Document LpJv0MaBQ0YNbRGvx2MbBVB5b

formation of polychlorinated DIBENZO-p-DTOXINS AND DIBENZOFURANS FROM PYROLYSIS OF SILICONE DIELECTRIC Fl.L'lD SPIKED WITH POLYCHLORINATED BIPHENYLS by Jon V, Swanson Dow Corning Corporation and Thonaa 0, Tlarnan, Ph.D, Bratua Laboratory Wright Stata Unlvaralty For Presentation At The EPRI PCDF Workahop December 6, 1964 HONS 215548 1 INTRODUCTION Sine* 1982 Dow Corning Corporation ha* had a prograa of ratrofllllng PCB containing transformer*, ualng It* alllcona fluid RatroSll1" procaaa. Tha concarn* over the cosibuatlon of PCS'a haa lead to the daalra to batter understand the by-producta formed during the pyrolyala of alllcona flulda containing email amounts of PCS'a which may remain after the RetroSll"" procaaa. Thus a contract to study this matter was Initiated with the Brehn Laboratory of Wright State University under the direction of Dr. Thoma* o. Tlernan. The study had 2 objectives. One was to determine amount a and apecles of polychlorinated dlbenzodloxlns (PCDD'a) and polychlorinated dlbenzofurana (PCDF's) formed, when a alllcona fluid aplked with various level* of PCB wa* pyrolyzed. In addition, tha effect of pyrolyala temperature on this aaaie ayatem waa to be evaluated. This paper dlacuaaea the reaults of theae studlea. "FTHODO!OCY All spiked samples were prepared at Dow Corning. Pvrolvsts studies were conducted at the Brshm Laboratory, at Wright State University under tha direction of Dr. Thomas Tlernan. Sample* of 50 centlatoke polydlmethylalloxane (DOW CORNING 561 Fluid) were aplked with 5, SO, 500 end 11,300 ppm of PCB from Arochlor 1260. The 11,300 ppm aaiaple waa Initially atudlad at 6 different temperatures to determine which temperature produced the maximum amount of PCDD'a and PCDF*a. Thla was determined to be 600*C. Then the aaaplea aplked with varloua amount* of PCB were pyrolyzed at thla temperature and tha by-producta analyzed. The pyrolyala apparatus la shown In Figure 1. It consisted of a high temperature tube furnace containing a quartz tub* reactor, a cooled sample Injection ayatem and a aeries of three traps to collect the volatile combustion products. Five ml, of sample ware used for each test with an average residence tlaw In the pyrolyala zone of two seconds. Table I summarizes the weights of the combustion products based upon the 5 ml (A,7 g,) sample Injected, The combustion products Include any water produced In tha pyrolysis reaction. Thus the total weight of the combustion products could, and lndaed did, at times exceed the amount of sample Injected. Preparation of Che pyrolyala products for analysis conslstad of soxhlat extraction in toluene for 16 hours and a series of washing steps ualng KOH, H_S0^ and anhydrous sodium sulfate, concluding with a hexane elution through a column packed with allies, and then concentrated by evaporation of the hexane, Thla waa then transferred to a laboratory prepared liquid chromatographv column packed with Woelm basic alumina which was previously activated at 600'C. After conducting a series of methylene chlorlde/hexane elutlona, the samples were evaporated tn dryness and stored In a freezer until analvsla. Theaa samples were then reconstituted by adding trldacane before injection of the appropriate allqnot into the analytical teat Instrument. MONS 215549 Analyses of the extract vets done using a Perkin Elmer Sigma III Caa Chromatograph coupled through a custom fabricated Interface Including a single-stage glaaa Jet separator to a Kvato* MS-25 Mass Spectrometer equipped with a DS-50SM Data Syatem. RESULTS OF ANALYSES The results of the analyses of the combustion products from the pvrolvsis tests for chlorinated dlbenzo-o-dloxina (CDDs) and dlbentofurana (CDF*) are summarized In Tables II and III. Also shorn ara the results of analyses of a method blank, of a sample of polydimtthylalloxane containing 11,300 ppm of PCBa (which was not pyrolyted), and a sample of the polydimethylilloxane containing 130 ppm PCB which had been previously subjected to electrical arcing, Aa can be saen from the data presented In Table IT, the Initial aerlaa of teats, in which samples of polvdimethylsiloxane fluid containing 11,300 ppm of PCBa ware pyrolyred at temperatures of 100*C, 600*C, B00*C and 1000*C respectively, revealed th.rt c' r 1. at ed di.ber.;ofuror, was formed in the i00* end COO' tests. in particular, the tetra-, pent a- and hexa-CDFe wera produced ar these temperatures. All data is presented in the concentration of total CDD $ CDFs In nanograme/gram of total combustion products or Perce Per Billion. Minimum Detectable Concentrations are given In parentheaaa. While the pretence of several laomera of each of cheat chlorinated elaaeaa can ba discerned from the mass chromatograms obtained In the analyses, It la noteworthy that the tetra-CDFa appear to Include 2, 3, 7, B-TCDF (or at laaat a TCDF Isomer which eoelutea with the 2, 3, 7. 8-TCDF). No CDDa other then 0CDD were detected In the combuetlon products from any of thaaa Initial testa. There were essentially no detectable CDFa In the products from the B00*C end 1000*C pyrolyeea. It was clear from the Initial series of pyrolyele teats that the quantities of CDFs produced from pyrolyele of the PCB-contamlnated silicons fluid were greater when the pyrolyele temperature waa 600*C. After these reaulce had been reported to Dow Corning and reviewed, the Brehm Laboratory was requested to conduct severs! additional pyrolyele testa in which each of three samples of the silicons fluid containing 500 ppm, 50 ppm end 5 ppm. respectively, of PCBa, were to be pyrolyted at a temperature of 600*C, The reaulta of analyzing the combustion products from these tests for CDDs/CDFa era' shown in Table III. As these data indicate, tecta-, pants- and hexa-CDFs were detected in all of these combustion products, although the quantities of these CDFs war* clearly dlrmttly related to the concentration of PCBa In the fluid pyrolyted. That la, the greater the concentration of PCBa in the field, the greater the concentrations of CDFa In the pyrolysis products. This trend Is, of course, sa expected If the PCBa ara the precursor* of the CDFs which are formed in the combuition process. Table III aleo ehow* that the second series of tests trmnd no CDDs (except some OCPP, which mv be questionable) when the silicone fluids containing 500, 50 and 5 ppm of PCBs were pvrolvzed. The analytical results obtained for the polydtmethvlsiloxane fluid containing 11,300 ppm PCB* which was not pyrolyted, a* well ae for the fluid sample containing PCBa which had bean previously subjected to electrical arcing, indicate MOMS 215550 - 3- that theae samples contain essentially no detectable CDDa/CDFa. It will be noted that the detection Halts reported for theae aaaplee are for the fluid samples theaaslves (and not for coabuatlon products), since theae saaplea were not pyrolyted. _ CONCLUSION Pyrolysis studies of polydlmethylalloxana spiked with various levels of PCB show that no polychlorinated dlbenzo-p-dloxlna are detected but that tetra, pents and hexa apeclea of chlorinated dlbentofurana are detected. A sample of PCB spiked polvdlmethylelloxsne fluid which had been subject to electrical arcing showed no detectable PCDO's or PCDF's formed. Maximum formation of PCDF's appeared to occur at 60Q*C, MOMS 215551 ( VI HI I I 10 HOOD PYROLYSIS TESTING APPARATUS CMA*CO*l ua'V - t*tf I*- &tm --|- MtCOOLM <Hm-T ttWCx/QUMTZ OWMtTZ t tMKOSML CCMIWTKM SO.Stw TMCX'IOMC kwutu TCMJUTunc tlM CUMACC I- nci* -h 10m -I N.COOIM SCI MJCCTMt OCThl KCOW v*pom^Kw ZWC in^coolm INLtl LlOUCM, All* DETAIL Of SAMPVC INJECTION SYSTEM ulm* raw (1/2* S1 COOLING STWMOf MOUNTID 0CV.CCIN0inCA4.CT conmxtto Stmucc runr INO or MJCCTON LOCATCD VS WAT INTO VATONIZATKlN ioki iioo*ci 1 combustion CAS M4.IT m,i toscali APPARATUS FOR TESTING PYROLYSES OF SELECTED CHEMICALS FOR PRODUCTION OF CDOs/CDFs MQNS 215552 TABLE I WEIGHTS OF COMBUSTION PRODUCTS FORMED FROM PYROLYSIS OF POLY DIMETHYLSILOXANE FLUIDS CONTAINING POLYCHLORINATED BIPHENYLS4 PCB Level (ppm) 11,300 Pyrolysis Temperature 400*C Total Weight of Combustion Products (gram*) 4.6 11,300 600* C 5.3 11.300 11,300 800* C 1000eC 4.6 5.1 500 600* C so 600*C 4.6 4.6 5 600 *C 4.5 ** In each experiment, 5.0 mL of the silicone fluid wore pyrolyzed, Since the density of the fluid le 0,9428 g/mL, 4.7 g. of fluid vert pyrolyzed In each ceee. The combustion product* Include any veter produced In the pyrolysis reaction, ee well ee incompletely pyrolyzed fluid In ceeea where combustion was not complete (lower temperatures). MOWS 215553 TABLE II THE EFFECT OF TEMPERATURE OH THE PYROLYSIS PRODUCTS OF PDMS SPIKED WITH 11,100 ppa PCB (Data In parts per billion) " Recmry Recovery Pyrolyala Teap. TCDFs TCDD* 2.3.7,8TCDD Internal Standard PCD Fa PCDDa HxCDFa RxCDDe HpCDFs HpCDDe of 37C1 - 1.2.3,4?6, 7,8- HpCDD Internal Standard OCDF OCDD Recovery of 13C12-OCDD Internal Standard Method N.D. N.D. 542 N.D. N.D. N.D. N.D. N.D. N.D. 412 N.D. N.D. 562 Blank (0.10) (0.05) (0.20) (0.20) (0.40) (0.30) (0.20) (0.20) (0.40) (0.10) Not Pyro- N.D. N.D. lyted (0.10) (0.45) 612 N.D. N.D. N.D. (2.0) (2.0) (2.5) N.D. (2.5) N.D. (2.5) N.D. (2.5) 622 N.D. N.D. (2.5) (14) 482 400* C 3,670 N.D. (2.5) 782 1,260 N.D. 1,600 N.D. N.D. N.D. (17) (10) (7.5) (7.5) 512 N.D. 10 (44) 412 600 *C 46,235 N.D. 682 11.500 N.D. 3.600 N.D. N.D. N.D. 822 N.D. 8.0 812 (24) (20) (10) (10) (10) (30) 800*C N.D. (4.0) N.D. (4.0) 502 N.D, N.D. N.D. (5.0) (10) (25) N.D. (15) N.D. (5.0) N.D. (5.0) 882 N.D. N.D. 812 (42) (10) 1000'C N.D. (2.0) N.D. (2.0) 732 N.D. N.D. N.D. (10) (10) (20) N.D. (20) N.D. (10) N.D. (5.0) 592 N.D. 12 (44) 4 52 HONS 215554 TABLE III PTROLTSIS PRODUCTS OF PPMS SPIKED WITH VARIOUS CONCEHTRATIOHS OF PCB (Data In parts per billion) ' PCB Level (PP) Pyrolysis Teap. TCDFs 11,300 600 *C 46,235 Recovery TCDOa of 37C1. 2.3.7.Ri TCDD Internal Standard PCDF* N.D. (24) 68Z 11,500 PCDDa HxCDFs N.D. (20) 3600 500 600*C 2,778 N.D. (7.0) 60Z 4 SO N.D. (15) 142 50 600*C 138 N.D. (4.0) 5 12 N.D. (4.0) N.D, (30) 5 600 *C 12 N.D. (5.0) 63X 2.0 N.D. N.D. (4.0) (10) Arced Hot Pyro- N.D. N.D. 53X N.D. N.D. N.D. lyied (12) (20) (4.0) (10) (25) HxCDOs N.D. (10) N.D. (15) N.D. (30) N.D. (10) N.D. (25) HpCDFe N.D. (10) N.D. (10) N.D. (10) N.D. (10) N.D. (12) Recovery HpCDDa of "ci_ 1.2.3,4.6. 7,8- HpCDD Internal Standard N.D. (10) 82Z N.D. (10) 100Z N.D. (10) 86Z N.D. (10) 71X N.D. (12) 17Z i MOWS 215555