Document MGkDo31qEGqzrGBo7wN8DdkDj

Methods for atmospheric vinyl chloride measurement have been reviewed. The lowest detection limits and most specific measurement are achieved by scrubbing atmospheric samples with activated charcoal, desorbing the vinyl chloride, and assaying it by gas chromatography (GC). NIOSH currently recommends collecting samples using tubes packed with 150 mg of coconut shell charcoal, desorbing v/ith carbon disulfide, and an3!y2ingbyGCequippedw:thflame-ioni23t;ondetectioni?IO);th9methodiscapableof detecting less than 1 ppm vinyl chloride and has an apparent recovery of about 90%. With thermal desorption techniques, the detection limit can be reduced to the ppb level with no loss of accuracy or precision. Some field methods, such as infrared analysis and conductivity measurement, are capable of detecting 1 ppm or lower but are subject to interferences by other contammants; they could be useful for evaluating sources of vinyl chloride leaks and for continuous monitoring. Permeation tubes are superior to gravimetric or volumetric methods for generating atmospheres of known vinyl chloride concentration. Measurement of atmospheric vinyl chloride SHELDON S. LANOE Center for Chemical Hazard Assessment. Syracuse Research Corporation, Merrill Lane. Syracuse. NY 13210 1 introduction Occupational exposure to vinyl chloride has describes methods for preparing standard vinyl often resulted in a comprehensive syndrome chloride samples for determining response which is now termed as "vinyl chloride disease." curves and calibrating measurement procedures. Prolonged exposure ultimately ends in fatal hepatic angiosarcoma.111 The chief commercial use for vinyl chloride is the production of polyvinyl chloride. Until occupational exposure studies traced a high incidence of hepatic angiosarcoma among vinyl chloride polymeriza tion workers, its occupational standard permitted dangerous levels of exposure. Since the hazard has been delineated, the standard has been lowered and, as a result, a number of recent studies have developed improved methods for its measurement. This article reviews the recent literature on measurement of atmospheric vinyl chloride. Measurement methods have been designed primarily to monitor whether or not vinyl chloride concentration is in compliance with the required occupational standards, and to identify sampling grab sampling and liquid scrubbing Grab sampling and liquid scrubbing are relatively unimportant collection methods for vinyl chloride. Virtually no information was available which evaluated liquid scrubbing.121 While grab sampling has been used with various evacuated containers, including glass bulbs, stainless steel canisters, gas syringes, and Tcdlar bags, it does not appear well suited for occupational sampling.(J,J> Among its disadvantages are its inability to collect a representative sample over a full work shift or large fraction of a shift, cumbersome equipment, and its higher detection limit than collection by adsorption. unusually high vinyl chloride concentration in a plant for corrective action. The literature which adsorption tubas is herein reviewed includes methods for sample The most popular measurement method is collection and analysis and, in addition, collection of vinyl chloride by adsorption and Copyright 1379, Amorieon Induatrial Hygiono Association SS Am. Ind. Hyg. Assoc J[AO) February 1979 RSV 000640? 'est jr.d i -jf i ith i no ind t to nyl t to |ide i i inyl nsc ires. ; for was ious jibs, ;dlar for its Jt a f: or tent, n by a is and ' 1973 subsequent gas chromatographic analysis. The measurement procedure begins with collection of a field sample by means of a solid sorbent packed into a glass or stainless steei tube. As the atmospheric sample is pumped through the packed t uhe at a known flow rate and time span, vinyl chloride (as well as other substrates) is adsorbed. The collected sample is then desorbed either by being dissolved in a suitable solvent or by thermal treatment.(2,4'5) Collection by adsorption has several advantages over direct field measurements or grab sampling: (1) when combined with gas chromato graphic measurement, the measure ment is the most sensitive; (2) it is capable of sampling over a larger fraction of a shift; (3) it is easily applied to occupational compliance measurements and per sonal sampling; (4) samples are easily transported and stored. The National Institute for Occupational Safety and Health (NIOSH) has devised a specific measurement procedure for evaluating compliance. Sampling requires adsorption onto activated charcoal and desorption with carbon disulfide (CS?).15-7' NIOSH suggests sampling with commercially available tubes which contain 150 mg of coconut shell charcoal (20/40 mesh) packed into a 4 mra I. D. glass tube. The charcoal is packed into two sections separated by a polyurethane plug: the front section contains 100 mg of charcoal, and the back-up section, which is used to evaluate breakthrough of vinyl chloride from the front section, contains 50 mg of charcoal. If the back section contains more than 20% of the amount found on the front section, the test is not valid. NIOSH(4,sl recommends that 5.0 L of air should be collected at 0.05 L/min with a personal breathing zone sampler, but if high humidity or high concentrations of contaminants are suspected, a 2.5 L sample should be used. However, high humidity or high contaminant concentrations are neither defined nor characterized. After sampling, the tube should be sealed with standard plastic caps. If sample shipment and other factors require more than one day between collection and analysis, NIOSH suggests that the capped tubes should be packed in dry ice for transport and then stored at --20C until the analysis. Each charcoal section should be analyzed >epa:dtei> The vinyl chloride is dejorbed b> adding the charcoal section to a 2 ml. viai containing l .0 mL CS*: and sealing v. ith a septum cap. The CS: samples for GC analysis must be withdrawn by a microliter syringe. NIOSH specifies that the injection should consist of a 5 jiL aliquot withdrawn when the charcoal and carbon disulfide have been in contact between 30 and 60 minutes. Several groups have evaluated the overall NIOSH recommendedprocedureand its various stages (collection, storage, and desorption). Two studies have evaluated the complete NIOSH recommended procedure.'7'11 Hill, eta!, reported that the coefficients of variation for samples (5 L) at 7.2 and 71.3 mg/ m3 were 0.076 (27 samples) and 0.075 (29 samples), respectively.(7> They noted that evaluation of the procedure's accuracy is also subject to the experimental errors from preparing the synthetic vinyl chloride atmosphere (vide infra). NIOSH estimated the recovery by comparing the concentrations detected by direct GC injection and by the proposed method. For atmospheres containing 2.6 to 64 mg/m3, the values agreed within 6%. Purcell evaluated the method by sampling 1.5 L of an 8 ppm vinyl chloride atmosphere (23 /igvinyl chloride).11' He reported an analytical recovery of 91.5%, with less than 0. 2% on the back section of the sample tube, and a relative standard deviation of 10.6% (for six samples). Three groups -- Hill, et al.,i6> Severs and Skory,1" and Cuddeback, et al.{il) -- examined breakthrough characteristics in complementary studies. While results generally agreed, the studies did report some conflicting data and conclusions. Cuddeback, et al. measured breakthrough with commercial tubes sold by Mine Safety Appliance, Table l."11 Hill, et al. looked at 21 sorbents (of which 6 were activated charcoals and 15 were gas chromatographic packings) packed into 15 mm sections in 4 mm 1. D. glass tubes, Table II.<7) Severs and Skory examined several activated charcoals in a series of tests which included packed beds and/or tubes and some commercial tubes. While Hill and co-workers4" and Severs and Skory1'1 concurred that breakthrough volume American industrial Hygiene Association JOURNAL (40) 2/79 97 RSV 0006408 TABLE I Retention Data for Vinyl Chloride on Charcoal Tubes* Vinyl Ch'oride Cone Ip pm) S s 5 25 25 25 50 50 50 31-npl* Rata liTit 'min) SO 100 ISO SO 100 ISO 50 100 150 Mm vcM F jw Rate i mini 0 639 1 278 1 915 3.19 536 9 58 6.38 12.78 19.18 R etsntion Volume iUl*' 100 98 29 3 7.9 22 0 20.5 9.0 18.1 14.8 Retention Tim* (mini-' 200 93 195 158 228 137 180 181 98.7 Toiel Mjii i-SI* 127.9 125.2 373.6 504 1456_ ... 1312 1285 2311 1691 `Standard (150 mg charcoal) tubes from Mine Salary Appliance. *At 10% breakthrough from from section of tube. `Three samples averaged 99 7 mg (16%) m 16.5 mm (10.9%) columns. decreased with increasing flow rate, the data of Cuddeback and co-workers lacked any correlation. Channeling in the commercial tubes was suggested as a possible cause for the inconsistent behavior.112' If channeling is a real effect in commercial tubes, it is a potential source of error. It would seem, then, that commercial tubes should receive more extensive study. While most of the data reported by Hill, et n/.m and Severs and Skoryllfl> agreed, the two groups disagreed on interpretation. Severs and Skory mistakenly challenged the claim of Hill, et at. that commercially available (coconut shell TABLE H Sorbent Screening Experiments with Vinyl Chloride1'1 Sorbent 1. Chromosorb 101-108 2. Chromotorb 106-107 3. Tenax GC 4. Silica Gel 5. Silica Gal w/1% AghO 6. Molecular Sieve. 5A 7. Carbopak A 8. Carboptk B 9. Carboiieve 8 10. Oow Carbon. XF4175L 11. Dow Carbon, XF4J75L 12. Dow Carbon. XF417SL 13. Oow Carbon. XF417SL 14. Petroleum Charcoal. SKC-104 15. Coal Charcoal. BPL. 16. Coconut Shell Charcoal. MSA-6 17. . Coconut Shell Charcoal. MSA-6 18. Coconut Shell Charcoal. MSA-8 19. Coconut Shell Charcoal. MSA-8 20. Coconut Shell Charcoal. MSA-8 21. Coconut Shell Charcoal. MSA-8 22. Coconut Shell Charcoal. MSA-6 23. Coconut Shell Charcoal. SKC-105 24. Coconut Shall Charcoal. PCS Meeh 60/80 60/80 35/60 20/40 20/40 30/40 45/60 45/60 45/60 20/40 20/40 20/40 20/40 20/40 20/40 20/40 20/40 20/40 20/40 20/40 20/40 20/40 20/40 *20/40 Concentra tion (jjg/L] 500 500 500 130 130 500 500 500 500 6.5 6.5 6.5 6.5 6.5 6.S 500 SCO 500 130 6.5 66 6.5 65 65 Sampling Rats (L/min) 1.00 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.15 0.10 0.05 0.10 0.10 1.00 0.20 0.05 0.20 020 0.10 0.05 0.10 0.10 `Value* *r* tingle experiments unless otherwise indicated. 'Average of two experiments. `Average of three experiments. 'Average of four experiments. Breakthrough Volume* (L) <0.1 <0.1 <0.1 <0.1 <0.1 2.0' <0.1 <0.1 2.8 5.7* 11.1 15.0 21.2* 7.7C 8.2* 0.9 2.4* 52 3.4 5.7 10.3* 10.7 10.6* 8.1* Am. M. Hyg. Assoc J(40) February. 1979 ftSV 0006409 TABLE II! Effact of Charcoal Tuba Storage Under Various Conditions on Vinyl Chloride Recovery'11' S toraga OiyS 0 Recovery (%) Total recovered in back section |%) Hi ft C 3nc6ftt rati an 31 3 Vinyl Chlo/xJ* Cappeb toas 'C 43 22 4 Fused TuPes 'C 43 22 4 99 69 69 89 99 89 000 000 LosV Concentration: 2.6 u0 Vinyl Chio'dO Capped Tubes 9d 43 22 4 Fused Tubes 'C 43 22 4 85 65 85 65 85 as 000000 7 Recovery |%) Total recovered in back section (St) 91 71 87 69 79 79 22 14 3 19 21 3 87 84 77 72 86 90 23 14 0 27 27 10 15 Recovery (%) Total recovered in back section (%) 78 98 85 60 81 85 28 20 11 28 23 IS 76 67 81 64 69 69 28 18 6 22 19 1 charcoal) tubes could collect 5 L of 206 ppm vinyl chloride at 0.05 L/min without significant breakthrough. They did find breakthrough at 1.5 L for these sampling conditions, but they used commercial tubes containig a petroleum based charcoal (SKC Lot 104) which has a lower adsorption capacity than coconut shell charcoal.'10' A strong point of disagreement between the two groups was the extent that high humidity and/ or high concentrations of other contaminants affect breakthrough volume. Both groups agreed that their effect was to reduce breakthrough volume. While Hill and co workers have suggested that a 2.5 L instead of a 5.0 L sample was sufficient to prevent breakthrough in the recommended commercial 150 mg sampling tube. Severs and Skory have disagreed. They still considered breakthrough a potential problem even with 2.5 L samples and have recommended a commercially available 600 mg charcoal tube for vinyl chloride collection. No information was available to resolve their conflicting conclusions, so this is an important research need. The effects of storage conditions on the recovery of vinyl chloride collected in commercial tubes were evaluated."1' The parameters were the tube sealing method (standard caps or fusing the ends), storage temperature (4, 22, and 43*0, storage time (up to 15 days), and the mass of vinyl chloride collected on the front section (2.5 and 31.9 jig). Table III summarizes the results. Among its salient observations, the study demonstrated that the standard caps were sufficient for retaining the vinyl chloride in the tube, and that loss by degradation and/or polymerization appeared negligible. When the tubes were stored for one week at room temperature, a minimum of 14% of the total recovered vinyl chloride migrated to the back section. Cooling the samples to 4C retarded but did not stop diffusion. Cooling to --20C seemed sufficient to stop any significant migration. Also, no migration to the back section occurred when 3 Hg of vinyl chloride were stored in the front section of a standard charcoal tube at --20 C over a 19-day storage period.11' Gas chromatographic packings are capable of adsorbing vinyl chlo^ide.,7,u",,' Hill and coworkers reported that Molecular Sieve 5A and Carbosieve B adsorbs vinyl chloride as well as charcoal. Table II.`7' Bellar.e/ and Russell11" also found good adsorption characteristics for Carbosieve B. Bellar and co workers evaluated adsorbents by collectisgvinyl chloride (and other organics) from aqueous samples. Their procedure was to purge a 5.0 mL sample with ISO to 400 mL of nitrogen at 200 mL/min and collect organics on 13 cm columns of test adsorbents. They obtained quantitative recovery of 50 ng of vinyl chloride with silica gel and Carbosieve B. Hill and co-workers, however, have reported that silica gel is a poor adsorbent for vinyl chloride. Table II. The conflicting claims of the two groups cannot be resolved with the evidence at hand. Chromosorbs and Poropaks exhibit good adsorption characteristics for vinyl chloride at ambient temperature.In addition, Ives has reported good vinyl chloride ^adsorption on Tenax-GC cooled to dry-ice. temperature.''4' American Industrial Hyfitna Association JOURNAL (40) 2/7$ w RSV 0006410 W??r. samples were collected at 0.85 L/min for ' i -- -u:es. recoveries averaged 909c at b ppl> _i: .-0% it 60 ppb. V.r.\l cbiorias recovery from the NIOSH rtzc--; mended C'S: desorption method ranges from 80 to 90%. ; Potential sources of loss include gas escape during mixing of CS2 and charcoal (a highly exothermic process), vinyl chloride partial pressure in the bead space of the sample vial, and vinyl chloride adsorption to container and syringe walls. The NIOSH method gave an average recovery of 85% (80 to 90% range) with samples containing 13 /ig of vinyl chloride on 100 mg of activated charcoal (charged with 5 L of a 2.6 iAg/L vinyl chloride atmosphere). NIOSH also evaluated desorption recovery with 0.5 mL of CS2 (instead of 1.0 mL), mixing by adding solvent to the charcoal (instead of adding charcoal to CS2), and mixing at 0C (instead of mixing at ambient temperature). They found no significant differences for changes in the CS2 volume or temperature, but found a slightly better recovery when the charcoal was added to the solvent.171 Better vinyl chloride recoveries were reported when CSi was added to charcoal at temperatures below ooc.<lul,> Cuddeback, et al. desorbed the vinyl chloride by placing the charcoal (100 mg) in a 2 mL glass vial, sealing it with a septum cap, and injecting 0.5 mL of CS2 while the vial was cooled in dry ice."11 The vial was then warmed to room temperature and held for 5 minutes before withdrawing samples. Desorption efficiency was 85 and 89% with 2.55 and 31.9 ^g vinyl chloride adsorbed, respectively. Lao, et al recommended adding 15 mL of cooled ("I5Q CS2 to the charcoal (1 g) in a 25 mL reacti-flask fitted with mini-inert valves.1 m The flask was thennostated at +15 C for 15 minutes and then samples were withdrawn for analysis. They reported that head space vinyl chloride concentration never exceeded 2% of the liquid concentration. Recovery ranged from 95 to 97% for 14.1 to 126 ng vinyl chloride and 88% for 2.7 ng vinyl chloride. Going111' and Severs and Skor^" achieved a somewhat greater recovery than the NIOSH workers1*'1'7' by adding charcoal to CS2 cooled with dry ice. Severs and Skory desorbed I g charcoal samples (containing 0.8 ng vinyl chloride) with lOmLCSi.1''Their procedure was to c?oi th- CS; (dry-ice temperature), slowly the tu the cold CS:. agitate the siu.--; ` 30 n:ni:'.e> while -cold, and withdraw aItqu.r.-. analysis while cooleu in a wet ice baih : achieved a 98% average recovery (range ji101%). Gomg also proposed a method in which the charcoal and CS2 were mixed at dry-ice temperature.,l3, The method was to slowly add the charcoal (450 mg) to 2 mL CS2 in a 3.5 mL vial while it was cooled in dry ice, then seal the vial with a septum cap. Samples (5 mL) were subsequently withdrawn with a microliter syringe while the vial was in dry ice. He recommended a 30-minute contact (recovery was constant from 25 to 140 minutes). For nine samples containing II to 13 vinyl chloride, recovery averaged 96.6 7.3%. At lower vinyl chloride loadings, the recovery was less (58 to 79% recovery for 154 to 768 og vinyl chloride). Other solvents can be used to desorb vinyl chloride from charcoaL An 88% recovery is reported for desorption with tetrahydrofuran; experimental details were not available.170' Vinyl chloride has been desorbed with a brominehexane mixture; in the procedure, vinyl chloride was convened to 1,2-dibromo-l-chloroethane (DB-VC), which has significantly greater sensitivity to electron capture detection (vide infra). The procedure consisted of stirring (magnetic stirring bar) 1 g of the activated carbon, 0.5 mL Br2, and 11 L of hexane in a screwcapped flask at --30 C for 5 minutes. The DB-VC was concentrated by column chromatography on silica gel with hexane followed by evaporation of the solvent. Neither recovery nor precision for the brominehexane desorption was described. DB-VC preparation could be useful as a derivative to confirm apparent vinyl chloride residues. Thermal desorption has been considered as an alternative to CS2 desorption for desorbing vinyl chloride from charcoal. Disadvantages of the CS2 method include the following: CS2 is very . volatile, toxic, and inflammable; vinyl chloride concentration is diluted by a factor of 200; and impurities in the CS2 can interfere with and/or lengthen analysis time.1*'23' The disadvantage of thermal desorption is that it permits odiy one analysis per sample, whereas solvent desorption allows multiple analyses. 100 Am. Ind. Hyg. Assoc. J (401 February. 1973 RSV 0006411 j i j ret, Me col per the < of 1 vin ads 0.8 (12 1 ana sue the pea sen: app met vin} van ther atta and onl> colic desc (0.6 Carl durit to 29 recov Ab collet char* by 3 tube with steel mesl carri rate by hr two of 9f addc Amenca add tiL : the vers iter He very nine -ide, inyl 8 to :de). inyl y `s ran; inyl uneride lane atcr \-\de ring ated in a The i to n <ane .ther rane ition firm is an nnyl t the very jride and d' or 2e of one uion . '.979 Three approaches to thermal desorption have been evaluated. Method A desorbed vinyl chloride in an inert gas stream, and all effluent y.as collected. In Method B. vinyl chloride is eluted onto an analytical GC column and retained through the desorption period. In Method C. the vinyl chloride is eluted into a collection chamber. Then, after the desorption period is completed, the sample is injected onto the analytical column. Severs and Skory reported the only evaluation of Method A.''4 They recovered 93% of a 25.6 ng vinyl chloride sample by heating the charcoal adsorbent at 430 C in a nitrogen flow of 0.5 to 0.8 L/min. The large effluent volume collected (12 L) is a tedious experimental problem. The selection of analytical column and analysis conditions are critical for Method B success. Unless the vinyl chloride is retained near the column inlet during the desorption period, peak broadening can occur and reduce sensitivity. Four groups have evaluated this approach.1**15'1*'25' Russell evaluated general methods for analyzing organic vapors, including vinyl chloride.'15' Vapors were collected on various GC packings and were subsequently thermally desorbed. The collection tube was attached to the GC inlet by a Swagelock fitting and wrapped with heating tape. Russell reported only preliminary results for vinyl chloride. It was collected on Carbosieve B and subsequently desorbed (5 minute heating at 270C) onto a 2 ft. (0.61 m) by 0.125 in. (3 mm) column packed with Carbosieve B. The column was held at 80 C during the desorption period, then programmed to 290 C at 20 C/min. Russell did not report the recovery. Ahlstrom, et al. evaluated vinyl chloride collection and analysis using 150 mg of SKC charcoal (Lot 104), packed into a 5 in. (127 mm) by 3/16 in. (4 mm) U-shaped stainless steel tube."4' The tube was attached to a GC equipped with a 6 ft. (1.8 m) by 1/8 in. (3 mm) stainless steel column packed with Poropak QS (80/100 mesh). The oven temperature was 90 C, and the carrier gas (nitrogen) was maintained at a flow rate of 0.20 L/min. Vinyl chloride was desorbed by heating the U-tube at 400 C (pulse heater) for two minutes. They reported apparent recovery of 96 to 102% for 5.8 to 17.3 Mg vinyl chloride added to the charcoal. Relative standard deviation of four samples charged with 1 ppm vinyl chloride was 1.2%. The average retention time for vinyl chloride was 6.50 min. P'jrccll end (j,or(Jano<,'",1 desorb .v ;r.. chionde from commercial active;.d . tubes. They transferred the charcoal . ^ the vinyl chioridc from the samphne tun - ti standard glass liner built for the Pcrkin-Eim.r Senes 900 GC system.l*'5il The analytical column was a 1.5 ft. (1.6 m) by 1/8 in. (3 mm) stainless steel column packed with Chromosorb 102 (80/ 100 mesh). The conditions were a 70C oven temperature and carrier gas flow (helium) at 0.040 L/min. They desorbed vinyl chloride by heating the charcoal at 260-300 C for four minutes. During the desorption period, the column was held at ambient temperature (by keeping the oven door open). At the end of the heating period, the oven door was closed and the column was rapidly heated to the operating temperature (70C). Purcell and Giordano reported recovery of 90 4.1% for 717 ng of vinyl chloride. Retention time for vinyl chloride was 3.2 minutes. Bellar and co-workers developed an apparatus and a procedure for desorption of organics collected on silica gel and Carbosieve-B (vide supra). They designed two desorber units. Desorber No. 1 works with the trap fit within the gas chromatograph injection port and is heated by the injection port heating block. They designed Desorber No. 2 for gas chromatography units where Desorber No. 1 will not fit into the injection port. Desorber No. 2, which is equipped to heat the trap by means of a heater wire insulated with asbestos, is attached to the gas chromatograph by a needle through the injection septum. Both Desorbers No. 1 and 2 are equipped with a backflush flow controller which purges the trap with a gas source (other than the GC carrier gas). The analytical column was a 6 ft. (1.8 m) by l/8in.t3 mm) stainless steel column packed with Chromosorb 101 (60/80 mesh) at 90 C or programmed from 90 to 200 C at 10 Cj min, and the carrier gas was nitrogen at a flow rate of 0.060 L/ min. They held the GC at ambient temperature during the desorption period (oven door left open). During desorption, the trap was heated to 150 C with 0.020 L/ min nitrogen flow for four minutes. Theti the backflush flow controller was removed and the GC temperature program w.as initiated. Bellar American Indvst/iil Hygiene Association JOURNAL (40) 2/73 lot RSV 000641 and Lichtenberg reported quantitative recovery of 50 ng of viny! chloride from either Carbosteve B or silica gel."11 Vinyl chloride desorption by Method C has been evaluated in two studies - Meyers, ?/ at'., :1> ar.d hes.'1'*' Meters and co-workers prepared sample tubes by packing activated charcoal <60 30 mesh) :n columns 6.5 cm long inside 5 mm O.D. glass tubing. The tubes, which were exposed to 1 L of J ppm vinyl chloride, were heated at 300 C for 30 seconds in a Burrell Pyrolizer. Then the desorbed vinyl chloride was passed through a 1/8 in. (3 mm) stainless steel tube and onto the analytical column 6 ft. (1.8 m) by 1/8 in. (3 mm) Chromosorb 101, which was operated at 100 C. They reported an average vinyl chloride concentration of 1.25 .037 ppm for 13 replicate samples. They found that 20 ppb could easily be detected with a 5 L sample and suggested that as low as 1 ppb could be detected by increasing the attenuation of the GC. Ives studied collection and desorption with TenaxGC.*'4) The tube, which was wrapped with heating tape, was heated at 150-180*0 for five minutes. Then desorbed vinyl chloride was injected onto the GC through a Hamilton syringe valve in the injection port. The GC column was a 6 ft. (1.8 mm) by 4 mm glass tube packed with Poropak QS and heated at 120 C. The system was tested with 6.5 cm columns of Tenax-GC packed into 5 mm I.D. glass tubes and exposed to 2.975 L of a standard vinyl chloride atmosphere. With 60 ppb vinyl chloride, the recovery was 100% (three samples), and for 6 ppb it averaged 89.6 4.7% (thirteen samples). automated sample collection An automated sampling system was developed for monitoring an industrial plant.*2** Each sampling station collects atmospheric samples through a cartridge filter which is protected from rain by an inverted 6 in. (15 cm) funnel. The sample flows from the sampling station to the laboratory through 1/4 in. (6 mm) stainless steel tubing. Samples are analyzed by means of gas chromatography (0.5 mL injections) and require 2 minutes for the complete analysis. The system can collect and analyze 20 samples per hour-* 19 plant sites and l calibration sample (10 ppm vinyl chloride in nitrogen). Baker and Reiter claimed reproducibility within 0.1 ppm fora [ ppm sample when the system was set to measure the range 0-25 ppm. analysis gas chromatography Gas chromatographs ;s the most sensin', e, accurate, and dependable method now available for vinyl chloride analysis. Salient features in the GC procedure are the analytical column, detector to remove potential interferences (substrates detected along with the vinyl chloride), and procedures to optimize the rapidity and cost for an analysis. Optimal cost and speed of analysis depend upon methods to remove interferences, vinyl chloride retention time, and either backflushing or increasing oven temperature to remove substrates of high retention time from the analytical column.11'2** The important considerations in the choice of a detector include detection limits, linearity of response, specificity, and expense. Table IV compares the specificity and lower limits for several detectors. The Ratne ionization detector (FID) which is recommended by NIOSH14'71 is an excellent general purpose detector. Its advantages include a low vinyl chloride detection limit (100 pg), a good linear response range, and inexpensive operation. Its major disadvantage is its nonselective response. Microcoulometry can be operated specifically for organohalides. The method requires substrate pyrolysis, then assays of the released halide ion. The Hall detector is reportedly the most sensitive of the available microcoulometric detectors for chloride. Its reported detection limit for vinyl chloride of0.07 ng is slightly better than the FID detection limit.*26'27* Its major advantage is its sensitivity to organohalides. Since the detector is virtually blind to nonhalogenated substrates, it eliminates interferences from potentially troublesome, low molecular weight hydrocarbons and other organics, such as acetaldehyde (see Table IV). Although mass spectrometry - specific ion monitoring (SIM) offers excellent selectivity and the lowest detection limit (20 pg), its expense places it out of reach of most laboratories. Vinyl chloride is monitored by its parent ions, m/ e 62 (2,C1) and 64 (J7C1). The individual ions are not 102 Am. Ind. Hyg. Assoc. J (40) Fe&nii. 1979 0006413 a e a d c. P sc sc th to h\ hs fo an T\ br ]-< de; res Pg- spe bas exc ozo com chlo olefi and yieli limi A beet sum tern; Mos chlo Amefic TABLE iV Comparison of Gas Chromatography Detectors for Vinyl Chloride Analysis OstBctsr fid--.e lorwiat'On detector Eleciron capl-'e Microcoolgmetry iHall defector^ Chemiluminescence Mass spectroscopy specific ion monitoring Specificity non* fidi'des halides otelins m/e 62 and 64 ions A osioiinate Oc;e':icn Limit 0 1 10 2 0 x io ` 70 X 10 20 x lO"* 10-20 x to"1' always specific for vinyl chloride.For example, SO? appears at m/e 62, and CF? appears at m/e 64. Since the natural isotope distribution of I5C1 to I7C1 is 3:1, vinyl chloride can be specifically monitored by measuring the parent ion isotopic ratio. Electron capture detection (ECD) is not as selective as microcoulometry, but is more selective than FID. Its major disadvantages are that the response for vinyl chloride is low relative to aromatic halides and to multiple chlorinated hydrocarbons (for example, trichloroethylene has a sensitivity of 0.020 ng compared to 2.3 ng for vinyl chloride), temperamental operation, and a relatively small dynamic response range.12' The ECD limit for vinyl chloride was reduced by brominating vinyl chloride to yield 1,2-dibromo1-chloroethane (DB-VC) (vida supra). The detection limit for DB-VC was 15 pg and the response curve was linear in the range of 50-300 pgThe chemiluminescence detector, which is specific for olefins, is a relatively recent system based upon the chemiluminescence of the excited reaction products of olefins and ozone.|21`2,,2,, McClenny and co-workers compared the detector response for vinyl chloride with several low molecular weight olefins (ethylene, propene, butene, butadiene, and 1,2-dichloroethylene) and found that it yielded the poorest response,1311 Its detection limit was about 20 ng. A number of GC analytical columns have been found adequate for the separation; TableV summarizes some of the packings and operating temperature ranges for vinyl chloride analysis. Most of the columns and conditions yield vinyl chloride retention times ofless than five minutes. Baker and Reiter utilized three columns for analysis. The first two columns consisted of 20% Igepal CO-880 on Chromosorb P and the third column was 20% Octoil S on Chromosorb P.'2*' After higher boiling components of a sample were removed on the first column, the second column separated vinyl chloride from most of the remaining interfering components. The final separation was achieved on the third column. A major analytical problem is to insure that other substances do not interfere with vinyl chloride measurement. Potential interferences include light hydrocarbons. Freon 11 and 12, acetaldehyde, and SOi.|,',1',4*23'2wn Specific interferences vary between columns. Table VI summarizes relative retention time data for potential interferences with Chromosorb 102, Poropak Q, and Carbowax 1500 on Carbopak A. Acetaldehyde is probably the most common interference for vinyl chloride. A specific method was developed to quantitatively remove acetaldehyde interference from the vinyl chloride analysis.1101 The method uses a 6 in. (15 cm) by 1/8 in. (3 mm) stainless steel column packed with a mixture of sodium bisulfite and glass wool ahead of a 6 ft. (1.8 m) by 1/8 in. (3 mm) Poropak Q column. The acetaldehyde forms a bisulfite complex on the precolutnn. athar analytical mathods Infrared analysis has been used to monitor vinyl chloride concentration directly in the field. Portable meters equipped with 20.25 m foldedpath absorption cells yield a detection limit of about 1 ppm at the 941 or917cnf' bands.12*13'14' However, many other substances can interfere with the analysis. The potential interferences American Industrial Hygiene Association JOURNAL (40) 2/79 103 RSV 000641** Common Packing* and Oven Temperatures for the Vinyl Chloride Analysis Column Packing Ovan Temperature <;C) 10-20 ,3-6 m) SE-3Q Oi'romosorO G. Cruc'''osor& W. or Aruk.'ovn A3S Ambient ;o 90 20' (6 m' 10% FPAP on C'romosorb W 65 S' (1 3 m| Poropak Q 100-135 5-6' (1 5-1 8 m| Poropak OS 70-1 20 6' <1 8 m| ChromosorfrJOl---. ._____ -- 90-100 1.5-6' <0.5-1.8 m> Chromosorb 102 70-145 6* (18 m) 10-20% OC 200 on Chromosorb W or Supeicoport 80 Relemnce 7 25.26 7 11.30.32 9.14.16 13.23 8.13.19.33 9.13 20' (6 m| Carbowax 4000 on Supeicoport 80 9 6' (1 8 m) 0.4% Carbowax 1500 on Carbopak A Ambient 13.28 6' <1 8 m) 5% 0V-101 on Chromosotb W Ambient 32 6* (1.8 m] 10% Apiezort M on Chromosorb W 6* (1.8 m) Silica gel 300' {90 m| Open tubular column, coatad with dibutyl malaata 16' (4.8 m| 16.7% triseyano ethoxypropana on Chromosorb W Ambient 30 0 Programed from 50 to 170 32 34 34 31 include aliyl chloride, ethylene, Freon 12, methacrylonitrile, perchloroethylene, styrene, tetrahydrofuran, trichloroethylene, and vinylidene chloride/1' While infrared analysis appears useful for surveys to monitor vinyl chloride spills in high-risk areas, it does not appear to be sufficiently sensitive or specific for routine monitoring. interferences can be removed by passing the airstream through a water scrubber prior to the degradation; these removable interferences include sulfur dioxide, carbonyl sulfide, hydrogen sulfide, and chlorine. Interferences which are not removed by the water scrubber include nitric oxide, carbon disulfide, methyl chloride, and trichloroethylene. A continuous system for vinyl chloride monitoring using conductivity measurement has been developed.1'7' As the air is sampled, the vinyl chloride is continually degraded as the airstream is passed through a 6 in. UV lamp operated at 253.7 nm. The degradation efficiency for this procedure was 80%. Vinyl chloride concentration was then measured by conductivity with a Davis Instrument Series 11- Preliminary results of Stark spectroscopy for field measurement of vinyl chloride were reported/311 The system used either a carbon dioxide or carbon monoxide laser with an absorption ceil consisting of two 40 cm stainless steel Stark electrodes spaced 1 mm apart. The method, which is capable of measuring vinyl chloride at the ppm level, is currently in the experimental stage. 7000 Sulfur Dioxide Meter. With a 2.5 minute hold-up time in the Davis Meter, response vinyl chloride standard samples reached 90% of the final reading. Confer found a Standard samples are required for calibration lower detection limit of 0.05 ppm and a linear purposes in vinyl chloride collection and response up to 25 ppm/'7' The method does have analysis. The most important standard samples l several potential interferences. Some are for use in the NIOSH recommended method I Q of collec analysis proeedu determini solutions curves. S temperati by volurc are comn atmosphe known vo in a gasbs Standard volumetri suggested chloride b vinyl chlo gain, and solvent) t and Skor chloride ir ampul in t TABLE VI Relative Retention Timas of Potential Interference* to Vinyl Chloride ' -- ! R elatlvs Retention = 1 01 Compound M sifiq ng Ethane Elhn 1.1 -DiMuofothylna Propone Propn Meihylscetylene Methyl chloride 1,1-Difluoroethane Chlorodifluoromethane Cyclopropane Formaldehyde 1 -Chloro-1,1 -diiluoroethane Acetaldehyde Freon 114 Isobutene Isobutylene Methanol 1.3-Butadiene 1-Butene Vinyl methyl ether Trant-2-butane Ethyl chloride Cis-2-butene Vinyl bromide 1.1 -dichloroethylene Chromoaorb' 102 0.15 0 21 0.21 - 0.S4 0.63 - - 0 93 -- 1.22 1.37 -1.57 1.43 1.57 1.70 1.73 2.00 Chromaaorb* 102 0 33 0.33 0.62 - 0.51 053 0 92 t .21 -- 1.25 .. 1.27 1 30 1.36 1 39 .. 1 43 1.85 -- Porppak Q 0 05 0.46 0.52 0.56 0.57 __ .. 0 59 0.62 .. 0.95 .. .. .. .. .. .. .. ~ .. - 0 4% Crbowi" ' 500 on Carbopak A 020 0.29 0.26 0.63 0.63 0.63 .. 0.45 __ .. .. .. 0.77 .. .. .. 1.38 .. .. .. 2.92 1.54 -- - -- A6` X 1/8" Chromosorb 102 (80/100 m#eh| at 100C1> * X 1/8" Chromosorb 102 (80/100 mesh| at 145C'* c6` x 1/8" Poropak Q (80/100 meant at 100*C" D6` x 1/8" 0.4% Carbowax 1500 on Carbopak A at ambient temperature11 of collection on activated charcoal and GC analysis.*4*5* The NIOSH recommended procedure uses standard atmospheres for determining analytical recovery and standard solutions (in CSj) for preparing GC response curves. Since vinyl chloride is a gas at ambient temperature, the standard samples are prepared by volumetric and gravimetric methods which are common in gas handling.**-*'7*'* Standard atmospheres have been prepared by diluting a known volume of vinyl chloride with an inert gas in a gas bag made of material such as Tedlar.*1*4' Standard solutions have been prepared by volumetric and gravimetric methods. NIOSH suggested either adding known volumes of vinyl chloride by gas syringe to the solvent or adding vinyl chloride to toluene, measuring the weight gain, and then diluting the sample with CS2 (the solvent) to the final concentration.'1*4* Severs and Skory sealed a known weight of vinyl chloride in an all-glass ampul and then broke the ampul in the solvent.*** The permeation tube for generating synthetic vinyl chloride atmospheres is an excellent alternative.11*-41' While precision for preparing a I ppm vinyl chloride atmosphere by the gas dilution method is about 10%, the permeation tubes yield a 1 ppm atmosphere with a standard deviation of 2%.|40) The permeation tube consists of a plastic-walled tube filled with vinyl chloride and plugged at both ends. The vinyl chloride diffuses through the plastic walls of the tube (Fick's Law of Diffusion) at a rate which is constant and also temperature dependent (a temperature variation of 0.1C will create a 1% error in vinyl chloride concentration). FEP (fluorinated ethylene copolymer) tubes have generally been* chosen;11*"*1* vinyl chloride diffusion remains constant for at least eight months with FEP tubes. The vinyl chloride permeation rate can be determined either by measuring the weight of vinyl chloride lost from the tube or the vinyl chloride released with time,'To measure Amencsn Industrial Hygiens AssocLifan JOURNAL m 2m RSV 00064X6 10S diffusion rate by weight loss from the tube, about three weeks are required to create sufficient accuracy. Cedergren and Fredrickson have developed a rapid (about one hour), accurate method for measuring the released vinyl chloride.'39' The diffused vinyl chloride is combusted at high temperature to yield hydrogen chloride (HCl), which is assayed by coulometric titration with silver in 70% acetic acid. They compared the gravimetric and coulometric methods for the permeation rate measurement.139' While the gravimetric method yielded a mean vinyl chloride diffusion rate of 0.464 .002 ug/ hour, the coulometric method value was 0.463 .001 >ig/hour. Synthetic vinyl chloride atmospheres are generated by placing the permeation tube in a thermostated chamber which contains two openings. The dilutent gas is metered into the chamber in one opening and the standard atmosphere is sampled through the second. Cedergren and Fredrickson prepared a 1.000 .002 ppm vinyl chloride atmosphere (calibrated by coutotnetric titration).(,,> acknowledgement This work was supported by the National Institute for Occupational Safety and Health, Contract No. 210-76-0167. references 1. Federal Register 35:2342 (April 5, 1974). 2. U.S. Environmental Protection Agency: Measurement Techniques. Scientific and Technical Assessment Report on Vinyl Chloride end Polyvinyl Chloride. EPA-600/6-75-004. p. 7. U.S. EPA. Office of Research and Development, Washington, DC <1975). 3. U.S. Environmental Protection Agency: Interim Method for Sampling and Analysis of Vinyl Chloride in Waste Water Effluents end Air Emissions. Preliminary Assessment of the Environmental Problems Associated with Vinyl Chloride and Polyvinylchloride, (appendices). PB-239-110. p. 17, National Technical Information Service, Springfield. VA (1974). 4. Crabte. J. V. end O. G. Teyfor Vinyl Chloride in Air. P4CAM 178. NIOSH ManualofAnalyticalMethods. Washington. DC (1974). 5. Vinyl Chloride in Air: P&CAM 178. Rev., National Institute for Occupational Safety and Health, Cincinnati, OH (1976). , 6. White, l. D.. O. G. Taylor, P. A. Mauer and R. E. Kupel: A Convenient Optimized Method for the Analysis ol Selected Solvent Vapors in the Industrial Atmosphere Am. Ind. Hyg. Assoc. J 31 225 (1970). 7. Hill. R. H., Jr., C. S. McCimmon, A. T. Saalwaechter. A. W. Teess end W J. Woodfin: Gas Chromatographic Determination of Vmyf Chloride m Air Samples Collected on Zh.ircoai Anal. Chem. ^3.1 395 (19761'. 3 Purcell, J. 5.: Gas Chromatograohn- 5.j - Vinyl Chlor oe Ant Lab. 7(7/.99 |137c. 9 Severs. 1. W. and l. Kr Sko.-y: f/.ixorir,- Personnel Exposure to Vinyl Chloride. Vuiviidin-. Chloride and Methyl Chloride m an Industrial Work Environment Am. Ind. Hyg. Assoc J. 36669 (1975). ^ 10. Severs, L.: Personal Communication. Dow Chemical Co.. Midland. Ml (1977). 11. Cuddebeck, J. E.. W. R. Burg end S. R. Birch: Performance of Charcoal Tubes in Determination of Vinyl Chloride. Environ. Sci. Tech. 3:1168 (1975). 12. Burg. W. R.: Personal Communication. Dept, of Environmental Health, University of Cincinnati, Cincinnati. OH (1977). 13. Going. J. E.: Sampling and Analysis of Selected Toxic Substances -- Task Ilf -- Vinyl Chloride. Secondary Sources. PB-252-966, National Technical Information Service, Springfield. VA (1976). 14. Ivee. N. F.: Sensitive Trapping and Gas Chromatographic Method for Vinyl Chloride in Air Samples. J. Assoc. Offic. Anal. Cham. 53:457 (1974). 15. Russell, J. W.: Analysis of Air Pollutants Using Sampling Tubes and Gas Chromatography. Environ. Sci. Tech. 9:1175 (1975). 16. Ahlstrom, O. H-. R. J. Kilgourends. A. Uebman: Trace Determination of Vinyl Chloride by a Concentrator/Gas Chromatography System. Anal. Chem. 47:1411 (1975). 17. Seller, T. A.: The Determination of Volatile Organic Compounds at the mg/L Level in Water by Gas Chromatography. EPA-670/4-74-009. U.S. Environmental Protection Agency, Cincinnati, OH (1974). 18. Beliar, T. A., J. J. Lichtenberg and J. W. Eiehelberger Determination of Vinyl Chloride at mg/L Level in Water by Gss Chromatography. Environ. Sci. Tech. 19:926 (1976). 19. Lao. R. C.. R. S. Thomas and J. L. Monkman: Improved Methods for Sampling and Analysis of Vinyl Chloride. Am. Ind. Hyg. Assoc. J. 37:1 (1976). 20. Keenan, R. R. in U.S. Environmental Protection Agency: Measurement Techniques. Scientific and TechnicalAssessment Report on Vinyl Chloride and Polyvinyl Chloride. PB-239-110. p. 8. National Technical Information Service. Springfield. VA (1975). 21. Hoffman. D., C. Patrianakos and K. D. Bmnnemenn: Chromatographic Determination of Vinyl Chloride in Tobacco Smoke. Anal Chem. 48:47 (1976). 22. Purcell. J. E. end B. D. Giordano: The Gas Chromatographic Analysis of Vinyl Chloride. In Gas 104 Am. Ind. Hyt Assoc J (40) . February. 1979 RSV 0006417