CA2767615C

CA2767615C - Multi reactor ethylene oligomerization process with recycle

Info

Publication number: CA2767615C
Application number: CA2767615A
Authority: CA
Inventors: P. Scott Chisholm; Stephen J. Brown; Eric Clavelle; Kamal Serhal
Original assignee: Nova Chemicals Corp
Current assignee: Nova Chemicals Corp
Priority date: 2012-02-08
Filing date: 2012-02-08
Publication date: 2019-01-15
Anticipated expiration: 2032-02-08
Also published as: WO2013116922A1; CA2767615A1

Abstract

A multi reactor system is used for the oligomerization of ethylene in the presence of a chromium/P-N-P catalyst system. The reactor system includes a mixed reactor and a tubular reactor. A portion of the reaction liquid is recycled back to the mixed reactor. The mixed reactor preferably contains a gas/liquid jet to facilitate the mixing of ethylene into the reaction liquid.

Description

MULTI REACTOR ETHYLENE OLIGOMERIZATION PROCESS WITH RECYCLE
FIELD OF THE INVENTION
This invention relates to the oligomerization of ethylene using a Cr catalyst having a so-called "bridged diphosphine ligand" in a process that uses at least two different types of reactors in series, namely at least one mixed reactor and at least one tubular reactor, with the provisio that part of the oligomer product that is discharged from the tubular reactor is recycled to the mixed reactor.
BACKGROUND OF THE INVENTION
Alpha olefins are commercially produced by the oligomerization of ethylene in the presence of a simple alkyl aluminum catalyst (in the so called "chain growth"
process) or alternatively, in the presence of an organometallic nickel catalyst (in the so ' called Shell Higher Olefins, or "SHOP" process). Both of these processes typically produce a crude oligomer product having a broad distribution of alpha olefins with an even number of carbon atoms (i.e. butene-1, hexene-1, octene-1 etc.). The various alpha olefins in the crude oligomer product are then typically separated in a series of distillation columns. Butene-1 is generally the least valuable of these olefins as it is also produced in large quantities as a by-product in various cracking and refining processes. Hexene-1 and octene-1 often command comparatively high prices because these olefins are in high demand as comonomers for linear low density polyethylene (LLDPE).
Technology for the selective trimerization of ethylene to hexene-1 has been recently put into commercial use in response to the demand for hexene-1. The patent literature discloses catalysts which comprise a chromium source and a pyrrolide ligand as being useful for this process ¨ see, for example, United States .. Patent ("USP") 5,198,563 (Reagen et al., assigned to Phillips Petroleum).

\\chclients\IPGrounµRrnfficrscna,mi9no1Can docx Another family of highly active trimerization catalysts is disclosed by Wass et al.
in WO 02/04119 (now United States Patents 7,143,633 and 6,800,702). The catalysts disclosed by Wass et al. are formed from a chromium source and a bridged diphosphine ligand and are described in further detail by Carter et al. (Chem.
Comm.
2002, p 858-9). As described in the Chem. Comm. paper, these catalysts preferably comprise a diphosphine ligand in which both phosphine atoms are bonded to two phenyl groups that are each substituted with an ortho-methoxy group. Hexene-1 is produced with high activity and high selectivity by these catalysts.
Similar diphosphine/tetraphenyl ligands are disclosed by Blann et al. in .. W004/056478 and WO 04/056479 (now US 2006/0229480 and US 2006/0173226).
However, in comparison to the ligands of Wass et al., the disphosphine/tetraphenyl ligands disclosed by Blann et al. generally do not contain polar substituents in ortho positions. The "tetraphenyl" diphosphine ligands claimed in the '480 application must not have ortho substituents (of any kind) on all four of the phenyl groups and the "tetraphenyl" diphosphine ligands claimed in '226 are characterized by having a polar substituent in a meta or para position. Both of these approaches are shown to reduce the amount of hexenes produced and increase the amount of octene (in comparison to the ligands of Wass et al.). Other bridged diphosphine ligands that are useful for the selective oligomerization of ethylene are disclosed in the literature. The formation of polymer as a by-product is a general problem with many of these ligands.
The oligomerization of ethylene is highly exothermic. The performance of the Cr bridged diphosphine catalysts is quite temperature dependent. Preferred operating temperatures are from 50 ¨ 150 C; especially from 60 ¨ 90 C. Sudden temperature changes (especially temperature drops) have been observed to lead to the formation of by-product polymer ¨ which is highly undesirable. In addition, the selectivity of the

2 dm( catalyst has been observed to change with temperature. Accordingly, good temperature control is highly desirable and an isothermal (as opposed to adiabatic) reaction is especially preferred.
It is also important to ensure that the ethylene feed and catalyst feed are well mixed. Ethylene concentration gradients (leading to localized high ethylene concentrations) have been observed to lead to polymer fouling. Likewise, localized catalyst concentration gradients in a poorly mixed reactor are also believed to lead to polymer formation. It is also believed that the relative ethylene/catalyst ratio is important and that very low catalyst concentrations (at a given ethylene concentration) lead to excessive polymer formation.
Thus, a first preferred condition is to provide a well mixed reactor in order to minimize gradients in reactor temperature, ethylene concentration and catalyst concentration.
Finally - and most interestingly - kinetic studies have shown the oligomerization reaction to be "mixed order" in ethylene concentration when using some bridged diphosphine catalysts, particularly catalysts that enable selective tetramerizations.
While not wishing to be bound by theory, we believe that two reactions are taking place simultaneously ¨ a selective trimerization reaction (which appears to be first order in ethylene) and a selective tetramerization reaction (which appears to be second order in ethylene). Thus, if it is desired to maximize octene production, a second preferred condition ¨ namely high ethylene concentration ¨ is desirable.
It is possible to satisfy both of the first preferred condition (i.e. a well mixed reactor, as noted above) and the second preferred condition ¨ namely high ethylene concentrations ¨ by operating a well mixed reactor (such as CSTR) at a high ethylene concentration and with low ethylene conversion. It will be appreciated by those skilled in

3 t1rhrliontql1P(Irnim \crntt1SrAnsarl9012001Can.docx . __________________________________________________ ¨
the art that the ethylene concentration in the discharge from a well mixed CSTR will correspond to the ethylene concentration in the bulk. Thus, a problem with the use of a CSTR in this process is that the product discharge will contain large amounts of unreacted ethylene.
It would appear that this problem might be mitigated by using a tubular reactor.
In a tubular reactor, the ethylene is converted as the reaction mixture flows through the tube. Thus, the desired high ethylene concentrations are provided at the start of the tube and lower ethylene concentrations are present at the tube discharge.
However, it is exceptionally difficult to provide well mixed conditions at the feed end (or start) of a reaction in a tube, particularly with a highly active catalyst and/or when rapid mass transfer is required. Both conditions are required with the present process and the resulting mixing problems could be expected to lead to gross polymer fouling.
The present invention mitigates these problems.
SUMMARY OF THE INVENTION
In one embodiment, the present invention provides:
A process for the oligomerization of ethylene in at least two reactors, wherein said process comprises a) providing ethylene to a mixed reactor in the presence of an oligomerization catalyst under oligomerization conditions, thereby producing an initial oligomer product;
b) discharging said initial oligomer product from said mixed reactor and directing it to a tubular reactor;
C) forming additional oligomer product in said tubular reactor;
d) discharging from said tubular reactor an oligomerization stream containing said initial oligomer product and said additional oligomer product; and

4 \t^h-"---"Dra--",e¨H`cren¨:"'012001Cartclocx e) recycling a portion of said oligomerization stream from the discharge of said tubular reactor back to said mixed reactor;
with the provisio that said catalyst system comprises a chromium catalyst having a bridged diphosphine ligand.
For clarity, the present invention must include a recycle flow from the discharge of the tubular reactor to the mixed reactor. In this manner, a semi-batch process is enabled (whereby ethylene is added to the reactor on a continuous/semi continuous basis, but the oligomer product is only withdrawn in a batch or semi batch manner).
It will also be appreciated that doing a batch/semi-batch operation, all of the discharge from the tubular reactor may be recycled back to the mixed reactor.
This would continue until the reactor becomes full or the reactor is terminated for a different reason.
Conversely, in a continuous process, some of the oligomer product is withdrawn during the reaction, as explained in more detail below. Thus, the term "a portion" (when referring to the oligomer product that is being recycled) is inclusive of "some or all", as the context requires.
The use of additional reactors is also contemplated. Most notably, the reactor system may be fitted with two or more tubular reactors. This would facilitate operating the process at low rates (with one tubular reactor) and higher rates with the additional tubular reactor.
BRIEF DESCRIPTION OF THE DRAWING
Figure 1 is a process flow diagram that illustrates a preferred embodiment of the invention.
DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS

5 \\chclients\IPGrouo\Scott\SCSoec12012001Can.docx Figure 1 illustrates a process flow diagram of a preferred embodiment of the present invention.
The mixed reactor 1 receives fresh ethylene via feed line 5 and catalyst feeds via feed line 6. It is especially preferred to also add hydrogen to mixed reactor 1. Thus, hydrogen may be added in the ethylene feed line 5 or an alternative feed line.
Recycle from the tubular reactor 2 is also added to the mixed reactor, via line 4.
In a preferred embodiment, mixed reactor '1 has a liquid 7 level that defines a gas space 10 above the liquid reaction mixture.
In one embodiment, mixed reactor 1 is a continuously stirred tank reactor .. (CSTR) and mixing is provided by an agitator.
In an alternative (preferred) embodiment, mixed reactor I is equipped with a gas/liquid ejector 8. The recycle liquid from reactor 2 is directed into reactor 1 in the form of a liquid jet through gas/liquid ejector 8 by way of recycle line 4'.
The liquid jet flows through a zone of reduced cross sectional area, thereby forming a zone of especially fast liquid flow (which, in turn, produces low pressure). The gas/liquid ejector 8 has an opening in the gas space 10 that communicates with the low pressure zone of the gas/liquid ejector 8 and this allows ethylene to be entrained in liquid and mixed in the jet flow. The flow from the ejector ¨ which consists of the reaction liquid and entrained/dissolved ethylene ¨ is directed into the liquid of mixed reactor 1.
Such gas/liquid ejectors are known in the art and are also commonly referred to as Venturi mixers and/or jet ejectors (as well as gas/liquid ejectors). Reactors equipped with such an ejector are commonly referred to as "gas circulation" or "jet loop"
reactors.
It should be noted that the ejector 8 is only shown in a symbolic or representative manner (as are pump 3 and reactors 1 and 2) with much detail omitted. For example, the liquid flow channel for the jet and the gas flow channel (or channels) which allow

6 st,i,,i;..f.uortr".,nm=-=-miQr`e---11012001Can.docx gas to be entrained in the liquid are omitted. This type of detail will be readily understood by persons of ordinary skill in the art of ejector design. Such detail is available in the literature (see for example, Ullman's Encyclopedia of Industrial Chemistry, Fifth Edition, ed.: Elvers et at.; ISBN 0-89573-539-3; vol B4, p.297). It should also be noted that the Figure does illustrate the preferred down flow of the liquid jet.
It is preferred that the liquid flow through the jet is at least 10 meters per second, m/s (and especially at least 20m/s) in order to efficiently entrain the ethylene.
It is preferred to use a pump 3 to provide the propulsion that is required to circulate the flow through the jet; the mixed reactor 1 and the tubular reactor 2. For clarity: reaction liquid circulates from the discharge of mixed reactor 1, through pump 3 and reactor 2 and then back to reactor 1. Thus, a combination of the gas/liquid ejector 8 and the pump 3 providing mixing in reactor 1 is a preferred embodiment of this invention.
In a preferred embodiment, the cooling system for the present invention is provided as an external cooling system (i.e.: it is preferred to avoid the use of internal cooling coils).
It is especially preferred to provide an external cooling shell on the tubular reactor and external cooling coils on the top of the well mixed reactor (to provide cooling for the gas space in mixed reactor 1.
As previously noted, it is difficult to mix high levels of ethylene and/or catalyst in a tubular reactor. Accordingly, it is especially preferred to add catalyst and ethylene to only the mixed reactor (i.e. to avoid adding fresh ethylene and/or catalyst to the tubular reactor). The term "tubular reactor" as used herein is meant to convey its conventional meaning, namely a reactor with a high length/diameter (or LID) ratio. A single tube or

7 µkpi.,,li..,muorz,...nkc,ni+kQi'Q--,\1012001Can.docx -----multiple tube bundle may be suitably employed. Tubular reactors typically do not have internal agitators and that is the case in the present invention, with pump 3 preferably providing the force to move the reaction liquid.
The ethylene is most preferably added as a gas although a portion of the ethylene may be added as a liquid (thereby cooling the reactor as the liquid ethylene flashes to a gas). It is especially preferred to add hydrogen with the ethylene. In one embodiment, a portion of the ethylene may be added below the liquid level (although this is not necessary). In another embodiment, the ethylene is added to the gas space in reactor 1.
10 The process of the present invention has the additional advantage that it facilitates the start-up of the oligomerization. At start-up, the reactor system will be cleaned/purged according to good engineering practice. Start-up liquid and (optionally) hydrogen may then be added to the reactor. The amount of start-up liquid is preferably low (25-35% of the volume of the mixed reactor). The start-up liquid may be any liquid that facilitates the reaction (such as an aliphatic, an aromatic, or even oligomer product from a previous reaction). The pump preferably starts to circulate the liquid through the reactor system when adding the catalyst, thereby proving well mixed catalyst.
Ethylene is then gradually added to the mixed reactor to provide "light off' (or initiation of the reaction).
In one embodiment, the start-up liquid is a very good solvent for the catalyst system (such as monochlorobenzene).
As the reaction progresses, the liquid level in the system increases as liquid oligomer is produced. The liquid oligomer is then removed from the process.
This may be done continuously (via a slip stream) or in a batch/semi-batch manner by way of product discharge line 9, through valve 11. The process may be operated with

8 mrhetiontquPrzr,..nlc,votmrq,,012001Can.docx . _ additional solvent being added (for example, with the ethylene) or ¨
alternatively ¨ the proces may be operated without additional solvent being provided after start up.
Additional details are provided below.
PART A CATALYST SYSTEM
The preferred catalyst system used in the process of the present invention must contain three essential components, namely:
(i) a source of chromium;
(ii) a diphosphine ligand; and (iii) an activator.
Preferred forms of each of these components are discussed below.
Chromium Source ("Component (i)") Any source of chromium that is soluble in the process solvent and which allows the oligomerization process of the present invention to proceed may be used.
Preferred chromium sources include chromium trichloride; chromium (III) 2-ethylhexanoate; chromium (III) acetylacetonate and chromium carbonyl complexes such as chromium hexacarbonyl. It is preferred to use very high purity chromium compounds as these should generally be expected to minimize undesirable side reactions. For example, chromium acetylacetonate having a purity of higher than 99%
is commercially available (or may be readily produced from 97% purity material ¨ using recrystallization techniques that are well known to those skilled in the art).
Liqand Used in the Oliqomerization Process ("Component (ii)") In general, the ligand used in the oligomerization process of this invention is defined by the formula (R1)(R2)-P1-bridge-P2(R3)(R4) wherein R1, R2,R3 and R4 are independently selected from the group consisting of hydrocarbyl and heterohydrocarbyl and the bridge is a moiety that is bonded to both phosphorus atoms.

9 = = = ¨ -.4r..'^-'2,0012001Can.docx Another family of suitable ligands uses a fluorocarbyl oxide (especially the aromatic group ¨ C6F5) for the R4 group ¨ with R1 to R3 being as defined above.
The term hydrocarbyl as used herein is intended to convey its conventional meaning ¨ i.e. a moiety that contains only carbon and hydrogen atoms. The .. hydrocarbyl moiety may be a straight chain; it may be branched (and it will be recognized by those skilled in the art that branched groups are sometimes referred to as "substituted"); it may be saturated or contain unsaturation and it may be cyclic.
Preferred hydrocarbyl groups contain from 1 to 20 carbon atoms. Aromatic groups ¨
especially phenyl groups ¨ are especially preferred. The phenyl may be unsubstituted (i.e. a simple C6I-15 moiety) or contain substituents, particularly at an ortho (or "o") position.
Similarly, the term heterohydrocarbyl as used herein is intended to convey its conventional meaning ¨ more particularly, a moiety that contains carbon, hydrogen and heteroatoms (such as 0, N, R and S). The heterohydrocarbyl groups may be straight .. chain, branched or cyclic structures. They may be saturated or contain unsaturation.
Preferred heterohydrocarbyl groups contain a total of from 2 to 20 carbon +
heteroatoms (for clarity, a hypothetical group that contains 2 carbon atoms and one nitrogen atom has a total of 3 carbon + heteroatoms).
It is preferred that each of R1, R2, R3 and R4 is a phenyl group (with an optional substituent in an ortho position on one or more of the phenyl groups).
Highly preferred ligands are those in which R1 to R4 are independently selected from the group consisting of phenyl, o-methylphenyl (i.e. ortho-methylphenyl), o-ethylphenyl, o-isopropylphenyl and o-fluorophenyl. It is especially preferred that none of R1 to R4 contains a polar substituent in an ortho position. The resulting ligands are useful for the selective tetramerization of ethylene to octene-1 with some co product mhaentqmpnrniirACrntfiSCAn0M2012001Can.docx hexene also being produced. The term "bridge" as used herein with respect to the ligand refers to a moiety that is bonded to both of the phosphorus atoms in the ligand ¨
in other words, the "bridge" forms a link between P1 and P2. Suitable groups for the bridge include hydrocarbyl and an inorganic moiety selected from the group consisting of N(CH3)-N(CH3)-, -B(R6)-, -Si(R6)2-, -P(R6)- or -N(R6)- where R6 is selected from the group consisting of hydrogen, hydrocarbyl and halogen.
It is especially preferred that the bridge is -N(R5)- wherein R5 is selected from the group consisting of hydrogen, alkyl, substituted alkyl, aryl, substituted aryl, aryloxy, substituted aryloxy, halogen, alkoxycarbonyl, carbonyloxy, alkoxy, aminocarbonyl, carbonylamino, dialkylamino, silyl groups or derivatives thereof and an aryl group substituted with any of these substituents. A highly preferred bridge is amino isopropyl (i.e. when R5 is isopropyl).
In one embodiment, two different types of ligands are used to alter the relative amounts of hexene and octene being produced. For clarity: the use of a ligand that produces predominantly hexene may be used in combination with a ligand that produces predominantly octene.
Activator ("Component 'Hp") The activator (component (iii)) may be any compound that generates an active catalyst for ethylene oligomerization with components (i) and (ii). Mixtures of activators may also be used. Suitable compounds include organoaluminum compounds, organoboron compounds and inorganic acids and salts, such as tetrafluoroboric acid etherate, silver tetrafluoroborate, sodium hexafluoroantimonate and the like.
Suitable organoaluminium compounds include compounds of the formula AIR3, where each R
is independently Ci ¨C12 alkyl, oxygen or halide, and compounds such as LiAIH4 and the like. Examples include trimethylaluminium (TMA), triethylaluminium (TEA), tri-ll "^h^,"^`^^^-^",012001Can docx _ _ _ _ _ isobutylaluminium (TIBA), tri-n-octylaluminium, methylaluminium dichloride, ethylaluminium dichloride, dimethylaluminium chloride, diethylaluminium chloride, ethylaluminiumsesquichloride, methylaluminiumsesquichloride, and alumoxanes.
Alumoxanes are well known in the art as typically oligomeric compounds which can be prepared by the controlled addition of water to an alkylaluminium compound, for example trimethylaluminium. Such compounds can be linear, cyclic, cages or mixtures thereof. Commercially available alumoxanes are generally believed to be mixtures of linear and cyclic compounds. The cyclic alumoxanes can be represented by the formula [R6A10]s and the linear alumoxanes by the formula R7(R8A10)s wherein s is a number from about 2 to 50, and wherein R6, R7, and R8 represent hydrocarbyl groups, preferably Ci to Ce alkyl groups, for example methyl, ethyl or butyl groups.
Alkylalumoxanes especially methylalumoxane (MAO) are preferred.
It will be recognized by those skilled in the art that commercially available alkylalumoxanes may contain a proportion of trialkylaluminium. For instance, commercial MAO usually contains approximately 10 wt % trimethylaluminium (TMA), and commercial "modified MAO" (or "MMAO") contains both TMA and TIBA.
Quantities of alkylalumoxane are generally quoted herein on a molar basis of aluminium (and include such "free" trialkylaluminium).
Examples of suitable organoboron compounds are boroxines, NaBF14, trimethylboron, triethylboron, dimethylphenylammoniumtetra(phenyl)borate, trityltetra(phenyl)borate, triphenylboron, dimethylphenylammonium tetra(pentafluorophenyl)borate, sodium tetrakispis-3,5-trifluoromethyl)phenyliborate, trityltetra(pentafluorophenyl)borate and tris(pentafluorophenyl) boron.

mtutuantsmPnmiln\srnitncsnPrA9012001Can.docx - ¨
Activator compound (iii) may also be or contain a compound that acts as a reducing or oxidizing agent, such as sodium or zinc metal and the like, or oxygen and the like.
In the preparation of the catalyst systems used in the present invention, the quantity of activating compound to be employed is easily determined by simple testing, for example, by the preparation of small test samples which can be used to oligimerize small quantities of ethylene and thus to determine the activity of the produced catalyst.
It is generally found that the quantity employed is sufficient to provide 0.5 to 1000 moles of aluminium (or boron) per mole of chromium. MAO is the presently preferred activator. Molar Al/Cr ratios of from 1/1 to 1500/1, especially 300/1 to 900/1 are preferred.
PART B PROCESS CONDITIONS
The chromium (component (i)) and ligand (component (ii)) may be present in any molar ratio which produces oligomer, preferably between 100:1 and 1:100, and most preferably from 10:1 to 1:10, particularly 3:1 to 1:3. Generally the amounts of (i) and (ii) are approximately equal, i.e. a ratio of between 2:1 and 1:2.
Components (i)-(iii) of the catalyst system utilized in the present invention may be added together simultaneously or sequentially, in any order, and in the presence or absence of ethylene in any suitable solvent, so as to give an active catalyst.
For example, components (i), (ii) and (iii) and ethylene may be contacted together simultaneously, or components (i), (ii) and (iii) may be added together simultaneously or sequentially in any order and then contacted with ethylene, or components (i) and (ii) may be added together to form an isolable metal-ligand complex and then added to component (iii) and contacted with ethylene, or components (i), (ii) and (iii) may be added together to form an isolable metal-ligand complex and then contacted with docx ethylene. Suitable solvents for contacting the components of the catalyst or catalyst system include, but are not limited to, hydrocarbon solvents such as heptane, toluene, 1-hexene and the like, and polar solvents such as diethyl ether, tetrahydrofuran, acetonitrile, dichloromethane, chloroform, chlorobenzene, acetone and the like. A
preferred solvent is the oligomer product that is produced by the present process or some fraction thereof ¨ such as hexene, octene or a mixture of the two.
For further clarity: the catalyst components may be mixed together in the oligomerization reactor, or ¨ alternatively ¨ some or all of the catalyst components may be mixed together outside of the oligomerization reactor. In general, it is preferred to mix the catalyst components outside of the reactor (due to comparative ease of control) then add the catalyst to the reactor shortly thereafter (because "aged"
catalyst may suffer from some loss of activity). This method of catalyst synthesis is illustrated in the examples. The solvent that is used to prepare the catalyst is preferably the olefinic product that is produced by the reactor (or some portion thereof). We have found that the use of octene generally works well. However, some catalyst components have comparatively low solubility in octene. For example, MAO that is made solely with trimethylaluminum (as opposed to "modified MAO" which also contains some higher alkyl aluminum, such as triisobutyl aluminum) is less soluble in octene than in some cyclic hydrocarbons such as xylene or tetralin. Accordingly, when one or more catalyst components are mixed together outside of the oligomerization reactor, the use of toluene, xylene chlorobenzene, or tetralin as the solvent may be preferred.
The xylene may be a mixture of ortho, meta and para isomers ¨ i.e. it is not necessary to use a pure isomer.
A variety of methods are known to purify solvents used in the oligomerization process including use of molecular sieves (3A), adsorbent alumina and supported 1\chcliants IPGrounlScott\SCSopc12012001Can docx _ de-oxo copper catalyst. Several configurations for the purifier system are known and depend on the nature of the impurities to be removed, the purification efficiency required and the compatibility of the purifier material and the process solvent. In some configurations, the process solvent is first contacted with molecular sieves, followed by adsorbent alumina, then followed by supported de-oxo copper catalyst and finally followed by molecular sieves. In other configurations, the process solvent is first contacted with molecular sieves, followed by adsorbent alumina and finally followed by molecular sieves. In yet another configuration, the process solvent is contacted with adsorbent alumina. When alpha olefinic solvents are used in the process, the preferred purifier system consists of molecular sieves, followed by adsorbent alumina and finally followed by another set of molecular sieves.
The catalyst components (i), (ii) and (iii) utilized in the present invention can be unsupported or supported on a support material, for example, silica, alumina, MgC12 or zirconia, or on a polymer, for example polyethylene, polypropylene, polystyrene, or poly(aminostyrene). If desired the catalysts can be formed in situ in the presence of the support material, or the support material can be pre-impregnated or premixed, simultaneously or sequentially, with one or more of the catalyst components.
The quantity of support material employed can vary widely, for example from 100,000 to 1 gram per gram of metal present in the transition metal compound. In some cases, the support material can also act as or as a component of the activator compound (iii).
Examples include supports containing alumoxane moieties.
Oligomerization reactions can generally be conducted under solution phase, slurry phase, gas phase or bulk phase conditions. Suitable temperatures range from

10 C to +300 C preferably from 10 C to 150 C, especially from 20 to 80 C.
Suitable \ ehrliantellPnrni \ Crrift \ Cr.Cnor17012001Can.docx pressures are from atmospheric to 800 atmospheres (gauge) preferably from 5 atmospheres to 150 atmospheres, especially from 10 to 100 atmospheres.
Irrespective of the process conditions employed, the oligomerization is typically carried out under conditions that substantially exclude oxygen, water, and other materials that act as catalyst poisons. In addition, the reactor is preferably purged with a nonreactive gas (such as nitrogen or argon) prior to the introduction of catalyst. A
purge with a solution of MAO and/or aluminum alkyl may also be employed to lower the initial level of catalyst poisons. Also, oligomerizations can be carried out in the presence of additives to control selectivity, enhance activity and reduce the amount of polymer formed in oligomerization processes. Potentially suitable additives include, but are not limited to, hydrogen or a halide source (especially the halide sources disclosed in U.S. patent 7,786,336, Zhang et al.). Other (optional) additives include antistatic agents (such as the polysulfone polymer sold under the trademark Stadise) and/or fluorocarbons to mitigate reaction fouling; or amines to alter the hexene/octene ratio of the product oligomer (as disclosed in U.S. application 20090118117, Elowe et al.). The use of hydrogen is especially preferred because it has been observed to reduce the amount of polymer that is formed. It is within the scope of this invention that an oligomerization product might also serve as a solvent or diluent. The preferred catalysts of this invention predominantly produce hexene and octene (as shown in the examples) but smaller quantities of butene and Cio+ olefins are also produced.
The crude product stream may be separated into various fractions using, for example, a conventional distillation system. Mixtures of inert diluents or solvents also could be employed. The preferred diluents or solvents are aliphatic and aromatic hydrocarbons and halogenated hydrocarbons such as, for example, isobutane, pentane, toluene, xylene, ethylbenzene, cumene, mesitylene, heptane, cyclohexane, methylcyclohexane,

11,=h,l012001Can.docx 1-hexene, 1-octene, chlorobenzene, dichlorobenzene, and the like, and mixtures such as lsoparTM.
Techniques for varying the distribution of products from the oligomerization reactions include controlling process conditions (e.g. concentration of components (i)-(iii), reaction temperature, pressure, residence time) and properly selecting the design of the process and are well known to those skilled in the art.
In another embodiment, a catalyst that produces ethylene homopolymer is deliberately added to the reactor in an amount sufficient to convert from 1 to 5 weight%
of the ethylene feed to an ethylene homopolymer. This catalyst is preferably supported.
The purpose is to facilitate the removal of by-product polyethylene.
The ethylene feedstock for the oligomerization may be substantially pure or may contain other olefinic impurities and/or ethane. One embodiment of the process of the invention comprises the oligomerization of ethylene-containing waste streams from other chemical processes or a crude ethylene/ethane mixture from a cracker as more fully described in co-pending Canadian patent application 2,708,011 (Krzywicki et al.).
The feedstock is preferably treated to remove catalyst poisons (such as oxygen, water and polar species) using techniques that are well known to those skilled in the art. The technology used to treat feedstocks for polymerizations is suitable for use in the present invention and includes the molecular sieves, alumina and de-oxo catalysts .. described above for analogous treatment of the process solvent.
Reactor Control The control systems required for the operation of mixed reactors and tubular reactors are well known to those skilled in the art and do not represent a novel feature of the present invention. In general, temperature, pressure and flow rate readings will provide the basis for most conventional control operations. The increase in process temperature (together with reactor flow rates and the known enthalpy of reaction) may be used to monitor ethylene conversion rates. The amount of catalyst may be increased to increase the ethylene conversion (or decreased to decrease ethylene conversion) within desired ranges. Thus, basic process control may be derived from simple measurements of temperature, pressure and flow rates using conventional thermocouples, pressure meters and flow meters. Advanced process control (for example, for the purpose of monitoring product selectivity or for the purpose of monitoring process fouling factors) may be undertaken by monitoring additional process parameters with more advanced instrumentation. Known/existing instrumentation that may be employed include in-line/on-line instruments such as NIR infrared, Fourier Transform Infrared (FTIR), Raman, mid-infrared, ultra violet (UV) spectrometry, gas chromatography (GC) analyzer, refractive index, on-line densitometer or viscometer.
The use of NIR or GC to measure the composition of the oligomerization reactor and final product composition is especially preferred.
The measurement may be used to monitor and control the reaction to achieve the targeted stream properties including but not limited to concentration, viscosity, temperature, pressure, flows, flow ratios, density, chemical composition, phase and phase transition, degree of reaction, polymer content, selectivity.
The control method may include the use of the measurement to calculate a new control set point. The control of the process will include the use of any process control algorithms, which include, but are not limited to the use of PID, neural networks, feedback loop control, forward loop control and adaptive control.
Catalyst Deactivation, Catalyst Removal and Polymer Removal In general, the oligomerization catalyst is preferably deactivated immediately downstream of the reactor as the product exits the reaction system. This is to prevent "^";,..+-"Dr,"¨oe^^000,"c^-.1-00120010an.docx polymer formation and potential build up downstream of the reactor and to prevent isomerisation of the 1-olefin product to the undesired internal olefins. It is generally preferred to flash and recover unreacted ethylene before deactivation.
However, the option of deactivating the reactor contents prior to flashing and recovering ethylene is also acceptable. The flashing of ethylene is endothermic and may be used as a cooling source.
In general, many polar compounds (such as water, alcohols and carboxylic acids) will deactivate the catalyst. The use of alcohols and/or carboxylic acids is preferred ¨ and combinations of both are contemplated. It is generally found that the quantity employed to deactivate the catalyst is sufficient to provide deactivator to metal (from catalyst + activator) mole ratio between about 0.1 to about 4, especially from 1 to 2 (thus, when MAO is the activator, the deactivator is provided on a ratio based on moles of Cr + to moles of Al).
The deactivator may be added to the oligomerization product stream before or after the volatile unreacted reagents/diluents and product components are separated.
In the event of a runaway reaction (e.g. rapid temperature rise) the deactivator can be immediately fed to the oligomerization reactor to terminate the reaction. The deactivation system may also include a basic compound (such as sodium hydroxide) to minimize isomerization of the products (as activator conditions may facilitate the isomerization of desirable alpha olefins to undesired internal olefins).
Polymer removal (and, optionally, catalyst removal) preferably follows catalyst deactivation. Two "types" of polymer may exist, namely polymer that is dissolved in the process solvent and non-dissolved polymer that is present as a solid or "slurry".
Solid/non-dissolved polymer may be separated using one or more of the following types of equipment: centrifuge; cyclone (or hydrocyclone), a decanter equipped with a skimmer or a filter. Preferred equipment include so called "self cleaning filters" sold under the name V-auto strainers, self cleaning screens such as those sold by Johnson Screens Inc. of New Brighton, Minnesota and centrifuges such as those sold by Alfa Laval Inc. of Richmond, VA (including those sold under the trade name Sharpies).
Soluble polymer may be separated from the final product by two distinct operations. Firstly, low molecular weight polymer that remains soluble in the heaviest product fraction (C2o+) may be left in that fraction. This fraction will be recovered as "bottoms" from the distillation operations (described below). This solution may be used as a fuel for a power generation system.
An alternative polymer separation comprises polymer precipitation caused by the removal of the solvent from the solution, followed by recovery of the precipitated polymer using a conventional extruder. The technology required for such separation/recovery is well known to those skilled in the art of solution polymerization and is widely disclosed in the literature.
In another embodiment, the residual catalyst is treated with an additive that causes some or all of the catalyst to precipitate. The precipitated catalyst is preferably removed from the product at the same time as by-product polymer is removed (and using the same equipment). Many of the catalyst deactivators listed above will also cause catalyst precipitation. In a preferred embodiment, a solid sorbent (such as clay, silica or alumina) is added to the deactivation operation to facilitate removal of the deactivated catalyst by filtration or centrifugation.
Reactor fouling (caused by deposition of polymer and/or catalyst residue) can, if severe enough, cause the process to be shut down for cleaning. The deposits may be removed by known means, especially the use of high pressure water jets or the use of rhnhants \ I PGmi \ Srtritt1SC:Sne.r1701200ican.doex a hot solvent flush. The use of an aromatic solvent (such as toluene or xylene) for solvent flushing is generally preferred because they are good solvents for polyethylene.
The use of the heat exchanger that provides heat to the present process may also be used during cleaning operations to heat the cleaning solvent.
6 Distillation In one embodiment of the present invention, the oligomerization product produced from this invention is added to a product stream from another alpha olefins manufacturing process for separation into different alpha olefins. As previously discussed, "conventional alpha olefin plants" (wherein the term includes i) those processes which produce alpha olefins by a chain growth process using an aluminum alkyl catalyst, ii) the aforementioned "SHOP" process and iii) the production of olefins from synthesis gas using the so called Lurgi process) have a series of distillation columns to separate the "crude alpha product" (i.e. a mixture of alpha olefins) into alpha olefins (such as butene-1, hexene-1 and octene-1). The mixed hexene-octene product 16 .. which is preferably produced in accordance with the present invention is highly suitable for addition/mixing with a crude alpha olefin product from an existing alpha olefin plant (or a "cut" or fraction of the product from such a plant) because the mixed hexene-octene product produced in accordance with the present invention can have very low levels of internal olefins. Thus, the hexene-octene product of the present invention can be readily separated in the existing distillation columns of alpha olefin plants (without causing the large burden on the operation of these distillation columns which would otherwise exist if the present hexene-octene product stream contained large quantities of internal olefins). As used herein, the term "liquid product" is meant to refer to the oligomers produced by the process of the present invention which have from 4 to 26 (about) 20 carbon atoms.

In another embodiment, the distillation operation for the oligomerization product is integrated with the distillation system of a solution polymerization plant (as disclosed in Canadian patent application no. 2,708,011, Krzywicki et al.).
If toluene is present in the process fluid (for example, as a solvent for a MAO
activator), it is preferable to add water to the "liquid product" prior to distillation to form a water/toluene azeotrope with a boiling point between that of hexene and octene.
The liquid product from the oligomerization process of the present invention preferably consists of from 20 to 80 weight% octenes (especially from 35 to 75 weight%) octenes and from 15 to 50 weight% (especially from 20 to 40 weight%) hexenes (where all of the weight% are calculated on the basis of the liquid product by 100%.
The preferred oligomerization process of this invention is also characterized by producing very low levels of internal olefins (i.e. low levels of hexene-2, hexene-3, octene-2, octene-3 etc.), with preferred levels of less than 10 weight%
(especially less than 5 weight%) of the hexenes and octenes being internal olefins.
In-Situ Polymerization One embodiment of the present invention encompasses the use of components (i) (ii) and (iii) in conjunction with one or more types of olefin polymerization catalyst system (iv) to oligomerize ethylene and subsequently incorporate a portion of the trimerisation product(s) into a higher polymer.
Component (iv) may be one or more suitable polymerization catalyst system(s), examples of which include, but are not limited to, conventional Ziegler-Natta catalysts, metallocene catalysts, monocyclopentadienyl or "constrained geometry"
catalysts, phosphinimine catalysts, heat activated supported chromium oxide catalysts (e.g.
"Phillips"-type catalysts), late transition metal polymerization catalysts (e.g. diimine, 11,11rhantel I DrInsun1Crntil crcn.Alni)nnt r n do cx c -diphosphine and salicylaldimine nickel/palladium catalysts, iron and cobalt pyridyldiimine catalysts and the like) and other so-called "single site catalysts" (SSC's).

Ilek,i,,,õ4,,ion"õ,,,µc,,aocro÷.,m17not Can docx

Claims

The embodiments of the invention in which an exclusive property or privilege is claimed are defined as follows:

1. A process for the oligomerization of ethylene in at least two reactors, wherein said process comprises a) providing ethylene to a mixed reactor in the presence of an oligomerization catalyst under oligomerization conditions, thereby producing an initial oligomer product;
b) discharging said initial oligomer product from said mixed reactor and directing it to a tubular reactor;
c) forming additional oligomer product in said tubular reactor;
d) discharging from said tubular reactor an oligomerization stream containing said initial oligomer product and said additional oligomer product; and e) recycling a portion of said oligomerization stream from the discharge of said tubular reactor back to said mixed reactor;
with the provisio that said oligomerization catalyst comprises a chromium catalyst having a bridged diphosphine ligand.

2. The process according to claim 1, wherein said bridged diphosphine ligand is defined by the formula (R1)(R2)-P1-bridge-P2(R3)(R4) wherein R1, R2,R3 and R4 are independently selected from the group consisting of hydrocarbyl and heterohydrocarbyl and the bridge is a divalent moiety that is bonded to both phosphorus atoms.

3. The process of claim 2 which further comprises the addition of an aluminoxane activator to said mixed reactor.

4. The process according to claim 2 wherein said bridge is -N(R5)- wherein R5 is selected from the group consisting of hydrogen, alkyl, substituted alkyl, aryl, substituted aryl, aryloxy, substituted aryloxy, halogen, alkoxycarbonyl, carbonyloxy, alkoxy, aminocarbonyl, carbonylamino, dialkylamino, and silyl groups.

5. The process of claim 1 wherein said mixed reactor has a gas space.

6. The process of claim 1 wherein said mixed reactor contains a gas/liquid ejector.

7. The process of claim 1 wherein said tubular reactor has external cooling means.

8. The process of claim 1 wherein a pump is provided between said mixed reactor and said tubular reactor.