Document MGym2aQELg9z2Gk0vkJbvX0GV

\\ 2488 ' Ktfe. NICUIIED APR 22 1959 This Report iWThe Property of The Dow Chejlucdl Company \W'-. .< EP P -654 "ST0283B78 CDP'tU - Fes t Vinyl-Vinylldene Chloride Copolymers by Conoflovr for Vinyl Asbestos Floor Tile Part II. Semi-Plant Scale Up by . W. G. MacPhei*son - C. D. Parker F. M. Poindexter Report No. EP P-654 March 15, 1959 L CENTRAL RESEARCH INDEX 0 B.E.R.MAR 26-1959 ',,. VV-> Date Book Ref. No: Charge No: EP-934 -826 -944 -916 -917 1254 pstiypjute4Zju& Page 1 ? SUMMARY J This work is concerned with a study of polymerization variables and resin properties of vinyl-vinylidene chloride copolymers in a com position range of 85# to 91>0# vinyl chloride. Resins were pre pared in both 50 gallon and 750 gallon Pfaudler glass lined reac / tors. Polymerization variables, composition distribution and poly mer granulation data are presented and discussed. Physical proper ties of these resins, both alone and in floor tile formulations are compared with VYHH, the leading commercial resin in this field. INTRODUCTION SUC8Z01S. The Initial work in the development of a vinyl-vinylldene chloride copolymer resin for floor tile applications in 20 gallon Pfaudler type reactors has been presented in Saran Polymerization report EP P-5^7* It was hoped that a single resin composition could be produced having properties that would satisfactorily fit It Into both fields of application. As a result some difference of opin ion existed among various Coatings Technical Service groups as to the exact resin composition. It was decided to initiate a scaleup In both the 50 gallon and 750 gallon reactors to obtain the following Information: 1) To gain information on possible scale-up problems. 2) Prepare resins having a composition of 85# and 88# vinyl chlor ide-15# and 12# vlnylidene chloride in the 750 gallon reactor for a complete evaluation and a final decision on polymer com position. 5) To obtain resin in sufficient quantity to have a plant floor tile run made and prepare 5000 sq. ft. of floor tile. This tile was to be placed in service on the floors of the C. C. Kennedy Research Laboratory. 088C 8Z01S . Page 2 EXPERIMENTAL Resins were prepared by suspension polymerization in glass lined 50 gallon and 750 gallon Pfaudler reactors- All resins were pro duced using the technique of automatic addition of mixed monomers by pressure. Resins were evaluated by the following lacquer and floor tile tests as recommended by the Asphalt Tile Institute. Plasticizer Absorption or P.A. 152 grams of resin with 10 grams of Tribase E were premixed on a Kltchenaide mixer. 68 grams of plasticizer (18 pts. dioctyl phthalate, 50 pts trlcresyl phosphate) was then added at slow speed over a period of 20 seconds, then stopped and the mixture checked for wetness. Resin was then mixed at high speed for an additional 20 seconds and checked again for wetness. A plasticizer absorption No. 1 = dry free flowing mix. A plasticizer absorption No. 6 = wet, paste mix. Baker Perkins Flux The following materials were premixed on a Kltchenaide Mixer: 152 grams Resin 68 grams Plasticizer (18 pts. D0P-50 pts. TCP) 10 grams Trlbase E This mixture was then placed in a Baker Perkins mixer at a temper ature of 275F and mixed until fused. The time to fuse this mix ture is known as the initial flux time. To the above then was ad ded the following: 56 grams Piccopale 320 grams Lesamite 115 grams No. 18 graining sand 5^ grams TiOs (No. 510) The mixture was then milled at the same.temperature of 275P until fluxed. The time to fuse this mixture is known as the final flux time. Page 3 Two Roll Mill Test The following materials are dry blended. 300 grams resin 6 grams lead stearate The rolls of a two roll compounding mill are adjusted so both are at a temperature of 200F. The dry blend Is then placed on the rolls and milled until It completely fused and banded on the rolls. The time to fuse and band is recorded as the flux time. 180C82O1S Floor Tile Recipe 210 27 27 228 477.6 130 60 18 1.2 1.2 grams resin tr plasticizer (DOP) M plasticizer Monoplex S-38 ii asbestos shorts (7 nun) ti Lesamite (CaC03) ii No. 18 graining sand ir TiOa (R 510) n Metasap No. 624 ii aluminum stearate it paraffin wax The above materials are dry blended and the plasticizer added. It then is fluxed on a two roll mill at a temperature of 280F (front roll) and 230F on the back roll. The fused material is then given two passes through a calender and then cut Into 6 inch squares for testing. Indentation This test utilizes the McBumey Indentation Tester. The tile is subjected to a 30 lb. weight on a 1/4" ball for periods of one minute and ten minutes. The resulting indentation of the tile is measured and recorded In mils. Impact In this test a 1" diameter ball weighing 0.143 lbs. is dropped through a distance of 10 Inches or 20 Inches striking the tile. * ST0283M2 Page 4 The number of drops required to produce a fracture is recorded as the Impact strength. Dimensional Stability A 6" x 6" tile is conditioned at 77P for 15 minutes and then care fully measured. It then is placed on a steel plate and exposed to a temperature of l8oF for a period of 6 hours. After which it is cooled to room temperature. After again being conditioned at 77 F for 15 minutes it carefully is measured a second time. A change in tile dimensions, either plus or minus, is recorded as the di mensional stability in mils. Heat Distortion These tests were determined on a Boyer apparatus on flash molded samples of resin. Tlnlus Olson Flow This test was made on numerous samples of resin to try and deter mine if it were a more realistic test than the Baker Perkins in determining flux characteristics of the resins. A pelletizer was made up of stainless steel to prepare polymer pellets for testing. This consisted of a 5/8" hole bored into a block of steel and a plunger of the same dimension. The block is filled with polymer, the plunger inserted and this subjected to a pressure of 7600 lbs/ sq. in. in a Preco Press. The compressed pellets then are run on a Tinius Olson at temperatures of 150, 160 and 170C at a con stant pressure of 500 lbs/sq. in. Graphs were prepared, plotting time in seconds to flow 1" vs. temperature. Brookfield Solution Viscosity 58 grams of resin are dissolved in 152 grams of a mixture of 2 pts methyl ethyl ketone to 1 pt. toluene. Solutions are held at a constant temperature of 25C for a period of 16 hours, and the viscosity measured with a Brookfield Viscosimeter. ` ST0283S83 Page 5 SECTION I DISCUSSION A total of eight polymerization runs were made in the 50 gallon reactors, none of which gave evidence of any scale-up problems. Polymerization and processing data on these runs are presented in Table I. The variables in these runs were in the granulator recipe, catalyst system, agitation and the percent monomer added to Initiate polymerization. Although each run was charged the same l.e., 91# vinyl chloride-9# vinylidene chloride, the anal yzed compositions varied from 80# to 92# vinyl chloride. This may be due to errors in the analysis, in weighing the monomer charge, incomplete mixing of the monomers or a combination thereof. Absolute viscosities averaged approximately 1.05 with two excep tions, B5-13^5 and B5-13^6. These two runs had an average vis cosity of approximately 0.97 which Is considerably lower than the rest of the series. Data in Table I indicates two possible causes. 1. Higher vinylidene chloride content by analysis. This would result in lower resin viscosity as vinylidene chloride is a good chain transfer agent. 2. Slower agitation speed. This may effect the resin viscosity if the agitation is slow enough to result In poor heat transfer in the reactor. This could result in localized overheating or hot spots which would tend to reduce the resin viscosity. Flux times on both the Baker Perkins and the two roll mill were quite uniform except for run B5-13^7. This run had a flux time on the two roll mill of 2 minutes as compared to approximately 30 seconds for the others. It is suggested that the high 0-D-B viscosity of this resin is due to a composition distribution con taining high vinyl chloride fractions. If true, this would result in a higher viscosity and longer flux time at the same average composition level. ST0283MU Page 6 P.A. or plasticizer absorption is a function of polymer granulation and appears to be greatly Influenced in this recipe by agitation. Referring to Table No. 1, runs B5-13^5 and B5-13^6 were made at an agitation speed of 120 RPM. Each had a P.A. of No. 4 indicating a wet resin-plasticizer mix. In comparison runs B5-13^7 thru B5-1352 which were prepared at 150 RPM had P.A. values from 1 to 3 indicat ing a relatively dry, free flowing mix. Figures 1 and 2 present graphically the Tinius Olson Flow and Heat Distortion data on these resins. With one exception, flow viscos ity and heat distortion values are comparable to those obtained on resins prepared in the 20 gallon reactors at the same composition level. (Refer to Saran Polymerization Report No. EP P-5^7) It is interesting to note that resin B^-13^7 has a much higher flow viscosity and higher heat distortion values. This previously has been found to be due to the effect of higher vinyl chloride com positions or a broad composition distribution. This data appears to correlate with the high 0-D-B viscosity and longer flux time of this resin which would substantiate the observations made earlier in this discussion. SECTION II Scale-up in the 750 gallon reactors Involved the preparation of a number of both B8800 and B8500 compositions. Polymerization and processing data are summarized in Table II. As will be noted there was little or no variations in the polymerization recipe in either the B8800 or B8500 composition runs. The exceptions were in the amount of monomer added to Initiate polymerization and the use of vinyl chloride to relieve the vacuum on the reactor prior to ad ding the initial shot of monomer. The addition of vinyl chloride at this point was made to shift the monomer composition in the water phase toward a higher vinyl chloride composition. Inasmuch as vinylidene chloride enters the polymer chain faster than vinyl chloride the initial polymer formed should be of a composition closer to the desired or charged monomer composition. This was discontinued however because it was found that the vinyl chloride was not being added as a vapor but as a raw liquid monomer. The Page 7 addition of liquid monomer would result in at least a portion being dissolved in the water phase. This could result in the Initiation of polymerization in which the initial polymer formed would be polyvinyl chloride. This could be one explanation for some of the variation in resin viscosities, solubilities, etc. S89C8Z01S > Runs R5-8049 through R5-8052 in which vinyl chloride was not added to relieve the vacuum, were very uniform in both resin and proces sing properties. However, there was another variable here that is not shown. In explanation, when a conoflow or continuous addition run is made more monomer is mixed up in the monomer weigh tank than is added to the reactor. This results in a monomer heel left in the weigh tank. In the preceding runs, this heel was left in the weigh tank and additional monomers added to make the succeed-, lng run. It was felt that this procedure could contribute greatly to a variable monomer composition and therefore to variable polymer composition and properties. This variable was eliminated from the above mentioned runs by dumping this monomer heel and preparing a fresh monomer mix for each run. Although the actual composition of the monomer mixtures in these runs is unknown it is indicated here that an analytical method is needed to adjust monomer mixtures to a constant composi tion. Then and then only can reproducibility of resin composition and properties be attained. It is suggested that the method cur rently being used at Production Plant based on refractive Index would be excellent. A comparison of B85OO and B88OO compositions indicates the former is a softer resin resulting in shorter flux times, higher indenta tion and lower Impact values. Solutions prepared in a 65/55 methyl ethyl ketone-toluene solvent system, were quite uniform In vlscosIty, producing very clear solutions that were free of Insolubles. Page 8 T.O. Flow and Heat Distortion Figure 3 through 6 graphically present the Tinius Olson Flow and heat dlstrotlon data on both the B8800 and B85OO compositions. In general these values compare very favorably with those obtained In both the 50 gallon and 20 gallon reactor batches at equivalent resin compositions. R8047 (B8500) and R8o4l (B8800) appear to fall outside the general grouping in melt viscosity, indicating a higher vinyl chloride composition or a broad composition distri bution. This appears to correlate well with the poor solubility and resin processing data on these resins. ai ST8283M6 Specific Gravity and Melting Point A comparison of specific gravity and melting points of the various vinyl chlorlde-vlnylidene chloride compositions with VYHH is pre sented in Table III. Specific gravity of the resin decreases with increasing amounts of vinyl chloride in the polymer, with VYHH being the lowest. This could reflect a cost advantage for VYHH if these resins were compared on a volume basis. B85OO has the lowest melting point, which in conjunction with its good plasticizer absorption properties accounts for its fast flux ing properties. Polymer Granulation A comparison of the polymer granulation of B8500 and VYHH may be observed in Figures 7 and 8. VYHH consists of polymer particles on an average smaller than B8500, however, it does have some large particles which are probably agglomerates. On the other hand B8500 appears to have some very large spherical particles with a large amount of smaller agglomerates. These small irregular shaped ag glomerates greatly increase the polymer surface area which accounts for plasticizer absorption properties comparable to VYHH. A criti cal comparison of these resins with VYHH as a lacquer resin was made by the Lacquer Section of C.T.S. and will be found in Coatings Technical Service Report L-815. L88C0ZftiS Page 9 Following is a reprint of the summary of this report: "Experimental Resin X2716 (B8500) a vinyl chloride-vinylidene chlo ride copolymer produced by a new manufacturing method was evaluated as an industrial maintenance lacquer resin. Solution and film properties were compared with those of three leading commercial all-purpose vinyl resins. Experimental resin X2716 was found to yield solutions with exceptional clarity in vinyl lacquer solvents. Viscosity is in the medium range and viscosity stability is equi valent to the commercial resins. Sprayability is equal to or bet ter than the control lacquers. Films of this resin are characterized by outstanding resistance to chemicals and solvents and an exceptionally low WVTR. This is en hanced by a high response to plasticizers; elongation being achieved at low plasticizer levels. This response also yields films of extra strength at desirable elongation values. The films were equivalent to commercial resins in heat and light stability. Adhesion to steel and wash primer was considered poor. It is concluded that Experimental Resin X2716 will perform excep tionally well as an industrial maintenance lacquer resin." A plant trial run with X2716 (B8500) was made by Flintcote in Chi cago, Illinois, in which 5000 square feet of floor tile was pre pared. In a discussion of the plant trial with these people at a later meeting, it was indicated the resin performed very well! They did mention that at one point in their process they had some trouble with the tile mixture falling off their compounding rolls. However, inasmuch as they were reluctant to give out any informa tion concerning this problem it would be only a guess as to its cause. The floor tile prepared in this plant trial is now in service on the floors of the C. C. Kennedy Research Laboratory (677 Building). ITABLE 50 G a llo n S em i-P lant Batches Page 10 ` ST8283W8 S S l?tij Sffu; iH oo oo v)vO 30'Oil OV^" 1 0> r--< CO CC'HH 0) O p ri 73 -H C-P (m ^ H (S O C UT\ O CM O ^job xriLi 3 Tui --------- 4----------`--------- TOJrMSW T:S "I #03 `0STA *u,ios OVCTiOO OVCU vpvOVOvO t-- CM CO K\CM-4r-i H H CM oo OMT\-=r in CHM rHH OH r<---ll-=i--ti in CM b--CM lA OV CbiCO CArrHH LfMA in cm in .4* in moving Iho m in in r-t oin cm h- HH CM O 8 8 iT\ O C c a\ oc H -p cS = rH nJ > <D 4) rH S -H n O: c VO \OlflNN VO vovOvOvO C b-Ob b-CO CTV H w -mP oC'-mm *^ CM i--l-=J- -=d- K\fT\ rH ClJ i--I t--l rH rH > in o) <--i vo* cm vo a\HHOO l i--l t--f (--l rH p--l pHco co co rm = -i c - - - b- n in in m in oinIianon imn in in =r o m c-- > >-ao-zt ovavrHvo nmo vo o\m CM rH-4" b-- ino mn in In rH Tinhrin-i CM page 11 TABLE I I uo-p^djoeqv aazTOT^s-eu 03H % I CM I CM X3 rH CM I CM rH rH <D rH rH rH OCFOvCMJ 3C0M 0^ ooo"H O CM O -I =I rt cmI +HHWH mmmrHcj- oo cm rH co m ov HooHoColiroinon rH -J* o03I %V Qoao1 voco c_ 3 m avb-ao o UT u,Ios #0*3 03 VO f<lDl<MiVbC O O f O O rH o O iCcnrsoosjA * sqv H H 7Q3) rH rH rH rH* rH 4) rH rH rH rH rH oo/suiS <h cm oco 6b30oiinnivnoovvo=rJH-cvoo O~ 0~0 iOrumnuin\ uv\--o in AqTSuaa in. in* rQHj in in in*vo* in CM CA rH tn CM Ov^- 00 CM CM invo 0\H-4- KV-4- -=t ooooooooo *** HHHHrlHrlHrl co vo vo vo vo-4-vo vo in tnc-~b-inb--=*-ovcacm b-vovo <nvo mnvovo o t-- inbO rH r-i rH STS^IvuV H cm 4> av b- in in-4- i?q TOA # * c............................ HOOHOOlH CM CO inoino^ a-4b- ovm- o\=t in nvo m 00CO 00 oo ovoo COCO CO ovoo rH a\co ovovco ocMOonoino i *ujw noMom i i o h in 4> m h in cnco 41 co o cm ovo rH CM KMn^- in-H m avavs- ao hoiooo --* rH rH rH rH saH rH rH rH tamUvmaauitQ ojromtDiooooo (uhjooojcccc 0)4)EC)O<D(llD >7 >> 0 >> t >J >S > bv qoqs isuiouow qsx *sqq 3O O O * * s * * * crvo rH CM irnno oooooo CMCMVOOOOOOO cMCMmmmmcMCMCM #05 do" min^f- -4ooo O ----------------- ^ (OH)DOW ooo O tn sa rH rH rH rH rH rH rH rH oooooooo n o M CO 0-. in CM inr cm aJ -P O o CM o H X> O HO rH ..in CM V0 t CO fir o p 1 iQ 03 i 53.K | H p fe 4) g^ o d-----------------------* *dmoQ rraqumj\f unu OOOOOOOO oooo o o o o GcoOoCoOcCoOcCoOcrooCoOoCcOoCcOo eQpQPQfqpQcQWcq vo b-co avo rn cm m nnn m-4- -4- -=t -=* cooaoocooooocpocooooocpo >irvni mt ini ini ini ini ini otonooinooinooinoionomooomoomooin COCOCOOOCOCOCOOOOO mcqcqfQPQpcjfqcQcq P- invo b~C0 OVO rH CM jit sr in n in SCoO CoO CoO 0o0o00oCOoCOoco IIIIIIIII ">|cinifunnn' -iA-- i1n-vH o K^O rH O b-O CO CT\ am C0 32 t^21 fl) 03 td p Sj*3, H o XJ p H X! scu: cou 6^0 f \ ** 03 c 3 (C r-i r-i TABLE I I I c-pu f-- tflP co P P C 1--1 T" 0) s o 0 m m tn P oo oo 1 m CT*V VO K\ tn rH p rH L C3 CTV VO VO .m P -=* CM -3" ft CO i-t rH r-4 rH KN =? rH o 2; i c 3 K C\ o m in oo O O co i oo i CTV m m tr; cc a> a; T3 p t-t nj Po c 0P O O J5 P <O P tn i--4 rH o > > a C e vH ^*H o >> o mt- HCO 0 a P u o rH <s> J5 "O OP fs a> o cp 0 T3 O H rH iH >4 >4 cc Pp >> O P Oo\ 0 T3 P u o p0 x: -o OP 0O CP 0 rC a o p pp >4 >4 CC PP >> (MIX) POO <D tH O P0 J2 X> OP U 0O CP 0 JC -O O p pp >4 >4 cc pp >> in in PCO 0) *T* ft s >4 E-i > oo o CVJ X c~~ vO pp c-- t-- CM CVI XX Page 12 ST02B3990 Cl to X to TO TMC CM. CO :r ro T " t. c m . 33Gt-t4 t> 0 10 TO T H E C M . 3 5 0 - 1 4 ST0283896 00 ( R ..0: I". ,ni. lsOX Figure 8 ST0283897 VYHH Magnif - l80X ST028389B Page 15 Conclusions and Recommendations Resin compositions of (85# vinyl chloride-15# vinylidene chloride) are comparable to VYHH in resin processing and basic floor tile- prop erties. This composition has solution properties comparable and in some respects superior to VYHH as a lacquer resin. Agitation in the range of 120 to 140 RPM appear near the optimum, however this may present a problem in scale-up to a 5500 reactor. These agitation speeds are not readily obtainable in the larger reactors. An absolute viscosity of 1.04 to 1.05 appears to be the range for good resin properties. This is obtained by polymerizing at a tem perature of 65C and a constant pressure of 125 psig. Heat distortion and Tinius Olson flow tests correlate very well with resin processing. The Tinius Olson flow test is recommended as a production control test. An analytical method to adjust monomer mixtures to a constant com position is recommended. This would result in greater reproduci bility of resin compositions and properties. C. C. KENNEDY RESEARCH LABORATORY The Dow Chemical Company Typed 5-12-59 WGM/CDP/FMP/lm By: By: By: