Document joKGrpkGwNvR0Ld0RaJyOyOZ

INTERNAL CORRESPONDENCE CHEMICALS ANO PLASTICS To(Nn) R. N. Wheeler Chemicals & Plastics u>atton South Charleston, W. Va Copy to W. B. Ackart - BB/98 L. E. Brydla - BB/200 0. M. Garty - BB/200 R. A, Gregory - BB/200 G. T. Kwlatkowskl - BB/200 R. W. Lasher - NY/33 G. S. Peacock - BB/98 F. G. Wllleboordse - BB/98 RIVER ROAD, BOUND BROOK, NEW JERSEY 08805 April 2, 1979 Ortflrutlng o*pt. Answarff* Subnet d*t R&D Determination of VCM In PVC Resins The Items discussed In our phone conversation this morning are attached. The reprint of the article from the Toumal of the Association of Official Analytical Chemists emphasizes the necessity of confirming peak assignments. I have underlined sections on pages 813, 615, 816, 818, and 819 which state the importance of a mass spectrometer In proving the presence of residual monomer. It should be noted that this article,which was published by Food and Drug Admini stration personnel, discusses vinyl chloride homopolymers that are less complex than copolymers. The other attachment Is a partial listing of compounds noted in the literature as being ones that could Interfere in the determination of residual monomer In vinyl chloride homopolymer or copolymers. This data was obtained from the American Society for Testing and Materials compilations and various scientific journals. If we can be of any further help In this matter, let us know. TBG/fp Att. T. B. Gibb UCC 061588 of the Food The Nctheril suggestions il>t. lit at the re* ennary Pub- nc. Of. Anal. C A H, k pi. Cham. 35, . (1973) Z. (7-140 ngesat, J. M. (51 075) 12th Ed., as E., Vshaft. 'rth. Inti. Pat- if. OS- Anal. urn. N. L., A ir. OS- Anal. i try II. 1171. m DSNNIMN MT AL 1. AMOC. OTT ANAL. CHBM (VOL. ft. NO. 4. INI) \ INDUSTRIAL CHEMICALS 813 Headspace Sampling and Gas>Solid Chromatographic Deter mination and Confirmation of 21 ppb Vinyl Chloride Residues in Polyvinyl Chloride Food Packaging J. LAWRENCE DENNISON, CHARLES V. BREDER, TIMOTHY McNEAL, ROGER C. SNYDER, JOHN A. ROACH, and JAMES A. SPHON Food and Dmg Admimttratton, Division oj Chemiitry and Phytic*, Wathmgton, DC tOtOi The determination and confirmation of real vapor phase*. A large aliquot of the headapace dual vinyl chloride (VC) in polyvinyl chloride (PVC) food |tackafinf materiala arc daacrikrd. PVC packafinf material* ara dlaaolvcd tat*# acthylaeetamide (DMAC). VC la sparged from elution with helium gaa and collected In aealad viala of ethanol. VC la detected and quantitated hy oaing a headapace earnpiinf technique and was withdrawn for G8C or GSC-mnsa apectrometne (MS) injection. Apparent lenaitivity haa been further increated by concentrating VC from the polymer aolution into another tolvent, using a sparging technique. Thia waa auggested by Beliar and atandard gat-olid chromatography (CSC) with Liehtenberg (8) and hae lieen uaed by Japaneae Same ionisation detection. CSC peak height! of mientiati* for aeveial years. ahoot 5% foil scale deflection are obtained for We have modified the sparging and headspace *ample* of PVC containing 1 ppb VC Polymer sampling technique* to obtain the described pro tamplet taken from taking, blood bags, food cedure* which permit the determination and packaging films, bottles, and nnproeeaeed reain confirmation of 1 ppb VC in PVC material*. were analysed. Reaidual VC level* ranged from OJ to 913 ppb. SEC oea confirmed hr CSC-coim -Aplymer sample* of tubing, bottles, blood bags, ontmamry, m*ng aeJacted ion rrcerdina food packaging films, and unprocessed PVC ,aa/ M ana o4 m conjontiion ajdjJgBj# resins were analysed using flame ionisation de tection (FID) and! The detection and quantitation of vinyl chlo ride (VC) in various matrices have been con tinuing problem* for the past several year*. Since the implication of VC as a cause of angioMrtoma of the liver (1) and several other dis eases (2), there has existed a need for a highly sensitive and specific analytical method for de termining and confirming residual VC in PVC food packaging materials. Previous work in this laboratory (3) was limited to quantitative de termination of VC from PVC solutions at the 1 ppm level by direct gac-eolid chromatographic (GSC) injection. Recent work haa shown that headapace sam pling techniques can increase sensitivities to the tow ppb range (4-7). PyC materials or PVC solutions were held in sealed viala and heated for fast equilibration between liquid or solid and METHOD Caution: Since VC i* a gas at room temperature and ha* been designsted carcinogen (1. 9. 10) by inhaistion, special handling precaution* must be taken to preserve the integrity of the standsrds as well as to ensure the safety of the analyst. Only the septasealed viala deecrilied under Apparalu* were found satisfactory for handling VC. Itearam* (a) Vinyl chloride.--(7) 3 lb.cylinder (99.9% pure) (Matheeon Gaa Product*, BBSS Amherton Dr, Dorsey, MD 31337, or equivalent). (2) VC gaa atandard (0.48 ppm v/v) containing 135 ng VC/mi nitrogen (MG Scientific Cases, Kearny. NJ 07039, or equivalent). (b) Ethanol.--Abaolutc (Publicker Industries, Inc, 1439 Walnut 8t, Philadelphia. PA 19102, or equivalent). Before use. test to show absence of oo<eiuting materials at retention time of VC under condition* of analyst*. C. V. Hfntor twina H*f*n* an ladinet AiMitiv** From Fowl FiHtoam. Thin pmiiar wu iimmud ftt tht flit Annuo) Mooting ml Ite AQAC. Oct. IT-*, 1977, it Woofci^ton, DC. 1 Anntytpui mrtboH attarhe'd r- Dor if, IOTA k*tfrr from (hmiflfw ftokrtitt Co.. Ltd. Ml Part At*. N>* York. NY IMM. to llonrtng Clerk. Kool iml Drug AdiiMiuatrottoo, MM Fmhtvo Lm, Rmkvtllo. MD MU 0004-5756/78/6104-0813302.10 Association of Official Analytical Chemisti, Inc. DKXNIHON KT AI. 1. ASHOT OFF, ANAL CHEM. (VOL. M. NO. 1W) y)tV ' J?l; j '-V-J v . Fta. |--Co d*tii: Iwagwial cap m**nind fr*m oim km* (1>A ast-St-Ust), Mir*ad*d M Mm* unflgurotMa w glass **tt1*. OrMM, and Uepd In Hi* t*w t* KcanwnOtl* SA In. lw*(*Mi kwitnstd Attlng. (C) Dimethylncetamtdt (DMAC).--3 kg bottle (Kastman Prartieal Grade, Eastman Chemical Product*, PO Box 431, Kingsport, TN 37663. or equivalent). Pre-sparge entire reagent bottle with helium (or 3-3 davs before uae. ttefore ine, tt t to show alwenee of co-chiting material* at retention time of VC under rendition* of analysis. 4/t|MrM (a) Scrtu^eapped boltlet.--I ox Barrow mouth, 3505 ml (Ace Scientific Supply Co., 1430 Ea*t Linden Ave, Linden. NJ 07036, No. 104256. or equivalent). 8imilar bottle* with 2 ox capacity. Cap* with Teflon-lined epta (Alltech Associate*. 303 Campu* Dr. Arim(ton Height*, IL >0004, No. >533, or equivalent). (b> Gat tyring*.--3, 5. and 10 ml (Preeinon Sampling Corp., PO Box 15110, Baton Rouge, LA 70815, Prwaure-Lok, Sene* A*3, or equivalent). (c) Mim-Vialt--5 ml. with cap* and Teflonlined aepta (Alltech AmocmUx, Inc., No, 9500, or equivalent). (d) Collection i*nl._Pcrkin-EImcr F-42 headH|wec annlyxer vials with gray aepta anil aluminum exit. Capper (P-E No. 105-0100) and deenpper (P-E No. 1054)107) for via!*, or equivalent. (a) Sparging apparatus.--Set deaoriptive draw ing! in Figa 1 and 3. For temporary u*c and pre liminary text*, a etandard metal cap may Ire uaad by removing liner, drilling hole through top, re moving Airfare paint with acetone, and mft-eoldering in a bulkhead fitting with sparge tube attached. Sparge tulte i* attached to bulkhead fitting by drilling hole slightly larger than in. downward at angle into liorr of tilting (ee Kig. 1). Plata of nut portion and upper major part of threada should be avoided. After sparge tulic i* inserted to proper length (enough to reaeh liottoni o( bottle), tube is silver-soldered in plarc. Thick gasket is then cut to fit inside o( rap. Threads on standard metal cap* are not sturdy enough for continual use and often strip oil after a short time. Plastic raps are unsatisfactory for even trm|x>rary usage. (f) Forced droit oven.--Precision Scientific Co, 3737 W Cortland St. Chicago, IL 60747, Thelco Model 18, or equivalent. (g) Got chromatograph. _ Hewlett-Packard Model 7630A, or rquivileot. equipped with tem perature programmer and flame ionisation detec tors (using OAl in, id flam* jets). Operating con ditions : Temperatures (*C)--injection port 300, detector 180, column 110 (after VC elution (ea 7 nun). |>mgrim column to 180*C at fastest rste anil hold for 30 min): gas flows (ml nun)--helium car rier .75 (0.7 rotameter), helium makeup ,70, oxygen 345 (335 rotameter), hydrogen 35 (0.75 rotameter); electrometer setting. 3 X 10-* amp full scale (AFS). Rowed Anew 21. IS7T. Ampted K-Hmsry It. tv?*. F n * o :Jr DEXNISON *f AL i. ASSOC- OKK. ANA).. ( HEM. (VOI. (I. NO. . Ifpll S -1 at SIS im and soft-eolder:c tube attached bred fitting by ,a in. downward 1'i'g 11 Flat* of ( thread* ahould .sorted to proper of bottle), tube gasket i then n standard metal intmual uee and Plastic cape are ; usage on Scientific Co., II, 50747, Thelco Hewlett-Packard nipped with tero- loniiation detec-1, Operatin* eonijoction port 300, VC elutioo (ca 7 nt faatrat rate and mini--helium enrmtikcup 30, oxygen I tO.75 rotameter); amp full acale I fhniafy 1W Imjltn rial PfTAlL* no. i--Iperga tram Oetn: mqulrea 7 X It In. auMeaa ateal bMig U M It la mgpsr tiaMng, ana m tap pa i*we In ns- 1. ; mint > ft X M in. (h) CkromntognphU column--4 ft x 44 in. od and calculate VC concentration either in w/w or coiled ftainlea* ateel, filled with 80-100 meah Pora- in w,'v unite (ca 50 ppm) This aoiution ia similarly pak ; retention time of VC ca 7 min ---------------- - dilutes! to yield solution containing 1-5 ppm. Make (0 &* cAronnU>9Tanh-mam tpaelrometer-t final dilution# into 10 ml ethanol in collection vial. min Model 3300K ele_ ctron impact quadnipol* Prepare working standard* in 1-50 pph range. If ___~jipertrmnmmeetteerr eeqquuiippppeedd wwiitthh FFiinnttuuggaann 4866000> refrigerated. these working standard* are stable for gas chromatoernuh and FinnTgan Model 6100 data * up to 1 week. Multiple septum puncture* ahorten i eastern Kull mam ecao* ami aelteted ion recording (Vorking life of atandards. pi m/s S3 and Of were oEtam*3"under computer | Determine actual VC concentration in head- cakbol. fj>nenitinit conditions" Temperatorm i space of shove working atandards under conditions --injector 170, column 155, separator 315, traasfer 1 of analysts by direct comparison to standard gas HEJI0JsPcr_Vt!_^PP14jr^iiuu'Jlsr_vcnfrT^^i (0.48 ppm tn nitrogen) containing 128 ng VC/ml ^uerammed to 190*C to Hear solvent from col- ^ Calculate average partition coefficient, uaing the am): helium earner gas 35 ml'min: aamplc size following equation m ml heedsnac* ga* at Mat unicrj Partition coefficient = 0, fC| (j) Pint bottles.--Flint glam, IS o* narrow mouth, perew^ap (Fiaher Scientific Co., 7733 Fen - (0.13/lt ml)/(057/10 ml) =0.11 (/) where C, = concentration in vapor (headspace); ton St. Silver Spring, MD 30810, No. 2-883DD, or C, = concentration in liquid; 0.13 = I3r/r deter equivalent), with foil-lined caps mined to he in headspace; 14 ml = volume of Preparation o/ Standards headspace (13 ml) plus volume of syringe (1 ml); 057 = 87% determined by difference to remain in Accurately weigh 3 os narrow mouth bottle, liquid; 10 mi = volume of ethanol standard. Mini-Vial cap, and ceptum. Add 50 nil abaolut* ethanol to weighed bottle. Cap, and reweigh. In hood, prepare VC stock aoiution in this bottle by Preparation of Soiree!* end Semple* Check alieolutc etlinnol to be used in collection quickly uncapping bottle and adding 05-3 g vials for materials eo-eluting with VC hy carefully liquid VC from inverted freesor-cooled cylinder pi|iettiBg 10.00 ml into one of the vials and sealing f VC. Immediately cap bottle tightly and mix the vial with vial eap|>cr. Analyse as described ell by disking. Rcwcigh and calculate VC concen under Annlynr. tration by weight (c* 10.000-40,000 ppm). Dilute Drill 3 holes in cap of 3 kg reagent bottle of this stock solution by witlidrawing aliquots through dimcthylnectaniide (DMAC) and run 44 m aeptum with svringc and injecting into weighed, Teflon or-stainlcm steel tube to Imttnm of bottle slot! container (1 os bottle) of ethanol. Rewcigh, through 1 of the holes. Sparge DMAC, using slow lI4 I 'I (i ucc 061591 a * 816 DENNISON cr Al: l. ASSOC. OFF. ANAL. CHEM (VOL. M. NO. . Ifft) flow of helium (5--15 ml/mio) for seversl dsy* st room temperature tieforc uae. After DMAC free of matcrnita co-eluting at retention time for VC lint lieen oLI:iiii<hI (ilrtmnmnl try analyzing re agent blank), mill 200 ml to I pt fmttie along with 2 in. Teflon-conteil magnetic itirnng liar. Cut thick PVC aamplea such aa tubing and bottles into small piecca which will fit through neck of narrow mouth pint bottlea. Prepare film by rolling aample into long narrow tube which ran be cut into ahort arctiona to fit into bottle. Accurately weigh 20.00 g PVC material and add to bottle containing 200 ml DMAC and stirring bar. Tightly cap bottle, uaing cap with foil liner, and atir content* until PVC it completely dieaolved. Complete aolution requires atimng from 2 hr to overnight, depending on PVC formulation involved. Highly plasticized samples dissolve more readily than do more rigid smnples. (Note Opaque solution* are generally obtained when the PVC composition contains impact modifiers or pig ments.) Analyiit After PVC sample is dissolved in pre-sparged DMAC. add 10.00 ml absolute ethanol to etch of 2 collection vials. Interconnect vials with short lengths of Ms in. stainless steel tubing as shown in Fig. 2, and cool in thick Dry Ice-ethanol shiny for ,10 min before uae. Fit liottie containing PVC solution with tight cap equipped with Ms in. stain less steel sparge tube extending to bottom of bottle. The apimratua ta shown in Fig. 1. Immerse pint bottle containing PVC aolution in steam bath and add helium aiiarge gaa via needle valve through M* in. atainleaa steel sparge tube at ca fifi ml.'min. Quickly leak check apparatus by applying Snoop (Potomac Valve and Fitting, Inc., 1500 E Jefferson St, Rockville, MD 20852) to alt joint* and briefly plug final outlet tut>c to build up jireaeuri-. After leek cheeking, allow sparge gaa to freely pass through PVC aulution, through bulk head fitting, into effluent tulie and into collection vials immersed in Dry Ice-ethanol slurry. Sparging apimratua is shown in Fig. 2. After sparging for 2-2V4 hr, disconnect collection vials while still cold, decap using decapper, and remeal using fresh unpunctured septum. Analyte vials immediately or store under refrigeration for not more than 2 days. For analysis, heat vials to 90*C in either water bath or forced draft oven and let equilibrate at this temperature for 1 hr. Heat gas-tight syringe having iotegral valve to S0*C in forced-draft oven. Wear glovea to uae pre heated gas-tight syringe and withdraw 1.00 ml headapaee. Close syringe valve and return syringe containing sample to oven for S min to vaporise any condensed solvent. Remove hot syringe and immediately inject aample into gaa chromatograph. Make standard VC gas injection (1.00 ini, 1JS ng) Iwitli licforc and after cttrli lieailaiuicc aaiuplc injection. Quantitate unknown using the following equation: VC in PVC, ppb s total ng VC in aample/20 g = [(ng/ml in HS) x 14 ml + (ng/ml in liq.) x 10 mll/20 g <*) combined with Eq. (/>, C./C| = 0.I1, where C| = CV0.11, then VC in PVC. ppb = 1C, x 14 ml + (C./0.11) x 10 mll/20 g = IC.04 ml+ 10 ml/0.11) 1/20 g = 525 C, The remainder of the sample in the collection villa after GSC determination of residual VC is fcrf gcpzfflrwT&e nwmee o4 VC br Q8C-M8 Equilibrate samples at S0*C for 1 hr in constant temperature bath and withdraw 2.00 ml aliquot of headspace with preheated gas-tight syringe hiving integral valve. Inject sample quickly intoj^ggf fc> minimise cooling of syringe and condensa tion of vapor. IBwsTjriSrtfnitTaia. CitaUiTiatos 7W m/s OB and 84~from reanonne areas _ retention timt'3T"WTat3""^^ii5_wit^hS^r Results and Diseussion Previous unpublished work in this laboratory indicated that quantitative determinations of VC in PVC at very low levels would require a com bination of headapaee sampling and sparging. Headspace techniques have the advantage of al lowing larger portion* of the VC present in the sample to be injected into the gaa chromato graph than eould be done by direct injection of solution!. Undesirable components can often be reduced or eliminated. Unfortunately, the low ppb levels of VC in solid food )iaekaging mate rials could not be determined accurately by us ing headspace sampling alone. This is due to dif fering and unknown equilibration times from sample to sample. Direct headspace sampling has been shown to reliably detect low levels of VC in several food simulating solvents (6). If the polymer is dissolved in a suitable solvent, a favorable partitioning of VC into the headapaee may be obtained. Our desire to use a relatively dilute polymer solution to ensure rapid equilibra tion reaulted in a low total VC concentration in the heads) > ppb level i step w;ii in transferrin: trntion to i bination <>. technique' , DMAC i eause of itlow volatil. collection iDry Ice ten for VC. II because of The spnt available chosen for in 200 ml DMAC iniat the tlin machining r of expanse., in the cap eould be u lower levelquired. Fi-_ Fig. 2 shov GSC an PVC sani)n that VC tv other com] Silar 10C 1 columns, t Porapak T vided into I was mnrgu The l'oraji eellent re.-, interferenr. VC which tion. GSCnent was i columns, fi fore the FI for the nr\ turn-aroun eessful Th. tor and th tor were i interferenr tore were lack of re ucc 061592 VIDC i-sc- mn( ctive I'M! heo* DENNISOS XT Ah. J. AIWOC. OFF. ANAL. dlEM. (VOL. M. NO. . INI) 817 the headspace for samples containing VC at the ppb level. Therefore, m additional concentration step was needed. Sparging provided a means of transferring VC from a solution of low concen tration to one of higher concentration. The com bination of sparging and headspace sampling techniques produced the desired sensitivity. DMAC was chosen to dissolve the PVC be cause of its excellent solvent properties and its low volatility. Absolute ethanol was choeen as collection solvent liecsuse it remains a liquid at Dry lee temperature* and is an excellent solvent for VC. Helium was chosen as the sparge gas because of its low solubility in ethanol. The sparge train was designed to use readily available components, the pint bottle being ehosen for its convenience in holding 20 g PVC m 200 ml DMAC without entrainment of the DMAC into the gas stream. Some bottles broke at the threads, possibly due to overtightening, machining mismatch, or differences in coefficient of expansion between the glass and brata used in the cap construction. A larger bottle (1 qt) could be used for the determination of VC at lower levels where a larger sample would be re quired. Figure 1 shows the details of the cap and Fig. 2 shows the assembled sparging apparatus. CSC analysis of the headspaeee of several PVC samples, following sparging, demonstrated that VC could not be adequately rcaolved from other components by using Carbowax 20M or Silar IOC liquid pitases in conventional packed columns. Chromoeorb 101, Chromoeorb 108, Porapak T, porapak R, and Tenax GC also pro vided inadequate resolution. Chromoeorb 104 waa marginally useful but if not recommended. The Porapak N column waa (elected for ita ex cellent resolution of VC from all encountered interferences except one component eluting after VC which sometimes precluded baseline resolu tion. GSC-MS analysis showed that thi* compo nent was isobutane. With the porous polymer columns, from 45 min to 1 hr was required be fore the FID baseline was considered adequate for the next analysis. Attempts to shorten these turn-around times by bnckflushing were unsuc cessful The Hall electrolytic conductivity defec tor and the Dohrmann microeouiomctric detec tor were evaluated in an attempt to provide interference-free detection of VC. Three detec tors were difficult to use nnd suffered from a lack of reproducibility on a day-to-day basis. Since VC interferences were not eliminated by using these detectors, the FID remained the de tector of choice. We were unable to obtain gyringet capable of continued use at 90*C. Syringe leakage problem* were occasionally encountered due to sustained use at 90*C. Syringes were tested by pulling beck the plunger, closing the integral valve, and pushing on the plunger while bolding the syringe under water. A stream of bubbles indicated leak age and the syringe waa dacarded. Pre-eparging of the DMAC satisfactorily re moved poesible interference* contained therein. No attempts were made to clean up the absolute ethanol. If materials co-eluting with VC were present in the chrometogramt of the blank eth anol, a new container was opened and checked. In general, interference* were not a problem. Before attempting the analysis of unknown PVC materials, a complete analysis of a reagent blank should be obtained by using the sparging apparatus described. A known amount of VC should be sparged from the apparatus to check for recovery, leakage, and the CSC retention time of VC under operating conditions. The partition coefficient waa determined by analysing 17 VC-ethanol solution* rsnging in concentration from IS to 12 ppb (polymer basis). These solutions (10.00 ml) were prepared in collection vials, analysed under the conditions of the analysis, and quantitated by direct com parison to the etendard gas (0.4S ppm in N,) which contained 1.25 ng VC/ml, By using Eq. I, the partition coefficient was calculated to be 0.11 with a standard deviation of 0.015. Spiking and recovery studies at levels of 0.72(1 ppb VC in DMAC and in 10% PVC polymer solution! in DMAC were carried out for the en tire procedure. The avenge recovery for 6 an alyses of DMAC containing no PVC was 91% with a standard deviation of 3.7. The avenge recovery for 21 polymer eolutions containing 10% PVC was 93% with a standard deviation of II. 7. The PVC for the recovery studiea waa free of residual VC. (Note: PVC containing no VC was obtained by sevenl precipitations from tetnhydrofunn (THF) into methanol. About 50 g PVC from tubing or film was dissolved in a minimum amount of THF. This solution waa slowly poured into a vigorously stirred beaker containing methanol. The ntio of the THF solu tion to methanol was about 1:10. The precipi- i! ucc 061593 818 DEN\1RON ir 41.. i AaAOC, off. ANAL. CHEM. (VOL. II. NO. . Il) TiMa 1. naaldual VC IMM la variant FVC MffRti &ed*A WPbVC. Atttdefcvttf. aVCrcepowel Sampla Sampla daacriptnn VC lovM* PPb 1 rifid calandarad land grada anaat 1.2 2 thin plaaticizad food Mm 21 J thin plaaticizad food turn 21 4 plaaticizad Mood sag 1.2 i tnm plaaticizad copolymer toad film 2.1 i thin plaaticizad lood him * (7 plaaticizad load and milk tuMng p.l* I plaaticizad MvarafA tuPmc A. 7* 0 rigid vegetable oil bottla 120 10 ri|id Franon water bottla 111 11 Franch bottla molding compound 0.0* 12 thick plaaticizad canning ahaet 0.4* 13 plaaticizad blood bag 0.1* 14 raagant Monk NO* * Quantitated by tangent akimming VC pack and maaiunng paak haigni--aingie datarmination. * Eat mated value--balow mathad quantitation limit, ' Nona dataetad. ****-* avalut ill (Im *a< ton current profile! Id* readily apparent in the eelected ion cur-8 ^nt profile of th gampfc (JVSj , l hlwtnri inn recording of m/ ft? and 64 wiwM d40in1(d1 to nmuin/finmni fthher ntvttmn--1r Vt C. Hie riiaracJ i, jonotie 3:1 rotto of thcae iota at tim mrmioff ' rTl . 1 -VC iiPod' uec^'hr the nat! ural"abundance !C1 in the moiofubtr ton of W.* w. 1 the ococifteitv of the melhotT fhia rat# j i Jim wprodwMhlo to within 2<E nf the theoretical' g ^ I-- whea the level of raMdual a the* alymer woo 1 mik HKnitorine the m/t 8B aiig 68 refin etwtnef tin onfv compounds containing one chtattoe t toot end thuethc eletaentil composition t? vfayl chloride as r molecular ion or fragment iy f--ibjc interierpnees to the GSd'-MS confirms* CltfafoUetie, epichlomhydrin, and dichloO (ated PVC tvno filtered onto a large Buchner thine arc tow molecular weight compoumt- funnel and air-dnrd. For proceiling larger quin titles of PVC, a large blender vu used for the1 t produce a fragment at m/t 632 connttaaiinniinngg** ghjfinc atom and ire notenriar*mte3S- ' precipitation. Generally 2-3 precipitutiono pro i. Allude dt maaftm of epiehlentfrwdrar* vided clean PVC if the otarting PVC contained no more than about 1 ppm VC. A variety of PVC matenaIs were analyzed using the method described and reeults are given in Table 1. A typical FID chromatogram k shown in Fig. 3 along with the chromatogram of the reagent blank. Thirteen aamplet were analyzed and containec^|CaiJi!V^ningjng from 0.3 to 913 ppb/fcenttty ^waooztKnnea bet AaA apectrowetry. it proved invaluable intnc 'jKetnJmonroijiGf^Mgjljividcniifjjrjgc^J jwwentt eluting near tlw* retention time of vintg jfhtondc. Thii etnured in the final procedure that* iMre were noi eompooznientz ci o-efuting with _v_i_n_vT \ chiondti that would mvihdate the CSC quaflti- I j# tition. Full eeani (m/i 16-200) of 12. g f the wmtuica ihftmiitogrujlied ort Poeimk m j hf Wrrc obtained to identify the eomnonenti elut-1 am near vinvl chtoride. Figure ie the total ian j grrent Profile of a typical lample recorded fron# > tie elution rf carbon dioxide until the elutwjh , etlianol. The reagent blank and the tunnies $ interned cirbonvl ilflde. nronene. teeter, end . aeetaldchvdc. Qiloromcthanc. obutane. ai4l butene were identified el the mwipIA. MfPtificatWW of Ott eowipcmenti wai baaed art ipierpretatmn and eomnarieon with ref-* enence eunnwwda inalvted under identical eon-' dfcnt. The tuple mweoented by Fit. 4 con- Tim (mu) no. I Typical gaa anmmatagremai A, 1 a apaaa at raagant bianhj % 1 ad haadapata at lampla J cantalnlng 2.1 ape VC, palymar knh (0.4 ag VC 1--Tat t.l c ratal and dichlm after VC . elute liter The lac I mide qu.u Good quui control of i itruetion tnalyted ai ndered red O) MiJto, An 7 (J) Thom IJCC . 06^594 rtorw trofilc i cur- DSNNisoN ir al..- i. amoc. orr. anal. crsm. <vol. si. no. , i*s> m 84 *i 819 S4 P) : se ef m/s M 2.1 pee VC, P*>rm*t I i *m *t i SOW NLTBCT no. A--'Titol ton limn steMe at tempt* t hMMif M ppM VC, heala <& n* VC HntaS); rMmM Mm* it VC tweisstaS My tflvu. i ffichlorpethane dwnoMtwted that iter churn er VC. The heavier chlorobutenca would tbrf i later thaa VC (111. _ The lack of interfering eo-eiutinf materials nude quantitation by GSC-MS unnecemary. Good quantitation by GSC-MS require etrict control of the analytical condition! and the eonetruetion of an analytical curve from standards analysed sitnultanoualy. Such an effort eu eontidered redundant in this case. RaruttNcs* (1) Maltoni, C., A Lefemine, G. (1974) fnwo*. Am. 7. 357-405 (2) Thomas, L. B. Popper, H, Berk, P. D, Bali* koff, I, * Falk. H. (197S) Ntm ing. J. Med. 292,17-22 (3) Breder, C. V, Dennison, J. L, A Brown, M. . (197$) J. Amoc. Of. And. Chem. it, 1214-1220 (4) Rosen, J. D-, Moreno, J. R-, Parties, 8. R,, Giaein, J. It, A Gilbert, 8. G. (1975) /. Amoc Of, And. CAtm. $9. 700-702 ($) Van Lierop, J. B. A Suit, W. (1979) J. Chromatogr. 121,183-197 (0) Diachenko, G. W, Breder, C. V, Brown, M. E., A Denaiaon.J. L. (1977) J. Amoc. Of. And. Chem. 00, 570-57$ (7) Steichen, R. I. (1975) And. Chem. 41, 1398- 1402 (1) Bellar, T. A, Liefctenberv. J. J. (Dec. 1974) J. Am. Water Work* Amoc. 739-744 (9) Viola, P. L. (1970) Tenth Intemationd Can cer Conference, Houston, TX, Seanon 51, p. 742 (Abstract 29) (10) Viola, P. L,, Biiotti, A, A Caputo, A. (1971) Cancer Ret. 31,515-522 (11) 8upina, W. A. (1974) The Recked Column m Get Chromatography, Supaleo, lac., Belle- foots, PA ucc 061595 POSSIBLE INTERFERENCES WITH VCM Porous Polymer toe Butane Cyclobutane Isobutane 2,2-Dlmethyl propane Ethyl chloride Methanol Methyl chloride Propane Ethylene Oxide Butene-1 Butene-2-els Isobutene Propylene 2-Methyl butane Polvether type Columns Toluene 2.3-Pentane dlone 2-Ethyl-3-methyl dloxolane 3-Hexanone Dl-lsopentyl ether 1-Butene-2-ol 2.4-Dlmethyl-l, 3-dloxane 2-Methyl propanol 3-Methyl-2-butanol 2,3-Dlmethyl butanol n-Butyl acetate 2-Methyl pentanol 3-Pentyl acetate 2-Pentyl acetate 3-Methyl-4,6-dloxanone 3-Pentanol 2,5,8-Trimethyl-4,6-dloxanone 2-Methyl decane 2-Pentanol Isobutyl propionate 3-Methyl-3-pentanol 3-Ethyl nonane ucc 061596