5 Background of the Invention:

This invention relates to a continuous process for the polymerization of ethylene and α-olefins with a metallocene catalyst system using dilute α-olefin feed.

Olefin polymerizations, particularly ethylene/α-olefin copolymerizations, can b 0 broadly differentiated as occurring in solution, suspension, or in the gas phase.

Within the continuous solution polymerization category, operating conditions can var quite broadly depending on such variables as the concentration of the reactants in the total feed, the nature of the catalyst system employed, the desired molecular weight of the polymer, and the desired monomer/comonomer ratio within the final 5 polymer.

When concentrated ethylene and α-olefins, e.g., propylene, feeds are copolymerized with conventional Ziegler-Natta catalysts, it is known, as described in U.S. Patent Nos. 3,912,698 and 3,637,616, to conduct such polymerizations continuously to obtain an ethylene copolymer, dissolved in the solvent, which is o continuously removed and isolated by known means. Unreacted monomers leaving the reactor are recovered and recycled to the reactor along with fresh monomers to replace those polymerized.

It is also known, as in EPA 270,339, to conduct continuous ethylene/α-olefin copolymerization under highly dilute conditions at atmospheric pressure using 5 conventional Ziegler-Natta catalysts. These processes suffer the drawback that the catalysts exhibit low productivities and produce polymer product exhibiting broad molecular weight distributions, high ash content, and number average molecular weights too large to be useful as lubricant additives. Consequently, if low molecular weight polymers are desired either hydrogen must be used to keep the molecular o weight of the product low, e.g. less than about 15,000, or the catalyst concentration has to be increased to extremely high levels to obtain low molecular weights. The hydrogen treatment at least partially saturates the terminal double bonds in the product, thereby significantly reducing or destroying the polymer's utility for most functionalization reactions, e.g., those used in the production of dispersants. 5 Saturated polymers have limited applicability for use in subsequent fuctionalization techniques (e.g., by "ene" reaction with maleic anhydride) which rely on a high terminal double bond content to achieve functionalization.

In contrast, recent developments in the catalyst and ethylene/α-olefin (EAO) copolymer art disclose that metallocene catalyst systems yield low molecular weight polymers with high terminal ethenylidine (vinylidene) content directly, without hydrogenation, as well as other advantageous properties (see, EP Publication Nos. 129,368; 440,504; 440,505; 440,506; 440,507; 440,508; 441 ,548; PCT Publication Nos. WO 91/11488; WO 90/01503; and U.S. Patent Nos. 5,017,299; 5,128,056; 5,151 ,204; 4,704,491; 4,668,834; 4,888,393; and 4,542,199).

More specifically, it is known in the art to employ high pressure/high temperature systems, as in U.S. Patent No. 5,084,534 and EP Publication 260,999, which utilize pure or nearly pure feeds and metallocene catalyst at pressures up to 2,500 bar and temperatures up to 300°C. Such systems are designed to produce high molecular weight polymers at high catalyst productivities (i.e. grams of polymer produced per gram of catalyst used). These systems suffer a number of drawbacks when applied to low molecular weight polymer production; most notably, expensive pure feeds and specialized equipment resulting in high fixed costs of production. Also, such systems operate with a single phase to allow efficient mixing of the reactants and, therefore, homogeneity of the product.

A single phase system is achieved by operating at temperatures and pressur sufficiently high to compress the ethylene and make it dense enough to dissolve the polymer product therein. This produces a homogeneous solution of polymer in reactant. To achieve high temperature and reduce the size of the reaction zone, the process is run adiabatically (heat is not removed), making temperature control difficult. Since the molecular weight of the product is directly related to temperature, failure to maintain constant temperature throughout the reaction process results in increased polydispersity (or Molecular Weight Distribution, MWD). Temperature control becomes increasingly more difficult at higher conversions in an adiabatic system. Consequently, conversions in the high temperature/high pressure process are kept to a minimum. For polymers having molecular weight of 100,000 or more, variations of + 1 ,000 or so have little effect on MWD. For polymers on the order of 10,000 molecular weight and below, however, such variations are extremely disadvantageous.

Moreover, the use of pure feeds is another limiting factor on the rate of conversion. As the conversion rate in a pure feed system is increased, the concentration of polymer in the reactor increases until it becomes extremely difficult or impossible to mix and pump the reactants efficiently. This problem is exacerbate at a low reaction temperature where polymer viscosity increases further. The limitations on conversion induced by pure feeds applies to essentially all polymerization processes.

Typically, a low conversion system requires the recycling of ethylene and comonomer out of the product stream and back into the reactor. Such a recycle system is disclosed in PCT Application No. EP92/00377 (Publication No. WO 92/14766) wherein unreacted ethylene must be separated out of the product stream and then repressurized into the reactor vessel. However, since catalyst is present i the recycle and ethylene is extremely self-reactive, the problem of ethylenic polymerization arises, thereby necessitating the use of a catalyst "killer" that can suppress polymerization in the recycle stream. Such a "killer" is not needed in the present invention since the improved mass transfer associated with large quantities diluent ensure that at least 90 percent of the ethylene will be converted. Hence, littl or no ethylene appears in the product output stream. The only process akin to a "recycle" in the present invention is the use of a reflux condenser in the preferred boiling reactor embodiment. This still poses no difficulty, as metallocene catalyst systems will not vaporize into the vapor space of a boiling reactor and therefore do not appear in the reflux.

Others have attempted to prepare low molecular weight EAO at low temperature and pressure, with metallocene catalyst as described in U.S. Patent No 4,704,491 to Mitsui Petrochemical Industries and U.S. Patent No. 4,668,834 to Uniroyal. The process described in Mitsui '491 operates with high catalyst concentrations, e.g., 10~2 moles/liter, pure undiluted vaporized feeds, at atmospheri pressure, extremely short reactant residence time (e.g., about 0.5 seconds), with no recycle of unreacted reactants. The high catalyst concentrations are needed because the mass transfer of the reactants into solution is poor and, consequently, low concentrations of reactants appear in solution. Low conversions are the result. The Uniroyal '834 patent operates at super atmospheric pressure with a compressor driven cooling system and pure undiluted feeds.

Methods employing dilute reaction mixtures and utilizing batch processes are known in the art. Typically, dilution of the reaction mixture occurs as a result of employing a metallocene catalyst system in a diluent, usually toluene. However, th use of a dilute feed of α-olefin is not found in this art. Moreover, rapid introduction o reactants into solution is often accomplished by introducing the pure reactants directly into the vapor space of the reactor instead of the liquid phase, or by bubblin the reactants up through the reaction mixture at pressures too low to provide effecti dissolution therein. Such processes are also conducted at very low monomer conversions.

KAMINSKY, et al., U.S. Patent No. 4,542,199, describes a batch process wherein pure ethylene and an α-olefin are introduced into a pressure vessel containing a metallocene dissolved in toluene.

LUKER, U.S. Patent No. 5,023,388 refers to a batch process, wherein the metallocene is dissolved in diesel oil in the presence of large quantities of α-olefin and ethylene and hydrogen gas at 7 bar. The molecular weight distribution of the product is reported to be 2.8. SLAUGH, et al., EP 366, 212 published May2,1990, teaches continuous or batch processes, though the examples offered are all batch. The feeds used are pur and the reaction mixture is highly concentrated. The process produces polymer wherein 80 percent of the product has less than 20 carbon atoms per molecule.

TSUTSUI, et al., EP 447,035 published September18, 1991 , refers to a series of batch processes, wherein ethylene is first polymerized or copolymerized with α-olefin in a first batch under concentrated or dilute conditions; the product is isolated; and then the product is introduced into a subsequent batch process with ethylene or an α-olefin. The process may be continued to a third round of batch processing. Reactants may be relatively concentrated in one batch, yet relatively dilute in the next or vice-versa.

Another approach, as described by HIROSE, et al., JP 2-173,110 disclosed July 4, 1990, is to recycle massive amounts of ethylene and propylene gas through a solvent-containing reaction vessel. The feeds are pure and the quantity of reactants to solvent is very high. The ratio of ethylene to α-olefin is necessarily very low in order to prevent polyethylene formation. Polymers formed by this process have ethylene contents less than 10 percent by mole.

It is also known in the art to cool polymerization reactors by evaporation and removal of unreacted monomers from the vapor space, such monomers being optionally cooled and recycled to the reactor. Reactors cooled in this manner are referred to as evaporatively cooled reactors or boiling reactors. Polymer is recovere from the reaction mixture by withdrawing polymer solution from the reactor and separating unreacted monomers which are usually recycled to the reactor. Also, as a general proposition, as the concentration of polymer in solution increases, and/or the molecular weight of the polymer increases, the viscosity of the reaction mixture increases. This in turn reduces the mass transfer of ethylene from the gas into the liquid phase and reduces the heat transfer properties of the reaction mixture thereby making it more difficult to cool the reaction mixture. As indicated above, failure to maintain a stable reaction temperature leads to fluctuations in the molecular weight o the polymer and a broadening of the molecular weight distribution. While evaporative cooling reactors improve heat transfer by removing the exothermic heat of reaction, and can maintain stable reaction temperatures, they have the disadvantage that ethylene concentration in solution in the reactor is usuall less than its equilibrium value (i.e., ethylene is continually being quickly removed and

recycled). Thus, as a general proposition, in order to produce a copolymer containi a particular proportion of ethylene in evaporatively cooled reactors , it is usually necessary to recycle a larger amount of ethylene in the reactor off-gas (to obtain the cooling benefit) than would be the case if a sealed reactor were employed and the concentration of ethylene in solution in the reactor achieved its equilibrium value. Economically, this increase in recycle volume means greater expense than would otherwise be the case. See U.S. Patent No. 3,706,719. Moreover, if the reaction temperature is increased (e.g., above the critical temperature of ethylene at 9.2°C (48.5°F)), the ethylene mass transfer problem becomes more acute since the solubility of ethylene will be more difficult, thereby reducing gas/liquid phase mixing. In addition to ethylene imbalance in the vapor space and mass transfer problems, evaporatively cooled reactors also lead to the associated problem of reactor fouling and polyethylene segment formation. More specifically, because ethylene and α-olefins possess different reactivities, they co-polymerize at different rates. Moreover, because ethylene reacts with itself so much faster than with α-olefins, the copolymerization of ethylene with α-olefins can result in polymers having large crystalline polyethylene segments randomly interspersed with occasional α-olefin moieties.

These phenomena not only make it difficult to control the ethylene content in the polymer, reduce the solubility of the polymer in the reaction mixture, and consequently lead to reactor fouling, but also more importantly, they limit the utility o the polymer in applications extremely sensitive to crystallinity, such as making dispersants for lubricating oil compositions.

The conventional solution to controlling polymer ethylene content, when usin Ziegler-Natta catalysts, has been to regulate the concentrations of ethylene and α-olefin in the reaction mixture. For example, to obtain a copolymer of ethylene and propylene having approximately 50 mole percent of each monomer in the copolymer, it has been considered that a large excess of propylene, e.g., greater than 10:1 mole ratio, is necessary in the catalyst-containing solution in the reactor. In contrast, a copolymerization conducted in a solution containing about equal amounts of ethylen and propylene, produces a copolymer so high in ethylene content, that under ordina Ziegler-Natta polymerization conditions, e.g., about -20° to about 80°C, it would not be soluble in the saturated hydrocarbon solvents used as the polymerization mediu However, when ethylene and propylene, for example, are polymerized in a reactor having both liquid and vapor phases, the mole or weight ratio of propylene to ethylene in the vapor phase is typically far less than the corresponding propylene to ethylene ratio in the liquid phase because of the greater volatility of ethylene. For example, if the propylene to ethylene mole ratio in the liquid phase is about 10:1 , the

propylene to ethylene mole ratio in the vapor phase above it may be only about 1 :1 t about 3:1.

Uniformity of ethylene monomer incorporation, known as "compositional distribution", is also a function of the mass transfer of ethylene into the reaction zone i.e., uniform mixing of the co-monomers. However, as discussed above, in those reactor designs which employ recycle of the vapor phase, e.g., using a reflux condenser, the reflux condensate returning to the reactor will typically have sufficiently high ethylene concentrations that reactor agitation of fresh and recycled ethylene alone will not suffice to prevent insoluble polymers having randomly high ethylene content from forming and clogging up the system. Consequently, it has been conventional in the art to attempt to introduce process steps for reducing the ethylene content in the recycled condensate, e.g., by removing ethylene from the condensate before introduction into the polymerization reactor. See U.S. Patent Nos 3,706,719 (col. 5, line 68 et seq.); 3,637,616; and 3,912,698. Such steps are costly and inefficient.

Separate and distinct from the need to control monomer ratio in the recycle stream are the mass transfer problems associated with employing pure feeds, particularly mixed pure feeds, even when supplied to reaction zones employing a solvent which dilutes the pure feed as it is introduced into the reactor. For example, the introduction of pure feeds into liquid reaction mixtures necessarily creates a higher concentration gradient of monomer at its point of introduction relative to the remainder of the reactor. Thus a finite amount of time will be required to achieve uniform mixing of the monomer into the reaction mixture. As long as this higher concentration gradient exists, there will be a propensity to form higher molecular weight polymer species relative to the molecular weight of polymer species formed at monomer equilibrium concentrations, since molecular weight is a function of monome concentration. Broadened MWD and non-uniform compositional distribution are a result.

In view of the above, there has been a continuing need to develop more cost efficient processes for preparing EAO copolymers with metallocene catalyst systems; the present invention was developed in response to this need.

SUMMARY OF THE INVENTION

A process for continuously producing copolymer comprising monomer units derived from ethylene and α-olefin in the presence of a metallocene catalyst system and in a reaction zone containing liquid phase which comprises:

(A) continuously providing a dilute liquified α-olefin feed stream comprisin at least one α-olefin reactant and diluent admixed therewith wherein the amount of diluent in said feed stream is at least 30% weight percent thereof;

(B) providing a feed stream comprising ethylene in liquid, vapor, or liquid/vapor form;

(C) admixing, the feed streams of steps (A) and (B) in amounts sufficient t provide a reactant feed stream having an α-olefin/ethylene weight ratio effective to yield a co-polymer containing between about 5 to about 70 weight percent monome units derived from ethylene; (D) continuously introducing reactant feed stream derived in accordance with step (C) and metallocene catalyst system into the liquid phase of the reaction zone in a manner and under conditions sufficient to:

(i) polymerize the ethylene and α-olefin to polymer product having a number average molecular weight of not greater than about 15,000; (ii) obtain an α-olefin conversion of at least 30%;

(iii) obtain an ethylene conversion of at least 70% (E) continuously withdrawing copolymer product from the reactor.

Several advantages result from the instant invention: (1 ) the use of dilute monomer feeds results in a lower concentratio gradient at the point of monomer introduction into the reactor and, consequently, le time is required to achieve uniform monomer mixing and less time is available for higher molecular weight species formation at the input port;

(2) the use of dilute feeds enables the process to operate at high conversion rates of ethylene/α-olefin without the attendant buildup of mass transfer resistance attributable to polymer formation in pure feed systems;

(3) in the preferred embodiment of the process of the present invention employing a boiling reactor and dilute feed, ethylene in the vapor space a in the liquid reaction mixture are in equilibrium. This is achievable because the reaction mixture gives essentially no mass transfer resistance at the liquid/vapor interface because uniform mixing is easily obtained;

(4) as a consequence of high ethylene conversions obtainable wit a metallocene catalyst system, the amount of unreacted ethylene entering the vapo space and in equilibrium with dissolved ethylene is minimized; minimization of ethylene vapor buildup facilitates recycle of ethylene and does not hinder recycle of the α-olefin;

(5) still further improvements are made possible by employing a fe diluent, such that the major constituents of the diluent boil at about the same

temperature as the α-olefin to be copolymerized with ethylene. Accordingly, ethyle content in the vapor space is further diluted by the α-olefin feed constituents, a maj portion of which is diluent. Thus, evaporative cooling does not depend on recycle o high amounts of ethylene in the vapor, ethylene buildup in the reflux is further minimized, and mass transfer resistance to ethylene mixing is further reduced;

(6) fouling caused by ethylene build-up at the reflux return port and attendant polyethylene formation is easily avoided since the more volatile ethylene i also more reactive than the α-olefin monomer, the proportion of olefin monomer to ethylene is typically greater in the vapor space than in the feed (occasionally may b equal to or only slightly less than that in the feed); and the concentration of ethylen in the vapor space is typically less than that in the feed;

(7) uniformity of polymer is greatly enhanced without the need for manipulation of the condensed vapor to alter its compositional distribution;

(8) the combined use of dilute feed and high conversion facilitates removal of catalyst (deashing) residue and quenching of the polymer/catalyst mixtur since it is easier to mix the polymer with deashing and quench media;

(9) use of dilute α-olefin containing feeds and high conversion not only permits adaptation of metallocene chemistries to evaporatively cooled reactors but it also allows for a significant improvement in the overall economics of the process because such dilute feeds can be readily obtained at very low cost as by¬ product waste streams derived from other commercial sources.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 is a flow diagram of the dilute feed polymer process. Figure 2 is a schematic view of the reflux boiling reactor system. Figure 3 is a schematic view of preferred dilute feed polymer production process. Figure 4 is a graphic representation of the mass flows of the process of Figure 3. Figure 5 is a graphic representation of the mass flows of a variation on the process of Figure 3.

DETAILED DESCRIPTION OF THE INVENTION

Polymers produced in accordance with the process of the present invention are polymers comprising monomer units derived from ethylene, and at least one α- olefin. Such monomers are characterized by the presence within their structure of a least one ethylenically unsaturated group of the structure >C=CH2- In addition, the are highly reactive at low catalyst concentrations. Metallocene catalyzed polymerizations are particularly adaptable for use with α-olefin monomers which: a)

have at least one hydrogen on the 2-carbon (hence, isobutylene polymerizes extremely poorly); b) have at least two hydrogens on the 3-carbon; c) have at least one hydrogen on the 4-carbon.

Accordingly, suitable α-olefin monomers include those represented by the structural formula wherein R ¹ is straight chain or branched chain alkyl radical comprising 1 to 18 carbon atoms and wherein the polymer formed therefrom contains a high degree of terminal ethenylidene unsaturation. Preferably R ¹ in the ^• above formula is alkyl of from 1 to 8 carbon atoms, and more preferably is alkyl of from 1 to 2 carbon atoms. Therefore, useful comonomers with ethylene include propylene, butene-1 , pentene-1 , 4-methylpentene-1 , hexene-1 , octene-1 , decene-1 , dodecene-1 , tridecene-1 , tetradecene-1 , pentadecene-1 , hexadecene-1 , heptadecene-1 , octadecene-1 , nonadecene-1 and mixtures thereof (e.g., mixtures of propylene and butene-1 , and the like).

The ethylene content of the polymers prepared in accordance with this invention is preferably in the range of between 5 and 70 wt. % (e.g., 11 to 70 wt. %), and more preferably between 15 and 50 wt. % (e.g., 15 to 45 wt. %).

The process of the present invention is controlled to make polymer having a number average molecular weight of not greater than 15,000 and typically from 300 t 15,000 (e.g., from 300 to 10,000), preferably from 900 to 8,000; more preferably from 1 ,000 to 5,000 (e.g., from 1 ,000 to 3,000). The number average molecular weight fo such polymers can be determined by several known techniques. A convenient method for such determination is by size exclusion chromatography (also known as gel permeation chromatography, GPC) which additionally provides molecular weight distribution information (see W.W. Yau, J.J. Kirkland and D.D. Bly, "Modern Size Exclusion Liquid Chromatography", John Wiley and Sons, New York, 1979).

The polymers produced in the process of this invention preferably exhibit a degree of crystallinity such that they are essentially amorphous.

The polymers produced in this invention are further characterized in that up to about 95% or more of the polymer chains possess terminal ethenylidene-type unsaturation. Thus, one end of such polymers will be of the formula POLY- CO ^"1 )=CH2 wherein T ¹ is C*ι to Cιβ alkyl, preferably C-] to Cβ alkyl, and more preferably C-j to C2 alkyl, (e.g., methyl or ethyl) and wherein POLY represents the polymer chain. The chain length of the T ¹ alkyl group will vary depending on the comonomer(s) selected for use in the polymerization. A minor amount of the polyme chains can contain terminal vinyl unsaturation, i.e. POLY-CH=CH2, and a portion of the polymers can contain internal monounsaturation, e.g. POLY-C(T ¹ )=CH(T ² ), wherein T ¹ and T* ² are as defined for J - above. The polymer products of this inventive process comprise chains which can be saturated by hydrogen, but

preferably contain polymer chains wherein at least 30, preferably at least 50, more preferably at least 60, and most preferably at least 75 percent (e.g. 75-98%), of whic exhibit terminal ethenylidene (vinylidene) unsaturation. The percentage of polymer chains exhibiting terminal ethenylidene unsaturation may be determined by Fourier Transform Infrared (FTIR) spectroscopic analysis, titration, proton (H)NMR, or C ¹³ NMR.

The molecular weight distribution (Mw/Mn) of the copolymers will be typically less than 5, preferably less than 4, and most preferably less than 3, e.g. between 1.5 and 2.5. The preferred α-olefin monomers are derived from dilute α-olefin containing refinery streams, such as Raffinate-2, and are butene-1 and propylene. Most preferred is the copolymerization of butene-1 with ethylene for the production of ethylene/butene-1 polymers (hereinafter referred to as EB or EB-1 Polymers) useful in the manufacture of dispersants. The advantages of such copolymers may be foun in commonly assigned U.S. Serial Number 992,192, entitled "Polymers Derived from Ethylene and Butene-1 for Use in the Preparation of Lubricant Dispersant Additives" filed on an even date herewith, (Docket No.: PT-944).

The ethylene content of the most preferred EB polymers is typically in the range of from 15 to 50 (e.g. 17 to 50), more preferably from 15 to 48 (e.g. 30 to 45), and or more preferably from 17 to 45 (e.g. 17 to 30) weight percent. The balance of the polymer content is substantially based upon units derived from butene-1. Thus, the butene-1 content is typically from 95 to 50 (e.g., 90 to 50), more preferably from 95 to 55 (95 to 65), and most preferably from 90 to 65 (e.g. 90 to 70) weight percent based on the total polymer weight. However, the EB polymers of this invention may optionally contain minor or trace amounts (e.g. up to 5 wt. %) of olefins other than ethylene and butene-1 since the reactant streams which may be used to prepare the EB polymers may contain such olefins as isobutylene, isopentene, butene-2, and butadiene. Similar considerations apply for other than EB copolymerization as well. The process of the present invention utilizes a metallocene catalyst system. Such metallocenes are extremely unreactive with non-terminal olefins, and terminal olefins which lack at least one hydrogen atom on the second carbon (e.g., isobutylene), at least two hydrogens on the third carbon (e.g., isopentene), or at least one hydrogen on the fourth carbon (e.g., 4,4-dimethylpentene-1 ). Hence, as described hereinafter, many of the components in refinery streams, such as Raffinate-2 (e.g. 2-butenes, and isobutylene) are essentially non-reactive in a metallocene system and become suitable diluents for use in the present process which need not be separated from the feed. Other constituents such as butadiene

are made non-reactive or non-poisonous by pre-saturating the double bonds with hydrogen.

Metallocene Catalyst System - General Description Metallocene Catalyst Systems

The process of this invention can be utilized with catalyst in which the catalyst is preferably a bulky ligand transition metal compound. The bulky ligand may contai a multiplicity of bonded atoms, preferably carbon atoms, forming a group which may be cyclic with one or more optional heteroatoms. The bulky ligand may be a cyclopentadienyl derivative which can be mono- or polynuclear. One or more bulky ligands may be bonded to the transition metal atom. The transition metal atom may be a Group IV, V or VI transition metal ("Group" refers to an identified group of the Periodic Table of Elements, comprehensively presented in "Advanced Inorganic Chemistry," F.A. Cotton, G. Wilkinson, Fifth Edition, 1988, John Wiley & Sons). Other ligands may be bonded to the transition metal, preferably detachable by a cocatalyst such as a hydrocarbyl or halogen leaving group. The catalyst is derivable from a compound of the formula

[L] _m M [X] _n wherein L is the bulky ligand, X is the leaving group, M is the transition metal and m and n are such that the total ligand valency corresponds to the transition metal valency. Preferably the catalyst is four coordinate such that the compound is ionizable to a 1 ⁺ valency state.

The ligands L and X may be bridged to each other and if two ligands L and/or X are present, they may be bridged. The metallocenes may be full-sandwich compounds having two ligands L which are cyclopentadienyl groups or half-sandwich compounds having one ligand L only which is a cyclopentadienyl group.

For the purposes of this patent specification the term "metallocene" is defined to contain one or more cyclopentadienyl moiety in combination with a transition metal of the Periodic Table of Elements. In one embodiment the metallocene catalyst component is represented by the general formula (Cp) _m MR _n R'p wherein Cp is a substituted or unsubstituted cyclopentadienyl ring; M is a Group IV, V or VI transition metal; R and R" are independently selected halogen, hydrocarbyl group, or hydrocarboxyl groups having 1-20 carbon atoms; m = 1-3, n = 0-3, p = 0-3, and the sum of m + n + p equals the oxidation state of M. In another embodiment the metallocene catalyst is represented by the formulas:

(C5R'm)pR"s(C5R ^, m)MeQ3. _p .χ and

R CsR mte eQ ¹

wherein Me is a Group IV, V, or VI transition metal CsR'm is a substituted cyclopentadienyl each R', which can be the same or different is hydrogen, alkenyl ar alkaryl or arylalkyl radical having from 1 to 20 carbon atoms or two carbon atoms joined together to form a part of a C4 to CQ ring, R" is one or more of or a combination of a carbon, a germanium, a silicon, a phosphorous or a nitrogen atom containing radical substituting on and bridging two CsR' _m rings or bridging one CsR'm ring back to Me, when p = 0 and x = 1 otherwise x is always equal to 0, each Q which can be the same or different is an aryl alkyl, alkenyl, alkaryl, or arylalkyl radical having from 1 to 20 carbon atoms or halogen, Q' is an alkylidene radical having from 1 to 20 carbon atoms, s is 0 or 1 and when s is 0, m is 5 and p is 0, 1 or and when s is 1 , m is 4 and p is 1.

Various forms of the catalyst system of the metallocene type may be used in the polymerization process of this invention. Exemplary of the development of metallocene catalysts in the art for the polymerization of ethylene is the disclosure of US. Patent No. 4,871 ,705 to Hoel, U.S. Patent No. 4,937,299 to Ewen, et al. and EP- A-0 129 368 published July 26, 1989, and U.S. Patent Nos. 5,017,714 and 5,120,867 to Welborn, Jr. These publications teach the structure of the metallocene catalysts and include alumoxane as the cocatalyst. There are a variety of methods for preparing alumoxane, one of which is described in US. Patent 4,665,208. For the purposes of this patent specification, the terms "cocatalysts or activators" are used interchangeably and are defined to be any compound or component which can activate a bulky ligand transition metal compound. In one embodiment the activators generally contain a metal of Group II and III of the Periodi Table of Elements. In the preferred embodiment, the bulky transition metal compound are metallocenes, which are activated by trialkylaluminum compounds, alumoxanes both linear and cyclic, or ionizing ionic activators or compounds such as tri (n-butyl) ammonium tetra (pentafluorophenyl) boron, which ionize the neutral metallocene compound. Such ionizing compounds may contain an active proton, or some other cation associated with but not coordinated, or only loosely coordinated to the remaining ion of the ionizing ionic compound. Such compounds are described in EP-A-0520 732, EP-A-0277 003 and EP-A-0 277 004 published August 3, 1988, and U.S. Patent Nos. 5,153,157; 5,198,401 and 5,241 ,025. Specific teachings directed t bis(cyclopentadienyl) ionic activated systems can be found in U.S. Patent No. 5,153,157. Further, the metallocene catalyst component can be a monocyclopentadienyl heteroatom containing compound. This heteroatom is activated by either an alumoxane or an ionic activator to form an active polymerization catalyst system to produce polymers useful in this invention. These types of catalyst systems are described in, for example, PCT International Publicatio

WO 92/00333 published January 9, 1992, U.S. Patent Nos. 5,057,475; 5,096,867; 5,055,438 and 5,227,440 and EP-A-0 420 436, WO 91/ 04257. In addition, the metallocene catalysts useful in this invention can include non-cyclopentadienyl catalyst components, or ancillary ligands such as boroies or carbollides in combination with a transition metal. Additionally, it is not beyond the scope of this invention that the catalysts and catalyst systems may be those described in U.S. Patent No. 5,064,802 and PCT publications WO 93/08221 and WO 93/08199 published April 29, 1993. The catalyst system of the invention may be, optionally, prepolymerized or used in conjunction with an additive or scavenging component to enhance catalytic productivity; a scavenging component such as an organoaluminum or alkyl aluminum reagent.

The Reaction Process

The process of the present invention is continuous, employs a dilute feed, and is operated to achieve a high conversion of α-olefin and ethylene as defined herein. Within these parameters, the ethylene/α-olefin product is controlled to have a numbe average molecular weight of not greater than 15,000 using a metallocene catalyst system as described above. For purposes of this invention "continuous" means that monomer feed is continuously introduced into the reaction zone and resultant produc continuously withdrawn.

The advantages of employing a dilute monomer feed are described above. The diluent can be any non-reactive (under the conditions employed) material which preferably is: (i) capable of being liquified under reaction conditions; (ii) capable of dissolving at least the α-olefin monomer employed; and (iii) capable of dissolving or at least suspending the polymer product under reaction conditions such that viscosity buildup is sufficiently minimized to the extent that the mass transfer rate of the ethylene needed to homogeneously distribute ethylene throughout the reaction zone is at least equal to and preferably is greater than, the reaction rate at which ethylene is consumed in the polymerization reaction. Suitable but less preferred diluents include such solvents as alkanes, aromatic hydrocarbons, and nonreactive alkenes. It is contemplated that the non-reactive diluents comprise typically at least 30, preferably at least 40, and most preferably at least 50 weight % of the α-olefin feed stream and the diluent can range typically from 30 to 90 preferably from 40 percent t 80, and most preferably from 50 to 60 weight % of the α-olefin feed stream before admixture with ethylene.

It is a particular advantage of the present invention that the preferred diluents are naturally present in various refinery streams containing α-olefin monomer

reactants; to be useful such streams must contain at least one α-olefin as the reacti constituent. However, these streams typically will contain non-reactive constituents which have a similar carbon number to the α-olefin. The similarity in carbon numbe causes the non-reactive constituents to have similar boiling points to the α-olefin. Consequently, the non-reactive constituents will vaporize together with the α-olefin and not only dilute the α-olefin in the vapor space, but also the ethylene. This dilution effect decreases the mass transfer resistance of the reactive monomers in t ' vapor space, particularly ethylene.

Accordingly, a preferred diluent will contain components comprising typically least 50, preferably at least 75, and most preferably at least 95 weight %, and typically from 50 to 100, preferably from 75 to 100, and most preferably from 95 to 100 weight % thereof, having a boiling point at the reaction conditions of typically within +20°C, preferably within +15°C, and most preferably within +10°C of the average boiling point of the α-olefin constituents of the feed. Representative of suc refinery streams are those which contain butene-1 , propylene or C5 α-olefin.

Preferred butene-1 containing streams are referred to herein as Raffinate-2 streams Such streams typically have isobutylene content significantly lowered in relation to the stream from which they are derived. Raffinate-2 is typically derived from either butane/butene catalytic cracking refinery streams (BB-streams) or Raffinate-1 which in turn, is derived from butadiene crude produced by steam cracking plants. The composition of Raffinate-2 can vary widely, depending upon the source, e.g., (weig %):

Crude Raff-2 Raff-2

Component Butadiene From Crude BB From BB Raff-1

BUTADIENE 43.5+20 0 - 5 0.3+.15 0.4+0.2 0.1 +.05

ISOBUTYLENE 25.2+10 0 - 5 12.6+6 0.2+0.1 44.6+20

BUTENE-1 15.5+8 49.5+25 13.6+6 15.4+7 27.4+15

CIS-BUTENE-2 2.0+1 6.4+3 9.0+4 10.2+5 3.5+1.5

TRANS-BUTENE-2 6.2+3 19.6+10 13.8+6 15.6+7 10.9+5

N-BUTANE 4.6+2 14.7+7 10.5+5 12.0+6 8.1+4

ISOBUTANE 2.9+1.5 9.4+4 36.7+15 42.1+20 5.2+2.5

^♦ OTHER 0.1 +0.5( ^a ) 0.2+0. l( ^b ) 3.5+1.5 4.1+2 0.2+.1

*Other: (a) includes propane, propene, pentanes, pentenes, water, trace other hydrocarbons.

(b) Raffinate-2 derived from MTBE production (using BB-stream or

Raffinate-1 ) will include traces of MTBE, methanol, di-methyl ether, and tert-butyl alcohol.

Typical commercially available butene-1 concentrations in Raffinate-2 range from about 15 to about 55 weight %. The above butene-1 containing refinery stream are preferred for making ethylene/butene-1 copolymer (EB-1 ) which has been found to be highly effective as a backbone for the production of lubricants, oil dispersants, and viscosity modifiers. The instant invention may also make use of BB streams and Raffinate-1 directly, since isobutylene is almost entirely unreactive with metallocene catalyst systems. Hence, depending upon shipping costs, convenience, or whatever other factors may affect the decision-making process, the practitioner has the option of either acquiring Raffinate-2 and running it through the process of the instant invention or first acquiring either Raffinate-1 or a BB stream, running it through the process, and then shipping the resultant isobutylene-enriched stream on to an MTBE plant or other end use. The use of Raffinate-2 is the more preferred. The use of crude butadiene streams directly is not desired since it would waste butadiene durin hydrogenation. It is not necessary that refinery streams be used and, in fact, it is contemplated that dilute α-olefin containing streams can be prepared by separately combining pure α-olefin and one or more pure diluents, e.g. pure isobutane, such as those typically found in the above refinery streams.

It will also be seen that this invention is useful in the production of virtually an ethylene/α-olefin copolymer and may therefore be used in the processing of other dilute refinery streams, such as dilute propene and pentene streams common in the industry. Dilute refinery propene streams, known in the industry as "C3 streams", an dilute refinery pentene streams, known as "C5 streams", are also derived from steam and catalytic cracking and generally can be represented to comprise the following components (ranges, weight %): For C3 streams: Propylene = 55 + 20; Propane = 3 + 15; Ethylene ^* = 2 + 1 ; Ethane = 8 + 4; and Other = 1 +.5 (Other includes methane, acetylenes, propadiene trace C4's and C5's, and trace polar compounds such as water, carbonyl sulfide, methyl mercaptan, and hydrogen sulfide). For C5 streams composition is more complex than that of C3 and C4 streams:

Component Ranqe (weiαht %)

2-METHYL-BUTENE-1 9.0 ± 4

3-METHYL-BUTENE-1 1.6 + 1

PENTENE-1 5.1 ± 2

2-METHYL-BUTENE-2 14.9 + 7

PENTENE-2 15.4 + 7

ISOPRENE 0.7 ± -3

ISOPENTANE 36.2 + 15 n-PENTANE 5.5 ± 2

CYCLOPENTANE 0.6 ± -3

CYCLOPENTENE 1.5 + .75

PIPERYLENE 0.9 ± -4

C6 OLEFINS 1.5 + .75

C6 ALKYLS 3.5 + 1.5

C7s AND C8s 2.0 ± 1

^♦ OTHERS 1.6 + 1

^♦ Others include benzene and polar compounds.

Pentene-1 and cyclopentene are the most reactive components of a C5 stre in the presence of a metallocene catalyst system and are easily separated from ea other by distillation and are easily concentrated. Whether a constituent, e.g. of the refinery stream, qualifies as a diluent und reaction conditions depends on whether it is non-reactive which in turn depends on the type of pretreatment to which the feed is subjected. "Non-reactive" when used i conjunction with diluent is meant that less than 5 wt.%, preferably less than 3 wt.%, and most preferably less than 1 wt.% of the constituent present in the feed is incorporated into the polymer product and the constituent does not totally deactivat the metallocene catalyst system. Typically any saturated hydrocarbon constituent qualify as diluent as well as highly unreactive (in the presence of a metallocene catalyst system) unsaturated constituents such as butene-2 and isobutylene. Materials such as butadiene tend to deactivate the catalyst. Hence, it is preferred that they be removed or at least partially saturated by hydrogenation. Once saturated, the butadiene becomes part of the diluent as butane, butene-2, or reacti butene-1.

The process of the invention is controlled to achieve high ethylene and α-olef conversion. Conversion is directly proportional to monomer concentration, catalyst concentration and residence time. Accordingly, the above parameters are controlle to achieve an ethylene conversion of typically at least 70%, preferably at least 80%, 5 and most preferably at least 90% and can range typically from 70% to 100%, preferably from 80% to 100% and most preferably from 90% to 100% (e.g., 90-95%). The α-olefin conversion is controlled to be typically at least 30%, e.g., at least 40%, preferably at least 50%, and most preferably at least 60% and can range typically from 30% to 95%, preferably from 40% to 90% and most preferably from 50% to 90 0 Monomer conversion (%) can be determined by either of the following equations:

- wt/hr of monomer incorporated into polymer _x -| oo or wt/hr of monomer in feed ₅ - wt/hr monomer in feed-wt hr monomer not reacted y I QQ wt/hr monomer in feed

The particular α-olefin conversion employed depends in part on the ethylene content sought to be imparted to the polymer and hence on the ethylene 0 concentration in the mixed feed. For example, at low ethylene content the α-olefin conversion typically will be lower than for high ethylene content feeds. While high conversion can be achieved by any combination of process conditions affecting conversion, it is preferred to maintain a low catalyst concentration and low monomer concentration and attain high conversion with a long residence time. Preferably the 5 ethylene conversion is controlled in a manner such that the ratio of the weight % of ethylene in the vapor phase to the weight % of ethylene in the reactant feed stream i typically not greater than 1.2:1 , preferably less than 1 :1 and most preferably from 0.1:1 to 0.7:1 (e.g. 0.1:1 to 0.5:1). The monomer in the reaction mixture is kept low through the use of the diluent in the feed and operating at high conversions. o The catalyst concentration is typically held just above the poison level due to cost of the catalyst. Preferably the feed is treated to remove most if not all catalyst poisons. Minor poison contamination can be accommodated by increasing the catalyst system concentration with the excess used to remove the poison by reaction therewith. Accordingly, while any effective catalyst concentration can be employed i 5 is contemplated that such effective amounts will be sufficient to achieve a weight rati of metallocene catalyst system to polymer product of typically from 1 x 10 ^** 4.1 to 7 x 10-4:1.

The residence time is determined from the equation:

total true volume of liquiud in reactor Residence time = total volume/time of liquid exiting reactor

wherein gas bubble volume in the liquid is subtracted from apparent volume of liquid in reactor to obtain true volume. Accordingly, typical residence times can vary typically from 0.1 to 5 hrs., preferably from 0.5 to 4 hrs., and more preferably from 1 3 hrs.

Reaction temperature and pressure are preferably controlled to liquify the diluent and α-olefin. However, the reaction temperature is typically selected to be above the critical temperature of ethylene but below the critical temperature of the α-olefin feed and/or diluent. Accordingly, while any effective temperature can be employed it is contemplated that such effective temperatures for a feed containing butene-1 will range typically from 30 to 150°C, preferably from 50 to 120°C, and mo preferably from 60 to 110°C. For the dilute refinery streams of propylene having propane as the major diluent, the critical temperature of propylene and propane are 92.42°C (198.36°F) and 96.7°C (206.06°F) respectively, so the typical range of reaction temperatures would be 30 to 96, and preferably from 60 to 92°C. The critic temperature of the feed components in the reactor places an upper limit on temperature when using a boiling reactor since the reflux mechanism becomes useless if nearly all or all of the feed flashes into the reactor vessel and there remai no liquid phase to reflux. In less preferred embodiments the operation above the critical temperature of the major reactor constituents must be compensated for by assisting or eliminating the reflux mechanism altogether and relying on alternative cooling means, such as jacket cooling or internal cooling coils. Neither of these solutions is as effective nor as efficient as reflux cooling in maintaining homogeneity of temperature throughout the reaction solution. As indicated above, the boiling reactor represents the preferred method for temperature control. Variations on the boiling reactor configuration include internal reflux, e.g. using cooling coils inserted into the vapor space or an external system wherein vapor is removed from the vapor space and introduced to an external reflux apparatus, the vapor condensed and the condensate returned to the reactor and/or feed. Alternative non-reflux temperature control means include pumparound cooling where liquid is removed from the reactor cooled, and then returned to the reactor. Pumparound cooling offers the added advantage of being able to return cooled liquid to the reactor using high pressure pumps to also provide mixing of reactor contents with high speed jets.

Reactor pressures are typically controlled to maintain the diluent and α-olefin in liquid form at the selected temperature. In boiling reactors the pressure is selecte to obtain boiling of the diluent/α-olefin reactor constituents at the reaction temperature. Accordingly while any effective pressure can be employed it is 5 contemplated that such effective pressures for butene-1 feeds will range typically from about 2.4 to about 39 Atm, preferably from about 4.4 to about 28 Atm, and mos preferably from about 5.6 to about 23.5 Atm.

The reaction mixture is preferably vigorously mixed by any suitable means such as impeller, jet pump, or vigorous boiling or combinations thereof. Baffles and o strategic placement of feed input can be employed to further facilitate mixing.

More specifically, the molecular weight distribution (Mw/Mn) of the polymer is broadened by variations of: temperature, monomer concentration, and catalyst concentration.

5 The Integrated Process

Referring now to Figure 1 , there is depicted a schematic diagram of the overa process scheme. A dilute propylene, butene, or pentene refinery feed is piped to a diene removal system 50. It is desirable to remove or saturate any dienes that may o contaminate the feed since dienes tend to either poison metallocene catalyst systems, crosslink polymer chains, or both. Alkynes are also poisonous and must b removed or saturated. However, this step in the process is an option since dilute refinery streams vary widely in composition and, therefore, may have virtually no diene contamination or, in fact, none at all. 5 The next step in the process is to carry the refinery stream to a concentration system 51 where more volatile ("lights") and less volatile ("heavies") components as well as catalyst-poisoning polar compounds, such as water, sulfides, alcohols, and nitrogen, sulfur, and oxygen derivatives in general may be removed so as to bring th weight percent of the olefin monomer in relation to the nonreactive diluents to within o the range desired. In cases where it is desired to recycle unreacted olefin monomer and/or to further dilute a refinery feed having excess concentration of olefin monome a recycle stream 52 may be combined with the input stream 53. The source of this recycle stream will be discussed below in connection with the diluent removal syste 65. 5 After adjusting the concentration and removing polar poisons, it is preferable t pass the feed through a scavenger bed system 54 that removes any remaining catalyst poisons and filters out any particulates. The resulting pre-treated dilute

refinery feed is piped to the preferred boiling reactor system 55 depicted in more detail in Figure 2 and co-polymerized with ethylene.

More specifically, referring to Figure 2, there is provided a reactor vessel 1 having an external reflux condenser 2 in fluid communication therewith via gas space 6, conduit 8, and condensate return conduit 10. Analyzer means 9 monitors the concentrations of unreacted gaseous reactants flowing from the gas space to condenser 2 through line 8. Within the reactor vessel 1 , there is provided an agitator 3 having one or more blades 3a and driven via a rotating shaft 3b affixed to a motor 4. During the continuous reaction, there will exist a gas space 6 over a liquid reactio phase 5 within the reactor vessel 1. The gas/liquid interface 7 is depicted by a wavy line. Baffles 13 are provided to increase turbulent mixing and eliminate vortexing of the liquid phase by creating axial mixing patterns. The metallocene catalyst system i fed into the reactor through line 12. Ethylene in liquid, vapor, or liquid/vapor form is fed through line 14 and liquified dilute α-olefin is fed through line 15. The olefin feed are mixed at a juncture 29 prior to injection into the reactor vessel through line 16, th purpose being to dissolve the pure ethylene feed into the diluent prior to contact with the catalyst, so as to prevent the ethylene from unduly reacting with itself rather than the α-olefin. For the same reasons, a pre-cooler (not shown) may be employed at this point to aid in dissolving gaseous ethylene in the diluent. Note that pre- dissolving may be further enhanced by running the reflux line 10 to the juncture 29 vi an extension 11 rather than injecting the reflux directly into the reactor vessel. As stated above, a unique aspect of the process of this invention is that the reflux is usually more dilute and higher in olefin monomer concentration relative to ethylene concentration, than in the feed. For start-up, a heat exchanger 17 may be employed to bring the reaction mixture up to operating temperature by injecting a heated, dilute, α-olefin feed through line 16a. When the reactor fills and comes to the desired operating temperature the catalyst and admixed feeds are injected. Also provided is an emergency quench vessel 27, filled via a quench feed 26 and pressurized by an inert gas line 28, which may rapidly inject its contents into the reactor in the event of a runaway reaction. The products dissolved in diluent, as well as unreacted reactants and initiator, exit through line 18 past isolation valve 19 and are quenched by a quench feed 20 driven by a pump 21. Since the pressure in line 30 is generally lower than in the reactor, a cooling heat-exchanger 22 is employed to recondense any diluent that may have flashed. Gases in the reactor output would reduce the efficiency of the centrifugal pump 23 that drives the product stream line 24. Also, lowering the temperature of the product stream to produce a single liquid phase is desirable when mixing with a quench solution. For especially viscous products, such

as those with high ethylene content and/or high molecular weights, it may be desirable to add solvent via a solvent feed 25 to the product stream.

Returning to Figure 1 , also entering the reactor system 55 would be the metailocene catalyst complex which, if not shipped to the process plant premixed, would be mixed on site in a catalyst mixing system 56. After quenching, the quenched polymer solution is passed through a deashing system 59 where metallic components from the spent catalyst and possibly the quench solution are removed. Note that if a supported catalyst system is used, the quenching and deashing steps would be replaced with a catalyst removal, reactivation, and recycle system. The mixture/solution of diluents, polymer product, and unreacted reactants is carried to a diluent removal process 65 where preferably three streams result: the first, nearly pure polymer product 60; the second, unreacted olefin monomer dissolved in diluent 62, preferably more dilute than the refinery feed entering the concentration system 5 for the situations where the dilute refinery feed entering the concentration system is already more concentrated in olefin monomer than desired; and a third 64 composed of "superlights" such as trace unreacted ethylene, methane, nitrogen, and the like. The second stream 62 may be used to dilute excessively viscous polymer solution resulting from the reactor system 55 as described above. Where this is not necessary, this stream may be diverted as an unreacted olefin monomer recycle 52 o otherwise disposed of, 63. In cases where olefin monomer conversion is so high that recycling is unnecessary, the unreacted olefin monomers may be diverted for other uses, such as to a refinery for processing. Since virtually all of the ethylene is reacted, the third gaseous stream 64 containing only trace amounts of this material and other "lights" may be used as fuel gas or sent to a flare, absent some other use. The nearly pure polymer product at this point will nevertheless contain traces of diluent and unreacted olefin dissolved within it as well as small amounts of extremely low molecular weight "light" polymers. Also, trace quantities of water will be present where the quench was in aqueous solution. Hence, the product stream is sent to a stripping process 66 via Line 60 which eliminates the last of the water, reactants, and solvents. Preferably, the "light" polymers are sent via Line 67 to the refinery pipestills to be reintroduced into the cracker and the now nearly pure polyme product emerges from line 68 to be piped to a holding vessel.

The Preferred Embodiment

Referring now to Figure 3, there is seen a preferred embodiment of the process of Figure 1 , here shown for the specific case of a Raffinate-2 feed. The overall industrial process may be broken down into eight subsystems — a Feed

Hydrogenation System, a Feed Concentration System, a Feed Treater, a Reactor System, a Catalyst Preparation System, a Deashing System, a Debutanizer, and a Stripper System.

The dilute α-olefin stream, referred to herein as Raffinate-2, enters the Feed Hydrogenation System 101 through input Line 102 and is immediately passed through one or more coalescing units which cause emulsified water to coalesce into large water droplets. The stream is then heated to about 55°C and pressurized with hydrogen at about 17 atmospheres from a hydrogen feed in the presence of a catalyst inside one or more hydrogenation vessels. The result is that the 1 ,3 butadienes in the raw feed are reduced to n-butane and butene-2. Of course, some of the valuable butene-1 is also lost, but since the dienes only make up less than 1 percent of the raw feed and are so much more reactive than butene-1 , only a fraction of 1 percent of the butene-1 is lost. Should the hydrogenation prove excessive, a carbon monoxide input may be provided to lower the activity of the catalyst in the vessels. Analyzers measure the input and output concentrations of butadiene and regulate the H2/CO mixture. The treated raw feed is passed through a particulate filter and exits the Feed Hydrogenation System via Line 103.

The hydrogenated feed from the Hydrogenation System is fed into the Feed Concentration System 104 through Line 103 and introduced into a first fractional distillation tower wherein the fractionation process removes the "light" materials (i.e., the materials more volatile than the C4's). During this process, a water trap or "boot" on a distillate drum removes any water from the hydrogenated feed. Gaseous lights such as methane, ethylene, ethane, propylene, propane, and insignificant amounts o hydrogen, carbon monoxide, isobutane, isobutylene and butene-1 are vented via Lin 105. Some water vapor will also exit this route. Liquid isobutane having methane, C2's, and C3's dissolved therein along with small amounts of the C4's and methanol, di-methyl ether, and trace water is recovered via Line 106.

The effluent from the bottom of the first distillation tower will have virtually no detectable amounts of organic materials having less than 4 carbon atoms. This stream is introduced into a second distillation tower which removes the "heavy" components, e.g. primarily cis and trans butene-2, small amounts of the other C4's, and virtually all of the pentanes, pentenes, MTBE, and tert-butyl alcohol which are removed via Line 108.

The parameters of the concentration system are adjusted to the composition o the hydrogenated feed; hence, analyzers are preferred on the input and reflux lines o the fractional distillers to monitor the C4 compositions. The output of the Concentration System is preferably nearly 100 percent C4's, 30 percent to 50 percen comprising butene-1 , and is fed via Line 109 through one or more treatment vessels

in the Feed Treater 110 and exits via Line 111 to the Reactor System 112. The purpose of the feed treater is to guard against the possibility of unexpected or unknown contaminants in the material stream that could conceivably make it through the Feed Concentration System. The Reactor System 112 is as described with regard to Figure 2. The metallocene catalyst system is supplied to the Reactor System 112 from Catalyst System 113. In the catalyst system, metallocene catalyst component is fed into a mixing vessel. Cocatalyst (activator) stored in one or more cocatalyst vessels is fed into the same mixing vessel by the force of a pressurized nitrogen line. The N2 gas i first passed through a drying vessel since water vapor will inactivate the preferred catalyst mixture of metallocene and MAO as well as many other catalysts. The nitrogen also pressurizes the mixture and provides an inert gas layer above the fluid level therein. The catalyst and cocatalyst mixture are mixed in the mixing vessel and the resulting catalyst mixture continuously fed into the Reactor System 112 by pump via Line 114.

After copolymer leaves Reaction System 112, it enters the Deashing System 115 via Line 116. Line 116 receives a quench feed into Line 116, the quench comprising (for metallocene/MAO catalyst mixtures) aqueous base, such as NaOH. Base destroys the catalyst and cocatalyst and renders the metals (ash) water-soluble Upon entering the Deashing System 115, the quenched polymer mix flows through a first baffled orifice mixer. This creates turbulence in the fluid flow and causes the as to dissolve in the water. This mix then enters a first settling vessel, wherein the aqueous component settles to the bottom and the polymer and solvents float thereon The aqueous portion flows down to a disengaging drum, but about 95 percent of this flow is diverted for recirculation to the settling vessel. The hydrophobic phase flows out the top of the first settling vessel, is mixed with pure water, e.g., steam condensate, and passes through a second set of orifice mixers on its way to a secon settling vessel. Again, about 95 percent of the aqueous phase emanating from the bottom of this vessel is diverted for recirculation to the 2nd settling vessel, while the remainder joins the flow for recirculation to the first settling vessel. Residual ash- containing aqueous effluent from the disengaging drum is treated in a wastewater treatment facility.

The deashed polymer solvent mixture then exits the second vessel and enters the Debutanizer 117 via Line 118. The polymer/solvent stream entering the Debutanizer is heated to about 150°C by a heat exchanger and then injected into a debutanization tower. Here, unreacted butenes and C4 solvents vaporize upward (a well as trace amounts of very light low MW polymers) while the polymer product flow out the bottom. The polymer still has substantial amounts of C4's mixed with it, abou

3 to 5 percent by weight. Therefore, the bottom product flow is passed through one or more heat exchangers to increase the temperature to about 230°C and flashed int a flash drum. The hot C4's and light polymer are returned to the debutanizer and the nearly purified polymer product flows to the Stripper System 120 via line 119. The C4's vaporizing out the top of the debutanizer are fed to a condenser. Uncondensed gases are vented. The condensate flows into a reflux drum having a water trap to catch residual H2O from the deashing process. The condensate can be fed back int ' the debutanizer, or diverted to the Feed Concentration System for recycle and/or to the reactor outlet via line 125 for further dilution of viscous product such as high MW polymer with high ethylene content.

The nearly pure product enters a stripper tower 120 via Line 119. The last of the C4's and catalyst solvents are vaporized and recondensed in a condenser system. Gases are vented, while liquid hydrocarbons are returned to the refinery. The polymer is drawn out the bottom of the stripper by a pump, passed through a fin particulate filter, and tranferred to a storage vessel or to an additional processing system.

Referring to Figure 4, it can be seen how the relative masses of reactants and diluents would compare for the production of EB-1 polymer having an ethylene content of 20 percent by weight. The first column, labeled R-2, shows the composition of a hypothetical Raffinate-2 comprising about 17 percent by weight of butene-1 , 79 percent unreactive isobutane, n-butane, and 2-butenes, and 4 percent of other components including dienes, C3's and C5's. After passing through the hydrogenator, the stream would appear as in column H2 - essentially unchanged since the dienes in the raw Raffinate-2 comprise a small and usually fractional percentage of the whole. The feed entering the concentration system when combine with recycle from the debutanizer would appear as in column CF. For a butene-1 conversion of 50 percent, recycling is economical. The recycle in this example show the butene-1 content to increase from 17 wt % to 20 wt % in the CF stage. The increase only seems small because the graph depicts relative, rather than absolute, amounts. In fact, at a conversion of 50%, the amount of butene-1 at the CF stage is doubled by the recycle. The effluent from the feed concentrator is shown in column CO wherein the butene-1 fraction is within the desired range at 40 wt percent. After mixing with ethylene, the combined feed before entering the reactor is shown in column FD. Here, the feed is about 5 percent ethylene by weight and 38 percent butene-1. The composition of the vapor space, and hence the reflux, is shown in column FX. As can be seen, very little ethylene appears in the reflux. Column RX depicts effluent from the reactor assuming a 50 percent butene-1 conversion. About 45 wt percent of the mass is polymer product. Since the product is low in ethylene

content, it is anticipated that the product will not be further diluted. Hence, after a quench of aqueous base, as shown in column Q, the relative masses are virtually unchanged with water comprising about one tenth of one percent of the total mass. At such low water concentration, the effluent from the deashing process is virtually identical to that after the quench in terms of relative weights as shown in column DA. However, it can be seen in column DB that, after the product stream flows through th debutanizer, the product comprises about 98 percent of the flow with the bulk of the remainder consisting of light polymers of less than about 500 molecular weight. The effect of the stripping tower is seen in column ST, wherein the final product is pure. Figure 6 shows the relative masses of the production flow for the case where the product is 45 percent ethylene by weight and the butene-1 conversion is expecte to be 80 percent. This process differs from that in Figure 5 in that recycling is not employed so column H2 and CF are identical. Also, because of the high ethylene content of the product and attendant higher viscosity, solvents from the debutanizer will be recycled back to the reactor effluent to dilute the flow, as shown in column Q.

Examples

In a 200 gallon reflux reactor several runs were conducted, involving copolymerization of butene-1 and ethylene in accordance with the procedures described in connection with Figure 2. Dilute α-olefin feeds were prepared by mixing butene-1 with isobutane. The weight % of butene-1 in this mixture for each run is reported at Column 20 of Table 1. Ethylene gas was then pumped into this mixture. The weight % of ethylene based on the weight of butene-1 and ethylene in the feed i reported at Column 19. The dissolved reactants were passed through a precooler and injected into an evaporatively cooled reaction vessel equipped with external reflux. Isobutane was also injected as an agitator flush through the lower bearing of the agitator shaft. The 200 gallon vessel was equipped with a three-turbine agitator having two upper 12 inch diameter axial turbines having blades that forced the reaction solution downward and a bottom-most 12 inch diameter radial turbine havin blades that forced the solution outward and upward.

The catalyst system employed a metallocene component of dimethyl silanyl bi (tetrahydroindenyl) zirconium dichloride and an activator of methyl aluminoxane characterized as having a molecular weight of 1 ,000 and an aluminum content of 5.9 mole %. For all the examples, the methylaluminoxane (MAO) is provided as a 10 wt % solution in toluene to which is added the metallocene component to give a final mixture containing 89.87 wt % toluene, 0.15 wt % metallocene component, and 9.98 wt % MAO. When fed to the reaction mixture it provides a mole ratio of

Aluminum:Zirconium of about 500:1. The catalyst mixture was further diluted with isobutane and entered the reactor vessel through an input separate from that of the reactants. The results of each run and the physical process and product parameters are reported in Table 1 , the left most column indicating the Run number. Since the isobutane and butene-1 were pressurized into the reactor with N2, noncondensable nitrogen gas appeared in the recycle. This was recycled into the reactor since reactants were dissolved in the N2 gas. In actual industrial use, the N2 would either be disengaged from the feed or pumps would be used rather than gas compression.

The data reported in Table 1 for each Column Number is as follows:

Column No. Explanation

1 A descriptive term for each run. "Low C2= EB" for runs conducted with butene-1 exceeding ethylene in the feed by a factor of about 8 by weight

(4 by mole), "Mid C2= EB" for a ratio of 3 by weight (1.5 by mole), and "Hi C2= EB" for a ratio of less than 2 by weight (less than 1 by mole, so ethylene molecules outnumber butene-1 molecules).

2 Number average molecular weight of the polymer product determined by gel permeation chromotography (GPC).

3 Molecular weight distribution of the polymer product by GPC.

4 Weight percent ethylene content of the product by Fourier transform infrared spectroscopy (FTIR) (examples listed in order of increasing percent ethylene content by weight). 5 - 8 Percent by number of vinylidene, trisubstituted, disubstituted, and vinyl terminal unsaturated polymer chains respectively as determined by HNMR spectroscopy. 9 Feed rate of the metallocene catalyst system (cc/min.) prior to further mixing with the amount of isobutane in column 14. 10 - 12 Feed rates (Ib./hr.) of ethylene, butene-1 , and isobutane, respectively.

13 Amount of isobutane diverted from the amount in column 12 (Ib./hr. ) injected into the reactor through the agitator shaft bearing as a flush to prevent polymer product from collecting in and fouling the bearing.

14 Amount of isobutane diverted from the amount in column 12 (Ib./hr.) added to the catalyst system feed (shown in column 9) to facilitate mixing of the catalyst system into the reaction mixture.

15 Assumed rate of ethylene conversion.

16 Rate at which polymer product is formed calculated from an assumed ethylene conversion of 100 percent, the ethylene feed rate, and the percent weight of ethylene shown in column 4.

17 Conversion of butene-1 (%) calculated from the polymer rate of column 16, and the butene-1 feed rate of column 11.

18 Catalyst yield (thousands of pounds of polymer per pound of catalyst) calculated from the polymer rate of column 16 and the catalyst system feed rate of column 9.

19 Ethylene feed (wt.%) based upon the total flow of ethylene and butene- 1. Isobutane is not included in this calculation.

20 Butene-1 component (wt.%) in the mixed butene-1 /isobutane feed stream. Ethylene is not included in this calculation.

21 - 22 Temperature of the reaction mixture, °F and °C, respectively. 23 - 27 Calculated reactor effluents (wt.%) for ethylene, isobutane, butene-1 , polymer product, and catalyst system, respectively.

28 Calculated concentration (ppm) of metallocene catalyst in the reactor effluent.

29 Fraction of the reaction volume that is gas bubbles, known as the "gas hold-up". 30 Residence time in the reactor, taking into account bubble formation and effluent volume.

31 Catalyst system concentration (moles/liter x 10 ^-5 ) in the effluent based on the volumetric flow of the effluent.

32 Observed reaction mixture density. 33 Total reaction volume, (gal.) including bubbles.

34 Agitator motor power (amps).

35 Rotation rate (rpm) of the agitator.

36 Agitator energy (horsepower per thousand gallons of reaction mixture).

37 Pressure (psi) in the vapor space. 38 Temperature (°F) at the bottom of the reactor.

39 Temperature (°F) at the midpoint of the reactor (near, but still below, the top of the reaction mixture).

40 Temperature (°F) of the vapor space.

41 - 44 Concentration (mol%) of nitrogen, ethylene, butene-1 and isobutane ga in the vapor space, respectively.

45 Liquid reflux (Ib./hr.).

46 Noncondensed gas recycle (Ib./hr.).

- 50 Noncondensed gas recycle (wt.%) of nitrogen, ethylene, butene-1 , and isobutane, respectively. Amount of noncondensable gas (Ib/hr.) vented from the reflux system. Ratio of ethylene to isobutane in the vapor space as measured by gas chromatography. Ratio of butene-1 to isobutane in the vapor space as measured by gas chromatography. - 57 Weight fraction in the vapor space of nitrogen, ethylene, butene-1 , and isobutane, respectively. Volume of gas (ft3/hr.) passing through the vapor space, based on a compressibility factor of Z = 0.83. - 60 Estimated kinematic viscosity (cSt) and the absolute viscosity in (cp) of the reaction liquid, respectively. - 63 Mole fractions of isobutane, butene-1 , and polymer respectively in the reactor effluent. Weight ratio of butene-1 to ethylene in the feed. Weight ratio of butene-1 to ethylene in the vapor space. Ratio of the weight % of ethylene in the vapor space to that in the feed.

TABLE 1

Vapor

RX RX Space Liquid Gas

Run Dens Total Agit Agit HP/ Press Bot Mid Vap N2 C2= C4=l iC4 Reflux Recycl

No. lb/ft3 Vol gals amps RPM KGAL psig degF degF de§F mol% mol% mol% mol% Ib/hr lb/hr

32 33 34 35 36 37 38 39 40 41 1 42 43 44 45 46

1 31.5 139.3 2.44 300 9.6 153.5 143.8 143.9 146.6 13.7 5.6 22.5 58.1 36.3 80.6

Sfi 2 32.3 136.5 2.26 300 9.1 205.2 175.3 176.6 178.0 6.9 3.0 22.0 66.6 369.9 94.5 c CrD 3 31.9 137.9 2.44 300 9.5 178.5 158.4 159.3 160.9 11.1 4.5 22.4 60.8 104.2 85.6 sα 4 31.7 138.1 2.36 300 9.4 178.4 158.3 159.2 160.7 10.7 5.7 26.6 56.5 87.2 85.6 c 5 25.0 146.1 3.83 425 25.1 224.6 176.6 177.8 178.1 9.0 7.0 31.4 52.1 103.0 «

— m . 6 32.0 137.5 2.37 300 9.1 183.0 161.6 162.9 163.6 7.5 8.2 30.1 53.9 124.2 89.0 oo l—i e ^* χn 7 27.4 134.9 2.17 300 9.5 221.5 176.2 175.7 180.0 7.5 3.6 19.1 68.2 105.6 1 m 8 25.6 138.4 3.66 425 29.2 275.3 203.9 203.6 206.6 4.9 2.2 11.1 78.9 101.7 q 9 34.5 111.6 3.80 425 35.0 320.3 207.8 207.4 209.6 8.7 4.6 17.5 68.1 119.3

33 10 37.4 108.6 2.52 300 12.3 318.3 206.7 201.5 207.5 10.4 4.6 15.1 68.5 121.4 m 11 33.4 113.9 4.37 425 33.9 321.8 204.7 201.4 206.5 9.3 7.1 20.8 61.7 115.6

12 37.7 107.8 2.43 300 12.2 322.1 203.6 199.5 204.4 11.5 7.0 18.9 62.6 121.4

^ ^■ ϋ»

13 36.6 111.4 4.41 425 34.5 315.8 200.0 195.9 201.7 131.1 4.9 13.3 67.1 119.6

14 29.4 130.6 4.09 425 27.4 322.2 205.1 203.5 206.7 2.8 16.2 27.4 52.7 120.8