Document RjmpEwoaKDYe4q1EEwDBYE0wV

BFG TECHNICAL DOCUMENT -/ BFC CHEMICAL CO.ANALYTICAL . . r * , !>c . ' SIT" O* f| REPORT k 04 030- m* / iP 1 1 J. V. N'UORA tc e* Al.TC/PSC * * ")* : . i mo o > c ' so *C * * * R. M. BURTON BRECKSV I LLF. 3 PHO1F6 v n r-.-------------------------- ------------------------------------------------------------- GC/MS ANALYSES ASSOCIATED WITH TEMPREL Qy HOSE FUMES A problem at BFG Marion involving Temprel hoe stock . n,, which causes throat and eye irritation has existed since a 1"rrt 1976 (see attachment). In 1976, the fumes generated during the manufacture of the Temprel hose were investigated by the Environ mental Health Department. While the fumes wore not specifically characterised, certain observations were reported by the .investi gators. Diallyl phthalate (DAP), triallyl isocyanurate (TAIC) which has since been replaced by triallyl trinellitate (TATN) and peroxide related materials could be detected in air saaples. Howevei belief was expressed (see attachment) that no simile Temprel ingredient, when heated, could produce the irritating tunes associated with Temprel stock. GC studies related the fumes to poroxide decomposi tion products and to allyl phthalate but these could not be confirmed in plant samples. Norm recently, customers using Temprel hoses have complained about the irritating fumes resulting from the catting of these ' materials with hot knives which can approach the pyrolytic decom position tsmperature of these materials. These considerations and the feme problem at BFC Marion resulted in e pleased study of the Tsmprel stock floats and the ferns geeerttsd from the ingredients see* likely to ho responsible with T6A-DTA and GC/MS. S. N. Newell etOdn definite lessoning of the fmmt whom Ohiflex 300 (a polyglycol . eetar derived from a mixture of 1,2-propane diol (major), 1,3-butanedftaS* 1,4-butnno diol and adipic acid'T)) waa not uaod xn the Teaprel A seri studie*. formulations with dicumyl peroxide, lead mollitste. Uniflex 300, diallyl phthalate, tod polyotkyloao were prepared to cover tho ingredients uaod in Teaprel hoses recipe. Of those, three compounded saaples, and tho plasticisers with and withoxaained by GC/MS-headspace techniques, aaterials identified in these GC/MS- The emftfxsls of the UUiflex 300 natorial.will he reported sepOBBtmiy by J. 1. Dorsch based on lJC NNt measurements (ear attachment). J. C. MeCool communication to J. A. Nikere, 1/21/tB. C O r * N.J s*' Trr (is Xnalyscs Associated with Temprcl Hose Fumes experimental The following compounds were heated in standard, evacuated headspace vials at 180C for 20 minutes: CPE BFM-1* BFM-3b TATMC Uniflex 300d DAP* DQA* 3 100 so 12 3.6 10 -- _7 _ 100 30 12 3.6 10 -- _1_ 1G u 30 12 10 Ml lead silicate, 181 Desical and 221 CPE. 491 DiCup (401 dicunylperoxide on calcium carbonate) and 601 CPE. c 741 triallrl Militate on Microcel. ^ Polyester based on a .fixture of diols (prinarily 1,2-propane diol) and adipic acid(*J. Diallyl pbtbalate. { Di-2-ethylberyl adipate. In addition to these compounds, the two MSterbatches, BFM-1 and IM-3, aad the throe plasticisers with and without dicunyl per oxide were exaaiaed under the sane conditions. Pure dicunyl peroxide was obtained by extracting DiCup with diethyl ether aad renoring the ether by rotary evaporation. (Interestingly, too anch 1FM-3 wasterbatch in a headspace vial ruptured the septun at 1MC and the escaping fane* ganoidlad Iron the dicunyl peroxide were irritating to ton. none, throat. and oyoa.) niante-lMC heating period, the as See gas tight syringe and injected onto poched with It/109 neeh Porapak OS. chrenategraphy were identified froa by conparieon with reference spectra. A series of K/W total ion current profiles (TICP's) were obtaiMd free the hdndepace noses geMrated at 1B0C. Pigusoa 1 and 2 show the notorials evolved from BFM-1 and BFM-S. The only volatiles expected frost BW-3, which contains dicunyl peroxide and chloriMted polyethylene, were expected to rosult fron the peroxide decoaposition. The following were identified: ?' O > o '1' ;MS Analyses Associated with Tcmpr cl Hose Kumcs Results and Discuss ions - (continued) methane ethane water methylchloride acetaldehyde iso-butene glycolaldehyde acetone (CPE?) methylethyl ketone ben7ene toluene C2-ben'enes Cj-benrenes a-methyl styrene 4 -benrenes acetophenone At tines, ethylene was also evident. The aldehydes and acetophenone seeaed to be the aajor irri tants. This became evident when the pressure generated at 180C ruptured the headspace vial's septum causing the gases to escape into the laboratory. Bye. nose, and throat irritation required the evacuation of the laboratory for about IS ainutes. The volatile gases from BFM-1 (Figure 1) at 180C can not be readily explained. None were expected since this aasterbatch con tains lead silicate, Desical and chlorinated polyethylene. BFM-i was analyzed several times and even after steam stripping the coluan and obtaining a water-blank TICP, the same volatile materials were evident from this saaple. These materials indicated that BFM-1 :ontained some dicumyl peroxide (possibly from preparing the aaterial oa the Brabeader). Figores 5-7 are the total ion current profiles obtained froa different plasticizers with and without star, was shown by 13 IM to be based am a mixture of dials (primarily 1,2-propanediol with 1,3 volatile When nixed were aateTials present: tentatively identified'as diallyi ether. When the allylphthalate was mixed uith the cnmyl peroxids, am new products were observed. What was obtained (Figaro i) was the superiaposition of the pre viously identified canyl peroxide volatile coaponents over those of the allylphthalate. O o u M'.-se Tunes oMilta and Discussions - (continued) Figure " shows the 180C volatile materials obtained from j mixture of the plasticizer dioctyl adipate with cumyl peroxide. In addition to the cumyl peroxide decomposition products, only .-ethylhexene and a small quantity of heptene and octene were identifiable. (The heating of these materials in vacuum and the nature of the analysis does not allow the clear distinction netweer. 2-ethylhexene and 2-ethylhexanol, which could dehydrate under these conditions.) Figures 8-10 present the GC/MS TICP's obtained from the headspace gases of compounds 3, T_, and 10_. Surprisingly, the volatile gases obtained from the tJTree plasticizer systems (with the excep tion of 2-ethylhexene) were no longer present at readily detectable levels. What was now found were the volatile materials related solely to the decomposition products of the dicuayl peroxide used in these systems. The following table presents a summary of the materials identified by these GC/MS-headspace techniques and the samples in which they were identified. In leaking at these results, the most obnoxious materials sees to be associated with the dicumyl peroxide decomposition pro ducts and the acrolein and allyl alcohol from the diallyl phthalate plasticizer. Since most of the irritating fumes are released during the lead stripping operation, it any ba possible to alleviate most of the eye and throat irritation problems with the installation of more ventilation in the stripping and milling areas. Should these problems persist, the ALTC GC/MS laboratory has three small vacuum pumps (lL/nia) for large scale area monitoring using charcoal and/or Taaax adsorbents to collect and concentrate organic vapors. John A. Nikora JAN/dm* Attachments cc: C. H. Lufter - P. Zakriski (BRC) b ti wriwr S. w. Newell (Merlon) R. F. Boebel J. Metserly (IRC) J. L. Dorsch R. Verge (BRC) J. B. Fausch - R. P. Lattlner (IRC) J. C. MeCool (Akron) B. B. Katzemeeyer, Jr. * H. W. Dietz (Akron) J. G. Quiseaberry ci O  -1 G O O O .'.O 1  '*1 1 1 1 1 llllllllllllllll r u i-^ 14 X -a / 4I * lid MMRI3 w I jo uiiioj qjiM poaBi.^ossv sos.\|ru\ s;  -I