i CHEMICAL ENGINEERING REVIEWS
PROCESSES REVIEW
I
II POLYMERIZATION I
THE
past \.ear has been one of vigorous activity on the commercial scale to bring into production many of the results of the previous years of work in the laboratory. Obviously a great share of this burden has fallen on the process and design engineers who have to convert the laboratory Jvork to a profit producing manufacturing plant. Linear polyethylene: all-cis polyisoprene, crystalline polystyrene: crystalline polypropylene and methylstyrene were only some of the products creating this activity. All this in addition to improvements on, and expansion of: existing polymers. T h e teamwork of the Chemist and Engineer has been suprrb and the future of polymers seems not only assured but brighter than ever. T h e four major parts of this revieiv arc concerned w.ith Processes and Equipment (as much as available on equipment and a concentration on techniques); Catalysts (specific on all types of catalysis); Chemistry of Polymerization (fundamental information on the chemistry of the various types of reactions); and Other Interesting Polymers (a quick look a t new polymers that are not just a modification of previously kno\vn materials).
Equipment, Design, and lnstrumentation Agitation in polymerization reactions has always been important but difficult. Some processes that have been described are: a continuous process for the polymerization of olefins in a series of stirred reactors using slurries of solid phosphoric acid suspended in liquid hydrocarbons ( 7 A ) ; a method of preparing soluble copolymers of unsaturated rubbers and styrene, consisting of processing in the presence of shearing agitation during the early stages of polymerization ( 7 A ) ; and a polymerization vessel consisting of a rotating d r u m containing loose balls for agitation (2A). Attendant equipment for the improve-
ment of polymerization processes were: Turbulent film evaporation for the continuous concentration of GR-S latexes. Tests sliowcd it to be superior in some respects to the standard concentration equipment (6.4). Continuous distillation for the removal of toluene from polyurethane slurries ( 5 4; X two-stage polyethylene recovery process (70-4) ; and Recovery of polymers from reaction mixtures by spraying the solution into a precipitant or freezing bath (72%). Sodium as a heat transfer medium in polymerization processes was discussed (73.1) and the use of a high temperature bed of coal to remove oxygen from inert %as blankets was patented (!?A). Pilot units, very useful for studies of the polymerization of propylene (3A) and emulsion polymerizations (75.4), were dcscribed. Design calculations usrful to polymerization process engineers were given for: heat transfer in very viscous media ( 7 6 A ) ; pressure drop in pipes carrying non-Xewtonian fluids (77A); and use of analog computers in the design of reactors ( J A ) . Instrumentation showed some progress in the use of infrared spectrometry to foilow styrene polymerization (74A) and the use of ultrasound to measure the cure of thermosets (8A).
A. F.
A number of other recent comniercial d e v e l o p m e n t s Lvere r e v i e w e d ( 7 7-4). Processes-Emulsion Suspension
and
Butadiene and st)-rene emulsion polymerizations have been aided by various publications. T h e viscosity characteristics of cold rubber latex Type X-667 were thoroughly investigated over a range of temperatures and solids content (76B). Below 25% solids the behavior is Newtonian in character. Above 25% it is non-Newtonian and can be described accurately by equations suitable for the calculation of pressure drop in pipes. h-eoprene Type 60 latex was similarly characterized (77B). Agglomeration of GR-S latex by the action of solvents was investigated and the effects of a niimber of variables were detrrmined ( I B ) . These results were compared )\it11 the natural agglomeration of latex a t intermediate stages of polymerization. Other investigations of the butadiene-styrene emulsion polymerization process were noted (IOB,
2 7B, 23B). Vinyl chloride and vinylidene chloride emulsion and suspension processes uere improved by patents on the recovery of monomer (7B) and a continuous process. T h e influence of a number of variables
ROCHE
The Dow Chemical Co., Midland, Mich. ARTHUR F. ROCHE is assistant supervisor, Plastics Production Laboratory of the DOW Chemical Co. Midland Division. Roche attended Michigan State cotlege and joined Dow in 1937; he served as group leader in the styrene polymerization laboratory and also in the physical research deportment before transferring to his present assignment in 1953. Extensive work on styrene polymer and copolymer development and their improvement has led to a number of patents in this fleld. Roche i s a member of the ACS.
VOL. 48, NO. 9, P A R T II
SEPTEMBER 1956
1643
UNIT PROCESSES REVIEW on the vinyl chloride emulsion polymerization was reported (7B, 6 B ) . Emulsifiers of various types were compared (3B),as to critical micelle concentration (73B, 24B)>and the particle size of the emulsion formed (25B). Other emulsifiers (I 7B, 22B) and suspending agents (923, 75B) were patented and reported. Modifiers for emulsion polymerization processes including a number of those that occur naturally in GR-S manufacture were investigated and patented
(74s:IQB,27B, 28B). Other process improvements in emulsion and suspension polymerization include: A continuous process for polyacrylonitrile (ZB),a semibatch process with characteristics similar to continuous methods (8B),a n improved method of monomer addition (ZOB), and variety of others (5B, 72B, 78B, Z9B) including a review of suspension polymerization
(26B). Processes-Condensation Cyclic dimethylsiloxane tetramer was polymerized by different methods and the results compared (ZC). Direct hydrolysis of the Grignard reaction mixtures rather than separating the monomers was used to prepare polysiloxanes (8C) and other process improvements for silicone manufacture were patented (3C-5C, 72C). Polyester-urethane foam propertiescomfor example, soft to rigid-were
pared with the formulations and processes used to prepare them ( 6 C ) . A process for the controlled addition of water to the diisocyanate-polyester reaction by means of water of crystallization on salts was patented (7C). Two revieivs of diisocb-anate pol\-merizations were published ( I C , 73C). Other condensation processes were covered by: a discussion of methods for converting alkyd resin plants to the production of butylated melamine fornialdehyde resins (9C); a proposal of the most satisfactor). conditions for the preparation of urea-melamine-formaldehyde resins ( I O C ) ; and a patent covering a method of preparing hiqh molecular weight polyethylene terphthalate (17C).
Processes-Solution Polymerization in solution was investigated for a variety of reaction types. Acrylic acid polymerizations were investigated to determine the effect of salt concentration (ZP) and a patent \vas issued on a process for recovering polyacrylate salts (5P). Solvent polymerization of ethylene in the presence of a nickel-cobalt catalyst was patented (41)). T h e use of high boiling alkylated aromatics as a solvent medium for the styrenation of drying oils to clear products was patented ( 7 P ) . An apparatus for the ionic polymerization of isobutylene was also patented ( 3 P ) .
Processes-Gas
Phase
Vinyl and vinylidene monomers polymerized in the vapor phase to low molecular weight compounds were patented (34). The alternate reversal of flow of olefins through a phosphoric acid catalyst bed was patented to prolong the life of the catalyst (7Q). The type of PO!?mer made by the styrenation of oils was controlled by the addition of styrene monomer as a gas to the reaction mixture (ZQ)and styrenated a!kyd resins of prcdetermined properties \yere prepared bv controlling the functionality of the alkyd
(4Q). Processes-Safety Some hazards of polymerization processes were noted in the determination of the toxicity of triethylenetetramine (ZR) and methyl acrvlate (3) and a warninq against disposal of waste amines and nitrocellulose in the same container (7RI.
Catalysts-Heterogeneous and Radiation T h e Ziegler catalyst (triethyl aluminum-titanium tetrachloride) for the polymerization of eth>-lene a t low pressures to essentially linear polymers ( 3 4 0 ) continued to create activity and discussion throughout the past year. Phillips Petroleum (260) has produced a pol>-ethylene (using hexavalent chromium
Courtesy Aiiied Chemical & Dye C N P .
Polyester resin building-bulk
1644
INDUSTRIAL AND ENGINEERING CHEMISTRY
and drum shipping facilities
POLYMERIZATION ~~~
oxide supported on silica-alumina) which is claimed to be similar to the Ziegler material but not identical. I t differs in the amount and type of branching and in the prevalence of trans configurations 1370). Heterogeneous catalysts are not necessary to produce such polymers, lioivevcr! if the correct conditions are med (very hiSh pressure) for the homoycneous, free radical polymerization 17 7 0 ) . Other heterogeneous catalysts \]sed in the polymerization of ethylene :,not exclusively to the above type) !\’ere mixed metal carbides and oxides ( 3 0 0 ) : nickel oxides and alkali boroh>-drides LjD), nickel oxide ( 7 3 0 ) , a n d molybdenum (720, 2 5 0 ) . A’atta a n d others published further details of the properties of polystyrene arid polypropylene made by heterogeneous catalysis (22D. 230, 240). During the year thc rcdiscovery (hlorton was first) of crystalline pol!.styrene was announced ( 3 3 0 ) . M a d e by the alfin catalyst system, the polymers: once in the amorphous form, can only be made to crystallize by special treatments, such as a hot solvent treatment. Essentially all-cis polyisoprme was reported by Goodrich-Gulf (70) and Good)-ear (80) b!- using the Zieglertype catalyst. Firestone ( 6 0 ) has made what appears to be an identical material by metallic lithium. This rubber seems to be equivalent to natural rubber and may be marketed a t about the same price. Isotactic polyvinyl ethers ( 2 7 0 )
~~
~
~
~
Jvere discussed, noting the critical conditions necessary for their preparation and the properties of poly-cu-methylstyrene, prepared by metallic sodium were rcported. polymerization (750), Control of the process for polymerizing prop!-lene to low molecular iveight products \vas improved by the discovery that the catalyst activity can be predicted by the results of a titration to determine the catalyst acidity (740). High energy radiation catalysis of polymerizations appears to be economical if performed a t the reactor site ( 2 0 ) . A number of monomers that were polymerized b>- high energy radiation of 1,arious types were acetylene (TO), ethylene ( 2 9 0 ) , acrylamide ( 3 0 , 2 8 0 ) , and methyl methacrylate ( 2 0 0 ) . Some others Ivould not ( 2 0 0 ) polymerize. Photoinitiated polymerization of methyl methacrylate (270, 3 2 0 ) , styrene ( 7 7 0 ) , and vinyl chloride ( 7 6 0 ) were studied. A number of compounds were evaluated as photosensitizers for the polymerization of various monomers ( 7 9 0 ) and a patent obtained on o-alkyl xanthate esters as photosensitizers (40). T h e photoinitiated decomposition of acetyl peroxide was investigated ( 7 8 0 ) . Ultrasound ma!- have some capacity for the initiation of acrylamide polymerization ( 9 0 ) and will form block copolymers of acrylamide and acrylonitrile from polyacrylamide dissolved in {vater and acrylonitrile (700).
Catalysts-Free
Radical
Relative activity of catalysts is of great importance to most process developments. A number of azo conipounds were compared and the activity of each correlated Lvith the composition of the substituent groups (73E: 27E). Various other catalysts were compared for the polymerization of methyl methacrylate and vinyl acetate (255E), for and for styrene ( 2 E ) . vinyl chloride (5E), Details of specific systems were studied in some cases. T h e chain transfer coefficient of benzoyl peroxide on poly(vinyl acetate) radicals was determined and by comparison \vith the known value for styrene and benzoJ-l peroxide the Q and e values of benzoyl peroxide Ivere calculated. These results ivere checked by calculating the coefficient for methyl methacrylate ( 2 E ) . T h e cumene hydroperoxide catalyzed styrene polymerization was studied in detail (37E). Also? the effects of polyfunctional peroxides on the molecular weight and distribution of methyl methacrylate were determined (77E). Various other systems investigated Ivere acetyl-benzoyl peroxide and vinyl acetate (23E), organic sulfur compounds and various vinyl monomers (70E, 20E), and a number of odd oxidation-reduction combinations for methyl merhacrylate in water (7gE). Patents were obtained on the use of chlorine in the &xylene pyrolysis process (72E), use of
Polyester resin kettle at Barrett Division plant, Toledo, Ohio VOL. 48, NO. 9, PART I1
a
SEPTEMBER 1956
1645
UNIT PROCESSES REVIEW perfluoroacyl peroxides for haloethylenes (32E),N-vinyl carbazole oxidation products (24E), halogenated propionyl pera n d ozone (9E). oxide (7E), Activated peroxide decomposition has been used to accelerate polymerization rates in many processes. By study (74E) of the decomposition rates of substituted benzoyl peroxides in the presence of dimethylaniline, it was determined that electron accepting substituents accelerate the reaction ; whereas electron donating substituents retard it (compared with unsubstituted benzoyl peroxide). T h e base catalyzed decomposition of a number of peroxides was investigated (4E) a n d shown to be a function of the strength of base and structure of the peroxide. Specific cases of activated peroxide systems studied were naphthoyl peroxide-amine-styrene (30E), benzoyl peroxide-dimethylaniline-styrene (78E), benzoyl peroxidedimethylaniline-vinyl chloride (77E). benzoyl peroxide-dimethylaniline-acrylonitrile (76E), benzoyl peroxide-diniethylaniline-methyl methacrylate (75E),a n d benzoyl peroxide-metal saltlinseed oil (27E). Emulsion polymerization catalysts of the oxidation-reduction type were studied by comparing the iron(I1) induced decomposition of various hydroperoxides in the presence of three different monomers (26E). T h e effects of the cumene hydroperoxide substituents were determined as a function of Hammett's constants. T h e effect of ionic strength of the medium. type of monomer, and the presence of methanol were also determined. Other work and patents on emulsion catalysts were noted (7E, 3E, 8E, 28E). Suspension systems were also of interest (6E, 29E).
Catalysts-ionic
(77F). Patents (7.427) were obtained on the use of ferric halogell-olefin oxide complexes for the polymerization of olefin oxides. Aluminum chloride -kctone complexes were patented 1751;) for the polymerization of biitencs. Propylene polymerization by silico tungstic acid was investigated and sho\\-n to produce higher conversions of the tctramer than the commonly used phosphoric acid (22F). Modified phosphoric acid catalysts were patented for the polynieriza2UF, 271;). tion of olefins (iF, I t was shown that loiv molecular weight polybutadiene made ivith sodium has a structure different than that niade by other methods ( S F ) . Polymerizations of dienes by sodium and ether promoters was patented ( 7 2 F ) . 'l'hc effccts of substituents on the polymerization of a number of acrylamidcs by lithium aluminum hydride were studied 7 9 F ) ,
Type
Condensation catalysts (amines) for urea-formaldehyde reactions, which are noncatalytic in the presence of water but become active as the composition is dried, have been shown (70F) to be particularly useful in the paper coating industry. T h e relative effectiveness of lithium, sodium, barium, and potassium hydroxides for the catalysis of the phenolformaldehyde reaction were shown to increase with decreasing ionic radius of the cation (7F). T h e effect of a number of different catalysts on the structure of phenol-formaldehyde resins was reported (75F). A wide variety of metal-organic compounds were patented for polyesterification reactions. These included compounds of lead ( 6 F ) , zirconium (5F), aluminum (4F),tin (33,titanium (ZF), a n d magnesium (23F). T h e polymerization of vegetable oils was found to be a
1 646
combination of condensation and DielsAlder addition reaction. Many compounds were investigated as catalysts for this reaction (78F). Epoxy resin cures were studied (77E;) and some specific catalysts were found. Boron trifluoride-amine complexes were patented for epoxy resins ( g F ) . Addition catalysts of the ionic type are often heterogeneous in nature and, therefore, might have been included in that section; however. for the purpose of this article, the more common types of ionic catalysts are discussed in this section. Boron trifluoridc in the presence of a proton-donor has been shoivn to bc a catalyst for the polymerization of stilbene (738'); but, under no conditions tried, was it a cis-trans isomerization catalyst. T h e polymerization kinetics of acenaphthylene catalyzed by boron trifluoride-ether complex \vas studied
inhibitors Quinone and hydroquinone were studied by Karasch and others (5G) to determine the mechanism of the short stopping reaction. A thorough investigation of low molecular Iveight products of aqueous, free radical-quinone reactions demonstrated that the reactions are very different than those previously postulated and that the mechanism is different for chloranil than it is for hydroquinonc and alkyl substituted hydroquinones. T h e end product of the quinone reactions is a disubstituted quinone or hydroquinone and the active intermediate is a quinhydrone. Quinone was studied by Bevington and others for styrene (2G) and for methyl methacrylate ( 3 G ) . A t 20' C., it was shown that termination of methyl methacrylate is chiefly by combination of the quinone radical end with the methacrylate
INDUSTRIAL AND ENGINEERING CHEMISTRY
radical end. .At 60' C., disproportionation betiveen the quinone end and the methacrylate end becomes more important. T h e end products of these reactions appear to be qiiinone diethers as previously shoivn by orhcrs. 1-liese compounds are very different froni the products of Karasch. Nine different inhibitors \v~'rc.rcstcd i n an azo-catalyzed, methyl metiidcr!.late polymerization a n d a n estirnatioti \vas made of the amount of inhibitor that copolymerized \b.ith the monoinrr ( 6 G ) . Other rvork covrred the inhihition of vinyl acetate by oxygen (7G) and patrnts on the use of alk\.l and aralkyl nitrites (-G)and substituted catechols ( i G j .
Chemist r y-Co
poly mers
Block- and graft-t),pe copo1yinci.s continue to he of interest, but havc: shown little commercial development to date, probably becaiise the yields arc. low. Block copolymcr!: have been prepared by using y , y '-azohis [y-cyanovalcric acid) initiator: thcrehy, producing a diacid pol>-mer Ivhich can br coupled with other diacid polymers by the usc of some diol (