Document QXbXVeZQJ4j8dLwjnbQQOBZvk

B. F Goodrich Chemical Company A DiVISION OF THE B. F. GOODRICH COMPANY DEVELOPMENT CENTER Staff Technical Report No. 126 4 a/ THE EVOLUTION OF(piBUTYLTINDICHLORIDE AND HC1 ) FROM RIGID PVC_COMPOUNDS STABILIZED WITH DIBUTYLTfNBIS(OCTYL THIOGLYCOLATE) By L.B. Crider J .E. Hartitz Closure Report for Project 2894-266-1 Distribution: Date Completed: June 22, 1966 Date Issued: Akron Gen. Chem. Fiant *R.N. Rylands-R.S. Reynolds Akron Legal Department E.K. Bean Develo pment Cente *L.F. Arnold M. Batiuk L .A. Benne tt *R.R. Bloor L .H. Conklin E .A. CoL1 ins L.E. Curry >C. E. Fleming J.E . Krause R. M. Kreager C.H. Lufter *v. c. Mast *R. J. Meyer J.R. McMillan C.E . Parks W.A. Reed G.G. Rothman J.W. Ryan N.H. Sherwood *J.A. TePas *G . L . Whee Lock *R.J. Wo L f T .'baical Count l1 C.T.F. O') Brecksvillfe Research Center V.L. Folt *J.E. Jansen D.E. Ley E.B. Newton *W.L. Semon J.J. Shipman Calvert City *W.E. Brodine Cleveland *W.F. Bixby R.J. Coffey *G.A. Fowles E.W. Harrington D.L. Kent M.W. Larson (3) *P.H. Lawrence R.E. Lynn *J.F. Maione *G.H. Metzger W.D. Parrish R.D. Scott-A. Vittone, Jr. B.M.G. Zwicker-J. Valentin Henry *C.B. Cooper-J.P. Piers Long Beach *A.R. Webber-W.D. F bb Louisville *L.G. Crunklatcr S.S. Michels Niagara Falls *T.R. Linak 20590001 BFG04105 TABLE OF CONTENTS page OBJECTIVES ............................................................................................ ........ 1 SUMMARY AND CONCLUSIONS............................................................. I FUTURE WORK...................................................................................... 2 INTRODUCTION ................................................................................................................ ... 3 DISCUSSION AND RESULTS ...................................................................................................... 5 BIBLIOGRAPHY...................................................................................... 11 APPENDIX.............................................................................................................................................. 1.2 20590002 i BFG04106 OBJECTIVE The Long range objective of this work is to improve the thermal stability and reduce the porosity of extruded pipe made from rigid PVC compounds stabilized with dibutyltinthioglycolate (S-8). Our immediate objective is to develop a method for measuring the relative rates at which dibutyltindichloride and HC1 are volatilized at elevated temperatures and to determine whether a relationship exists between porosity and d ibutyl t i1- d ichloride volatilization. The development of a sensitive -md reprodu-. ibl" method for measuring the rates at. which these components are volatilized would sLsu provide a useful tool for studing the syr.ergisti. effects of other compounding ingredients in PVC compounds sttbilized with organoti-thioglveolates. SUMMARY AND CONCLUSIONS The use of dibutyltin bis (octyl thi: glycolate) as a prc-cessi-g stabilizer in PVC serves two functions. It acts as a chlorine scavenger and it also apparently does some patching of the sites of unsaturation in the polymer back bone created by dechlorination. Chlorine split off by thermal decomposition reacts with the stabilizer to yield dibutytindichloride (DBTDC) plus an octyl thioglycolate radical. The dibutyltir.dichloride is evolved from the PVC compound as a volatile product which can be readily measured by mass spectrometry. The failure to observe octyl t h iogly :ola te as a volatile product is strong evidence that this radical must become attached to the polymer backbone. At elevated temperatures (450F.) the initial evoLution of PCI a volatile product is coincident with a sharp decline in the evolution of DBTDC. This event apparently signals the depletion of the effective st^bili: in the compound. The evolution of DBTDC from a typical rigid PVC compound is detectable after 15 seconds of exposure at 450F. The rate of volatilization is sharply increased at about 5 minutes and reaches a maximum at about 6 1/2 minutes. A sharp decrease in the rate for DBTDC is always accompanied by a sharp in:rnin the rate for HC1. Increasing the amount of stabilizer in the sample incrc ;ses thtime at which the maximum for DBTDC occurs. A similar effect can he noted by increasing the amount of calcium stearate in the compound. BFG04W7 20590003 -2 The results of this study did not yield any conclusive evidence that HC1 or DBTDC evolution is responsible for porosity in extruded pipe. It is suggested that calcium stearate may have a deleterious effect on both stability and porosity. It acts as an HCl scanvnnger but not as a polymer 1 stabilizer since the acid radical i produced by HCl attack does not I attach itself to the polymer ( backbone but is volatilized as i stearic acid, ii Ion Current (a rb itra ry u n its) Typical rates of volatilization for DBTDC and HCl from a PVC pipe ii compound (Geon 8750) are shown in the figure at the left. Time (Minutes) Figure A FUTURE WORK Further exploitation of this method for studying the synergestic effects of other compounding ingredients in rigid PVC is now in progress. Our initial efforts will be concentrated on further measurements of rates for HCl and DBTDC volatilization as a function of stabilizer concentration. These measurements will be made in the absence of lubricants. Similar rates will also be measured for other organotin stabilizers of the type [Bu2 Sn Y2] but containing different Y groups. These same tests will then be repeated with the addition of lubricants. These data, combined with experimental extrusion results will hopefully reveal some of the interplay that must occur between the ,olymer, stabilizer and lubricant at extrusion temperatures. 20590004 b?G0^08 -3- INTRODUCTION The recent introduction of a new concept in screw design for extruders used in the processing of rigid PVC cubes has greatly increased the rate at which pipe compounds can be extruded. At these higher rates the stock temperatures may be in excess of 450F. With our standard pipe compounds, such a Geon 8750 and 8759, our customers have experienced difficulties in obtaining quality pipe and phenomena suchas porosity and chicken tracking have forced them to lower output rates. The problem of porosity was rationalized as being related to the formation of volatile materials in the stock during the exposure to these higher temperatures in the extruder. At the beginning of this investigation our limited objective was to identify the materials that were being volatilized. Much to our surprise we learned that the initial evolution of volatiles was very low in HC1 and high in dibutlytindichloride. The emphasis in our work on this problem was then shifted to measuring the relative rates at which HC1 and dibutyltindichloride are volatilized with the hope that these results would reflect the cause of porosity. The dibutyltindichloride is formed in rigid PVC compounds from the reaction of chlorine with the dibutyltin bis (octyl thioglycolate) stabilizer. The failure to observe octyl thioglycolate as a volatile product from this reaction strongly supports the proposition that this radical must become attached to the polymer chain at the site vacated by the chlorine; \ H- C-H \ H-C- Cl H-C-H plus \ H-C- Cl / H-C-H \ H-C- Cl / H-C-H Bu Bu \/ Sn /\ ch2 I C-0 ' 0 I Oct ch2 I C=0 I 0 I Oct s H-C-H \ 0 II H-C- S--CH2--C-0-0ct / H-C-H \ + H-C-S--CH0--C-0-0ct / H-C-H \ 0 H-C- Cl / H-C-H \ Bu Eu \/ Sn /\ Cl Cl The work of Frye, Horst and Paliobagis(lH2) (3)0n the chemistry of organotin stabilizers (using radioactive tracer techniques) has done much to illucidate the role that these compounds play in the stabiliza ion of PVC. 20590005 BG0^9 -4- A set cf experiments was designed to show whether P7C is butylated by organotin stabilizers of the type [Bu2 Sn Y] under conditions normally used to study the thermal degradation of the polymer. Using tags in the Cl position of the butyl groups it was showa^' that it is unlikely that any butylation of the polymer occurs. Tagging the Y groups with suggested'3) that the stabilizing action of the organotin compounds arises from reactions wherein the polymers chlorine atoms are exchanged for the stabilizers Y groups. H3gn tagged compounds showed that most of the Sn is lost due to degradative reactior-3. The retention of a small amount of radioactivity was rationalized as being due to the existence of a coordinative linkage formed between the stabilizer's Sn atom and the donor atom (Cl) present in the polymer^ molecular structure. With this background, it appeared that the ability to measure the relative rates at which dibutytindichloride and E01 are volatilized would give us some further insight on the effects of adding*more (or less) stabilizer and of varying the amount and types of lubricants used in a rigid PVC compound. An attempt could be made to correlate these rate curves with porosity. The difficulties in doing this were apparent before we started, however. First of all, the volatilization rate is conditioned by two important factors i.e., the rate at which the volatile molecules are formed and the rate at which they are diffused out of the compound cube or particle. Reproducible results would then be dependant cm using cubes that have exactly the same physical dimensions. Cube-to-cube variations in composition (inhomcgeneity) presents another problem. In spite of these limitations, the results obtained from our initial measurement of rates of volatilization for HC1 and dibutyltindichloride show that the method is sensitive to stabilizer concentration and other differences in rigid compound formulation. 9000G90Z DISCUSSION AND RESULTS The procedure used to study the volatilization of dibutyltindichloride and HC1 in rigid PVC cubes is our standard mass spectrometer method for the analysis of volatiles in polymers. Since this procedure is quite well known I shall only describe it briefly in this report. The sample is contained in a small glaas bottle (0.5 dram). Since the cubes from production material are not uniform in size, some effort was made to select cubes that were 1/8" square. Eight cubes of this size are placed in the sample container which is then inserted into the vacuum chamber that has been preheated to 450F. The temperature is measured on the walls of the chamber and is not a measurement of the stock temperature. One minute is allowed to evacuate the chamber and to pre-heat the sample. The valve between the sample chamber and the ion source is then opened and a mass spectrum is scanned every 15 seconds. Separate sample runs are required for the measurement of the volatilization rates for HC1 and dibutyltirdichloride.' Both components ;ar r,ot be monitored from a single scan at 15 second intervals because of the gross difference in molecular weights. The ions that are monitored for HC1 sr.d dibutyltindichloride are monocomponent so that direct plots can be made of ion current vs. heating time. The identification of the ions is unequivocal since these are based on the natural abundance of the 37ci/35ci ratio for HC1 and on the combined Sn- Cl isotopes for dibutyltindichloride. (Calculating the relative abundance of the ions containing both tin and chlorine turned out to be a bit of a task because of the muliplicity of the Sn isotopes. This required writing a computer program to give us the reLati-vc abundance of ions containing SnCl SnCl2, SnClj and SnCl4.) Plotting the amount of ion current for each component at 15 second intervals gives a relative measurement of the rate of volatilization. To make this result quantitative would require a complete analyses of the total spectrum and a heat loss measurement at each 15 second interval. This amount of effort did not seem warranted since a relative rate can describe what is occuring in the sample cube. A comparison of relative rates of volatilization for HC1 a-.d dibutyltindichloride (DBTDC) from samples of Vyram 3000, Geon 8750 and Geon 8759 are shown in Figure 1, 2 and 3. 20590007 -6- Ion C urrent (a rb itra ry u n its ) ! j Ion C urrent (a rb itra ry u n its ) Time (Minutes) Figure i Time (Minutes) Figure 2 In all three cases, the evolution of HCI is retarded until the maximum for DBTDC is reached. This suggests that this event is coincident with the depletion of the effective stabilizer in the sample. The intensity of the maximum also reflects the concentration of tin stabilizer in the original sample. Figure 3 BFG04112 20590008 X-ray fluorescence analyses(4) for !n in Vyram 3000 and Geon 8759 are shown in Appendix I [run by Crobaugh Laboratories in Cleveland] The maxima for DBTDC in these samples reflects the higher Sn content of the Geon 8759: Sn Wt.% DBTDC Max. Vyram 3000 0.18 2950 Geon 8759 0.38 7000 -7- [The reproducibility of the results for Sn obtained by X-ray fluorescence is remarkable and suggests that this instrument could be used to study the rate of volatilization of DBTDC on a quantitative basis] The reproducibility of the rate curves for HC1 and DBTDC are shown in Figure 4 and 5. 20590009 BFG04H3 -1 Although there are some c ' . . tr. curve ? fc taiv:u fr -? rhrec the overall reproduribiTi ty is. good enough to reflect tLc gr ' ' 3 '. fe re. nee in composition between Vyram .3000 and Ceon 8759. The differ between Geon 8750 ;nd Geon 8759 world not b. as c1:arly c 'i-.d. .= - The he.it Loss values measured at tAe t: of the eight mi :* ."tpoH'tc at i50`>F. are as follows: ,.-i -n y -jn ;A 2 Ron A3 to hO Vyram 3000 Wt ,7. 3.40 3.23 V-e: U 750 W t. , :v 5. r 0 A , 95 f .06 ' 0: - - 8 W- , ' 1-/.t 1 J>, V u , ,.'4 4 , 39 The lowc r heat loss value iov v ^ ;m r 000 is. pa.rt, r a 1 ` d by the- l owe. r tin at. ibilioer content . Priv: f t u bear less rhiT.-it.s-. ' a- J i 5 O U ue to the volatilization of pre> . O S ? 1 -g aid r'd other omp >. i-g ingred unts. I.r. r iw becomes of iris rVbt t .) dr : :r. i. ne w^rth.-t i. b p r 7 of the so other compounding ingredit i.t n Kl.S , i ` 1 u-nee oi the r 1 i--. . < o u rve s for HCL and IBjCC through some syr 1, S t 1 ! ffects , Our efforts to measure tK f C fc ' - of V 1 : : ium s t ea t ~t te ! 3) i he tv lutive volatilization rates tor tICl - -d r- " gave resul t- f L a t. we r not roproducible. It is believed that t> : r la k of reproducibiliry is related to lack of homogeneity in the test c ,'v *. A comparison of the rates from duplicate runs are shown in Figures 6 a.ri 7 f r samples of Vyram 3000 (control) and from samples of Vyram 3000 containing an additional L part and 2 parts of calcium stearate. Although the results are rot reproducible, they do show that the addition of calcium stearate retards the evolution of DBTDC and PCI. The results irom run if! show that each additional part of calcium stearate added advances the maximum for DBTDC at least a full minute and that a corresponding effect is also evident for HC1. 20590010 BFG04114 -9- (a rb itra ry u n its) Time (Minutes) ^ ___ __________ Time (Minutes) Figure 6 Fi g-j - - ~ This data would support the theory that calcium stearate has some stabilizing effect and would act to conserve the tin stabilizer. The results from Run #1 would be difficult to interpret other than to suggest that there was poor mixing of the L-3 with the Vyram in preparing the samples on the mill. The suggested stabilizing effect of calcium stearate at milling temperatures may also have some adverse effects as related to porosity. The evolution of stearic acid as a volatile product is most evident in PVC compounds containing calcium stearate. The reaction of chlorine or HC1 with calcium stearate is unlike that with the tin stabilizer in that the acid radical that is split off does not patch the vacant site in the polymer backbone but is expelled as a volatile product. It's value as a stabilizer is limited to the single function of acting as a chlorine scavenger. One can visualize this as having a deleterious effect since the damage to the polymer backbone is not repairjas would have occured if the chlorine had reacted with the tin stabilizer. This school of thought is a good arguement for the proponents that calcium stearate can create problems of porosity and poor stability at the temperature encountered at high extrusion rates. A comparison of the relative rate curves for Vyram 3000 and Vyram 3000 plus 1 part of N,N' ethylene bis stearamide (L-9) shows an effect similar to calcium stearate. This data is shown in figure 8. The addition of 1 part of L-9 retards the evolution of both HC1 and DBTDC about 45 seconds. 20590011 BFG04115 __ Time (Minu te e) Figure 8 In summary, the results of this initial effort to use mass spectrometry to study the volatilization rates for HC1 and DBTDC has furnished much con firming evidence about the chemistry of dibutyltin bis(octylthioglycolate) as a stabilizer in PVC. We strongly support the findings of Frye, et.al., that the Y groups become attached to the polymer backbone and that the dibutyl tin is evolved as the dichloride. There is no evidence to support the claim of Kenyon^) that stabilization of PVC by tin compounds of the type [Bu2 Sn Y2] must be due to the ability of the stabilizer to donate a butyl group to a polymer radical. If this were true, one would expect that the volatile materials would contain some [Cl2-Sn-Y2] or [BuCl-Sn-Y2] and none can be found. The mass spectra is also absent of Bu-Sn-Cl^ and SnCl^. The demonstrated ability of the method to reflect the addition of lubricants in PVC compounds certainly warrants our continued exploratation of this technique in studying the synergestic effects of other.compounding ingredients with organic tin stabilizers. ACKNOWLEDGMENTS The authors wish to acknowledge the contributions of John Nikora for recording and tabulating the mass spectral data and to J.R. Smith for writing the computer program to calculate the relative abundance of ions containing tin and chlorine. BFG04116 20590012 BIBLIOGRAPHY Frye, Alfred H., Horst, Raymond W,,, and Paliobagis, Mark:* A,, "The Chemistry of Poly (vinyl chloride) Stabilisation, III, Organ-tin Stabilizers Having Radioactively Tagged Alkvl Groups", S', r, prlyir.. v Science, Part A., VoI. 2, p 1765-1784 (1964) F l ye . Alfred H ., Horst, Raymond W.., a <->. Poly (vi-y: chloride) Stabil i zat.ic - . '-.n ing Radio -.'lively Tagged Tin Atoms." .1. 2, p 1785-1799 (1964) Pal i b igis Mark A., "The Ch. mi..** t IV . Crganoti- S t ab v 1 i re ra J,, "f Polymer Sr ier*.--, Par* A, Frye, Alfred F., Horst, Raymond W,, ard Pali -:gis, Chemistry of Paly (vinyl chloride) Stabilize:ion V. Stabilizers Having Radioactively Tagged Cr ps," Science, Part A, Vol. 2, p 1801-1814 ('.964) Mark : A., "The Organotin J. of Polymer Laboratory Report R-7621, Crobangh Labor-utorit *, Clevela-d, Ohio, Ar-ril 6 1966 (attached as Appendix I) Kenyon, A.S,. Natl. Bur. Standards (7,8.) Cir No 525 91 (1953) 20590013 BFG04117 Laboratory Report CROBAUGH LABORATORIES RESEARCH ANALYSIS TESTING 3800 PERKINS AVENUE CLEVELAND 14. OHIO (AAKA COOK 2161 UTAH 1-7320 To, B. P. Goodrich Chemical Company Development Center Moore & Walker Roads Avon Lake, Ohio 1*4012 Rapvrtina l)u.... April 6, 1966 0. g 7621 Dale Keo-ived. March 21, 1966 7' Material................PVC Discs Marked .......... Please see below P.O. No. 136 AL-33-2 [Note: These identifications were not known to Crobaugh Laboratories] REPORT Sample No. Geon 87# 1.5 pt. TM-180 jon 87# 2.5 pt. CC-11 jeon 8759 2.5 pt. T-31 Vyram 3000 Vyram 3000 Geon 8759 2.5 pt. T-31 Goon 8759 2.5 pt. CC-11 Geon 8759 1.5 pt. TM-180 Geon 8759 2.5 pt. TM-180 71-1 71-2 7 L-3 71-4 71-5 71-6 71-7 71-8 71-9 Geon 3759 2.5 pt. T-31 Vyram 3000 Geon 8759 2.5 pt. TM-180 71-11 71-12 % Tin in P.V.C. 0.23 0.47 0.38 0.18 0.18 0.38 0.48 0.23 0.45 0.38 0.18 0.45 Respectfully submitted, CROBAUGH LABORATORIES Alex Sltkin BfG04l V8 20590014 0 r-4 1