Document jpB8BY2LwQJBR9j0X7ORdgxO

: 4 tt. F. Ooodrtch Chemical Company OW THC B. P. OOOOMICH COMPANY DEVELOPMENT CENTER Company THRMAL STABILITY OF POWDERED GEON-HYCAR NITRILE POLYBLENDS V: r UTILITY OF DYNAMIC THERMAL STABILITY TESTS hr M. E. Wood* Prefect No. 33B4-188 Oota Completed. August 14, 1974 Data bsuedb August X, 1974 ABSTRACT V This study sms InitfeOsd to develop a ysntlWiss tolnd to pndkt the prof--log " safety and safe of decomposition of flexible Goon campouedi lint ore modified with powdered Hycor nitrile tlwtOMn. Tfio standard Btabendor Dynamic Theneol Stability tost was shown to provide the nsrtwsy qualitative infen--tIon. Two Iwdl--tois of processing safety wsn dovolopod from this data. Two indicators of dagrodatl-- rota woro also obtained by analyzing the bahovioc of the caeyo jnds during the dsci--position phase of the tost. This tost was then used os a tool to stody the affects of changes in both fee erwepound rocipe and nitrile elastomer on the thermal stability of fee blonds. This feint study by fee Go-- and Hycor groups clearly demonstrated feet procambsg safety Is Inversely proportioned to the vieaarity of fee nitriW startowor. The role of decomposition was drown to be effected by fee PVC staMheri in the compound and fee stabillsstfee system wed in fee pewrfesed robber. These stages meulted in the development of e powdered caeycund of Geon end Hyoar far a easterner eat--fan opplkotioe. They also sneblsd us to eccwctofy define fee type of "ideal" pewderod nitrile slestoann, feat ero requisad to optimlee fee processIbllHy ef feero feeneoplestlc blends. DISTRIBUTION C R. Goats *tj*. . Schulte ITL - R.J. Meyer C.T#./lnt'.<4) G. Smell W ,H. YfeMtlngton 0 .G. Proser P J. Dorset E.J. Sehte-E.G. Schwaegerle N.G. Duke - J.P. Morrill trocksvilfe RAD Center K. Greens "I.J. Fawcett P Jl. Ho been *R.A. Krueger *W-Moorehouee J.ley-C.H.Luftor J.L. Nelson - P J). Terry H. (Col.) Tucker R.B. Osboroe - G.S. RAD Fife (2) RjD. Scott - I.L. Toole 0.1.fepmpeon BJM.d? Zwksker - M.E. 1.f. Welch-A. Kukic Louisville TTT 9.H. Lawrence W. Hein- KH. Sherwood-D .1. Wright RAO File (2) 398085 BFG26327 ^23561001 Tablo of Contanh ^ajor Objective Minor Obfoctivos introduction . . Conclusions...................................... Future Action................................ Enporfcnontal Description .... Results and Phcuwion.................... htatprotatlon of DTS Tost Data DTS LmoI of Hyaar NUrdo tabbor PAGE 1 1 1 1 10 13 17 II II Z00T9S C Z BFG26328 TABLE I List of Tables Compound Recipes. PAGE 3 TABLE II * Brobender Conditions for Dynamic Thermal Stability Experiments ... 3 T ABLE III - OTSAnalysis of Control and Rubber Modiffad Compound......................... 8 TABLE W - Indicators of Bland Thermal Stability and Decomposition Rato................ 9 TABLE V - DTSAnalysis of Stabilizer Variations........................................................11 TABLE VI - DTSAnalysis of Effects of Type and Level of Hycar Rubber......................... 12 T ABLE VII - Processing Mlcator* of Coeyoundi Modified with Linear and Crosslinked Nitrile Rubber.......................................................................12 T ABLE VIII - OTS/biolysis of Rubber and Stabiliser Verietiem.........................................14 TABLE IX - DTSAnalysis - Effect of Liquid Rubber an Confound Processability . . 15 TABLE X - Pretewing indicators of Compounds Bssd on Liquid and Powdered Hycar Elastomers.......................................................................................... 16 Uri of Figures FIGURE I - Typical Dynamic Thermal Stability Data 5 7 23561003 BFG26329 \. Mojof Objective To improve the thorns*I {(ability of powdered Geon/Hycar polyblends. Minor Objectives To determine Ho utility of dynamic thermal liability toiH at an indicator of compound processability. To uaa dynamic thermal liability tests to itudy (he effects of several nitrile rubber variables on (he procomability of potybiendi. Introduction Thk report is one of (be results of a joint marketing and do rolopomnl effort of both iba Goon and Hycar groups. Use pacific goal war la dovelop a powder compous of Geon resin and powdered Hycar nitrile robber for ime in a fleidble tbeneoplastfe ep licatian. The amtoemr, Coeley Products, equeried that die carepownd conlain thirty weight percent nitrile rubber. The first compound war based an Hycar 1422, a proem Indeed powdered nitrila dmt w limmlnpod far PVC mndlWrUnn. 0 * 2> When this compound was extruded in die customer's pient It ran neoeddy el the beginning of the deterred by this problem and requested drat we develop o cotepound which had a slows fw^d iv HU m IW ID NDCf DRCD ml^^ DDWCMQ nW OIWi * (heneel degradation. Thorofore, ere initiated a study to develop a moaeure of rale of degradation. Once diis wc* done, vie dren itudled die effect els of the study. Conelusions I Stability (DTS) tost con be weed to of < 2. The pracetsing seioty relative to e FVC compeend containing liquid mono meric plastic leer eon he determined by the ratio of hath the imbender steady stele league values end dm dneompositlon times of the rubber modified com* pound In the central. ^0019^2 BFG26330 2. 3. The rate of thermal degradation during processing ralotiv* to the PVC com pound containing liquid monomark plaiticizor con bo determined by compar ing the rote of torque increase end the rate of temperature increase of the rubber modified compound in the degradation region of the curve, 01 relative to the control compounds. 4. This OTS lest can be used In compound development to accurately determine the effects of compound 'rariationt os well as rubber changes on the processability of the blend. As a result, a compound wra developed by the Geon group and extruded successfully at both Avon trite and in the customer's plant 5. The use of thh technique revealed that compound processing safety was invessoly related to the viscosity of tho polyblond after fluxing. 6. The coeipound viscosity after fluxing was largely controlled by the viscosity of the nitrile rubber portion of tho bland. 7. The data indicated that the processing safety and degradation rates wore greatly improved by using low Mooney (30 or less) Hycar nitrile elastomers. 8. This study generated a now profeet aimed at do raloping powdered farms of very low Mooney viscosity nitrile elastomers far use m WC compounds. ra > __ a T1RPH RETIQII Continue to use fee OTS data to study fee effects of compound variations on pro cessability. Generate festoon malt theology data la study feo sffecfc of compound variations on feo overall flow bohmrioc of feeoe materials. PMMMH W MPPlity Of pMpVMg HI nOWW| powwi Of low moonay nyco iipinopwpi imcnHow A number of powdered dry Wossfe of Goon resin ao4 Hyoor nitrile aara prepared according to fee recipes rheum In Table I. Thera sratortali mesa then used In fee feabor Dynamic Thoitool Stability (PTS) tests under feu oondlHoas oatflnod fe Table It. The i hnngos bi bath tMgue mi stack taaperaturc ware nootdod during each DTS tun. This iOH pHCHI IOHV*mOnl O PO nj^H gsngm ram^fe mmri Pf PO 1^00 ^Ow ** subjected to OTS task to determine feelr effect on prooemfeg safety. * )0 T 9 G e Z BFG26331 TABLE t BASE COMPOUND RECFES* COMPOUND Geon 92 DDP Hycor Powder T ri Octyl Plwphot* Virryzene MG** Antimony Oxid* Calcium Carbonat* Stabilizer Lubricant PVC CONTROL 100 60 10 5 9 21 Variable Variable RUOBa MOOIFIED 100 26 72 10 5 9 21 Variable Variable Customer Con idontial ** Fungicid* 3. TABLE 1 AKfOa CONDITIONS FO DYNAMIC THOMAL STAMUTY EWOtIMENTS Roller Type BO Gm wMi turning 23561006 BFG26332 4. Results and Discussion Interpretation of OTS Test Data The Brobender Dynamic Thermal Stability test (DTS) i* frequently used during the development of PVC compound*. The data obtained from tbit fast are shown in Figure 1 and consist of plots of torque and stock temperature vs time. The tost results can be separated into three distinct phases as tha mixing time Increases. The first phase involves the fluxing of the compound ingredients into a uniform materia!. This is characterized by a steodily increasing temperature and a maximum in tha torque plots. Whan the fluxing stop is complete the material entors a itoady state zone of constant torque and tempera ture. Whan tha stabilizing systam is depleted, the compound then enters the last phase, the decomposition zona. This phase is characterized by amskad discoloration of the stock ond a steadily increasing torque and tompeiature. When this test was to the PVC contra) compound, (first recipe in Table i) we obtained the ciyoctad pattern of temperature and torque as disarmed above. However, when die rubber modified compound (second recipe in Table 1 with 72 parts Hycar 1432) wQI ralmKi| era O^WraB Pra ^I^RV^Rstra QITWRnOHs 1. Tha steady state torque values ware such greater than those of tha PVC 2. The duration of dm slaady state saw was eerch goallor. 3. The stock temperature hseraoeod during dm steady state tesqw region. 4. The slopes of bath dm tooquo-theo and tssspseetuso-Hmo curves In the I, dmrefare, docidod to quawHHss those differences to dateneine H msy of dm diffwencm would eamofofte wMh dse prahlem which erase during dm extortion run in the customer's pleat. The mautfc ware oalcslotsd by die following pracadura. and Hem liatioilad by M, T, end t respectively. The values of those parassotom fo dto Mdlf difo Mffoa oarry the subscript te, while these in the decomposition 23561007 BFG26333 TEMPERATURE 5. R.UXRMG STEADY STATE DECOMPOSITION TORQUE 23561008 TIME TYPICAL DYNAMIC THERMALSTAMUTY OATA FKMgf 1 BFG26334 A. The following informotion is recorded in the steady stote region. Note, I have defined the steady state region as one of constant torque. o. t,, - the time to reach steady slate b. My - the torque at steody state c. Tm - stock temperature at the onset of steady state likewise the following dote is tabulated for the decomposition region: a. tj - the time where the decomposition begins as evidenced by the increase of the torque above the steady slate value. b. Tj - the temperature at which the decomposition starts, i.e.the temperature at Id. c. Kate of change of torque with tiew dM^/dt. The slope of the linear torque-tfaw curve during decomposition. d. Kate of change of temperature with time, dTj/dt. The slope of the temperature Hem curve in the decomposition sane. The last stop is to calculate the following parameters: a. The time in die steady state regiom t^ - t * At. b. The rate of change of temperature during steady slats. _f- . T. T. * u - . Those paraswton were calculated far each of the compounds mentioned above and era tabulated la Table M. The JVC control compound required a long time to reach steady stale. If toptod to decompose at 37 minutes and toon entered a lass severe mode of decomposition at 41 sriautos. On toe other hand, the rubber modified compound behaved quite Mmtljf. It seochad steady state fairly quickly but entered the decom position sene sblr nine (f) minutes of total running thee vs 37 far the control. The torque at steady Mala eras four and a half Hews (4-1/2) that of too torque of the PVC control. This torque is directly proportional to the viscosity of toe stock vtotch in aim is proportional to shear induced heeling. The tewparatura at staedy state is 2tPf grantor than the central, and more importantly rises at a rale of almost flvs (5) degrees par minute during the constant torque period. All of these obserratlens Indicate that rubber modified compound k loss stable than too contra! because of toe peater melt viscosity end the resulting increase In stack temperatora duo to shorn hooting. 39808 23561009 JVNAMCplCRMALSTAMLITYTEST ANALYSE FIGURE 2 BFG26336 23561010 23561011 TAKE III 9. THe behovior of both compounds during decomposition is also illuminating . The control compound goes through two relatively mild decomposition stages. The rubber modified compound undergoes a strong decomposition in which the rote of torque increase is more than ten (10) times that of the control compound and the temperature increase is three timet that of the control. Therefore, bosed on the above discussion, this method for analyzing OTS data provides us with several indicoton of both the relative stability of the compounds plus the rote of decomposition. The suggested indicaton are shown in Table N. The two indicators of relative blend stability are in good agreement and show that the rubber modi fied blend has one quarter of the stability or processing nfety of the control compound. The decomposition rote indicaton show that the rubber modified compound degraded much more rapidly than the control. Therefore, the compound extruded at the customer's plant had two mayor shortcomings: it hod only ont quotfot of thi piocisinQ loftty of a com parable compound which contained liquid asonomoric plasticizer; and once it started to degrade, it decomposed at a tenfold greater rate. Since we felt that the above indicators gave us an occurate pichire of the process ing safety of flexible PVC blonds, we decided to uee this interpretation of DTS testing data in our Joint effort to derelop a loss sensitive compound. TABLE N INDICATORS Of BLEND THERMAL STAM.ITY AND DECOMPOSITION BATE Sample Control Bnlartre Stability fodlcalaw loloHro Decomposition Rate Indicators Tarqra* Thee to Decamp.* Torque Omngpe Temp. Change0 111-17-25-2 111-17-25-1 0.22 O^t 10.3 3 Steady state torque of central / Steady stale torque of compound b ^ eempored/*4 central e t%ll) ti--ptend / (dM^*) central d OI4/O/) dMpMnd / (dTj/dt) control ZT0T9SSZ 39808 BFG26338 ! C. DTS Evoluotion of Selected Stabilizer Systems The Geon group mode several stabilizer changes in the base recipe. The OTS acj'a are compiled in Toble V. It can be readily seen that none of these stabilizers pro duced significant reductions of the steady state torque and the resulting shear heating during steady state. The S-8 stabilizer did slow down the rate of torque increase (dM^/dt) by a factor of three (3). While this is significant, we did not feel that this decrease *os sufficient to warrant further evaluation given the reduced tj values of the S-8 stabilized recipes. DTS Analysis of Effects of Type and Level of Hycor Nitrile Rubber All of the previous wodc was done using Hycar 1422 powder in the recipe. This polymer was specifically designed to give good color stability in PVC compounds,'' and is also precross I inked to decrease the amount of nerve in the final compound. We decided to see if switching to a linear rubber would change the decomposition charac teristics of the final compound. We looked at this linear polymer, Hycor 1452P-50,ot two levels; 72 ports which would give the same amount of rubber os in the 1422 recipe, and 48 ports which should give the some compound hordness as the 72 part Hycor 1422 recipe.^) The results of the OTS analysis are given in Table VI; and the processing indicators are shown in Table VII. Thera is an interesting apparent contradiction in the data. First, the use of the linear polymer gives a significant reduction in steady state torque. This reducrion is increased when the lower level of I452P-S0 is used. This data is in agree ment with the data on the tkne to decomposition as shown in Table VII. However, once decomposition starts, the compounds containing the linear rubber decompose at one and a half to two times that of the coerpound containing the crassIirked rubber. This apparent contradiction is readily explained by the fact that the cross!inked 1422 was designed to give good thermal protectfan to PVC compounds. Therafora, it degrades at a lower rate than the linear rubber compound sdtich was not designed for that purpose. This means that we have two (2) alternatives to reduce the rate of decomposition and dtear hooting in the compounds. We eon eMwrmdce a color itoble uncrossltrked version of the Hycor 1422, or increase the ovetoH dWtlze system in the compounds containing the Hycor 1452P-50. Improvement fa PVSfahftfaotlon System Since it was easier to make a compounding change then to develop a new powdered rubber, we decided to improve the compound stability end use the linear Hycar 1452P-50 as the nitrile portion of the blend. The Geon group, therefore, made several modifications of the stabiltaetlon systems. To be thorough, we evaluated their recommsnded stabilisa tion iyitem in compoundi containing both the linear and crossI Irked powdered Hycar elastomers. $FG26339 23561013 OTS ANALYSIS OF STABILIZER VARIATIONS W0T95C2 o 1 tO fN Jk 11 3 o o -- o fto* $ a> y Q < a ! *5 ! -a t/l / rv o 8O to <> o no o o S -o o 5 i _a to ** r' oMil -- :n > MW* l 1 R i S c r * 1l }! T KK 11. BFG26340 T 12. 1 BFG26341 OTS ANALYSIS O f EFFECTS OF TYPE AND LEVEL OF HYCAR RUBBER 13. The dato ore given in Table VIII and the PVC control is also included "or comparative purposes. The results shoe that this high proportion of lead stobilizer-iubricant improved the compounds containing Hycar 1422 only a small amount. However, when this new stabilization system wos used in the Hycar 1452P-50 compound, several positive changes were obtained, first, the iteody state torque decreased by twenty five (25) percent, probably because of the extra lubricant. The decomposition behavior become complex with the first decomposition zone being followed by a second steady state zone before the final decomposition region was entered. This could reflect the action of eoch of the lead stabilizers. However, even more importantly, the first decomposition region is quite com parable with that for the PVC control. Therefore, the data indicated that this highly stabilized recipe (111 -17-139-7) would hove processing characteristics only slightly in ferior to the PVC control compound. A large sample of this dry blend was prepared for extrusion tests at Avon Lake. The extrusion trial on the 2-1/2" NRM unit was successful. The trial was stopped at the request of the customer because he wanted to run the remainder of the dry blend at his plant. The customer's plant evaluation was also successful. Thaw successful extrusion runs clearly demonstrate the value of the DTS test in predicting pro cessing safety of rubber modified PVC. Reduction of Compound Viscosity of Hycar 1422 Modified Compounds The successful extrusion of the compound mentioned in the previous section not the end of the protect, however, because bs Hycar 1457-50 was not as readily dry btendeble with the Geon coeipound. The bask reason far this is Oral the Hycar 1452P-50 consists of little chunks of rubber with on average particle size of 500 microns. When this material is dry blended with Geon 92 resin the rubber particles eon be easily teen with the naked eye.'*' Therefore, see decided that we would have to use e spray dried powdered rubber because of its Moeller (50 micron) ultimate powder particle size. These particles cannot be detected in the final dry blend. Since Hycar 1422 is a spray 4*od product, we decided to me if we could farther reduce the viscosity of dwre mopounds by replocing part of the powdered Hycar with liquid Hycar 1312. A series of dry blends wot prepared with different ratios of 1422 to 1312 as shewn in Table IX. The pseceming indicators are given in Table X. A comparison of the first three (3) compoinrh in Table IX dtowi that dm fallowing changes occur whan the ratio of Hycar 1312 to Hycar 1422 is increased die steady IvgN decreases, the time to ilegradaHon increases, and both the rates of torque and |j*pq--jfpi Increara during degradation decrease. These changes ore in X. If r^s^s^s t^^^st l^^^t ^^^tt^ts ^rr^r ^tr l^sss than those of dm fVC cMHol efas one fourth or mere of the powdered rubber is replaced by liquid rubber, the rheological Implications of these results will be discussed i.i the next section. 23561016 39808 BFG26342 DTS ANALYSIS OF RUBIER AND STABILIZER VARIATIONS e a 1 11 = .S 8~ I:i = i I <3 ^ a 1 J Ji > - Qail e .o u -r K H* O BFG26343 OTS ANALYSIS - EFFECT OF LIQUK) RUBBER O N COMFOUNO PROCESSABILITY % 15. BFG26344 23561018 PROCESSING IN D IC A OMIS f COMPOUNDS BASED O N U Q U AND POMJERED HYCAR ELASTOMERS 16. BFG26345 23661019 17 The practical application of these findings to the development of a safe processing compound for Cooley Products caused new problems. Because of the sticky nature of the liquid rubber, it could not be added directly to the dry bland. It was diluted with the plasticizer to thin it out prior to adding it to the Henschel. However, a sticky powder resulted. Therefore, we tried to incorporate the liquid rubber in the powdered rubber by blending the latices prior to spray drying. The blend containing 25% liquid rubber spray dried to give a marginal powder. The 50% blend was much too tacky and did not make an acceptable powder. The 25% liquid rubber powder did not give an acceptable dry blend with the Geon resin because of the "wetness" of the dry blend. Therefore, we will have to come up with another method of preparing a low viscosity powder to improve the processing safety of rubber modified Geon compounds. Rheological Considerations During Compound Development A comparison of the processing safety indicators in Tables IV. VII and X show that there was a definite inverse relationship between the compound steady state torque and the time at which thermal degradation occurred. This observation points to the importance of shear induced heating during the extrusion process. The data in this report also indicates that the shear heating is influenced matfcedly by the viscosity of the rubber used in the blend recipe. We have previously shown that the viscosity of rasitv/rvbbar blends is a fanction of the volume fraction of nitrile rubber in the blend, W and studies are currently in progress to menerrs the effect of rubber Mooney viscosity on blend flow properties. Wa can, howerrar, obtain an approximate picture of the Mooney effects by exom ining the data in Table IX in light of the data reported by P. H. Stanner.'^) This data showed that the Mooney Viscosity of a Hycar 1091 was reduced from 72 to 37 when it was blended with 25 parts of Hycar 1312. Ha farther reported that blenA containing more than 33% of ii^sid Hycar 1312 ware very toft and sticky and difficult to handle. This is in agreement with our observation that powdarad rubbers containing 25% 1312 ware marginal with respect to stickiness end powders based on 50% liquid rubbers were too sticky. It is ski jMdUl l adu on estimate of the Mooney viscosity of various solidliquid rubber Mfcdii Mg dta following equation: ^Mooney Hand - ^ ^ Mooney, dj >J Moonay2 whom and d 2 ** yolUNW fractions of components 1 & 2 39808 BFG26346 23561020 4' 18. If the vokw of the viscosity of lU Hycor 1422 is 81 and wo assume that do Mooney viscosity of the liquid rubber is soro (0), the 50/50 blond will have a Moonoy value of 20. The 79/29 blond would hove a value of 45. Those values are too low to be spray dried into an acceptable powder. Therefore, we have generated a new problem, namely to prepare a powdered from of a low viscosity nitrile rubber which contains the stabilization system of Hycor 1422. Efforts to prepare such a material are currently being evaluated. References 1. D. G. Frazer and M. E. Woods, Development Technical Report, "Thermal Stability of Powdered Goan Hycor Polyblends III The Influence of Hycor Polymerization Ingredients'*, April, 1971. 2. D. G. Frazer, M. E. Woods, and W. R. Ssento, Development Technical Report, "Thormol Stability of Powdered Goon - Hycor Nitrile Pofyblends Pf The Effect of Hycor Spray Owing Compound Ingredients", SoptenAer, 1971. 3. M. E. Woods and 0. G. frazor. Development Technical Report, The Modification of Goon WMi Powdered Hycor Nitrile Rubber III SPC Paper*, January, 1974. 4. G. Smell, Primte Communication. 5. P. H. Simmer, Ds'iolopmont Technioai Report, "lepra red Handling of Hycor 1312 Liquid Rubber I - Evaluation of Various Methods", March, 1974. I want to erirwmriedge dm contributions of dm fallowing Individuals to Ibis feint J. Gallo M - 4 Kdkic # Ml W. dlhftttnglon - Geon Technical Group n|W ISVI_W1_Vtn__^ Goan Mmbating Geon Technioai Hycor Technical