BE897868A - AGENTS PREVENTING REDUCTION OF MOLECULAR WEIGHT OF POLY-ISOBUTYLENE OR BUTYL RUBBER - Google Patents

Abstract

Protection against regression of molecular weight during a polymerization reaction producing polysiobutylene rubber or butyl rubber, obtained by the incorporation of minor amounts of certain alkyl metals, more particularly compounds of the lower aluminum type or aluminum lower alkyl hydrides, which serve as buffering, neutralizing or complexing agents against polymerization poisons and activators of external catalysts.

Description

       

    <Desc / Clms Page number 1>
 registered by the so-called company: EXXON RESEARCH AND ENGINEERING
COMPANY relating to: Agents preventing the reduction of the molecular weight of polyisobutylene or butyl rubber Qualification proposed: PATENT OF INVENTION Priority of a patent application filed in the United States of America on September 30, 1982 under the number NO 429. 139 on behalf of Kenneth W. POWERS and Ralph H. SCHATZ

  <Desc / Clms Page number 2>

 
The present invention relates to the polymerization of elastomeric homopolymer polyisobutylene rubber and elastomeric copolymer butyl rubber, more particularly the polymerization reaction necessary to produce the isobutylene-isoprene form of butyl rubber.

   More particularly, the subject of the invention is a process for protecting the polymerization process used to produce polyisobutylene and butyl rubbers against a regression in molecular weight and other undesirable effects caused by the presence of poisons and / or polymerization activators, by the addition of a complexing or neutralizing agent with a buffering effect to the polymerization zone.



   An important problem encountered during the implementation of the butyl rubber production process lies in the sensitivity of the reaction system to the presence of catalyst activators and polymerization poisons, which, even in amounts corresponding to traces, can affect the polymerization process in a very detrimental way, by reaction or complexion with the catalytic species, or by the propagation of carbonium ions, so as to deeply influence the stages of priming, propagation, transfer and completion of polymerization.

   The consequences of these harmful effects are manifested by a significant reduction in the molecular weight of the butyl rubber generated, activation or deactivation of the catalytic system and, frequently, rapid heating and clogging of the reactors.

  <Desc / Clms Page number 3>

 polymerization. Activators and polymerization poisons include oxygenated compounds such as alcohols, ketones and ethers; acidic compounds such as hydrochloric acid, chlorine, organic compounds containing chlorine and organic acids.



  Many of these polymerization activators or poisons are generated in the various areas of the butyl rubber production system where water is present and originate from hydrolytic reactions involving methyl chloride and / or methylene chloride. Certain activators or poisons are formed by the chemical reactions that take place in the alumina drying stage, of the process that is used to remove water from the diluent recycling section of a manufacturing facility. butyl rubber. Still other activators or poisons are introduced as impurities present in the monomer feed streams.

   As specific examples of activators and / or of polymerization poisons, mention may be made of methanol, isopropanol, acetone, dimethyl ether, diethyl ether, dimethoxy methane, t-butyl alcohol, tbutyl chloride, bis-chloromethyl ether, chlorine gas, hydrogen chloride, propionic acid and the like.



   The recognition of this problem by the prior art can be seen on reading the patent of the United States of America 3,005. 008 granted October 24, 1961 to Kelley et al., Whereby the polymerization poisons are removed according to one of the features of a process for treating the recycled diluent from the butyl rubber polymerization process. As far as has been brought to the knowledge of the applicant, no effective technique has been described to date for adding a complexing agent or buffer to the polymerization reaction mixture, in order to eliminate the regression

  <Desc / Clms Page number 4>

 molecular weight and to remedy other problems associated by the presence of activators or poisons of polymerization mentioned above.



   The use of aluminum alkyls to purify the hydrocarbon diluents used for the polymerization of alpha-olefins is described in British patent 920,513 (1963) granted to the company Eastman Kodak Company.



  Similarly, in British Patent 1,298. 909 (1972) to Monsanto Company, the use of trialkyl aluminum compounds has been described as scavengers for water, oxygen and traces of alcohol and other impurities, in connection with a Ziegler polymerization process of olefins, such as ethylene. Although some of the trialkylaluminum compounds used in this British Monsanto patent are used for the purposes of the present invention, the manner in which they are used according to the reference literature and in the polymerization process of this reference literature is different of the present invention.

   Similarly, in U.S. Patent 3,257. 332 granted June 2, 1966 to Ziegler et al., Excess amounts of trialkylaluminum are used to remove impurities in the feed ethylene used for carrying out an ethylene polymerization process. Similarly, in US Patent 3,442,878 granted May 6, 1969 to Gippin, dihydrocarbon aluminum chlorides are used to destroy impurities present in the isoprene feed and in the hydrocarbon solvent used for the polymerization of isoprene.



   According to the present invention, a pro-
 EMI4.1
 . yielded improved preparation of C 4 isoolefin homopolymers to polyisobutylene rubber, 4 c7p by polymerization catalyzed by Lewis acid or

  <Desc / Clms Page number 5>

 butyl rubber, by the Lewis acid catalyzed polymerization of a C4 - C7 isoolefin with a diene
 EMI5.1
 conjugate in an inert diluent, characterized conjugate in C e in that one introduces into the zone of the polymerization reaction up to approximately 1.0 to approximately 1.25 mole of a complexing agent or buffer per mole of catalyst , in an amount effective to protect the rubber produced against a regression in molecular weight, due to the presence of catalyst activators and of polymerization poisons,

   the complexing agent or buffer being constituted by certain alkyl metals, more particularly a lower trialkyl aluminum, the alkyl radical of this compound being a methyl or ethyl group, or a lower dialkyl aluminum hydride, the alkyl radical of this hydride being a C1 - C4 alkyl group. It has also been found that other alkyl metals, such as di-n-hexyl-magnesium, were also advantageous for carrying out the process according to the present invention, but that, on the whole, they were less effective and more expensive than preferred alkylaluminiums.



   It has been found that the complexing agent or buffer which it is particularly preferable to use for the purposes of the present invention is triethyl aluminum. Other preferred complexing agents or buffers are, as has been observed, trimethyl aluminum and diisobutyl aluminum hydride. According to a preferred embodiment of the invention, the alkyl buffering agent or complexing agent is added to the supply stream of cold monomer just before its introduction into the zone of the polymerization reaction.



   A whole series of substantial improvements come from the implementation of the method according to the invention. The polymerization reaction is effectively protected against the harmful effects caused by a sudden or unexpected variation in the level of traces of poisons

  <Desc / Clms Page number 6>

 polymerization that we talked about above. The alkyl compound serves as a buffering agent or complexing agent which maintains the free concentration of these poisons at low constant values. The problem of regressing molecular weight is substantially eliminated, allowing the desired molecular weight polymer to be produced with a relatively lower constant reaction monomer concentration, which represents a considerable economic benefit of the process.

   The buffering agent according to the present invention allows the presence of tolerable higher levels of these polymerization poisons. The need for extensive purification and recycling monomer, as well as diluent streams, is significantly reduced. Likewise, starting the polymerization system is greatly facilitated, since the need for an excessive initial amount of catalyst flow to remedy the poisons present has been remedied. Operation with the alkyl buffering agent also gives results by a reduced degree of heating of the reactor when the polymerization takes place.



   The term "butyl rubber" as used in this specification and the claims terminating it, denotes copolymers of isoolefins
 EMI6.1
 at C4 to C7 and conjugated dienes at 4 7 4% c14, which comprise about 5.5 to about 15 mole percent of conjugated diene and about 85 to 99.5 mole percent of isoolefin.

   By way of illustrative examples of the iso-olefins used for the preparation of butyl rubber, mention may be made of isobutylene, 2-methyl-1-propene, 3-methyl-1-butene, 4-methyl-1 -pentene and beta-pinene. As illustrative examples of conjugated dienes which can be used for the preparation of butyl rubber, mention may be made of isoprene, butadiene, 2,3-dimethyl butadiene, piperylene, 2,5-dimethylhexa -2, 4-diene,

  <Desc / Clms Page number 7>

 cyclopentadiene, cyclohexadiene and methylcyclopentadiene. The preparation of butyl rubber is described in US Patent 2,356. 128
 EMI7.1
 and is also described in an article by R. M.

   Thomas et al. in Industrial and Engineering Chemistry, volume 32, pages 1283 et seq., October 1940. Butyl rubber generally has an average molecular weight by viscosity of approximately 100,000 to approximately 800,000, preferably approximately 250,000 to approximately 600,000. and an iodine index Wijs of approximately 0.5 to 50, preferably 1 to 20. The expression "homopolymers of iso-olefins" as used in the present specification aims to define the homopoly-
 EMI7.2
 mothers of C4 to C iso-olefins, particularly polyisobutylene, which have a low degree of terminal unsaturation and which have certain elastomeric properties.



   The main commercial form of butyl rubber is isobutylene-isoprene rubber and both butyl rubber and polyisobutylene are produced during the implementation of a low temperature cationic polymerization process using acid type catalysts Lewis, typically aluminum chloride (AlCl); we know that BF3 is also interesting. The process widely used in industry uses the use of methyl chloride as a diluent for the reaction mixture at very low temperatures which are below about -90oC. Methyl chloride is used for a variety of reasons, including the fact that it is a solvent for the monomers and the aluminum chloride catalyst and a non-solvent for the polymer formed.

   Likewise, methyl chloride has freezing and boiling points which allow respectively polymerization at low temperature and efficient separation of the polymer and the monomers.

  <Desc / Clms Page number 8>

 not entered in reaction. Other diluents suitable for using for the implementation of the butyl rubber polymerization process are methylene chloride, vinyl chloride, ethyl chloride, propane, butane, pentane and hexane, but methyl chloride is the most preferred polymerization diluent.



   For the preparation of butyl rubber, about 70 to 99.5 parts by weight, preferably 85 to 99.5 parts by weight, of an iso-olefin such as isobutylene are reacted with about 30 to 0.5 parts of a conjugated diene, such as isoprene. Particularly preferred is the reaction product from 95 to 99.5 percent by weight of isobutylene over 0.5 to 5 percent by weight of isoprene. These monomers are cooled with about 1 to 5 volumes of inert diluent, such as methyl chloride, to a temperature which varies from about -60oC to about -130oC. The cold mixture is polymerized in a reaction zone by the addition of a Friedel-Crafts catalyst, preferably aluminum chloride.

   Generally about 0.02 to about 0.5 percent by weight of catalyst is used, based on the weight of the mixed olefins.



   Throughout the description, examples and claims, the expression "complexing agent" is used interchangeably with the expression "buffering agent". The Applicant makes no distinction whatsoever between these two expressions.



   When practicing the present invention, attention must be paid to a number of conditions and limitations in order to be able to take advantage of its advantages. These conditions and limits relate to the quantity of complexing agent or buffer used and to the manner of introducing it into the reactor, that is to say the adjustment of the injection temperature and the duration of

  <Desc / Clms Page number 9>

 stay before the entry of the buffering agent in the polymerization reaction zone.



   The amount of buffering agent used is mainly a function of the amount of catalyst used. It has been found that the amount of buffering agent should not exceed about 1.0 to about 1.25 moles, preferably not exceed about 1.0 moles per mole of polymerization catalyst. If a large molar excess of buffering agent relative to the catalyst is present, it is possible that a reaction of this buffering agent on the aluminum chloride-type catalyst or on the Lewis acid-type catalyst will occur, thereby causing significant reductions in catalyst efficiency.



  The efficiency of the catalyst is defined as the weight of polymer generated per weight of catalyst. By the device of controlled experiments, it has been found that the minimum effective amount of buffering agent is that which is in molar excess relative to the polymerization poisons present in the reaction mixture. Consequently, in relation to the implementation of the process according to the invention, the appropriate course of action consists in ensuring that the amount of complexing agent or of buffering agent does not exceed approximately 1.0 to approximately 1, 25, preferably not more than about 1.0 mole, per mole of polymerization catalyst, since the polymerization poisons, when present, are always present in proportions which are considerably less than the amount of catalyst used.

   A substantial molar excess of buffering agent relative to the polymerization poisons is not detrimental to the implementation of the method according to the invention. Generally, the amount of complexing agent or buffering agent to be used varies in the range from a molar ratio of about 0.1 to 1, to a molar ratio of about 1 to 1; the preferred amount

  <Desc / Clms Page number 10>

 fluctuates between 0.5 to 1 and 1 to 1 mole of buffering agent per mole of polymerization catalyst, such as aluminum chloride.



   The location of the injection site of the buffering agent and the injection temperature are also important factors for a successful implementation of the process according to the present invention. If the buffering agent is added to the cold incoming monomer feed stream, which contains monomers and inert diluent, this should be done immediately before the feed stream is introduced into the polymer reaction mixture. - station. When producing butyl rubber, the monomer feed stream typically contains a mixture of the monomers formed by isoolefin and conjugated diolefin.

   An appreciable length of time in the feed stream results in the appearance of undesirable side reactions with the monomer, which limits the effectiveness of the buffering agent. A residence time of more than several minutes in the cold feed stream has been found to be undesirable in many cases.



   Consequently, according to an embodiment of the process in accordance with the present invention, the buffering agent is injected into the polymerization zone in such a way that the duration of stay or of contact of the buffering agent with
 EMI10.1
 the monomers, prior to polymerization, are generally less than about 60 seconds, preferably less than 30 seconds, more advantageously still less than about 10 seconds. By using this technique, a process for introducing the buffering agent into the polymerization zone consists in introducing this buffering agent into the reactor at the start of the process, prior to the introduction of the monomer (s) and of the catalyst in the polymerization zone.

   The agent is then continual-

  <Desc / Clms Page number 11>

 LEMENT introduced as the polymerization progresses. As an alternative and, preferably, the agent is added directly to the cold monomer feed stream, immediately before its penetration into the reactor, paying particular attention to the residence time. According to another approach, the buffering or complexing agent is present in the polymerization zone before proceeding with the introduction of the monomer (s) and, then, when the monomers are introduced, the agent is incorporated into the current d supply of monomer in an appropriate amount to obtain the buffering or complexing effect.



  Other equivalent methods will obviously appear to those skilled in the art on the basis of a description of the important variables which will be described later in this memo, more particularly in the examples.



   The preferred butyl rubber embodiments to which the present invention can be applied are isobutylene-isoprene butyl rubbers which use an aluminum chloride catalyst and a methyl chloride diluent, the product having an average molecular weight per viscosity of approximately 250,000 to 600,000 and a Wijs iodine index of approximately 1 to 20.



   The present invention will now be further illustrated with the aid of the following examples which do not limit its scope in any way.



  Process and materials for Examples 1-11
A series of discontinuous polymerizations were carried out in a dry box, purged with nitrogen, in order to show the harmful effect of traces of an oxygenated base on the molecular weight of butyl rubber and to show the effectiveness of the buffering agents. according to the present invention. The entire series of tests was carried out

  <Desc / Clms Page number 12>

 serving as aliquots taken from a common feed mixture, so that the load or feed was the same for each test in the series. and the entire test series was also started using aliquots taken from a running batch of catalyst solution so that the catalyst was the same for each test in the series.

   The various tests in the series then differed only in the variables examined.



   Polymerizations were undertaken using 500-ml round bottom reaction flasks, 500 ml capacity, made of standard Pyrex glass, with necks fitted with ball joints. A standard bearing and a glass stirrer shaft were connected through the central neck and glass heat sinks were connected through two of the side necks. One of the heat sinks was for a thermometer to visually monitor the reaction temperature, while the other heat sink was for a thermocouple connected to a recorder. A jacketed and cooled addition funnel was mounted over the third side neck.

   Measured amounts of the catalytic solution were allowed to drip slowly from the addition funnel into the reactor, as desired. To carry out the test, the following general operating method was used:
1. An aliquot of the feed or feed mixture was weighed into the reaction flask which was then clamped in position, dipped to the neck in a cooling bath.



   2. The stirring shaft was connected to the agitator motor above the bath and the stirring was started. We put the thermometer and thermocouple in place in their respective wells.

  <Desc / Clms Page number 13>

 



   3. After the contents of the reactor had reached the desired temperature, the desired poisons, activators or buffering agents were introduced via a volumetric pipette, before mounting the funnel for the dropwise addition of the catalyst by above the third collar to then tighten it in position. The jacket of the addition funnel was filled with liquid from the cold bath and the desired catalyst solution was poured into the funnel.



   4. After the desired residence time, the catalyst solution was allowed to slowly trickle through the reactor until the desired volume was reached.



  The rate of addition of the catalyst was adjusted so as to retain only a small temperature difference between the bath and the reactor so as to carefully control the temperature of the polymerization.



   5. After the desired amount of catalyst has been added and the reaction time has elapsed, a measured volume of cold methanol is introduced into the reactor to serve as an abrupt coolant. The cooled reactor was then removed and removed from the bath and placed in the inlet of the dry box from which it was then removed and placed in a hood where it was allowed to warm up, so as to leave methyl chloride and the monomers which have not entered the reaction escape by boiling. Isopropyl alcohol containing 1 percent PX-441 was added as an inhibitor to the reaction flask as the volatiles escaped by boiling to prevent undue staining of the polymer and its alteration by remains or residues of catalyst.

   After it had warmed to room temperature, the polymer was recovered quantitatively.



  Kneaded in isopropyl alcohol to remove catalyst residue and then added 0.2

  <Desc / Clms Page number 14>

 percent of PX-441 by mixing before drying in a vacuum oven at 70oC. The dried polymer was weighed and subjected to analysis.



   The isobutylene used was purified by washing through anhydrous calcium chloride and anhydrous barium oxide in order to remove any moisture and was distilled just before use through a small fractionation column located in the dry box. The methyl chloride used was the Ansul product of high purity which had been obtained from the company Air Products Specialty Gas Division. The used isoprene was dried over white Drierite and distilled through a fractionation column before use. The catalyst solution used during the batch polymerization tests consisted of aluminum chloride in solution in methyl chloride.



  A saturated solution containing 0.9 g of Aid was prepared per 100 grams of solution by bringing methyl chloride to reflux on Aida and this solution was diluted to the desired concentration with methyl chloride. washed and filtered.



   A feed or feed lot was prepared for a series of tests by weighing the desired amount of cold isoprene-isobutylene and methyl chloride in a cold feed or load flask and then shaking and keeping the stoppered balloon in a cold bath.



  Aliquots of this feed or feed mixture were weighed in the individual reactors for carrying out the series of tests. In each test of Example 1 described hereinafter in this specification and the following examples, the reactor was charged with 230 g of a feed or charge mixture consisting of 0.75 g of isoprene, 24 , 25 g of isobutylene and 205 g of methyl chloride.

   Then we stirred the reactor and we

  <Desc / Clms Page number 15>

 cooled as described above to -96oC, the indicated amount of buffering agent or stabilizer and / or polymerization poison was added and the catalyst solution, consisting of 0.18 g of AlCl per 100 grams of methyl chloride, flow slowly drop by drop in order to initiate the polymerization. Over about 10 minutes, sufficient catalyst was added to convert about 25 percent of the monomer to polymer, while maintaining the reaction temperature at -960C. The reaction mixture was treated with methanol and the polymer was collected and dried as previously described.



   In the following examples, "INOPO" is a measure of the degree of unsaturation, also known as the iodine-mercuric acetate process, as described in the book Ind. Eng. Chem 40.1277 (1948). The abbreviation TEAL is used to denote triethylaluminium. The abbreviation Mv refers to the average molecular weight by viscosity of the polymers, based on a
 EMI15.1
 viscosity measurement of solution diluted in diisobutylene at 20oC.



  Example 1
 EMI15.2
 
 <tb>
 <tb> No. <SEP> test <SEP> MeOH <SEP> added <SEP> TEAL <SEP> added <SEP> Analysis <SEP> from <SEP> butyl
 <tb> Mv <SEP> INOPO
 <tb> 1 <SEP> nothing <SEP> nothing <SEP> 971 <SEP> 500 <SEP> 7, <SEP> 7 <SEP>
 <tb> 2 <SEP> 12.6 <SEP> ppm <SEP> nothing <SEP> 486 <SEP> 500 <SEP> 6, <SEP> 7 <SEP>
 <tb> 3 <SEP> 12, <SEP> 6 <SEP> ppm <SEP> 58.1 <SEP> ppm <SEP> 941 <SEP> 000 <SEP> 7, <SEP> 3 <SEP>
 <tb> 4 <SEP> nothing <SEP> 58.1 <SEP> ppm <SEP> 962 <SEP> 100 <SEP> 7, <SEP> 4 <SEP>
 <tb> 5 <SEP> nothing <SEP> nothing <SEP> 945 <SEP> 000 <SEP> 7, <SEP> 3 <SEP>
 <tb>
 
The results obtained clearly show that the addition of 12.6 ppm of methanol greatly decreases the molecular weight of the polymer from more than 900,000 to less than 500.

   000, while the addition of 58.1 ppm of triethylaluminum (TEAL) as a buffering agent or complexing agent has no appreciable effect on the molecular weight

  <Desc / Clms Page number 16>

 by itself, but that its presence prevents methanol from causing a regression in molecular weight (that is to say that the molecular weight of test 3 of the series carried out with the presence of both methanol and TEAL, is the same as that of controls 1 or 5 where methanol was absent).



  Example 2
 EMI16.1
 
 <tb>
 <tb> No. <SEP> test <SEP> Acetone <SEP> added <SEP> TEAL <SEP> added <SEP> Analysis <SEP> from <SEP> butyl
 <tb> Mv <SEP> INOPO
 <tb> 1 <SEP> nothing <SEP> nothing <SEP> 705 <SEP> 000 <SEP> 7, <SEP> 5 <SEP>
 <tb> 2 <SEP> 24.5 <SEP> ppm <SEP> nothing <SEP> 334 <SEP> 000 <SEP> 8, <SEP> 1 <SEP>
 <tb> 3 <SEP> nothing <SEP> 81.3 <SEP> ppm <SEP> 835 <SEP> 000 <SEP> 7, <SEP> 8 <SEP>
 <tb> 4 <SEP> 24.5 <SEP> ppm <SEP> 81.3 <SEP> ppm <SEP> 765 <SEP> 000 <SEP> 7, <SEP> 1 <SEP>
 <tb> 5 <SEP> nothing <SEP> nothing <SEP> 725 <SEP> 000 <SEP> 7, <SEP> 6 <SEP>
 <tb>
 
 EMI16.2
 1 The results show that the addition of 24.5 ppm of acetone greatly decreases the molecular weight from more than 700,000 to 334,000;

   the addition of TEAL alone at 81.3 ppm causes a slight increase in molecular weight and the addition of acetone to the buffered system containing TEAL has little effect on molecular weight. TEAL is able to counteract or prevent the regression of molecular weight caused by acetone just as it was effective with the methanol of Example 1. TEAL is an effective buffer with acetone, just as it is with methanol.



  Example 3
 EMI16.3
 
 <tb>
 <tb> No. <SEP> test <SEP> Ether <SEP> dimethyl <SEP> TEAL <SEP> added <SEP> Analysis <SEP> from <SEP> buadded <SEP> tyle
 <tb> INOPO
 <tb> 1 <SEP> nothing <SEP> nothing <SEP> 705 <SEP> 000 <SEP> 7, <SEP> 5 <SEP>
 <tb> 2 <SEP> 10.7 <SEP> ppm <SEP> nothing <SEP> 457 <SEP> 900 <SEP> 8, <SEP> 3 <SEP>
 <tb> 3 <SEP> nothing <SEP> 81.3 <SEP> ppm <SEP> 835 <SEP> 000 <SEP> 7, <SEP> 8 <SEP>
 <tb> 4 <SEP> 10.7 <SEP> ppm <SEP> 40.7 <SEP> ppm <SEP> 705 <SEP> 000 <SEP> 7, <SEP> 8 <SEP>
 <tb> 5 <SEP> nothing <SEP> nothing <SEP> 725 <SEP> 000 <SEP> 7, <SEP> 6 <SEP>
 <tb>
 
The results obtained show that dimethyl ether is an important poison with respect to molecular weight

  <Desc / Clms Page number 17>

 
 EMI17.1
 e in the absence of TEAL, but its effect is canceled in the buffered system containing TEAL.



  Example 4
A series of discontinuous polymerization tests were carried out, according to the procedures described above, using tertiary butyl alcohol as poison, obtaining the following results:
 EMI17.2
 
 <tb>
 <tb> No. <SEP> test <SEP> t-BuOH <SEP> added <SEP> TEAL <SEP> added <SEP> Analysis <SEP> from <SEP> butyl
 <tb> Mv <SEP> INOPO
 <tb> 1 <SEP> nothing <SEP> nothing <SEP> 830 <SEP> 000 <SEP> 7, <SEP> 9 <SEP>
 <tb> 2 <SEP> 25.1 <SEP> ppm <SEP> nothing <SEP> 650 <SEP> 000 <SEP> 8.3
 <tb> 3 <SEP> nothing <SEP> 65.0 <SEP> ppm <SEP> 850 <SEP> 000 <SEP> 7, <SEP> 8
 <tb> 4 <SEP> 25.1 <SEP> ppm <SEP> 65.0 <SEP> ppm <SEP> 817 <SEP> 000 <SEP> 7, <SEP> 6 <SEP>
 <tb> 5 <SEP> nothing <SEP> nothing <SEP> 800 <SEP> 000 <SEP> 8,

    <SEP> 0 <SEP>
 <tb>
 The results show that t-butanol constitutes
 EMI17.3
 e is a potent molecular weight regressant in the absence of TEAL, but has little effect in a buffered system containing TEAL.



   Examples 1 to 4 show that when a buffering agent such as TEAL is used, at the appropriate concentrations relative to the catalyst and the poisons, and that it is suitably added, it is capable of preventing all classes of oxygenated bases to adversely affect the polymerization of butyl. In these examples, TEAL was added in a molar excess relative to the poison, but in an amount of less than 1 mole per mole of AIC13 type catalyst present. TEAL was added to the cold feed mixture at -960C and allowed to form a complex with the poison before adding the catalyst to initiate polymerization. Only a short residence time (several minutes) was tolerated between the formation of the complex and the initiation of the polymerization.



  Example 5
This example is given to demonstrate that the buffering agent is not effective against regression agents

  <Desc / Clms Page number 18>

 molecular weight, unless present in a molar excess over the amount of the polymerization poison present in the system.
 EMI18.1
 
 <tb>
 <tb>



  No. <SEP> test <SEP> MeOH <SEP> added <SEP> TEAL <SEP> added <SEP> Moles <SEP> TEAL / Analysis <SEP> from
 <tb> moles <SEP> MeOH <SEP> butyl
 <tb> Mv <SEP> INOPO
 <tb> 1 <SEP> nothing <SEP> nothing-848 <SEP> 500 <SEP> 7, <SEP> 8 <SEP>
 <tb> 2 <SEP> 12, <SEP> 6 <SEP> ppm <SEP> nothing-469 <SEP> 400 <SEP> 7, <SEP> 6 <SEP>
 <tb> 3 <SEP> 12, <SEP> 6 <SEP> ppm <SEP> 8, <SEP> 2 <SEP> ppm <SEP> 0, <SEP> 18 <SEP> 381 <SEP> 700 <SEP> 7, <SEP> 3 <SEP>
 <tb> 4 <SEP> 12, <SEP> 6 <SEP> ppm <SEP> 16, <SEP> 5 <SEP> ppm <SEP> 0, <SEP> 37 <SEP> 442 <SEP> 000 <SEP> 7, <SEP> 6 <SEP>
 <tb> 5 <SEP> 12 <SEP> 6 <SEP> ppm <SEP> 41, <SEP> 0 <SEP> ppm <SEP> 0, <SEP> 87 <SEP> 863 <SEP> 600 <SEP> 7, <SEP> 3 <SEP>
 <tb> 6 <SEP> 12 <SEP> 6 <SEP> ppm <SEP> 57, <SEP> 4 <SEP> ppm <SEP> 1, <SEP> 28 <SEP> 883 <SEP> 800 <SEP> 7, <SEP> 8 <SEP>
 <tb> 7 <SEP> 12.6 <SEP> ppm <SEP> 121.7 <SEP> ppm <SEP> 2,

    <SEP> 75 <SEP> 924 <SEP> 000 <SEP> 7, <SEP> 7 <SEP>
 <tb> 8 <SEP> nothing <SEP> 123 <SEP> 5 <SEP> ppm-843 <SEP> 500 <SEP> 7, <SEP> 8 <SEP>
 <tb> 9 <SEP> nothing <SEP> nothing-810 <SEP> 000 <SEP> 7, <SEP> 4 <SEP>
 <tb>
 The above data shows that TEAL is ineffective
 EMI18.2
 effective in preventing regression of the molecular weight caused by methanol, unless about as many moles of TEAL are present as moles of methanol. The excess of TEAL is not detrimental provided (as shown in Example 7 later) that its molar ratio vis-à-vis Aid is not too high. TEAL is apparently effective in a molar proportion of approximately 1/1 relative to the polymerization poison.



  Example 6
The data from these experiments show that the buffering or complexing agent, in this case TEAL, is not effective if the residence or preservation time before the initiation of the polymerization in the polymerization reactor is too long. , or if the length of stay takes place at too high a temperature.

  <Desc / Clms Page number 19>

 
 EMI19.1
 
 <tb>
 <tb>



  No. <SEP> Me <SEP> OH <SEP> TEAL <SEP> Duration <SEP> from <SEP> con-Mv <SEP> INOPO
 <tb> test <SEP> added <SEP> added <SEP> servation <SEP> to
 <tb> - <SEP> 70oC <SEP>
 <tb> 1 <SEP> nothing <SEP> nothing <SEP> nothing <SEP> 877 <SEP> 000 <SEP> 7, <SEP> 8 <SEP>
 <tb> 2 <SEP> 12.5 <SEP> ppm <SEP> nothing <SEP> nothing <SEP> 576 <SEP> 000 <SEP> 8.4
 <tb> 3 <SEP> 12.5 <SEP> ppm <SEP> nothing <SEP> 3 <SEP> 1/2 <SEP> h <SEP> 612 <SEP> 000 <SEP> 8.7
 <tb> 4 <SEP> 12.5 <SEP> ppm <SEP> 58 <SEP> ppm <SEP> nothing <SEP> 864 <SEP> 000 <SEP> 9, <SEP> 0 <SEP>
 <tb> 5 <SEP> 12, <SEP> 5 <SEP> ppm <SEP> 58 <SEP> ppm <SEP> 3 <SEP> 1/2 <SEP> h <SEP> 414 <SEP> 000 <SEP> 8,

    <SEP> 4
 <tb>
 
These data show that methanol constitutes a powerful agent for regressing molecular weight in the absence of TEAL and that the conservation of the feedstock in the mere presence of methanol has no additional effect. Data from Test No. 4 show that methanol has no adverse effects in the buffered system containing TEAL, but data from Test 5 shows that TEAL completely loses its buffering ability when the filler or feed containing methanol and TEAL is kept for 3, 5 hours at -70 ° C. before the addition of the catalyst to initiate the polymerization.
 EMI19.2
 
 <tb>
 <tb>



  No. <SEP> Me <SEP> OH <SEP> TEAL <SEP> Duration <SEP> from <SEP> con-Mv <SEP> INOPO <SEP>
 <tb> test <SEP> added <SEP> added <SEP> servation
 <tb> 6 <SEP> nothing <SEP> nothing <SEP> nothing <SEP> 803 <SEP> 000 <SEP> 8, <SEP> 1 <SEP>
 <tb> 7 <SEP> 12.7 <SEP> ppm <SEP> nothing <SEP> nothing <SEP> 555 <SEP> 000 <SEP> 8, <SEP> 0 <SEP>
 <tb> 8 <SEP> 12.7 <SEP> ppm <SEP> 58 <SEP> ppm <SEP> nothing <SEP> 925 <SEP> 000 <SEP> 7, <SEP> 8 <SEP>
 <tb> 9 <SEP> 12, <SEP> 7 <SEP> ppm <SEP> 58 <SEP> ppm <SEP> 4, <SEP> 5 <SEP> h <SEP> to <SEP> 200 <SEP> c <SEP> 459 <SEP> 000 <SEP> 8, <SEP> 1 <SEP>
 <tb> 10 <SEP> 12, <SEP> 7 <SEP> ppm <SEP> 58 <SEP> ppm <SEP> 4 <SEP> h <SEP> to <SEP> -70oC <SEP> 395 <SEP> 000 <SEP> 7, <SEP> 8
 <tb> 11 <SEP> 12.7 <SEP> ppm <SEP> 58 <SEP> ppm <SEP> -1h-950C <SEP> 361 <SEP> 000 <SEP> 8,

    <SEP> 5 <SEP>
 <tb>
 
These data again show that methanol is a potent molecular weight regressant but that its deleterious effects can be completely avoided in the buffered system containing TEAL. However, the molecular weight is severely reduced if the buffered system is kept at 20oC, -70oC or even -950C before the addition of the catalyst to cause polymerization. These data show that TEAL forms an effective buffered system only when added to the load or cold feed immediately prior to initiation of polymerization.

  <Desc / Clms Page number 20>

 
 EMI20.1
 
 <tb>
 <tb>



  No. <SEP> Acetone <SEP> TEAL <SEP> Duration <SEP> from <SEP> con- <SEP> J1v <SEP> l <SEP> NOPO <SEP>
 <tb> test <SEP> added <SEP> added <SEP> servation
 <tb> 13 <SEP> nothing <SEP> nothing <SEP> nothing <SEP> 700 <SEP> 000 <SEP> 7, <SEP> 8 <SEP>
 <tb> 14 <SEP> 24.5 <SEP> ppm <SEP> nothing <SEP> nothing <SEP> 334 <SEP> 000 <SEP> 8, <SEP> 3
 <tb> 15 <SEP> nothing <SEP> 81.3 <SEP> ppm <SEP> nothing <SEP> 836 <SEP> 000 <SEP> 7.8
 <tb> 16 <SEP> 24.5 <SEP> ppm <SEP> 81.3 <SEP> ppm <SEP> nothing <SEP> 756 <SEP> 000 <SEP> 7, <SEP> 8 <SEP>
 <tb> 17 <SEP> 24.5 <SEP> ppm <SEP> 81.3 <SEP> ppm <SEP> 1 <SEP> h <SEP> to <SEP> -1000C <SEP> 255 <SEP> 000 <SEP> 8, <SEP> 4
 <tb>
 
These data show that acetone is a potent molecular weight regressant but has no harmful effects in a buffered system containing TEAL.

   However, the molecular weight of the buffered system is significantly reduced when the mixture or feedstock containing acetone plus TEAL is stored for 1 hour at -1000C before polymerization.
 EMI20.2
 
 <tb>
 <tb>



  No. <SEP> Acetone <SEP> TEAL <SEP> Duration <SEP> from <SEP> con-Mv <SEP> INOPO <SEP>
 <tb> test <SEP> added <SEP> added <SEP> servation
 <tb> 18 <SEP> nothing <SEP> nothing <SEP> nothing <SEP> 830 <SEP> 000 <SEP> 7, <SEP> 6 <SEP>
 <tb> 19 <SEP> 24.9 <SEP> ppm <SEP> nothing <SEP> nothing <SEP> 453 <SEP> 000 <SEP> 8, <SEP> 3 <SEP>
 <tb> 20 <SEP> 24.9 <SEP> ppm <SEP> 81.3 <SEP> ppm <SEP> nothing <SEP> 802 <SEP> 000 <SEP> 7, <SEP> 6 <SEP>
 <tb> 21 <SEP> 24.9 <SEP> ppm <SEP> 81.3 <SEP> ppm <SEP> 5 <SEP> h <SEP> to <SEP> 200C <SEP> 673 <SEP> 000 <SEP> 7, <SEP> 5 <SEP>
 <tb> 22 <SEP> 24.9 <SEP> ppm <SEP> 81.3 <SEP> ppm <SEP> 1 <SEP> h <SEP> to <SEP> - <SEP> 990C <SEP> 680 <SEP> 000 <SEP> 7,

    <SEP> 5 <SEP>
 <tb>
 
These data again reveal that acetone is a potent molecular weight regressor but that its deleterious effect is completely eliminated in the buffered system containing TEAL. However, if the buffered system is kept before polymerization, as is the case with tests 21 and 22, TEAL loses efficiency and the molecular weight is reduced.
 EMI20.3
 
 <tb>
 <tb>



  No. <SEP> Poison <SEP> TEAL <SEP> Duration <SEP> from <SEP> con-Mv <SEP> INOPO
 <tb> test <SEP> added * <SEP> added ** <SEP> servation
 <tb> 23 <SEP> 0 <SEP> nothing <SEP> nothing <SEP> 662 <SEP> 000 <SEP> 7, <SEP> 0 <SEP>
 <tb> 24 <SEP> 12 <SEP> ppm <SEP> nothing <SEP> nothing <SEP> 439 <SEP> 000 <SEP> 7, <SEP> 1 <SEP>
 <tb> 25 <SEP> 12 <SEP> ppm <SEP> 1, <SEP> 2 <SEP> nothing <SEP> 640 <SEP> 000 <SEP> 6, <SEP> 8 <SEP>
 <tb> 26 <SEP> 12 <SEP> ppm <SEP> 1, <SEP> 2 <SEP> 1 <SEP> h <SEP> to <SEP> -970C <SEP> 405 <SEP> 000 <SEP> 6, <SEP> 7 <SEP>
 <tb> 27 <SEP> 12 <SEP> ppm <SEP> 1, <SEP> 2 <SEP> nothing *** <SEP> 646 <SEP> 000 <SEP> 7, <SEP> 4 <SEP>
 <tb>
 * The poison consisted of 4 parts per million of methanol, 4 parts per million of acetone and 4 parts per million of dimethyl ether.

  <Desc / Clms Page number 21>

 



     ** TEAL was added at the rate of 1.2 moles per mole of the total poisons added.



   *** No shelf life before polymerization, but residence time of 15 minutes at -950C after initiation and before completion of polymerization.



   These data show that the molecular weight was greatly reduced by adding the mixed poisons to the load or unbuffered feed (test 24), but was not affected in the buffered load (test 25).



  The conservation of the charge or buffered supply prior to the initiation of the polymerization caused TEAL to lose its effectiveness (test No. 26); but tolerating a period of stay in the reactor after the initiation of the polymerization (as in the case of test No. 27), there was no adverse effect on the buffering ability. These data show that TEAL is an effective buffering agent when used appropriately.



   The data from all these experiments show that TEAL must be used appropriately in order to form an effective buffering agent. It is only effective when added to the cold feed or feed immediately before polymerization. It loses efficiency if the load is hot or if it is stored before polymerization, while TEAL is present there.



  Example 7
This example demonstrates that the molar ratio of the buffering agent to the catalyst affects both the molecular weight of the polymer and the efficiency of the catalyst at a level significantly greater than about 1: 1. Here, TEAL and of AlC13.

  <Desc / Clms Page number 22>

 
 EMI22.1
 
 <tb>
 <tb>



  Test <SEP> Report <SEP> molar <SEP> Weight <SEP> molecular <SEP> Efficiency <SEP> relaNo. <SEP> TEAL / AlCl3 <SEP> relative <SEP> with <SEP> TEAL <SEP> tive <SEP> from <SEP> catalyst
 <tb> "opposite <SEP> from <SEP> tee-with <SEP> TEAL <SEP> opposite <SEP> without <SEP> TEAL <SEP> screw <SEP> from <SEP> witness <SEP> without
 <tb> TEAL
 <tb> 1 <SEP> 0, <SEP> 16 <SEP> 1, <SEP> 04 <SEP> 1, <SEP> 0 <SEP>
 <tb> 2 <SEP> 0, <SEP> 20 <SEP> 1, <SEP> 11 <SEP> 0, <SEP> 9 <SEP>
 <tb> 3 <SEP> 0, <SEP> 42 <SEP> 1, <SEP> 01 <SEP> 0, <SEP> 8 <SEP>
 <tb> 4 <SEP> 0, <SEP> 62 <SEP> 1, <SEP> 08 <SEP> 0, <SEP> 8 <SEP>
 <tb> 5 <SEP> 1, <SEP> 05 <SEP> 1, <SEP> 13 <SEP> 0, <SEP> 7 <SEP>
 <tb> 6 <SEP> 2.01 <SEP> 0, <SEP> 09 <SEP> 0.15
 <tb>
 
The data show that adding TEAL to the feed or pure feed (as free of oxygen impurities as possible)

   has only a moderate beneficial effect on the molecular weight and a moderate negative effect on the efficiency of the catalyst until a quantity significantly greater than one mole of TEAL per mole of Aid is added, ratio after which the molecular weight and the efficiency of the catalyst drop rapidly. In this example, a 5% molar excess had not yet given a reduced molecular weight, but caused a 30% reduction in the efficiency of the catalyst. At a TEAL / AICI molar ratio of 2/1, the molecular weight is 10% lower than the molecular weight of the control and the efficiency of the catalyst has been reduced almost by a factor of six.

   The operating field on which TEAL is effective as a buffer is defined by the quantity of poisons present and by the quantity of AlCl-type catalyst used. Buffer becomes effective at somewhat lower level when the amount added on a molar basis just exceeds the total amount present with poisons and ceases to be effective at a somewhat higher level when the amount added on a molar basis
 EMI22.2
 e significantly exceeds the amount of catalyst of the AlC13 xce 3 type used.



  Example 8
This example is incorporated into this memo in the

  <Desc / Clms Page number 23>

 aim of establishing a comparison and of demonstrating the critical nature of the buffering agents used for the purposes of implementing the method according to the invention. In this example, it was found that tri-isobutyl aluminum (TIBAL) was ineffective as a buffering agent. The poisons added were, in the latter case, t-butanol (TBOH) and methanol.
 EMI23.1
 
 <tb>
 <tb>



  Test <SEP> TBOH <SEP> added <SEP> TIBAL <SEP> added <SEP> Report
 <tb> No. <SEP> moles <SEP> Mv <SEP> INOPO
 <tb> TIBAL /
 <tb> moles
 <tb> TBOH
 <tb> 1 <SEP> nothing <SEP> nothing-487 <SEP> 000 <SEP> 8, <SEP> 3 <SEP>
 <tb> 2 <SEP> 26 <SEP> ppm <SEP> nothing-249 <SEP> 000 <SEP> 8, <SEP> 4 <SEP>
 <tb> 3 <SEP> 52 <SEP> ppm <SEP> nothing-226 <SEP> 000 <SEP> 8, <SEP> 4 <SEP>
 <tb> 4 <SEP> 26 <SEP> ppm <SEP> 110 <SEP> ppm <SEP> 1, <SEP> 6 <SEP> 240 <SEP> 000 <SEP> 8, <SEP> 7 <SEP>
 <tb> 5 <SEP> 52 <SEP> ppm <SEP> 220 <SEP> ppm <SEP> 1, <SEP> 6 <SEP> 244 <SEP> 000 <SEP> 9, <SEP> 0 <SEP>
 <tb>
 These data show that t-butanol causes
 EMI23.2
 he significant regression in molecular weight with or without TIBAL.



   The results of a series of tests carried out with methanol as poison and TIBAL evaluated as a buffering agent are collated below:
 EMI23.3
 
 <tb>
 <tb> Trial <SEP> MeOH <SEP> TIBAL <SEP> Report <SEP> moles
 <tb> No. <SEP> added <SEP> added <SEP> TIBAL / moles <SEP> Mv <SEP> INOPO
 <tb> Me <SEP> OH <SEP>
 <tb> 6 <SEP> nothing <SEP> nothing-908 <SEP> 000 <SEP> 7, <SEP> 9 <SEP>
 <tb> 7 <SEP> nothing <SEP> 111 <SEP> ppm-619 <SEP> 000 <SEP> S, <SEP> 4 <SEP>
 <tb> 8 <SEP> 9 <SEP> ppm <SEP> nothing-522 <SEP> 000 <SEP> 8, <SEP> 2 <SEP>
 <tb> 9 <SEP> 9 <SEP> ppm <SEP> 111 <SEP> ppm <SEP> 1, <SEP> 0 <SEP> 547 <SEP> 000 <SEP> S, <SEP> 3 <SEP>
 <tb>
 
Methanol causes a significant regression in molecular weight whether or not TIBAL is present and, in fact, the data from Trial 2 reveal that the addition of TIBAL, by itself,

   a regression in molecular weight. The effect of TIBAL alone is also very different from the effect of TEAL alone, as shown by the data in Example 7. TEAL alone tends to increase molecular weight, while TIBAL alone reduces it.

  <Desc / Clms Page number 24>

 



  Example 9
This example is also incorporated into the present specification for the purpose of establishing a comparison and further demonstrates the critical nature of the buffering agents used for the purposes of implementing the method according to the invention. In this example, it was found that diethyl aluminum chloride (DEAC) was ineffective as a buffering agent, despite its similarity to the buffering agents used in accordance with the present invention. TBOH represents t-butanol; DME represents dimethyl ether.
 EMI24.1
 
 <tb>
 <tb>



  Test <SEP> Poison <SEP> (base) <SEP> Agent <SEP> Report <SEP> moles <SEP>
 <tb> No. <SEP> buffer <SEP> DEAC / mole
 <tb> base
 <tb> 1 <SEP> nothing <SEP> nothing <SEP> Trial <SEP> witness <SEP> 836 <SEP> 000
 <tb> 2 <SEP> nothing <SEP> 82 <SEP> ppm <SEP> Buffer <SEP> onlyDEAC <SEP> ment <SEP> 944 <SEP> 000
 <tb> 3 <SEP> 25 <SEP> ppm <SEP> TBOH <SEP> nothing <SEP> Poison <SEP> only <SEP> 481 <SEP> 000
 <tb> 4 <SEP> 25 <SEP> ppm <SEP> TBOH <SEP> DEAC <SEP> 1.5 <SEP> mole / mole <SEP> 436 <SEP> 000
 <tb> 5 <SEP> 11 <SEP> ppm <SEP> DME <SEP> nothing <SEP> Poison <SEP> alone-593 <SEP> 000
 <tb> ment
 <tb> 6 <SEP> 11 <SEP> ppm <SEP> DME <SEP> DEAC <SEP> 1.4 <SEP> mole / mole <SEP> 339 <SEP> 000
 <tb> 7 <SEP> 25 <SEP> ppm <SEP> acé-rien <SEP> Poison <SEP> alone-334 <SEP> 000
 <tb> tone <SEP> ment
 <tb> 8 <SEP> 25 <SEP> ppm <SEP> acé-DEAC <SEP> 1,

  1 <SEP> mole / mole <SEP> 408 <SEP> 000
 <tb> tone
 <tb>
 
These data prove that DEAC behaves, by itself, in a similar way to TEAL, in that it exerts a moderate beneficial effect on the molecular weight of the polymer, but that DEAC is completely ineffective as of buffering agent with various oxygen bases. Acetone, dimethyl ether and t-butanol all greatly reduce the molecular weight of butyl rubber, whether or not DEAC is present as a buffering agent. Therefore, DEAC, such as TIBAL, is not suitable as a buffering agent during the polymerization of butyl rubber.

  <Desc / Clms Page number 25>

 



  Example 10
This example proves the usefulness of trimethyl aluminum (TMAL) as a buffering agent, in accordance with the present invention.
 EMI25.1
 
 <tb>
 <tb>



  Test <SEP> Poison <SEP> added <SEP> TMAL <SEP> added <SEP> Report
 <tb> No. <SEP> mole <SEP> TMAL / Ev <SEP> INOPO
 <tb> mole <SEP> Poison
 <tb> 1 <SEP> nothing <SEP> nothing <SEP> Witness <SEP> 891 <SEP> 000 <SEP> 8.1
 <tb> 2 <SEP> nothing <SEP> 41.6 <SEP> ppm <SEP> TMAL <SEP> alone-969 <SEP> 000 <SEP> 7.8
 <tb> ment
 <tb> 3 <SEP> 12.8 <SEP> ppm <SEP> MeOH <SEP> nothing <SEP> MeOH <SEP> alone-763 <SEP> 000 <SEP> 8, <SEP> 1 <SEP>
 <tb> ment
 <tb> 4 <SEP> 12.8 <SEP> ppm <SEP> MeOH <SEP> 58 <SEP> ppm <SEP> 2, <SEP> 03 <SEP> 967 <SEP> 000 <SEP> 7, <SEP> 7 <SEP>
 <tb> 5 <SEP> 24.9 <SEP> ppm <SEP> Acé-rien <SEP> Acetone <SEP> 509 <SEP> 000 <SEP> 7, <SEP> 6 <SEP>
 <tb> tone <SEP> only
 <tb> 6 <SEP> 24.9 <SEP> ppm <SEP> Acé-56 <SEP> ppm <SEP> 1, <SEP> 84 <SEP> 790 <SEP> 000 <SEP> 7,

  4
 <tb> tone
 <tb>
 
These data show that methanol or acetone significantly reduces the molecular weight of the poly-
 EMI25.2
 e mother when added to the control polymerization, but the reduction in molecular weight is strongly avoided when these same poisons are added to the feed or buffered feed containing TMAL. TMAL is an effective buffering agent when added to cold butyl feed or feed immediately before polymerization.



   Another series of discontinuous polymerizations similar to those of Example 1 was carried out, with dimethyl ether as poison and TMAL as buffering agent. Some of the buffered feeds or charges were also kept before starting the polymerization, as in the case of Example 6, in order to determine whether the TMAL loses efficiency if the charges which contain it are kept before the polymerization.

  <Desc / Clms Page number 26>

 
 EMI26.1
 
 <tb>
 <tb>



  Test <SEP> DME <SEP> TMAL <SEP> Report <SEP> Duration <SEP> from
 <tb> No. <SEP> added <SEP> added <SEP> mole <SEP> store- <SEP> Mv <SEP> INOPO
 <tb> TMAL / vation
 <tb> mole
 <tb> DME
 <tb> 7 <SEP> nothing <SEP> nothing <SEP> Witness <SEP> nothing <SEP> 589 <SEP> 000 <SEP> 8, <SEP> 2 <SEP>
 <tb> 8 <SEP> 34, <SEP> 5 <SEP> ppm <SEP> nothing <SEP> DME <SEP> only-nothing <SEP> 315 <SEP> 000 <SEP> 7, <SEP> 9 <SEP>
 <tb> also
 <tb> 9 <SEP> 34.5 <SEP> ppm <SEP> 78.7 <SEP> ppm <SEP> 1.45 <SEP> nothing <SEP> 621 <SEP> 000 <SEP> 7, <SEP> 9 <SEP>
 <tb> 10 <SEP> 34.5 <SEP> ppm <SEP> 78.7 <SEP> ppm <SEP> 1.45 <SEP> 4 <SEP> h <SEP> to <SEP> 424 <SEP> 000 <SEP> 8.3
 <tb> - <SEP> 780C <SEP>
 <tb>
 
These data show that the addition of dimethyl ether to the control polymerization greatly reduces the molecular weight of the polymer,

   but that the reduction in molecular weight is prevented when dimethyl ether is added to the buffered charge containing TMAL. However, if the buffered charge is kept for long periods before polymerization, as in the case of test 8, the TMAL loses its effectiveness as a buffering agent and is no longer able to prevent dimethyl ether to reduce the molecular weight of butyl rubber. These data show that TMAL is an effective buffering agent for the polymerization of butyl rubber, but that, similarly to TEAL, it must be used appropriately to maintain its effectiveness.



  Example 11
This example demonstrates the advantage of diisobutyl aluminum hydride (DIBAL-H) as a buffering agent used in accordance with the present invention.

  <Desc / Clms Page number 27>

 
 EMI27.1
 
 <tb>
 <tb>



  Test <SEP> Poison <SEP> added <SEP> Agent <SEP> buffer <SEP> Report
 <tb> No. <SEP> added <SEP> mole <SEP> tam- <SEP> Mv <SEP> INOPO
 <tb> pon / mole
 <tb> poison
 <tb> 1 <SEP> nothing <SEP> nothing <SEP> Witness <SEP> 635 <SEP> 000 <SEP> 8, <SEP> 6 <SEP>
 <tb> 2 <SEP> nothing <SEP> 87.4 <SEP> ppm <SEP> DIBAL <SEP> H
 <tb> DIBAL <SEP> H <SEP> only <SEP> 753 <SEP> 000 <SEP> 8, <SEP> 6 <SEP>
 <tb> 3 <SEP> 24.2 <SEP> ppm <SEP> nothing <SEP> Acetone
 <tb> Acetone <SEP> only <SEP> 424 <SEP> 000 <SEP> 8.2
 <tb> 4 <SEP> 24.2 <SEP> ppm <SEP> 87.4 <SEP> ppm <SEP> 1, <SEP> 48
 <tb> Acetone <SEP> DIBAL <SEP> H <SEP> 671 <SEP> 000 <SEP> 8, <SEP> 6 <SEP>
 <tb> 5 <SEP> 24.2 <SEP> ppm <SEP> 63.4 <SEP> ppm <SEP> 1, <SEP> 33 <SEP> 668 <SEP> 000 <SEP> 7, <SEP> 8 <SEP>
 <tb> Acetone <SEP> TEAL
 <tb> 6 <SEP> 12.3 <SEP> ppm <SEP> nothing <SEP> Methanol
 <tb> Methanol <SEP> alone-487 <SEP> 000 <SEP> 8,

    <SEP> 5 <SEP>
 <tb> ment
 <tb> 7 <SEP> 12.3 <SEP> ppm <SEP> 87.4 <SEP> ppm <SEP> 1, <SEP> 60 <SEP> 860 <SEP> 000 <SEP> 8, <SEP> 1 <SEP>
 <tb> Methanol <SEP> DIBAL <SEP> H
 <tb>
 
These data show that methanol or acetone greatly reduces the molecular weight of butyl rubber when added to the control polymerization, but that reduction in molecular weight is prevented when these same poisons are added to a feed or feed. buffered containing a molar excess of DIBAL H or TEAL. DIBAL H in trial 4 is as effective in preventing the molecular weight loss caused by acetone as TEAL is in trial 5. The data also show that DIBAL H alone (trial 2) tends to compared to the control, to increase the molecular weight of the butyl rubber rather than reduce it.

   Both in terms of buffering efficiency and effect on molecular weight, DIBAL H behaves much more like TEAL than tri-isobutyl aluminum.
 EMI27.2
 
 <tb>
 <tb>



  Test <SEP> t-Butanol <SEP> DIBAL <SEP> H <SEP> Report <SEP> moles
 <tb> No. <SEP> added <SEP> added <SEP> DIBAL <SEP> H / moles <SEP> INOPO
 <tb> t-Butanol
 <tb> 8 <SEP> nothing <SEP> nothing <SEP> Witness <SEP> 745 <SEP> 000 <SEP> 8, <SEP> 5 <SEP>
 <tb> 9 <SEP> 23 <SEP> ppm <SEP> nothing <SEP> t-butanol <SEP> seu-237 <SEP> 000 <SEP> 9, <SEP> 0 <SEP>
 <tb> also
 <tb> 10 <SEP> 23 <SEP> ppm <SEP> 66.2 <SEP> ppm <SEP> 1, <SEP> 50 <SEP> 755 <SEP> 000 <SEP> 8, <SEP> 1 <SEP>
 <tb>
 

  <Desc / Clms Page number 28>

 
These data show that t-butanol significantly reduces the molecular weight when it is added to the control, but that it has no effect when it is added to the load or buffered diet containing an excess. DIBAL-H molar.

   DIBAL-H is indicated to be an effective buffering agent to prevent the deleterious effects of t-butanol on the polymerization of butyl rubber.
 EMI28.1
 
 <tb>
 <tb>



  Test <SEP> Ether <SEP> dimethy-DIBAL-H <SEP> Report <SEP> moles <SEP> Mv <SEP> INOPO <SEP>
 <tb> No. <SEP> lique <SEP> added <SEP> added <SEP> DIBAL-H / moles
 <tb> ether
 <tb> 11 <SEP> nothing <SEP> nothing <SEP> Witness <SEP> 589 <SEP> 000 <SEP> 8, <SEP> 2 <SEP>
 <tb> 12 <SEP> 34.5 <SEP> ppm <SEP> nothing <SEP> Ether <SEP> seu-315 <SEP> 000 <SEP> 7, <SEP> 9 <SEP>
 <tb> also
 <tb> 13 <SEP> 34, <SEP> 5 <SEP> ppm <SEP> 15.2 <SEP> ppm <SEP> 1, <SEP> 43 <SEP> 678 <SEP> 000 <SEP> 7, <SEP> 1 <SEP>
 <tb>
 
These data show that DIBAL-H is an effective buffering agent to prevent the deleterious effects of dimethyl ether on the polymerization of butyl rubber.



   All of these data show that DIBAL-H, like TEAL and TMAL, is an effective buffering agent for the polymerization of butyl rubber when used appropriately. These buffering agents counteract or prevent the deleterious effects that "poisons" have on the polymerization of butyl rubber. Their use results in the implementation of an improved polymerization process.



  Example 12
In order to confirm the advantages of the present invention under real industrial conditions, a series of tests were carried out in a large-scale butyl rubber reactor, to show that the harmful effects of traces of poisons and / or activators in the filler or feed for butyl rubber can be combated by the appropriate use of complexing agents or buffers. These tests or trials were carried out

  <Desc / Clms Page number 29>

 using a conventional large-scale butyl rubber reactor, which had been provided with auxiliary equipment to allow the injection of controlled and measured quantities of poisons, activators and buffering agents into the charge or reactor supply. The reactor was of the conventional draft tubular model.

   The contents of the reactor were circulated by a propeller pump with bottom penetration, centered in the draft tube. The reactor was cooled by circulating liquid ethylene through the liners where it vaporized to provide refrigeration. The reactor temperature was kept at the lowest possible value allowed by the installation and the reactor was warmed up slowly during a test as it became fouled.



  The rate of reheating is a function of the rate of fouling. All the tests or tests were carried out at a constant feed rate, with a residence time of approximately 40 minutes. The reactor effluent overflowed into a flash tank to which steam and hot water were added to recover the butyl rubber as a suspension in water. The single mothers who did not enter a reaction were flashed at the head for recycling purposes.



   The trials or tests were carried out by establishing constant state conditions and then starting the continuous injection of the desired buffering agent and / or activator / poison at a regulated and measured flow rate. After establishing a new constant "test" state, the effects of the injected agents were determined by comparing the constant state conditions of the reactor before, during and after the test injection. The constant monomer conversion tests or tests have usually been carried out by adjusting the flow rate of the catalyst to correct the effect of the agent injected on the activity of the catalyst, but all the others have been maintained.

  <Desc / Clms Page number 30>

 parameters as constant as possible during a series of tests, so that the effects of the agent injected can be clearly established.

   The control reactors were then operated simultaneously with the test reactor, so as to be able to observe the disturbances caused by uncontrolled variations occurring in the installation and not to erroneously attribute them to the injected agent.



   The charge or feed cooling accessories included two stages of ethylene-cooled heat exchangers. The accessories for the injection of the agents were constructed so that the poison or the activator was injected into the charge or feed of the reactor at the inlet of the first refrigerant of the ethylene charge, while the buffering agent was injected downstream of this point, at the inlet of the second refrigerant of the ethylene charge. This was done in order to comply with conditions that studies in laboratory dishes had found to be necessary for effective buffering.

   In this way, the buffering agent was injected into the cold "poisoned" charge with a contact time of only a few seconds before entering the reactor where the polymerization was carried out continuously.



   The data from a series of tests which show the harmful effects of the injection of methanol into the reactor charge and then the annihilation of these effects by the injection of TEAL downstream of the methanol, are summarized below. The feed rate of the reactor was 6655 kg / hour of a feed or feed containing 31.5% by weight of isobutylene and 0.82% by weight of isoprene in methyl chloride. During the test period, methanol was injected into the charge at a rate of 0.129 kg / h = 4.03 moles / h = 19 ppm by weight relative to the charge or

  <Desc / Clms Page number 31>

 food. During the TEAL injection period, the latter was injected downstream of methanol at a rate of 0.854 kg / h = 7.5 moles / h = 128 ppm by weight relative to the feed.

   The molar ratio of TEAL to methanol during the test was therefore 1.86. The data from the various constant state periods of the test series are summarized below. The reactor index is defined as the Mooney viscosity of the butyl rubber produced, divided by the isobutylene content of the liquid phase of the reactor.
 EMI31.1
 
 <tb>
 <tb>



  Results <SEP> summaries-Methanol <SEP> followed <SEP> from <SEP> the <SEP> series <SEP> of tests
 <tb> at <SEP> TEAL
 <tb> Before <SEP> test <SEP> + MeOH <SEP> + <SEP> MeOH <SEP> After
 <tb> ô <SEP> TEAL <SEP> test
 <tb> Viscosity <SEP> Mooney <SEP> from
 <tb> rubber <SEP> from <SEP> butyl <SEP> 60 <SEP> 49 <SEP> 60 <SEP> 60
 <tb> Hint <SEP> from <SEP> reactor <SEP> 11, <SEP> 7 <SEP> 9, <SEP> 8 <SEP> 10, <SEP> 7 <SEP> 10, <SEP> 5
 <tb> Isobutylene <SEP> no <SEP> entered
 <tb> in <SEP> reaction, <SEP>% <SEP> 5, <SEP> 1 <SEP> 5, <SEP> 0 <SEP> 5, <SEP> 6 <SEP> 5, <SEP> 7 <SEP>
 <tb> Conversion <SEP> from <SEP> isobutylene, <SEP>% <SEP> 87, <SEP> 3 <SEP> 87, <SEP> 3 <SEP> 86, <SEP> 0 <SEP> 85, <SEP> 8
 <tb> Debit <SEP> of A1C13'kg / h, <SEP> 1, <SEP> 374 <SEP> 1, <SEP> 642 <SEP> 1, <SEP> 79 <SEP> 1, <SEP> 25 <SEP>
 <tb> moles / h <SEP> 10, <SEP> 3 <SEP> 12, <SEP> 3 <SEP> 13,

    <SEP> 4 <SEP> 9, <SEP> 36 <SEP>
 <tb> Efficiency Help <SEP>, <SEP> p / p <SEP> 1330 <SEP> 1115 <SEP> 1003 <SEP> 1439
 <tb> Speed <SEP> from <SEP> warming,
 <tb> OC / h <SEP> 0, <SEP> 12 <SEP> 0, <SEP> 085 <SEP> 0, <SEP> 006 <SEP> 0, <SEP> 14 <SEP>
 <tb>
 
In this series of tests, the reactor was brought into constant state operation and the operating conditions were measured and recorded; then the injection of methanol was started and the flow rate of the catalyst was adjusted in an attempt to keep the conversion of isobutylene approximately constant. The reactor was balanced to a new constant state which was subjected to measurement and recording; then we started injecting TEAL downstream of methanol.

   We have adjusted again

  <Desc / Clms Page number 32>

 the catalyst flow rate to keep the conversion about constant and the reactor was balanced to a third constant state which was again subjected to measurement and recording; finally, the injections of both methanol and TEAL were stopped and the catalyst flow rate was again adjusted to keep the conversion constant and the reactor was balanced to a fourth constant state.



  The first and fourth constant states, both without intentional injection of materials into the supply or load, represent the basic case to which the test periods can be compared. Although the initial and final constant states represent the basic state without injection into the supply or load, they were not identical because the reactor had been in service for a long period by the time the fourth constant state was achieved. In the meantime, he had suffered considerable fouling and had been warmed by several degrees Celsius. Since it takes at least six hours for a reactor to reach a new constant state after making a modification, the reactor had been in production for at least 24 hours by the time the fourth constant state was reached .

   In general, when a reactor becomes clogged and heated, we can expect certain trends:
1.) The efficiency of the catalyst tends to increase slightly (note that the initial efficiency of the catalyst was 1330 and the final efficiency of this catalyst was 1440).



   2.) The heating speed tends to increase in an always accelerated way (note the initial heating speed of 0, 120C per hour and the final heating speed of 0.140C per hour).



   3.) The index of the reactor, which is simply the ratio of the Mooney viscosity to the unreacted isobutylene, tends to decrease slightly (note the initial index of

  <Desc / Clms Page number 33>

 11.7 and the final index of 10.5). The reactor index is a measure of the performance or behavior of a reaction system which is expressed in the absence of poisoning effects, a high index being preferred.



   It is important to be aware of these trends as the reactor operates to interpret the data obtained.



   A comparison of the period preceding the tests or tests to the period during the injection of methanol shows that methanol greatly reduces the Mooney viscosity (a measure of the molecular weight of the polymer) and the reactor index, as also the activity of the catalyst. These changes are undesirable and, if they occur as a result of stochastic fluctuations in the level of methanol in the feed or feed, would make it very difficult to adjust the reactor and produce butyl rubber to specification. The data from the simultaneous injection of methanol and TEAL show that the injection of TEAL counteracted the detrimental effect of methanol on the Mooney viscosity and the reactor index.

   A comparison of this constant state with the final constant state after stopping the injection of the two aforementioned agents shows that the TEAL completely prevented the methanol from reducing the Mooney viscosity or the reactor index. Obviously, the injection of methanol seriously affected the reaction mass but had little effect on the buffered reaction mass containing TEAL.



   TEAL does not counteract the detrimental effect of methanol on the efficiency of the catalyst because, as will be shown later, TEAL itself reduces the efficiency of the catalyst, but this is not a significant practical problem. It is even apparent that the introduction of methanol and TEAL exerts a favorable effect on the reduction of the rate of heating of the reactor, which is

  <Desc / Clms Page number 34>

 extremely desirable. This favorable effect of TEAL is further demonstrated in later examples.



   These data thus show that TEAL constitutes an effective buffering agent under conditions of production of butyl rubber on a large scale and that it is capable of counteracting the harmful effects of methanol on the molecular weight of the polymer and the index of the reactor. In addition, it can be seen that TEAL exerts a beneficial effect on the rate of heating of the reactor.



  Example 13
Another series of tests was undertaken in a similar manner to that described in Example 12, in order to confirm the effectiveness of TEAL as a buffering agent against the harmful effects of methanol in a reactor. polymerization of butyl rubber. Also in this series of tests, TEAL was injected into the reactor first and methanol was then injected, to assess the effects of methanol in the buffered reactor containing TEAL. The injection points were the same as those described in Example 12, the TEAL being injected into the feed or cold feed downstream of the methanol and immediately before entering the reactor.

   For this series of tests, the feed rate in the reactor was 6655 kg / h of a feed or feed containing 32.4% by weight of isobutylene and 0.84% by weight of isoprene in chloride methyl. As in Example 12, TEAL was injected into the feed or feed at a rate of 0.854 kg / h and methanol at a rate of 0.129 kg / h during the injection periods. The TEAL / methanol molar ratio was also 1.86 here.

   The data from the various constant-state parts of this series of tests are collated below:

  <Desc / Clms Page number 35>

 Summary results-Test series with TEAL followed by methanol
 EMI35.1
 
 <tb>
 <tb> Before <SEP> test <SEP> + <SEP> TEAL <SEP> + <SEP> TEAL <SEP> After
 <tb> 6 <SEP> MeOH <SEP> test
 <tb> Viscosity <SEP> Mooney <SEP> 60 <SEP> 62 <SEP> 59 <SEP> 58
 <tb> Hint <SEP> from <SEP> reactor <SEP> 10, <SEP> 5 <SEP> 10.9 <SEP> 10, <SEP> 5 <SEP> 10, <SEP> 0 <SEP>
 <tb> Isobutylene <SEP> no <SEP> entered <SEP> in <SEP> reaction, <SEP>% <SEP> 5.7 <SEP> 5, <SEP> 7 <SEP> 5, <SEP> 6 <SEP> 5, <SEP> 8 <SEP>
 <tb> Conversion <SEP> from <SEP> isobutylene, <SEP>% <SEP> 86, <SEP> 4 <SEP> 86, <SEP> 0 <SEP> 86, <SEP> 3 <SEP> 86, <SEP> 1 <SEP>
 <tb> Debit <SEP> d'AlCle, <SEP> kg / h <SEP> 1, <SEP> 13 <SEP> 1, <SEP> 68 <SEP> 1, <SEP> 70 <SEP> 0,

    <SEP> 96 <SEP>
 <tb> moles / h <SEP> 8, <SEP> 44 <SEP> 12, <SEP> 59 <SEP> 12, <SEP> 73 <SEP> 7, <SEP> 18 <SEP>
 <tb> Efficiency <SEP> from <SEP> AlCl3,
 <tb> p / p <SEP> 1655 <SEP> 1103 <SEP> 1095 <SEP> 1930 <SEP>
 <tb> Speed <SEP> from <SEP> warming, "C / h <SEP> 0.14 <SEP> 0, <SEP> 23 <SEP> 0, <SEP> 17 <SEP> 0, <SEP> 68 <SEP>
 <tb>
 
During this series of tests, the reactor was again balanced at four separate constant state conditions, as in the case of Example 12, adjustments of the catalyst flow rate being made after each change to keep the conversion approximately constant. The reactor was first balanced in normal operation, then with TEAL injection, then with TEAL and methanol injection and finally again with injection of the two agents stopped.



   The data show that injecting TEAL alone has a moderate beneficial effect on the Mooney viscosity and the reactor index, as in the case of laboratory tests, and causes a significant (although not significant) drop in catalyst efficiency. . Then, the injection of methanol into the TEAL reactor was started and this injection was practically ineffective. The buffered reactor into which TEAL was injected remained virtually unaffected by methanol.



  This is in sharp contrast to the effect of the injection of

  <Desc / Clms Page number 36>

 methanol in an unbuffered reactor, as in Example 12. The simultaneous elimination of TEAL and methanol remained without appreciable effect on the Mooney viscosity or on the index of the reactor, although the efficiency of the catalyst s 'obviously raised clearly. The heating rate of the reactor rose again strongly even when TEAL and methanol were removed, showing their beneficial effect on this important parameter. These data clearly show that TEAL is an effective buffering agent under large-scale polymerization conditions to prevent methanol from adversely affecting the performance of the reactor.

   Variations in the level of
 EMI36.1
 ethanol from a feed or methanol from in a buffered TEAL-containing reactor does not disturb the reactor in any way, while similar variations in the methanol level of the feed or feed entering a butyl rubber reaction system unbuffered, strongly disturb the reactor and cause the rubber to be produced to be out of specification and cause difficult adjustment of the reactor.



    Example 14
A series of large-scale tests, similar to those described in Examples 12 and 13, were carried out to confirm the effectiveness of TEAL as a buffering agent to counteract the detrimental effects of acetone in a polymerization reactor butyl rubber. In this series of tests, acetone was injected into the balanced reactor to measure its harmful effect. This was followed by a period in which TEAL and acetone were injected simultaneously to determine if TEAL could eliminate the harmful effects of acetone. The reactor feed rate for this series of tests was 6,542 kg / h of a feed containing 32.0% isobutylene and 0.83% isoprene by weight

  <Desc / Clms Page number 37>

 in methyl chloride.

   Acetone was injected at a rate of 0.22 kg / h = 3.77 moles / h = 33 ppm based on the load and TEAL was injected at a rate of 0.854 kg / h = 7.5 moles / h = 130 ppm compared to the load or supply during the injection periods. The TEAL to acetone molar ratio was 1.99.

   The data from the various constant state parts of this series of tests are collated below: Summary results-Series of acetone tests followed by TEAL
 EMI37.1
 
 <tb>
 <tb> Before <SEP> + <SEP> Acetone <SEP> + <SEP> Acetone <SEP> After
 <tb> test <SEP> ô <SEP> TEAL <SEP> test
 <tb> Viscosity <SEP> Mooney <SEP> 68 <SEP> 59 <SEP> 70 <SEP> 64
 <tb> Hint <SEP> from <SEP> reactor <SEP> 10, <SEP> 5 <SEP> 9, <SEP> 4 <SEP> 10, <SEP> 5 <SEP> 10.5
 <tb> Isobutylene <SEP> no <SEP> entered
 <tb> in <SEP> reaction, <SEP>% <SEP> 6, <SEP> 5 <SEP> 6, <SEP> 3 <SEP> 6, <SEP> 7 <SEP> 6, <SEP> 1 <SEP>
 <tb> Conversion <SEP> from <SEP> isobutylene, <SEP>% <SEP> 84.0 <SEP> 84, <SEP> 3 <SEP> 82, <SEP> 7 <SEP> 85, <SEP> 2 <SEP>
 <tb> Debit Help <SEP>, <SEP> kg / h <SEP> 1.14 <SEP> 1, <SEP> 40 <SEP> 1, <SEP> 74 <SEP> 0,

    <SEP> 95 <SEP>
 <tb> moles / h
 <tb> Efficiency <SEP> from <SEP> AlCl. <SEP> p / p <SEP> 1540 <SEP> 1256 <SEP> 990 <SEP> 1870
 <tb> Speed <SEP> from <SEP> warming, "C / h <SEP> 0.085 <SEP> 0, <SEP> 10 <SEP> 0, <SEP> 14 <SEP> 0, <SEP> 27 <SEP>
 <tb>
 
These data show that the injection of acetone into the balanced reactor significantly reduces the Mooney viscosity, the reactor index and the efficiency of the catalyst. A stochastic variation in the level of acetone in the feed or feed introduced into a butyl rubber reactor seriously disrupts the reactor and results in production out of specification. The injection of TEAL downstream of the acetone completely counteracts the harmful effects of acetone on the Mooney viscosity and the reactor index (but not on the effectiveness or efficiency of the catalyst as already explained).

   The simultaneous elimination of TEAL and acetone had no effect on the reactor index (and would not have had an effect on the Mooney viscosity if the conversion of isobutylene had - 37-

  <Desc / Clms Page number 38>

 been kept constant). It is again clear here that TEAL is an effective buffering agent to counteract the deleterious effects of acetone in a butyl rubber reaction system. The introduction of acetone into a buffered reactor containing TEAL would have little effect on the performance of the reactor or on the product obtained.



  Example 15
A series of large-scale tests, similar to those already described, have been carried out to confirm the effectiveness of TEAL as a buffering agent against the harmful effects of diethyl ether in a rubber polymerization reactor. butyl. During this series of tests, TEAL was first injected into the reactor and then diethyl ether was injected to assess the effects of ether in the buffered reactor containing TEAL. The injection points were as previously described, the TEAL being injected into the cold feed or feed downstream of the ether injection point and immediately before penetration into the reactor.

   The feed rate for this series of tests was 6542 kg / h of a feed containing 30.4% isobutylene and 0.79% isoprene by weight in methyl chloride. TEAL was injected into the feed or feed at a rate of 0.69 kg / h = 6.1 moles / h = 105 ppm relative to the feed or feed and diethyl ether was injected at a rate of 0.288 kg / h = 3.89 moles / h = 44 ppm relative to the load during the injection periods.



  The TEAL / ether molar ratio was 1.57. As in the previous tests, the reactor was again balanced at four separate constant state conditions, adjustments to the catalyst flow rate being made after each change to keep the conversion of isobutylene approximately constant. We first balanced the reactor

  <Desc / Clms Page number 39>

 in normal operation, then, however, while TEAL was injected into it, then with simultaneous injection of TEAL and diethyl ether, and finally with injection of the two agents stopped.

   The data from the four constant state parts of this series of tests are collated below: Summary results-TEAL series of tests followed by diethyl ether
 EMI39.1
 
 <tb>
 <tb> Before <SEP> + <SEP> TEAL <SEP> + <SEP> TEAL <SEP> ô <SEP> After
 <tb> test <SEP> ether <SEP> di- <SEP> test <SEP>
 <tb> ethyl
 <tb> Viscosity <SEP> Mooney <SEP> 64 <SEP> 68 <SEP> 64 <SEP> 65
 <tb> Hint <SEP> from <SEP> reactor <SEP> 11, <SEP> 6 <SEP> 11, <SEP> 7 <SEP> 11, <SEP> 2 <SEP> 11, <SEP> 4 <SEP>
 <tb> Isobutylene <SEP> no <SEP> entered
 <tb> in <SEP> reaction, <SEP>% <SEP> 5.5 <SEP> 5.8 <SEP> 5, <SEP> 7 <SEP> 5.7
 <tb> Conversion <SEP> from <SEP> isobutylene, <SEP>% <SEP> 85, <SEP> 5 <SEP> 84, <SEP> 3 <SEP> 84, <SEP> 5 <SEP> 85, <SEP> 1 <SEP>
 <tb> Debit <SEP> of A1C13'kg / h <SEP> 1, <SEP> 20 <SEP> 1, <SEP> 84 <SEP> 1.84 <SEP> 1,

    <SEP> 18 <SEP>
 <tb> moles / h
 <tb> Efficiency <SEP> from <SEP> AlCl3,
 <tb> p / p <SEP> 1410 <SEP> 912 <SEP> 915 <SEP> 1440 <SEP>
 <tb> Speed <SEP> from <SEP> warming, <SEP> oC / h <SEP> 0, <SEP> 09 <SEP> 0, <SEP> 05 <SEP> 0, <SEP> 07 <SEP> 0, <SEP> 13
 <tb>
 
These data show that the introduction of TEAL alone had a very moderate beneficial effect on the Mooney viscosity and the reactor index and caused a substantial drop in the efficiency of the catalyst. Then, the introduction of diethyl ether into the TEAL buffered reactor had practically no effect at all. TEAL very effectively buffered the reactor from the deleterious effects of diethyl ether, so its introduction did not in any way disturb the reactor.

   During a comparative test, the results of which are presented below, the introduction of diethyl ether into a reactor not protected by the buffering action of TEAL greatly reduced the Mooney viscosity and the index of the reactor and

  <Desc / Clms Page number 40>

 resulted in poor quality butyl rubber. The simultaneous removal of TEAL and ether again had no effect on the Mooney viscosity or on the reactor index, thus showing that TEAL prevented the ether from reducing these parameters. The efficiency of the catalyst obviously increased after the removal of the buffering agent and the poison.

   These data again show that TEAL is an effective buffering agent under large-scale polymerization conditions, so as to prevent diethyl ether from adversely affecting reactor performance. The TEAL-containing buffered reactor has been made insensitive to variations in the ether level in the feed or feed and would not be disturbed by the variations in the quality of the natural feed or feed which occur from time to time.



   The data from the comparative test showing the effects of the injection of diethyl ether into a normal unbuffered reactor are presented below.



  During this test, the flow rate of the feed or feed was 6542 kg / h of feed or feed containing 32.5% of isobutylene and 0.84% of isoprene by weight in methyl chloride. Diethyl ether was injected at a rate of 0.25 kg / h = 3.4 moles / h = 38.5 ppm relative to the feed.
 EMI40.1
 
 <tb>
 <tb>



  Results <SEP> summaries-effects <SEP> from <SEP> ether Diethyl <SEP>
 <tb> Before <SEP> the test <SEP> + <SEP> Ether Diethyl <SEP>
 <tb> Viscosity <SEP> Mooney <SEP> 55 <SEP> o
 <tb> Hint <SEP> from <SEP> reactor <SEP> 10 <SEP> 8, <SEP> 5 <SEP>
 <tb> Isobutylene <SEP> no <SEP> entered <SEP> in
 <tb> reaction, <SEP>% <SEP> 5, <SEP> 5 <SEP> 5, <SEP> 4 <SEP>
 <tb> Conversion <SEP> from <SEP> isobutylene, <SEP>% <SEP> 87, <SEP> 1 <SEP> 87, <SEP> 2 <SEP>
 <tb> Debit <SEP> d'AlC13'kg / h <SEP> 1, <SEP> 04 <SEP> 1, <SEP> 30 <SEP>
 <tb> moles / h
 <tb> Efficiency <SEP> from <SEP> AlCl, <SEP> p / p <SEP> 1780 <SEP> 1420
 <tb> Speed <SEP> from <SEP> warming,
 <tb> oC / h <SEP> 0.17 <SEP> 0,

    <SEP> 14 <SEP>
 <tb>
 

  <Desc / Clms Page number 41>

 
The data show that the injection of ether into the normal unprotected reactor caused a significant reduction in Mooney viscosity, reactor index and catalyst efficiency and resulted in the production of out of specification butyl rubber. .



   The results of these installation tests 12 to 15 revealed that TEAL was an effective buffering agent for all classes of oxygenated poisons (alcohols, ketones and ethers). This product effectively prevents these poisons from affecting a butyl rubber reactor, so that stable production can be maintained within specifications, despite fluctuations in the levels of these poisons in the feed or feed.

   Furthermore, these data show that an injection of TEAL into the reactor has the beneficial effect of reducing the rate of heating of the reactor, which has significant economic consequences, because a reduced rate of heating makes it possible to extend the duration. of service and / or of increasing the rate of polymer production without reducing the service length which can be reached to an unacceptably low value.



  Example 16
The purpose of this example was buffering against
 EMI41.1
 the effect of HCl in a polymerization reaction of butyl rubber. HCl is commonly referred to as a catalyst activator in the process for making butyl rubber. It is known that it reduces the Mooney viscosity and the index of the reactor and that it greatly increases the efficiency of the catalyst and the rate of heating of the reactor when it is present in the feed introduced into the reactor. Increased catalyst efficiency is frequently desirable, but rapid heating of the reactor is always undesirable because it limits service life and production of butyl rubber.

  <Desc / Clms Page number 42>

 



   A series of large-scale tests was carried out by repeating the procedures previously described. However, due to the extremely powerful effect that HCl exerts on the rate of heating of the reactor, it was not possible to carry out a prolonged test to obtain a certain number of conditions of constant state in the same reactor, as was the case with previous tests. On the contrary, the tests with HCl were carried out one at a time and, under certain circumstances, it was necessary to compare the reactor containing the HCl with a control reactor side by side, instead of comparing two different constant states in the same reactor, as was the case with the previous examples.



   The extremely harmful effects that HCl exerts on the polymerization of butyl rubber are shown by the results of the following experiment.



   Anhydrous HCl gas was injected into the feed or feed of a balanced reactor at a rate of 7.5 ppm by weight relative to the feed or feed. The injection point was the same as that used for the injection of oxygenated poisons, at the inlet of the first ethylene charge cooler. As soon as the injection of HCl gas was started, the reactor temperature began to rise rapidly. It continued its rapid rise, accompanied by a sharp drop in the rate of isobutylene not entering into reaction, in spite of all the efforts made to reduce the speed of the catalyst in the reactor. It proved impossible to stabilize the reactor to obtain a new constant state.

   In less than an hour, the temperature of the reactor had risen by more than 50C and the reactor collapsed, so that it had to be put out of production for cleaning. It is clear that a stochastic fluctuation in the level of HCl in the charge of this amplitude would be extremely harmful for the operators.

  <Desc / Clms Page number 43>

 
 EMI43.1
 polymerization rations and could lead to a runaway temperature of all reactors in service.



   Since the effects of HCl were so powerful that it was not possible to inject 7.5 ppm into the charge or feed of a reactor in service without causing its "roasting" and its clogging. Consequently, an attempt was made to initiate the polymerization, however, that HCl was already present in the feed or feed at the start. This technique allowed the operation of a reactor containing HCl for a sufficiently long time to obtain a constant state. However, it was not possible to remove the HCl and rebalance the reactor without HCl to compare the two constant states in the same reactor, due to the rapid rate of warming.

   Consequently, the reactor containing HCl must be compared with another reactor operating at the same time as a control reactor. The two reactors were operated in the same manner and for the same period of time, except that anhydrous HCl was injected into the feed or supply of the test reactor at a rate of 7.5 ppm relative to load or feed. The flow rate of the feed or feed in the two reactors was the same and amounted to 6542 kg / h of a feed or feed containing 33% of isobutylene and 0.86% of isoprene by weight in chloride of methyl.

   The constant state conditions in the two reactors are compared below:
 EMI43.2
 
 <tb>
 <tb> Reactor <SEP> test <SEP> Reactor <SEP> witness
 <tb> with <SEP> 7.5 <SEP> ppm <SEP> HCl <SEP> without <SEP> HCl
 <tb> Viscosity <SEP> Mooney <SEP> 44 <SEP> 70
 <tb> Hint <SEP> from <SEP> reactor <SEP> 9, <SEP> 4 <SEP> 12, <SEP> 5 <SEP>
 <tb> Isobutylene <SEP> no <SEP> entered <SEP> in
 <tb> reaction, <SEP>% <SEP> 4.7 <SEP> 5.6
 <tb> Conversion <SEP> from <SEP> isobutylene, <SEP>% <SEP> 89, <SEP> 8 <SEP> 86, <SEP> 8 <SEP>
 <tb> Debit <SEP> d'Alde, <SEP> kg / h <SEP> 0, <SEP> 22 <SEP> 1, <SEP> 31 <SEP>
 <tb> moles / h
 <tb> Efficiency <SEP> from <SEP> AlCl.

    <SEP> p / p <SEP> 8880 <SEP> 1430
 <tb> Speed <SEP> from <SEP> warming, <SEP> oC / h <SEP> 1, <SEP> 2 <SEP> 0, <SEP> 1 <SEP>
 <tb>
 

  <Desc / Clms Page number 44>

 
The data show that HCl significantly reduced the Mooney viscosity and the reactor index and caused a sixfold increase in catalyst efficiency. At the same time, the heating rate of the reactor increased by a factor of 12 up to 1.20C / h. During the five hour period required to obtain a steady state, the test reactor had warmed by more than 6oC. It became clogged and had to be put out of production for cleaning shortly after stopping the injection of HCl in an attempt to restore equilibrium without HCl.

   It is clear that the presence of HCl in the feed has a disastrous effect on the performance of the reactor.



   In another test, HCl and TEAL were injected simultaneously into a balanced reactor to confirm the effectiveness of TEAL as a buffering agent to prevent the harmful effects of HCl on the polymerization of rubber. butyl. During this test, the reactor was balanced under normal conditions and then the simultaneous injection of HCl and TEAL into the feed or feed was started to establish a new constant state. A third constant state was reached by interrupting the injection of the two agents. The feed rate of the reactor was 6542 kg / h of a feed containing 33.2% of isobutylene and 0.86% of isoprene by weight in methyl chloride.

   During the injection period, anhydrous HCl gas was injected at the location of the first ethylene refrigerant, at the rate of 0.042 kg / h = 1.15 mole / h = 6.4 ppm relative to charge or power.



  TEAL was injected downstream of the HCl at the inlet of the second refrigerant of the feed or ethylene feed, into the feed or cold feed, at a rate of 0.992 kg / h = 8.7 moles / h = 152 ppm compared to the load or supply. The TEAL / HCl molar ratio was 7.57.

   Data from these constant state periods in this series

  <Desc / Clms Page number 45>

 of tests are gathered below:
 EMI45.1
 
 <tb>
 <tb> Results <SEP> summaries- <SEP> Injection <SEP> simultaneous <SEP> HC1 / TEAL
 <tb> Before <SEP> test <SEP> HCl <SEP> and <SEP> TEAL <SEP> After <SEP> test
 <tb> simultaneous
 <tb> Viscosity <SEP> Mooney <SEP> 57 <SEP> 65 <SEP> 59
 <tb> Hint <SEP> from <SEP> reactor <SEP> 10, <SEP> 2 <SEP> 10, <SEP> 8 <SEP> 9, <SEP> 8 <SEP>
 <tb> Isobutylene <SEP> no <SEP> entered
 <tb> in <SEP> reaction, <SEP>% <SEP> 5, <SEP> 6 <SEP> 6, <SEP> 0 <SEP> 6, <SEP> 0 <SEP>
 <tb> Conversion <SEP> from <SEP> isobutylene, <SEP>% <SEP> 86, <SEP> 9 <SEP> 85, <SEP> 9 <SEP> 85, <SEP> 9 <SEP>
 <tb> Debit <SEP> d'AlCle, <SEP> kg / h <SEP> 1.23 <SEP> 1, <SEP> 23 <SEP> 1, <SEP> 23 <SEP>
 <tb> moles / h <SEP> 9, <SEP> 19 <SEP> 9, <SEP> 19 <SEP> 9,

    <SEP> 19 <SEP>
 <tb> Efficiency <SEP> from <SEP> AlCl, <SEP> 1541 <SEP> 1524 <SEP> 1524
 <tb> P / P
 <tb> Speed <SEP> from <SEP> warming, <SEP> oC / h <SEP> 0, <SEP> 2 <SEP> 0, <SEP> 2 <SEP> 9, <SEP> 1
 <tb> Methanol, <SEP> ppm <SEP> by <SEP> report <SEP> to <SEP> the <SEP> charge <SEP> 11 <SEP> 11 <SEP> 11 <SEP>
 <tb>
 
These data show that TEAL is an effective buffer for preventing the harmful effects of HCl on the polymerization reaction of butyl rubber. TEAL completely prevented HCl from reducing the Mooney viscosity and the reactor index, and also prevented HCl from increasing the efficiency of the catalyst and the rate of heating.



  In fact, the simultaneous introduction of HCl and TEAL had, in total, only very little effect on the reactor, in contrast to the disastrous effects that HCl exerted in the absence of the protective buffering action of TEAL. The slight increase in the Mooney viscosity and the reactor index, which occurred during this series of tests, when HCl and TEAL were introduced simultaneously, was due to the fact that a small amount of methanol was present in the feed or feed and that TEAL will counteract its effects as well as preventing HCl from affecting the reactor.



   TEAL is clearly an effective buffering agent against the harmful effects of HCl gas on poly-

  <Desc / Clms Page number 46>

 meritization of butyl rubber. Very small variations in the HCl level of the butyl rubber production charges occur during normal practice and significantly affect performance in the absence of TEAL, but have little effect in a buffered reactor containing TEAL.



  Example 17
Another powerful catalyst activator is anhydrous chlorine gas, like gaseous HCl. At very low proportions in the reactor feed or charge, it significantly reduces the Mooney viscosity and the reactor index and greatly increases the efficiency of the catalyst and the rate of heating of the reactor. The injection of 7.5 ppm of chlorine gas into the feed or feed of a balanced reactor causes effects very similar to those described in Example 16 with regard to the injection of 7.5 ppm of gaseous HCl. The reactor heats up at high speed, cannot be stabilized and clogs before it can reach a constant state.

   Since a test with chlorine injection only could not be performed, a test with simultaneous injection of chlorine and TEAL was carried out to confirm that TEAL was an effective buffering agent against the harmful effects of chlorine on the polymerization of butyl rubber. During this test, the reactor was normally balanced, and then the simultaneous injection of chlorine and TEAL was started to establish a new constant state. The injection points were the same as those described in Example 16. The feed or charge rate in the reactor was 6542 kg / h of a charge containing 33.5% isobutylene and 0.87 % of isoprene by weight in methyl chloride. Analysis of the feed or charge showed that it was contaminated with approximately 8.2 ppm of methanol.

   Chlorine was injected at the rate of 0.049 kg / h =

  <Desc / Clms Page number 47>

 0.7 mole / h = 7.6 ppm relative to the load and TEAL was injected at the rate of 0.74 kg / h = 6.5 moles / h = 113 ppm per
 EMI47.1
 2 was related to the charge. The TEAL / C12 molar ratio 9, 28.

   The data from the constant state parts of this test are collated below:
 EMI47.2
 
 <tb>
 <tb> Results <SEP> summaries- <SEP> Injection <SEP> simultaneous <SEP> from <SEP> chlorine <SEP> and <SEP> from <SEP> TEAL
 <tb> Before <SEP> test <SEP> Chlorine <SEP> and <SEP> TEAL <SEP> simultaneously
 <tb> Viscosity <SEP> Mooney <SEP> 51 <SEP> 53
 <tb> Hint <SEP> from <SEP> reactor <SEP> 9, <SEP> 1 <SEP> 9, <SEP> 5 <SEP>
 <tb> Isobutylene <SEP> no <SEP> entered <SEP> in
 <tb> reaction, <SEP>% <SEP> 5, <SEP> 6 <SEP> 5, <SEP> 6
 <tb> Conversion <SEP> from <SEP> isobutylene,
 <tb>% <SEP> 87, <SEP> 1 <SEP> 87, <SEP> 4 <SEP>
 <tb> Debit <SEP> from <SEP> AlCl.

    <SEP> kg / h <SEP> 1, <SEP> 39 <SEP> 0, <SEP> 91 <SEP>
 <tb> moles / h <SEP> 10, <SEP> 46 <SEP> 6, <SEP> 8 <SEP>
 <tb> Efficiency <SEP> from <SEP> AlCl3,
 <tb> p / p <SEP> 1367 <SEP> 2106 <SEP>
 <tb> Speed <SEP> from <SEP> warming, <SEP> oC / h <SEP> 0, <SEP> 1 <SEP> 0, <SEP> 2 <SEP>
 <tb>
 
These data show that TEAL is an effective buffering agent to prevent the disastrous effects of chlorine on the polymerization of butyl rubber. The simultaneous introduction of chlorine and TEAL resulted in a moderate improvement in Mooney viscosity, reactor index and catalyst efficiency, and had little effect on the rate of warming.

   Some of the moderate improvement was due to the fact that TEAL counteracted the negative effects of the small amounts of methanol also present in the feed or feed.



   In the absence of TEAL, the injection of a similar quantity of chlorine would have caused rapid heating of the reactor and clogging of the latter and would have greatly reduced the Mooney viscosity and the index of the reactor, just as it would have greatly increased the efficiency of the catalyst.

  <Desc / Clms Page number 48>

 



  Example 18
Tertiary butyl chloride is another very powerful activator during the polymerization of butyl rubber. Its harmful effects are similar to those of HCl and chlorine but are not as pronounced, so that it is possible to carry out a sequential test and to obtain several constant states in the same reactor in service. A series of large-scale tests, similar to those described in Examples 12-18, were performed to confirm the effectiveness of DIBAL-H as a buffering agent against the deleterious effects of t-butyl chloride in a polymerization reactor. butyl rubber. In this series of tests, t-butyl chloride was injected into a balanced reactor to establish a new constant state.



  Then, DIBAL-H was injected into the feed or cold feed downstream of the t-butyl chloride and immediately before entering the reactor to establish a third constant state with DIBAL-H counteracting the harmful effects of t-butyl. Catalyst flow rate adjustments were made after each change to keep the conversion as constant as possible. The feed or feed rate in the reactor for this series of tests was 6542 kg / h of a feed or feed containing 33% isobutylene and 0.86% isoprene by weight in methyl chloride . Tertiary butyl chloride was injected at the rate of 0.046 kg / h = 0.5 mol / h = 7 ppm relative to the charge. DIBAL-H was injected at a rate of 0.32 kg / h = 2.24 mo / h = 48 ppm relative to the feed or load.

   The DIBAL-H / t-butyl chloride molar ratio was 4.48. The data from the various constant state parts of the test are collected below:

  <Desc / Clms Page number 49>

 
 EMI49.1
 
 <tb>
 <tb> Results <SEP> summaries-Series <SEP> with <SEP> chloride <SEP> from <SEP> t-butyl <SEP> tracking
 <tb> of <SEP> DIBAL-H
 <tb> Before <SEP> test <SEP> + <SEP> t-BuCl <SEP> + <SEP> T-BuCl <SEP> 6
 <tb> DIBAL-H
 <tb> Viscosity <SEP> Mooney <SEP> 59 <SEP> 49 <SEP> 62
 <tb> Hint <SEP> from <SEP> reactor <SEP> 10, <SEP> 5 <SEP> 8, <SEP> 7 <SEP> 10, <SEP> 3 <SEP>
 <tb> Isobutylene <SEP> no <SEP> entered <SEP> in
 <tb> reaction, <SEP>% <SEP> 5.6 <SEP> 5, <SEP> 6 <SEP> 6, <SEP> 0 <SEP>
 <tb> Conversion <SEP> from <SEP> isobutylene,
 <tb>% <SEP> 87, <SEP> 0 <SEP> 87, <SEP> 6 <SEP> 85, <SEP> 8 <SEP>
 <tb> Debit <SEP> from <SEP> the AlCla, <SEP> kg / h <SEP> 1,

    <SEP> 23 <SEP> 0, <SEP> 375 <SEP> 1, <SEP> 43 <SEP>
 <tb> moles / h <SEP> 9.22 <SEP> 2, <SEP> 8 <SEP> 10, <SEP> 7 <SEP>
 <tb> Efficiency <SEP> from <SEP> AlCl3,
 <tb> p / p <SEP> 1520 <SEP> 5045 <SEP> 1295 <SEP>
 <tb>
 
These data show that the introduction of tertiary butyl chloride into the balanced reactor significantly reduces the Mooney viscosity and the index of the reactor and causes a more than tripled increase in catalyst efficiency. The introduction of DIBAL-H completely eliminates all the harmful effects of t-butyl chloride. The Mooney viscosity and the reactor index are restored to their pre-test levels and the catalyst activation effect is more than completely suppressed.



   These data show that DIBAL-H is a very effective buffering agent against the damaging effects of t-butyl chloride in a butyl rubber polymerization reactor. An increase in the level of t-butyl chloride in the feed or feed for the manufacture of butyl rubber would lead to the formation of a product no longer meeting the imposed specifications and would very significantly affect the performance of an unbuffered reactor, while the same increase in the proportion of t-butyl chloride in the feed has practically no effect in a reactor protected by

  <Desc / Clms Page number 50>

 effective buffering agents like DIBAL-H or TEAL.



  The ability of these agents to buffer the reactor against the serious consequences of including powerful activators in the feed or feed represents a great economic benefit.



  Example 19
As further confirmation of the effectiveness of TEAL as a buffering agent to protect butyl rubber reactors in operation against the harmful consequences of variations in the proportions of traces of poisons and activators in the feed or charge streams , a third type of large-scale test was carried out in addition to those already described.



  During normal operation of large-scale butyl rubber reactors, the feed components (methyl chloride, isobutylene, isoprene and isobutylene dimer) are continuously dosed at the desired ratios and flow rates, in a supply or mixing tank. feed where they are mixed and kneaded to prepare the feed or feed mixture which is then introduced in metered doses into separate reactors which may be in operation. A convenient way to demonstrate the effectiveness of a buffering agent against a poison or an activator is to put two reactors into production, fed from the same feed or feed mixture tank, only one of the reactors being protected by injecting the buffer agent into its feed or charge stream.

   Next, a quantity of poison or activator is injected into the supply or feed mixture tank, so that the poisoned feed or feed is introduced into the two reactors.



  The difference in response between the protected reactor containing the buffering agent and the normal reactor can then be used to assess the effectiveness of the buffering agent against

  <Desc / Clms Page number 51>

 this poison. This type of test was undertaken with propionic acid as an activator and TEAL as a buffering agent.



   This is not a constant-state test as in the preceding examples, but a test which measures the transient response of the reactors to a fluctuation in the rate of poison in the feed or feed. During this test with propionic acid, an amount of propionic acid was introduced into the feed or feed mixture to obtain 24.2 ppm by weight of propionic acid in the feed or feed from the test.



  This proportion then decreased exponentially in the feed or charge mixture tank, since a cleaner charge was continuously added. Computer calculations evaluated the rate of spoilage and showed that the level of propionic acid had dropped to 1 ppm approximately 107 minutes after the start of the test. The level of propionic acid in the reactors started at 0 rose during the period when the poisoned feed or feed was introduced into it and then slowly deteriorated again when the propionic acid was swept from the system. Computer calculations showed that the maximum level of propionic acid in the reactors reached 7.4 ppm approximately 39 minutes after the start of the test and had dropped to 1 ppm after 170 minutes.

   Reactor conditions changed continuously during this period, but there were no changes in the speed of the catalyst. The reactors were allowed to respond naturally to the change in the level of propionic acid. Propionic acid is a relatively powerful activator for the polymerization of butyl rubber, but weaker than the HCl, tbutyl chloride or anhydrous chlorine gas described in the previous examples.



   The feed or charge rate of the two reactors

  <Desc / Clms Page number 52>

 The weight during this test was 7870 kg / h of a load containing 33.1% isobutylene and 0.87% isoprene by weight in methyl chloride. TEAL was injected into the protected reactor at the rate of 0.724 kg / h = 6.35 moles / h = 92 ppm relative to the feed or feed. The introduction of the propionic acid spike into the feed or feed mixture had no significant effect on the reactor into which TEAL had been injected as a buffering agent. This protected reactor continued to produce butyl rubber meeting the imposed specifications, without sudden heating, or change in the rate of unreacted isobutylene.

   On the other hand, the control reactor containing no buffering agent suffered from a major disturbance due to the presence of the contaminated feed or feed of propionic acid. The reactor temperature rose 2.50C, the rate of unreacted isobutylene fell 2.5%, and the Mooney viscosity of butyl rubber in production fell 15 points. This reactor produced rubber outside the specifications imposed on it for more than 4 hours before finally recovering from the disturbance. There was therefore a very significant difference between the response of the protected reactor and that of the control reactor vis-à-vis the addition of propionic acid to the feed.

   TEAL acts displayly as a buffering agent to prevent the reactor into which it was injected from being affected in any way by the inclusion of the activator in the feed or feed, while the control reactor was severely disordered and produced a large amount of rubber unacceptable. The economic benefit of using the buffering agent is clearly shown by this test.



   It is clear that TEAL is a very effective buffering agent against the harmful effects of acid

  <Desc / Clms Page number 53>

 propionic on the polymerization of butyl rubber.



  Although this was an artificial test, in the sense that propionic acid was purposely injected into the charge, it nevertheless closely simulates a situation of real imbalance caused by poisons or activators present in the charge or food. These materials normally enter the load or feed when there is a certain imbalance in the recycling system (i.e. dryers or fractionation towers) or when a batch of contaminated raw materials (i.e. i.e. monomers or diluents) manifests in the filler.



  The poison or the activator is therefore naturally introduced in a manner substantially identical to that simulated during the present test. The very excellent stabilizing effect of TEAL appears easily and is extremely interesting.



  Example 20
Another test similar to that described in Example 19 was carried out to compare the transient response of a buffered reactor and of a control reactor to an inclusion of isopropyl alcohol in the feed or feed. Isopropyl alcohol is a relatively powerful activator in the feed or supply of butyl rubber. During this test, a large amount of alcohol was introduced into the feed or feed mixture tank to determine the reaction of the reactors to a large inclusion of activator.

   Sufficient isopropyl alcohol was injected at once into the feed or charge mixture tank to raise the alcohol level to 120 ppm and was allowed to decrease by
 EMI53.1
 exponentially as the fresh feed or feed is continuously added and the contaminated feed or feed is introduced into the reactors. Computer calculations have shown that the level

  <Desc / Clms Page number 54>

 of alcohol in the charge or feed mixture tank decreased to 1 ppm after 124 minutes. The proportion of isopropanol in the reactors rose slowly from 0 to 31.7 ppm in 33 minutes and then slowly fell to 1 ppm after 218 minutes.

   As in the case of the test of Example 19, the intention was to let the reactor protected by TEAL and the control reactor respond normally to the inclusion of the charge or feed contaminated with alcohol, but the charge contaminated by alcohol had such a significant effect on the control reactor that it was necessary to interrupt the flow of the catalyst in this reactor and to interrupt its operation; the TEAL-protected reactor was not significantly affected.



   During this test, TEAL was injected into the charge of the buffered reactor, at a rate of 0.724 kg / h = 6.35 moles / h = 92 ppm relative to the charge. The feed ratio of the two reactors in production was 7870 kg / h of a feed containing 33.1% of isobutylene and 0.87% of isoprene by weight in methyl chloride.



   There was no noticeable change in the buffered reactor which was protected by the injection of TEAL into its feed or feed. The temperature rose by a few tenths of a degree, but this is hardly more than the normal fluctuations which are obtained in experiments in a reactor; and there was no change in the rate of unreacted isobutylene, or in the Mooney viscosity of the rubber produced. TEAL was clearly an effective buffer.



   On the other hand, the witness, namely the unprotected reactor, was seriously disturbed by the charge or feed contaminated with isopropanol to the point of having to be taken out of production and cleaned. The reactor temperature began to rise rapidly about 8-10 minutes after

  <Desc / Clms Page number 55>

 the introduction of isopropanol into the feed or feed mixture and this temperature continued to rise rapidly. After 20 minutes, the flow of catalyst to this reactor was stopped to prevent overheating, but the temperature continued to rise. In 25 minutes, the reactor temperature was almost 50C higher than that of the start and continued to rise.

   The suspension was so hot at the time that the reactor began to clog and it had to be shut down to undergo a wash cycle to clean it. The concentration of unreacted isobutylene in the reactor had dropped by more than 3% and the Mooney viscosity of the polymer produced had also started to drop. It is clear that the charge or feed contaminated with isopropanol had strongly affected this normal reactor, causing the premature termination of the test in progress.



   These data show the effectiveness of TEAL as a buffering agent to protect a butyl rubber reactor in operation against a disturbance caused by variations in the levels of poisons and activators present in the feed or charge streams of the reactor. A TEAL-buffered reactor is likely to continue production of acceptable butyl rubber during periods of fluctuating feed quality or load, which would cause a normal unprotected reactor to produce unacceptable rubber or to be out of adjustment to the point of rapid fouling and agglomeration of suspension particles forcing the reactor to stop and the implementation of a cleaning cycle.

   The important economic advantages which come from this ability to keep reactors in stable operation despite variations in the quality of the supply or load, are obvious.


    

Claims (1)

 CLAIMS 1.-Method of polymerization catalyzed by Lewis acids for the preparation of homopolymers of iso-  EMI56.1  C4 olefins to copolymers of a 4% c 7 1 e iso-olefin C a in and of a conjugated diene an inert diluent, characterized in that the polymerization reaction is introduced into the zone at about 1.0 to about 1.25 moles of a buffering or complexing agent per mole of Lewis acid catalyst, in an amount at least effective for protecting said rubber produced against a regression in molecular weight, due to the presence poisons or polymerization activators, the buffering or complexing agent being (1) a lower aluminum trialkyl, the alkyl group being a methyl or ethyl radical, or (2) a lower dialkyl aluminum hydride,  the alkyl group of the hydride being an alkyl radical in  EMI56.2  C1 to C4. 1 4 * 2.-A method according to claim 1, characterized in that the buffering or complexing agent is triethyl aluminum.  3.-A method according to claim 1, characterized in that the buffering or complexing agent is trimethyl aluminum.  4.-Method according to claim 1, characterized in that the buffering or complexing agent is diisobutyl aluminum hydride.  5.-A method according to claim 1, characterized in that the amount of buffer or complexing agent reaches up to about 1.0 mole of this agent per mole of the catalyst in question.  6.-A method according to claim 1, characterized in that the amount of buffering agent or complexing agent represents at least a molar excess relative to the amount of polymerization poisons present in the polymerization reaction mixture.  <Desc / Clms Page number 57>    7.-A method according to claim 1, characterized in that the homopolymer is polyisobutylene and the copolymer is butyl rubber isobutylene-isoprene.  8.-A method according to claim 1, characterized in that the catalyst is aluminum chloride.  9. A method according to claim 1, characterized in that the inert diluent is chosen from the group formed by methyl chloride, methylene chloride, vinyl chloride, ethyl chloride, propane, butane, pentanes or hexanes.    10.-A method according to claim 9, characterized in that the inert diluent is methyl chloride.    11.-A method according to claim 7, characterized in that the butyl rubber has an average molecular weight, by viscosity, from 250,000 to 600,000, the diluent is methyl chloride, the catalyst is aluminum chloride and buffering or complexing agent is chosen from the group formed by triethyl aluminum, trimethyl aluminum or diisobutyl aluminum hydride.  12.-A method according to claim 1, characterized in that the buffering or complexing agent is introduced into said zone of the polymerization reaction by introducing said agent into the feed stream or charge of incoming monomer, cold, immediately before introducing this feed or feed stream into said polymerization reaction zone.  13.-A method according to claim 1, characterized in that said buffering or complexing agent is introduced into said polymerization zone before the moment of introduction of the monomers and the catalyst.  14.-A method according to claim 13, characterized in that said buffering or complexing agent is also introduced into the supply or charge stream of incoming monomer, cold, at a point immediately before  <Desc / Clms Page number 58>  introducing said feed or charge stream into said polymerization reaction zone.  15.-A method according to claim 1, characterized in that said buffering or complexing agent is in contact with the isobutylene monomer or the iso-olefin and the conjugated diene before the polymerization, during a period of time less than about 60 seconds.    16.-Method according to claim 12, characterized in that the iso-olefin homopolymer is polyisobutylene.    17.-A method according to claim 12, characterized in that the copolymer is a butyl rubber based on isobutylene-isoprene.