 130 131 132 Z x35 136 137 x38 139 146

L. F. Albright and C. Smith, AIChE J., 14 (1968) 325. C. Bawn and R. Symox, J. Polym. Sci., 34 (1959) 139. V. Nevzorov and G. Belov, Neftekhimiya, 29 (1989) 340; Chem. Abstracts, 112 (1990) 36489. M. Tsutsui, J. Aryoshi, and T. Koyano, Adv. Chem. Ser., 70 (1968) 266. A. Nakamura and M. Tsutsui, Oligomerizations, Principles and Applications of Homogenous Catalysis, John Wiley, New York, 1980,133. Y. Ishimura and K. Maruya, Chemistry Lett., (1981) 657. K. Maruya, T. Mizoroki, and T. Ozaku, Bull. Chem. Sot. Japan, 45 (1972) 2255. G. Wendt, J. Fenster, R. Schollner, and H. Siegel, New Horizons in Catalysis, T. Seiyama et al. (eds.), Elsevier, Amsterdam, 1981, 978. J. R. Sohn and H. J. Kim, J. CataI., lOl(1986) 428. I. E. Vflyand and A, D. Pomogailo, J. Mol. Catal., 55 (1989) 302. K. Urabe, M. Koga, and Y. Izumi, J. Chem. Sot., Chem. Commun., (1989) 807. J. R. Sohn and A. Ozaki, J. Catal., 59 (1979) 303. J. D. Fellman, R. R. Schrock, and G. A. Rupprecht, J. Am. Chem. Sot., 103 (1981) 5752. U. Dorf, K. Engel, and G. Erker, Agnew. Chem. Int. Ed. Engl., 21 (1982) 1982. Y. Okamoto, N. Ishida, T. Imanaka, and S. Teranishi, J. Catal., 58 (1979) 82. J. Skupinska, Chem. Rev., 91(1991) 613. A. Langer, ACS Pet. Chem. Preprints, 17 (1972) B119. G. Hen&i-Olive, and S. Olive, Coordination Polymerization, J. Chien, (ed.), Academic Press, New York, 1974, 291. E. Frietas and C. R. Gum, Shell Technical Bulletin, SC 335-8,1984. P. A. Galtier, A. Forestiere, Y. Glaize, and J. Wauquier,Chem. Eng. Sci., 43 (1988) 1855. R. Quirk ted.), Transition Metal Catalyzed Polymerization, Cambridge University Press, Cambridge, 1988,751. J. R. Zietz, G. C. Robinson, and K. L. Lindsay, Comprehensive Organometallic Chemistry, G. Wilkinson et al. (eds.), Pergamon Press, Oxford, vol. 7,1982,365. K. L. Lindsay, Encyclolpedia of Chemical Processing and Design, (J. McKetta et al., ed.), Marcel Dekker, New York, 1982,482. M. G. Dokurno and P. L. Douglas, Can. J. Chem. Eng., 62 (1984) 818. E. R. Tucci, Ind. Eng. Chem. Prod. Des. Dev., 5 (1966) 161. W. Keim, B. Hoffman, R. Lodewick, M. Peuckert, G. Schmitt, and U. Meier, J. Mol. CataI., 6 (1979) 79. P. W. Jolly, Comprehensive Organometallic Chemistry, G. Wilkinson, et al. teds.), Pergamon Press, Oxford, vol. 7, 1982,613. A. Behr and W. Keim, Arab. J. of Sci. & Eng., 10 (1985) 377. W. Keim, F. Kowaldt, R. Goddard, and C. Kruger, Agnew Chem. Int. Eng., 17 (1978) 466. W. Keim, Catal. Polym. of Olefins. Stud. Surf. Sci. & Catal., vol. 25, Elsevier, Amsterdam, 1986,201. M. Peukert and W. Keim, Organometallics, 2 (1983) 594.

111

141 142 143

144 145 146 147 148 149 150 151 152

E 155 156 157 158 159 160 161 162 163

Ei 166 167 168 169

M. Peukert, W. Keim, S. Storp, and R. S. Webber, J. Mol. Catal., 20 (1983) 115. W. Keim, A. Behr, B. Limbacker, and C. Kruger, Agnew. Chem. Int. Ed. Engl., 22 (1983) 503. U. Muller, W. Keim, C. Kruger, and P. Betz, Angew. Chem. Int. Ed. Engl., 28 (1989) 1011. A. Behr, V. Flabe, V. Freuendenberg, and W. Keim, Isr. J. Chem., 27 0986) 277. Y. V. Kissin, J. Polym. Sci., Polym. Chem. Ed., 27 (1989) 147. R. D. Knudsen, ACS Pet. Chem. (Preprints), 24 (1989) 572. S. J. Brown and A. Masters, J. Organometallic. Chem., 367 (1989) 371. J. Gehrke and R. Taube, J. Orgometallic Chem., 304 (1986) C4. G. Nestrov, V. Zakharov, G. Rink, and W. Fenzel, J. Mol. Catal., 66 (1991) 367. G. Nestrov, V. Zakharov, G. Rink, and W. Fenzel, J. Mol. Catal., 69 (1991) 129. J. Heveling, C. Nicolaides, and M. Scurrell, J. Chem. Sot., Chem. Commun., (1991) 126. J. Heveling, C. Nicolaides, and M. Scurrell,, Proc. of the ACS Symposium on Alkylation, Aromatization and Oligomerization of Short Chain Hydrocarbons, New York City Meeting, August 25-30, 1991. M. Peukert and W. Keim, J. Mol. Catal., 22 (1984) 289. W. Keim, Proc. 5th Int. Symp. on Homog. & Heterog. Catal., Y. Ermakov et al. (eds.), VNU Sci. Press, Utrecht, 1986,499. D. L. Beach and T. P. Kobylinski, J. Chem. Sot., Chem. Comm., (1980) 933. R. L. Espinoza, R. Snel, C. Korf, and C. Nicholaides, App. Catal., 29 (1987) 295. R. L. Espinoza, C. Nicolaides, C. Korf, and R. Snel, App. Catal., 31 (1987) 259. R.L. Espinoza, C. Nicolaides,C. Korf, and R. Snel, App. Catal., 29 (1987) 175. I. V. Elev, B. N. Shelimov, and V. B. Kazanskii, Kin. & Catal., 25 (1984) 955. V. B. Kazansky, I. V. Elev, and B. N. Shelinov, J. Mol. Catal., 21(1983) 265. A. W. Langer, J. Macromol Sci.-Chem., A4 (1970) 775. F. Dyachkovskii et al., Vysokomol Soedin., 20 (1978) 746; Chem. Abstracts, 88 (1978) 170586. L. Russiyan, et al., Vysokomol. Soedin, Ser. A, 21 (1979) 1831; Chem. Abstracts, 91(1979) 175769. G. Tembe and M. Ravidranathan, Ind. Eng. Chem. Res., 30 (1991) 2247. V. G. Popov and S. V. Visilchenko, Kin. & Catal., 29 (1988) 847. N. A. Byrikhina and E. L. Pechatnikos, J. App. Chem., 59 (1986) 332. V. N. Melnikov, P. Matkovskii, L. Russiyan, and Z. Bushiva, Kin. & Catal., 29 (1988) 107. V. Melmkov, P. Matkovskii, L. Russiyan, and Z. Bushiva, Khim. Prom. St., 6 (1986) 323; Chem Abstracts, 105 (1986) 153589. P. Longi, R. Jackson, and S. Maddock, Chim. Ind. (Milan), 55 (1973) 252. Chem. Abstracts, 79 (1973) 7923.

112

170 171 172 173 174 175 176 177 178 179 Z 182 183

I.86 187

ml 191

192 193 194

C. Attridge et al., J. Chem. Sot., Chem. Comm., (1973) 132. D. Young, J. Mol. Catal., 53 (1989) 433. R. He, J. Jiang, and Y. Wang, Fermi. Cuihua, 2 (1988) 283; Chem. Abstracts, 110 (1989) 232136. J. W. Ward (ed), Catalysis 1987, Elsevier, Amsterdam, 1988, 187. P. A. Jacobs et al. teds.), Structure and Reactivity of Modified Zeolites, Elsevier, Amsterdam, 1984,241. E. G. Derouane, J. P. Gilson, and J. B. Nagy, J. Molec. Catal., 10 (1981) 331. T. G. Kofke and R. J. Gorte, J. Catal., 115 (19891.233. E. G. Derouane, C. Lefebvre, and J. Nagy, J. Mol. Catal., 38 (1986) 387. 0. V. Bragin and L. M. Kustov, Kin. & Catal., 19 (1988) 1210. W. Reagen, ACS Pet. Chem (Preprints), 24 (1989) 583. J. Briggs, J. Chem. Sot., Chem. Commun., (1989) 674. A. G. An&its and W. K. Hall, Kin. & Catal. 29 (1988) 540. J. Hay, J. McCabe, and J. Robb, J. Chem. Sot., Farad. Trans. I, 69 (1972) 1227. F. Rodriguez, M. Abadie, and F. Schue, European Polymer J., 12 (1976) 17. G. Grassous and F. Schue, Polymer, 24 (1983) 1203. H. Magnin, F. Rodriguez, M. Abadie, and F. Schue, J. Polymer Sci., Polym. Chem. Ed., 15 (1977) 875. ChemSystems Intl., Chemical Process Economics, Report by ChemSystems, vol. 4, Processes No. 124 and 125, London, 1988. D. G. Demianiw, Encyclopedia of Chemical Technology, Kirk-Othmer (eds.), John Wiley, New York, vol. 16,1988,480. B. Reuben and H. Wittcoff, J. Chem. Education, 65,7 (1988) 605. W. Keim, Fundamental Research in Homogeneous Catalysis, M. Grazaini et al. (eds.), Plenum Press, New York, vol. 4, 1984, 131. E. F. Lutz, J. of Chem. Education, 63, 3 (1986) 202. Lummus Crest Inc., Report, Comparison of Ethylene Oligomerization Technology for the Production of Linear Alpha Olefins, Combustion Engineering, N.J., U.S.A. 1989. Lummus Crest Inc., Report, Alpha-olefins Technology, General Information, Combustion Engineering, N.J., U.S.A. 1989. R. Murray and T. Wenzel, ACS Pet. Chem. Preprints, 24 (1989) 599. J. Hagin, Chem. & Eng. News, 67,39 (1989) 42.

PAPl P2 P3 P4 P5

N. LeQuan et al., (IFP), Butene-1 by Ethylene Dimerization, European Patent No. 135,441(1985). D. Commereuc et al., (IFP), Butene-1 by Dimerization of Ethylene, French Patent No. 2,546,488 (1984). Y. Chauvin et al., (IFP), Pure Butene-1 from the Crude Product of Ethylene Dimerization, European Patent No. 200,654 (1986). B. Labatut et al., (IFP), Ethylene-Butene-1 Copolymers by Polymerization of Ethylene, European Patent No. 183,684 (1986). D. Commereuc et al. (IFP), Ethylene - Butene-1 Copolymers from ethylene, French Patent No. 2,565,591(1985).

113

I% P7 P8 P9 PlO Pll P12 P13 P14 P15 P16 P17 P18 P19

P21

P24

f2 P27

P29 P30

S. Pillai et al. (Indian Petrochem. Corp. Ltd.), Process for making Butene-1 from Ethylene, European Patent No. 221,206 (1987). M. P. McDaniel et al., (Phillips Petroleum), Ethylene Dimerization and Polymerization, European Patent No. 230,983 (1987). J. VybihaI (ChemopetroZ), Butene-1, Czechoslovakian Patent No. 179,726 (1979). J. Schoengut et al., (ChemopetroZ), Oligomerization of Lower Olefins, Czechoslovakian Patent No. 189,984 (1982). J. Vybihal, (ChemopetroZ), Butene-1, Czechoslovakian Patent No. 179,183 (1979). J. Vybihal, (ChemopetroZ), Butene-1, Czechoslovakian Patent No. 179,248 (1979). J. Vybihal, (ChemopetroZ), Separation of a Catalyst from the Reaction Product of Ethylene Dimerization to Butene-1, Czechoslovakian Patent No. 187,822 (1981). Mitsui Toatsu Chemicals, Dimerization of Ethylene, Japanese Patent No. 57 169,430 (1982). M. Kioka et al., (Mitsui Petrochem.), Oligomerization of Alpha-Olefins, Japanese Patent No. 62 72,631(1987). B. Schleppinghoff et al., (Erdoel-Chemie), Butene-1 by Ethylene Dimerization, German Patent No. 3,301,162 (1984). G. Pez (Allied Chem.) Lithium Dicyclopentadienyl Titanium Hydride Composition and their Use in Polymerizing Acetylene and Ethylene, European Patent No. 11,767 (1980). Chemische Werhe Huels, Preparation of Alpha-Butene from Ethylene, Belgian Patent No. 634,232 (1963). K. Azuma et al., (Tokuyama Soda), Low Polymers of Ethylene, Japanese Patent No. 69 04,961(1969). M. Izawa et al., (Toy0 Soda), Dimerization of Ethylene, German Patent No. 1,803,434 (1969). S. Izawa et al., (Toy0 Soda), Butene-1, Japanese Patent No. 71,43,366 (1971). K. Yamado et al., (Toy0 Soda), Preparation of Butene-1 by Dimerization of Ethylene, Japanese Patent No. 62,62932 (1987). S. Izawa et al. (Toy0 Soda), Dimerization and Co-Dimerization of AlphaOlefins, Japanese Patent No. 72 01,444 (1972). S. Yamada et al., (Toy0 Soda), Alpha-Olefins Dimerization, German Patent No. 2,026,246 (1971). W. Marconi et al., (Snamprogetti), Butene-1 by Catalytic Dimerization of Ethylene, German Patent No. 2,022,650 (1970). Mitsubishi Pet. Co., Butene-1, Japanese Patent No. 58,146,517 (1983). Mitsubishi Pet. Co., Butene-1, Japanese Patent No. 58,146,518 (1983). K. J. Cann et al., (Union Carbide), Process for Simultaneously Dimerizing Ethylene and Copolymerizing Ethylene with Dimerized Product, World Organization Patent No. 86 05,500 (1986); K. Ziegler, Volatile Ethylene Polymers, UK Patent No. 787,438 (1957). G. Belov et al., (Inst. of Chemical Physics), Butene-1, US Patent No. 3,879,485 (1974). G. Belov et al., (Inst. of Chemical Physics), Butene-1, US Patent No. 3,911,042 (1975).

114

P31

P33

P37 P38 P39 P46 P41 P42 P43 P44 P45 P46 P47 P48 P49 P50 P51 P52 P53

G. Belov et al. (Inst. of Chemical Physics), Butene-1, Russian Patent No. 485,102 (1975). G. Belov et al., @wt. of Chemical Physics), Butene-1, Russian Patent No. 493,454 (1975). P. Mathovskii, (Inst. of Chemical Physics), Alpha-Butene, Russian Patent No. 496,258 (1975). G. Belov et al., (Inst. of Chemical Physics), Alpha Butylene, Russian Patent No. 485,101(1975X G. Belov et al., (Inst. of Chemical Physics), Butene-1, US Patent No. 3,969,429 (1976). G. Belov et al., (Inst. of Chemical Physics), Butene-1, French Patent No. 2,274,583 (1976). V. Shukov et al., (Inst. of Chemical Physics), Dimerization of AlphaOlefins, German Patent No. 2,287,830 (1977). I. Campbell (Inst. of Chemical Physics), Process for the Production of Butene-1, UK Patent No. 1,44,812 (1976). T. Shevlyakov et al., (Inst. of Chemical Physics), Dimerization of AlphaOlefins, French Patent No. 2,34,540 (1977). V. Zhukov et al., (Inst. of Chemical Physic), Manufacture of Butene-1, World Organization Patent No. 89 08,091(1989). Inst. of Chemical Physics, Method for Producing Butene-1, UK Patent No. 1,447,811(1976). G. Belov et al., (Inst. of Chemical Physics) Butene-1 by Dimerization of Ethylene in an Organic Solvent, German Patent No. 2,462,771(1980). Sci. Ind. Enterprises, Dimerization of Alpha-Olefins, Japanese Patent No. 77,113,903 (1977). V. Zhukov et al., (Scientific Ind.), Alpha-Butylene, Russian Patent No. 600,132 (1978). V. Zhukov et al., (Scientific Ind.), Dimerization of Alpha-Olefins, US Patent No. 4,101,600 (1978). D. R. Burnham, (ICI), Ethylene Oligomer Mixture, German Patent No. 2,525,404 (1975). D. R. Burnham, (ICI), Catalytic Oligomerization of Ethylene, German Patent No. 2,538,142 (1976). BASF, Dimerization of Alpha-Olefins, European Patent No. 1,178,419 (1964). C. Lepetit et al., (Centre National Research and Science), Nickel Catalyst on a Si Support for Selective Dimerization of Ethylene or Propylene in Monoolefins, European Patent No. 209,436 (1987). A. Masters et al., (Commonwealth Scientific Organization), Catalysts for Olefin Oligomerization and Isomerization, World Organization Patent No. 83,02,907 (1983). International Synthetic Rubber Co., Catalysts for the Oligomerization and Cooligomerization of Ethylene and Vinyl Cyclic Hydrocarbons, French Patent No. 1,599,774 (1970). S. Ozaki et al., (Tokyo Inst. of Technology), Catalysts for Dimerizing Ethylene Selectively, Japanese Patent No. 74 05,516 (1974). S. Ikeda et al., (Tokoyo Inst. of Technology), Transition Metal ComplexPolymer Adducts, German Patent No. 2,412,105 (1975)

115

S. Koskimes et al., (Neste Oy), Selective Dimerization of Ethylene to Butene-l using Amine or Acetal Derivatives Promoted Nickel Chelate Catalysts, Finnish Patent No. 76,067,1988. S. Izawa et al., (7’0~0 Soda), Butene, Japanese Patent No. 72 22,205 (1972). R. Knudsen (Phillips Petroleum), Ethylene Dimerization, US Patent No. P56 4,487,847 (1984). C. Carter, (Phillips Petroleum), Surface Coditioning in Olefin P57 Dimerization Reactors, US Patent No. 4,538,018 (1985). C. Carter, (Phillips Petroleum), Olefin Dimerization in a Loop Reactor, P58 US Patent No. 4,242,531(1980). C. Smith et al., (Phillips Petroleum), Dimerization of Monoolefins with P59 Catalysts on Solid Support, US Patent No. 428,103 (1976). H. Dunn, (Phillips Petroleum), Nickel Complex Catalysts for Olefin P60 Dimerization, US Patent No. 3644,218 (1972). P61 Y. Eidus et al., (Inst. of Organic Chemistry), Butylenes, Russian Patent No. 235,016 (1969). F62 V. Kabanov et al., (Inst. of Petrochem. Synthesis), Dimerization of Olefins, Belgian Patent No. 852,471(1977). A. Antonov, (Inst. of Petrochem. Synthesis), Dimerization of Oleflns, Belgian Patent No. 852,472 (1977). I. Bocharov et al., (Inst. of Petrochem. Synthesis), Dimerization of Olefins, Belgian Patent No. 852,473 (1977). V. S. Feldblyum et al., (Scientific Research Inst.), Olefin Dimerization, UK Patent No. 1,X4,882 (1969). V. S. Feldblyum et al., 6Scientific Research Inst.), Catalytic Dimerization of Olefins, French Patent No. 1,588,162 (1970). P67 V. Feldblyum et al., (Scientific Research Inst.), Catalyst for Dimerizing Ethylene or Propylene, Russian Patent No. 686,753 (1979). N. Petrushanskaya et al., (Scientific Ind. Enterprises), Catalyst for the Dimerization of Olefins, Russian Patent No. 290,764 (1971). Z. Zulfugarov et al., (Inst. of Znorg. & Phys. Chem.) Butenes, Russian Patent No. 950,706 (1982). Butenes, Russian F-70 D. B. Furman et al., (Inst. of Organic Chemistry), Patent No. 981,307 (1982). F71 G. Tolstikov et al., (Bashkir Inst. of Chemistry), E-2-Butene, Russian Patent No. 539,860 (1976). P72 E. Drent et al., (Shell), Catalysts for the Dimerization of Aliphatic Monoolefms, UK Patent No. 2,209,288 (1989). P73 T. Mithell et al. (Mobil), Complexed Metals Bonded to Inorganic Oxides, US Patent No. 3,980,583 (1976). F74 1CZ, Butenes Preparation from Ethylene using Rhodium Trichloride Catalysts, French Patent No. 1,521,991(1968). P75 A. Ketley, (Grace), Dimerization of Alpha-Olefins, French Patent No. 1,499,833 (1967). P76 R. Arganbright, (Petro-Tex Chem.), Butenes by Catalytic Dimerization of Ethylene, German Patent No. 2,210,751(1972). Research, Olefin Dimers, Japanese Patent No. P77 Shell International 61,145,126 (1986).

116

International Research, Dimerization of Alpha-Olefins, NETH 294,637 (1965). E. Drent, (Shell), Prepartion of Butene by Ethylene Dimerization, I?79 European Patent No. 307,026 (1989). F%o S. H. Klein, (Shell Oil), Ethylene Dimerization to n-Butenes, US Patent No. 3557,236 (1967). L. Wideman, (Goodyear), Butene-1, US Patent No. 3,813,453 (1972). E T. Hata, et al., (Toray Znd.), Ethylene Oligomers, Japanese Patent No. 7103,162 (1971). G. Hata, (Toy0 Rayon), Preparation of n-Butenes, Japanese Patent No. P83 69 02,203 (1969). W. Butte, (Sun Oil), Lithium-Cobalt Catalyst useful for the Dimerization, Isomerization and Hydrogenation of Olefins, US Patent No. 3542,896 (1970). W. Butte, (Sun Oil), Lithium-Cobalt Catalyst useful for the DimerP85 ization, Isomerization and Hydrogenation of Olefins, US Patent No. 3542,899 (1970). R. Reusser, (Phillips), Catalyst for Converting Ethylene to Butenes, US P86 3,364,278 (1968). F. Tebbe, (Du Pant), Hydride Complexes of Metals of Groups IVB and P87 VB, US Patent No. 3,933,876 (1976). R. R. Schrock, (Du Pant), Cyclooctatraene Complexes of Niobium and Tantalum, US Patent No. 3,969,429 (1976). P89 R. R. Schrock, (MIT), Tantalum Catalysts or Catalyst Precursors, US Patent No. 4,197,419 (1980). P90 R. R. Schrock, (MIT), Tantalum and Niobium Catalysts or Catalyst Precursors, US Patent No. 4,231,947 (1980). Dimerization of Olefins with Chromium Halide P91 E. Zuech, (Phillips), Complex Catalyst Systems, US Patent No. 3,726,939 (1973). E. Zuech, (PhilZips), Dimerization of Olefins with Chromium Halide Complex Catalyst System, US Patent No. 3,627,700 (1971). ), Olefins via the Oligomerization of P93 S. Tanaka, et al., (Mitsubishi Ethylene, Japanese Patent No. 70 01,441(1970X K. Saeki, et al., (Mitsui), Lower Polymerization of Ethylene and Propylene, Japanese Patent No. 73 19,601(1973X K. Saeki, et al., (Mitsui), Lower Polymers of Ethylene and Propylene, Japanese Patent No. 73,22,602 (1973). V. Dzhemilev, et al., (Bashkir Inst. of Chemistry), Butene-1, Russian Patent No. 857,097 (1981). Proc.), Process and Catalysts for P97 A. Azizov et al., (Inst. Petrochem. Butene-1, Manufacture, Russian Patent No. 1,498,747 (1989). P98 A. A. Khanmetov, et al., (Inst. of Petrochem. Proc.), Butene-1, Russian Patent No. 1,361,133 (1987). D. M. Maschmeyer et al., (Dow), Mixture of Ethylene and Butene-1, US Patent No. 4484,016 (19841. PlcNl Dow Chemical Co., Mixtures of Butene-1 and Ethylene, Japanese Patent No. 61 122,230, (1986). Conversion of Ethylene to AlphaPlOl H. Fernald et al., (Chevron/Gulf), Olefins in the Presence of Benzothiazole, US Patent No. 3,444,264 (1969). I’78

Shell

117

P102 H. Fernald et al., (Chevron/Gulf), Conversion of Ethylene to AlphaOlefins in the Presence of an Organic Sulfide, US Patent No. 3444,263 (1969). P103 H. Fernald et al., (Chevron/Gulf), Diphenylamine Polymerization Inhibitor, Alpha-Olefins from Ethylene, US Patent No. 3,478,124 (1969). P164 H. Fernald et al., (Chevron/Gulf), Conversion of Ethylene to AlphaOlefins in the Presence of Pheothiazine, US Patent No. 3,441,631(19691. P165 H .B. Fernald et al., (CheuronlGuZf3, inhibitors for Reducing Polymer Formation in a Process for Converting Ethylene to Alpha-Olefins, US Patent No. 3,641,191(1972). P166 H. Fernald et al., (CheuronlGuZfl, Measuring Degree of Conversion of Ethylene to Alpha-Olefins, US Patent No. 3,477,819 (1969). P107 H. B. Fernald et al., (Chevron/Gulf), Determining the Extenet of Conversino of Ethylene to Alpha-Olefins, US Patent No. 3,557,236 (1970). Pl68 H. Fernald et al. (Chevron/Gulf), Removal of Aluminium from an Alpha-Olefin Product Stream, US Patent No. 3,477,813 (1969). P169 H. B. Fernald et al., KXeuronlGuZfl, Catalytically Converting Ethylene to Alpha-Olefins, US Patent No. 3,702,345 (1972). PllO Chevron/Gulf, Alpha-Olefins by Catalytic Polymerization of Ethylene, UK Patent No. 1,86,610 (1970). Plll H. B. Fernald, (CheuronlGuZf3, Straight Chain Alpha-Olefins, German Patent No. 1443,927 (1970). Pll2 H. B. Fernald, (Chevron/Gulf), Minimising Polymer Formation in Conversion of Ethylene to Alpha-Olefins, US Patent No. 3,502,741(1970X P113 H. B. Fernald et al., (Chevron/Gulf), Conversion of Ethylene to AlphaOlefins in the Presence of Trialkylaluminium Catalyst Dissolved in Diisobutene, US Patent No. 3,655,899 (1972). P114 C. W. Lanier (Ethyl), Olefin Production, US Patent No. 3,663,647 (1972). P115 C. W. Lanier, (EthyZ), Olefins, US Patent No. 3,906,053 (1975). P116 W. T. Davis et al., (Ethyl), Organoaluminium Compounds, US Patent No. 3,415,861(1968X P117 W. T. Davis et al., (Ethyl), Straight Chain Alpha Olefins, US Patent No. 3,391,219 (1968). P118 P. Kobetz, (Ethyl), Separating Olefins from Aluminum Alkyls by Forming Complexes of the Aluminum Alkyls which are Insoluble in the Olefins, US Patent No. 3,689,584 (1972). P119 J. K. Presswood et al., (Ethyl), Recovering Olefins from Mixtures with Aluminum Alkyls, US Patent No. 3,557,231(1971X P126 J. K. Presswood et al., (Ethyl), Olefins Recovery from Mixtures of Aluminum Alkyls, US Patent No. 3,400,170 (1968). P121 W. Shewbart et al., (Dow), Linear Alpha-Olefins Production, US Patent No. 4,314,096 (1982). P122 A. Fowler et al., (Dow), Linear Alpha-Olefin Producton using a Tank Growth Reactor, US Patent No. 4,380,684 (1983). P123 Y. Inoue et al., (Sumitomo), Alpha-Olefins, Japanese Patent No. 75 135,662 (1975). P124 S. Munemiya, (Sumitomo), Process for Prodcuing Higher Olefins, German Patent No. 2,228,262 (1972). P125 S. Munemiya et al. 6!3umitomo), Alpha-Olefins from Ethylene, Japanese Patent No. 74 20,104 (1974).

118

P126 G. Greenfield et al., (Continental Oil), Olefins, US Patent No. 938,012 ( 1975). P127 R Rose, (Conoco), Reducing Formation of Polyethylene during a Growth Process, US Patent No. 3,388,143 (1968). Pl28 R. L. Poe et al., (Conoco), Combined Process for Preparing Alumina and Alpha Olefins, US Patent No. 4,455,289 (1984). P129 K. Ziegler Alpha-Olefin Preparation of Polymeriztion of Ethylene, Belgian Patent No. 610,302 (1962). P130 J. McClure et al., (Shell Oil), Ethylene Oligomerizaton with Cyclopetadienyl Nickel as Catalyst, US Patent No. 3,424,816 (1969). P131 K. W. Barnett et al., (Shell Oil), Olefin Oligomerization, European Patent No. 307,206 (1969). P132 W. P. Glockner et al., (Shell Oil), Ethylene Oilgomerization, US Patent No. 3527,839 (1970). P133 K. W. Barnett et al., (Shell), Oligomerization of Ethylene, German Patent No. 2,001,102 (1970). P134 K. W. Barnett et al., (Shell Oil), Ethylene Catalytic Oligomerization, US Patent No. 3532,765 0970). P135 H. Van Zwet et al., (Shell), Ethylene Oligomerization in the Presence of Complex Ni-Fl Containing Catalyst, US Patent No. 3,644,564 (1972). P136 Shell Intl. Research, Catalytic polymerizaton of Ethylenes, Netherland Patent No. 70-16,039 (1971). P137 F. Farley (Shell Oil), Linear Alpha-Oleflns from Ethylene, US Patent No. 3,647,906 (1972). P138 D. M. Singleton (SheZZ Oil), Ethylene Oligomerization, US Patent No. 4472,522 (1984). P139 A. Kister et al., (Shell Oil), Oligomeriztion Reaction System, US Patent No. 4,020,121(1977). P146 F. Lutz (SheEZ OiZ), Ethylene Oligomerization in a Monohydric/Dihydric Alcohol Solvent Mixture, US Patent No. 4,528,416 (1985). P141 A. Berger (Shell Oil), Ethylene Conversion to Linear Mono-Olefins, US Patent No. 3,726,938 (1973). P142 R. F. Mason, (SheZZ Oil), Catalytic Production of Alpha-Olefins, US Patent No. 3,676,523 (1972). P143 E. Lutz (Shell Oil), Ethylene Oligomerization, US Patent No. 3,825,615 0974). P144 R. Mason (Shell Oil), Catalytic Production of Alpha-Olefins, US Patent No. 3,737,475 (1973). P145 R. Mason (Shell Oil), Catalytic Production of Alpha-Olefins, US Patent No. 3,686,351(1972). P146 R. Mason et al., (SheZZ), Catalyst for Making Linear Alpha-Olefins and Oligomerizing Ethylene, German Patent No. 2,264,088 (1973). P147 E. Lutz, et al., (Shell), Ethylene Oligomerization carried out in a Diol Based Solvent Containing Monohydric Alcohol or Water, European Patent No. 177,999 0986). P148 D. L. Beach et al., (Chevron /Gulf), Process for the Oligomerization of Ethylene, US Patent No. 4,293,727 (1981). P149 D. L. Beach et al., (Chevron), Oligomerization of Ethylene in Methanol and Water, US Patent No. 4,711,969 (1987).

119

P150 D. L. Beach et al., (Chevron/Gulf), Sulfonated Group VA Ylides, US Patent No. 4,375,529 (1983). Pl51 D. L. Beach et al., (Chevron I Gulff, Nickel Ylides containing Ylide Ligands with a Sulfonated Group V Comp., US Patent No. 4,529,554 (1985). P152 D. L. Beach et al., K’hevronlGuZfl, Group VA Salts, US Patent No. 4,337,528 (1983). P153 D. L. Beach et al., (Chevron/Gut~, Nickel Ylides Containing Sulfonated Group V Ligands, European Patent No. 46,330 0982). P15Q D. L. Beach et al., ~Chevro~lGu~~, Novel Group VA Salts and Process for Preparing s, European Patent No. 52,930 (1982). P155 D. L. Beach et al., (Chevron), Oligomerization of Ethylene using Ni YlideIalkyl Al Alkoxide Catalyst, US Patent No. 4,686,315 (1987). P156 D. Beach, et al., (Chevron/Gulf), Sulfonated Group VA Ylides & Process for Preparing same, US Patent No. 4,377,529 (1983). P157 D. L. Beach et al., ~Chevron/Gu~~, Process for Oligome~zation of Ethylene in Methanol, US Patent No. 4,310,716 (1982). P158 D. L. Beach et al., (Chevron/GuZ#, Process for Oligomerization of Ethylene in Methanol, US Patent No. 4,382,X3 (1982). P159 D. L. Beach et al., KhevronlGuZfl, Process for the Oligomerization of Ethylene, US Patent No. 4,288$X8 (1981). Pl60 D. L. Beach et al., ~Chevron/Gu~~, Use of Metal Modified Refractory Metal Oxide/Silica Supported Nickel Cluster Catalyst to Oligomerize Ethylene, US Patent No. 4,293,725 (1981). P161 K. W. Barnett et al., (Arshland), Catalyst for Oligomerization of Ethylene, US Patent No. 4,628,138 (1986). P162 K. W. Barnett IAshland), Bis(l,5Zyclooctadiene) Nickel(O) Catalyst for Oligomerization of Ethylene, US Patent No. 4,628,139 (1986). P163 R. E. Murray endow Carbons, Process for Alpha Olefin M~~acture by Ethylene Co-Oligomerization, US Patent No. 4,689,437 (1987). Pl64 P. 0. Nubel (Amoco), Zinc-Promoted Nickel Catalysts for Conversion of Ethylene to Linear Butenes, US Patent No. 4,788,373 (1988). P165 R. Knudsen et al. (PhiZZip$, Ethylene Oligomerization, US Patent No. 4,482,640 (1984). Pl66 H. Dunn ~PhilZi~sJ, Catalytic Olefin Dimerization, US Patent No. 3,5~,7~ (1971). P167 P. Haenssles, (Chemische Werke Huels), Ethylene Oligomerization, German Patent No. 3,027,782 (1982). P168 Furbwerke Hoechst, Oligomerization of Linear Alpha-Olefins, French Patent 1,576,X34(1969). P169 Furbwerke Hoechst, Catalysts for Oligomerization of Alpha-Olefins, Netherlands Patent No. 6,601,770 (1967). Pl70 Fa~bwe~ke Hoechst, Oligomerizatin of Alpha-Olefins, UK Patent No. 1,158,211(X%9). P171 J. Mertzwieiller et al., (EjcxonlEsso), Oligomerization of Olefins, German Patent No. 1,808,833 (1969). P172 Sentral Inst. Industriell Forskiuing, Monoalkenes, Netherlands Patent No. 6,601,770 (1966). P173 T. Arkawa et al.,
120

P174 K. Ziegler et al., Cyclic Process for Preparation of Alpha Olefins, German Patent No. 1,118,188 (1961). P175 A. W. Langer (ExxonlEsso), Linear Alpha-Olefins from Ethylene, UK Patent No. 1,118,987 (1968). P176 D. V. Favis (Exxon /Esso), Process for Making Linear-Alpha-Olefins, US Patent No. 3549,723 (1970). P177 D. V. Favis (ExxonlEsso), Polymerization of Ethylene to Low Molecular Weight Linear Olefins, US Patent No. 3,472,910 (1969). P178 R. Bearden et al., (Exxon /Esso), Linear Alpha-Olefins Production, German Patent No. 1,803,124 (1969). P179 S. Butter, (Mobil), Linear-Olefins by Ethylene Oligomerization, US Patent No. 3,981,941(1976). P180 Phung Nhu Hung et al., (ZFP and Sot. Nutl. du Petrole), Catalytic Composition for the Oligomerization of Ethylene, French Patent No. 2,217,064 (1974). of Ethylene to Linear-Alpha-Olefins, P181 H. Mix et al., Oligomerization German Patent No. 108,260 (1974). P182 B. Calcagno et al., (Societa Ital. Res.), Polymerization of Ethylene to Linear Olefin Mixtures, Italian Patent No. 710,124 (1966). P183 T. Arakawa et al., (Mitsui), Highere Olefins, Japanese Patent No. 74 02,083 (1972). P184 T. Arakawa et al., (Mitsui), Process for Oligomerization of Ethylene, Japanese Patent No. 73 23,046 (1973). P185 D. A. Young et al., (Exxon), Process and Zirconium Compound Adduct Catalyst for Preparing Linear Alpha Olefins by Ethylene Oligomerization, European Patent No. 295,960 (1988). P186 W. Langer, (Exxon), Linear Olefins Products, US Patent No. 4,434,313 (1984). P187 W. Langer et al., (Exxon), Linear Olefin Products, US Patent No. 4,361,714 (1982). P188 W. Langer, (Exxon), Preparation of Linear Olefins Products, US Patent No. 4,410,750 (1983). P189 W. Langer, (Exxon), Preparation of Linear Olefin Products, US Patent No. 4,434,312 (1984). P190 W. Langer, (Exxon), Preparation Linear Olefin Products, US Patent No. 4,486,615 (1984). P191 Y. Shiraki et al., (Zdemitsu), Catalysts for the Production of Linear Alpha Olefins, European Patent No. 328,728 (1989). P192 Y. Shiraki et al., (Zdemitsu), Catalyst for Oligomerization of Ethylene to Linear Alpha Olefins, European Patent No. 241,596 (1987). P193 Y. Shiraki et al., (Zdemitsu), Preparation of Alpha-Olefins, Japanese Patent No. 62 59,525 (1987). P194 Y. Shiraki et al., (Zdemitsu), Linear Alpha-Olefins, Japanese Patent No. 62 00,430 (1987). P195 Y. Nakashima et al;, (Zdemitsu), Production of Q-C18 Alpha OIefins with Improved Disposal of Polymeric Products, European Patent No. 320,571(1989). P196 Mitsubishi Petrochem. Co., Linear Alpha-Olefins, Japanese Patent No. 58 201,729 (1983).

121

P197 P. Longi et al., @fontecatini Edison), Catalytic Oligomerization of Ethylene, German Patent No. 2,044,343 (1971). P198 P. Longi et al., (Montecatini Edison), Catalytic Manufacture of Ethylene Oligomers, Italian Patent No. 871,096 (1970). P199 V. Khodakovskayaya et al., (Inst. of Chemical Physics), Catalyst for Oligomerization of Ethylene into Higher Alpha-Olefins, World Organization Patent No. 80 00,224 (1988). P209 G. A. Mamedaliev et al., (Inst. of Petrochem. Proc.), Ethylene Oligomers, Russian Patent No. 1,35,192 (1987). P2Ol E. Magoon et al., (Shell Oil), Catalytic Conversion of Ethylene to Linear Alpha-Olefins, US Patent No. 3,592,866 (1971). P202 P. W. Glockner (Shell), Oligomeriztion of Ethylene, German Patent No. 1,931,966(19701. P203 M. Isamoto et al., (Toruy Ind.), Ethylene Oligomers, Japanese Patent No. 7141,882 (1971).