Document yRZzeMdjMvD0LyJp1xbOnVM6

Internal Correspondence 3M Occupational Health and Environmental Safety Division To: C.E. Colton - OH&ESD - 260-3B-09 L.L. Janssen - OH&ESD - 260-3B-09 A.R. Johnston - OH&ESD - 260-3B-09 D.J. Larsen - OH&ESD - 260-3B-08 H.E. Mullins - OH&ESD - 260-3A-07 P.E. Olson - OH&ESD - 260-3B-09 J.B. Palazzotto - OH&ESD - 260-3B-08 Y.T. Shih - OH&ESD - 260-3B-08 From: R.A. Weber (7-4459) - OH&ESD - 260-3B-09 Subject: MONITOR ARTICLE ON LONG TERM SAMPLING Date: September 13, 1994 Don Larsen and I have been working on developing technical information that can support our diffusion monitors. If you have time, I would like you to review and comment on the enclosed article by September 26, 1994. At this time, the information will be sent out to customers on request, however, we are considering publishing if it is worthy. When you review, please do it with a critical eye. Thanks. RAW:llj/145 Attachment 3M 110727 t. J V\ j Long Term Sampling With Diffusion Monitors Donald J. Larsen and Robert A. Weber OH&ESD, 3M Company INTRODUCTION Historically diffusion monitors have been used for monitoring the working environment for full work shifts of 8 hours. Lab and field evaluations by 3M and others have been performed over the last 20 years to validate their use for 8 hour and STEL sampling.. Recent concern with indoor air quality, environmental emissions and hazardous waste disposal gives rise to situations where monitoring for extended periods is desirable. For indoor air quality investigations the concentrations may be very low, therefore long sampling times may be necessary to accumulate enough mass for analysis. Validation of a diffusion monitor must address the following factors - whether the sampling period is 8 hours , 15 minutes or several weeks: recovery, analytical sensitivity, capacity/reverse diffusion, linear uptake rate, orientation, temperature, face velocity, interference and storage. Validation of active sampling tubes like charcoal tubes must address these issues also. Reverse diffusion is the loss of previously adsorbed analyte during the sampling period. When this occurs with diffusion monitors the sampling rate will appear lower than predicted. Reverse diffusion effects can be studied by using a monitor that has a backup layer of sorbent and exposing the monitor to concentration levels that would not exceed the expected capacity of the primary layer. If the primary section on the monitor is not overloaded, any adsorbed material found on the backup section is due to reverse diffusion during sampling. In this study we evaluated the long term sampling performance of the 3M 3520 Organic Vapor Monitor. This report outlines results of five labortory experiments conducted with toluene, 1,1,1-trichloroethane, acetone, methyl ethyl ketone (MEK), and methylene chloride and a field test that evaluated the performance of the organic vapor monitors during exposure to a mixture of n-butanol and isopropanol. The compounds selected enabled us to test several classes of compounds that have a range of boiling points and a range of affinity for activated carbon. 3M 110728 EXPERIMENTAL DESIGN To evaluate the performance of the monitors the known air concentrations of analytes were generated in a laboratory generator/dilution system. The challenge concentration was also monitored by a portable infrared analyzer. A relative humidity of 30% and a temperature of 23 C. was maintained inside the system. The challenge concentration was passed through a Plexiglas chamber (5" x 5" x 21") containing the monitors. A challenge air flow of 90 Lpm created a face velocity of 19 feet per minute (fpm). In each of the five lab studies, the mass of the collected analyte was measured and compared to the mass expected. The mass expected is derived from our published sampling rates. A significant difference between the contaminant mass found and the mass expected would indicate contaminant loss and reverse diffusion. LABORATORY RESULTS Experiment 1 Twelve monitors were exposed to 2.3 ppm of toluene. After 2.75 days, six of the monitors were removed and analyzed. The remaining six monitors were exposed to 0 ppm toluene for 1, 2 or 3 days and analyzed. The average amount found on the monitors given additional exposure was the statistically the same as that of the monitors exposed to toluene alone (table 1). Based on a sampling rate of 31.4 cc/min. for toluene the predicted concentration on the monitors was 1.09 ppm whereas the average of the twelve monitors was 1.09 ppm +/- 0.064. These results indicate that reverse diffusion did not occur using the 3M 3520 Monitors under these conditions. TABLE 1 EXPOSURE TO 2.33 PPM TOLUENE FOR 2 3/4 DAYS PLUS 1, 2, OR 3 DAYS AT 0 PPM Set Exposure days at 2.75 ppm toluene 1 2.75 2 2.75 3 2.75 4 2.75 Exposure days at 0 ppm toluene 0 1 2 3 n Average Standard deviation 6 1.122 0.03 2 1.055 0.122 2 1.062 0.102 2 1.084 0.067 3M 110729 Experiment 2 Six monitors were initially placed in the chamber and exposed to 0.45 ppm of 1,1,1 trichlororethane. After 4 days, 2 were removed and 2 new monitors were started. Sampling continued for another four days. The two 4 day exposures showed nothing on the back-up section. Less than 1% of the total mass collected was found on the back-up of the monitors exposed for 8 days. The average amount found based on a sampling rate of 30.9 cc/min. was 0.83 mg compared to the amount expected which was 0.86 mg (table 2). This experiment indicates that the 3M 3520 can be used for 1,1,1-trichloroethane at low concentrations for extended periods of time. TABLE 2 EXPOSURE TO 0.45 PPM 1,1,1-TRICHLOROETHANE FOR 8 DAYS Exposure First 4 Days Last 4 Days All 8 Days mg Expected 0.45 0.41 0.86 mg Found A 0.394 0.445 0.393 0.384 0.890 0.842 0.772 0.817 mg Found B 0.000 0.000 0.000 0.000 0.005 0.005 0.004 0.005 mg Total 0.394 0.445 0.393 0.384 0.901 0.853 0.781 0.828 Average 0.830+/-0.044 8 day exposure (+/-5.3%) 3H 110730 Experiment 3 Six monitors were exposed to 2.86 ppm of acetone for 5 days. Based on the published sampling rate of 40. lcc/min. the amount expected over a five day period was 1.97 mg acetone. The mean acetone level was 1.77 mg with a standard deviation of .063 and a coefficient of variation of 3.55. This translates to an accuracy of 17.4%. For acetone, we found 15% of the total mass collected on the backup section (see table 3). If a single stage monitor had been used the amount on the back section would have been lost and this would have resulted in an accuracy of 40.9%. TABLE 3 EXPOSURE TO 2.86 PPM ACETONE FOR 5 DAYS ID mg A mg B mg Total 1 1.343 2 1.288 3 1.287 4 1.307 5 1.411 6 1.270 0.202 0.222 0.203 0.199 0.209 0.189 1.788 1.776 1.765 1.744 1.871 1.685 average mg: 1.767 standard deviation: 0.063 coefficient variation: 3.55% amount (mg) expected : 1.97 3N 110731 Experiment 4 Twelve monitors were exposed to a mixture of 0.17 PPM toluene and 0.03 PPM of methyl ethyl ketone. Six monitors were removed after 2 weeks, three more after 3 weeks and the last three after 4 weeks of exposure. The amount found was very close to the amount expected, (table 4). No toluene or MEK was found on the backup sections showing that reverse diffusion did not occur even after this extended sampling time. Figure 1 shows that the uptake rate was linear throughout the 28 day experiment. This experiment demonstrates that the 3M 3520 organic vapor monitor works accurately for long sampling periods. TABLE 4 28 DAY EXPOSURE TO MIXTURE OF 0.17 PPM TOLUENE AND 0.03 PPM MEK TOLUENE HOURS 356.3 500.2 673 AMT EXPECTED UG 438 618 827 AMT. FOUND UG 438+/-6.0% 615+/-4.7% 815+7-3.6% MEK HOURS 356.3 500.2 673 AMT EXPECTED UG 64 93 125 AMT FOUND UG 68 +/-6.9% 100 +/-3.9% 130 +7-4.3% FIGURE 1 TOLUENE & MEK EXPOSURE 3M OVM UPTAKE RATE + MEK .03 PPM A TOLUENE .17 PPM TIME (MRS) 3M 110732 Experiment 5 Twenty-one monitors were exposed continuously and intermittently to methylene chloride at 1.9 ppm for 1 to 5 days. Three monitors were removed each day and analyzed. Based on the 15 monitors exposed continuously over the five days, the uptake rate remained linear throughout this exposure. Analysis of the set indicated a coefficient of variation 5.3%, a bias of -3.2%, this translates to an accuracy of 13.8%. The remaining six monitors were exposed "day on day off' to the methylene chloride, for a total of 0 ppm methylene chloride for 48 hours and 1.9 ppm methlylene chloride for a total of 72 hours (table 5). This was done to simulate intermittent real life exposures. The intermittently exposed monitors collected 95% of the amount on the monitors exposed continuously. TABLE 5 METHYLENE CHLORIDE EXPOSURE AT 1.9 PPM FOR 5 DAYS ALTERNATING WITH 2 DAYS AT 0 PPM ( 72 hours at 1.9 ppm methylene chloride/ 4 days 0 ppm toluene) ID mg A mgB Total mg 1 0.558 0.185 0.964 2 0.542 0.211 1.006 3 0.569 0.180 0.965 4 0.557 0.172 0.936 5 0.570 0.199 1.007 6 0.539 0.195 0.968 FIELD TEST RESULTS Our field test was conducted at a 3M pilot plant. Monitoring was conducted for isopropanol and n-butanol from a coating operation over a 3 day time period. The coater operated intermittently. On day 1, it ran for 5 hours. On days 2 and 3 it ran for 12 hours. Monitors were exposed for 1, 2 and 3 days. We used this experiment as another measure of the effects of reverse diffusion. If the back up section contained significant concentrations and the monitors were not near their expected capacity, it would show that reverse diffusion was occurring. As a control we obtained charcoal tube samples on Day 1 and Day 3. Nine monitors were initially exposed to the n-butanol and isopropanol environment. Three monitors were removed each day and analyzed. The n-butanol results indicated nothing on the back section of any of the monitors. This shows that reverse diffusion of n-butyl alcohol did not occur. See table 6 for summary of n-butanol data. 3M 110733 TABLE 6 Time 1737 1737 1737 3255 3255 3255 4715 4715 4715 n-BUTANOL FIELD TEST mg ppm 0.025 0.027 0.025 0.164 0.172 0.163 0.053 0.052 0.056 0.182 0.180 0.194 0.077 0.076 0.076 0.184 0.180 0.181 Average of 3 0.166 0.185 0.182 The results of the isopropanol analysis were somewhat different. On day one we found 65 ug on the front and <1 ug on the back. On day two, 141 ug on the front and 6 ug on the back and on day 3,202 ug on the front and 10 ug on the back (see table 7). These results showed that reverse diffusion occurred with isopropanol although the effect was small, and that the use of a monitor with backup was able to provide accurate results. The average air concentration of isopropanol was 0.55 +/-0.07 ppm (+/-12.7%). 3M 110734 TABLE 7 Time mg A ISOPROPANOL FIELD TEST mg B Total mg ppm Ave of 3 1737 1737 1737 0.065 0.076 0.068 0.000 0.000 0.001 0.065 0.076 0.070 0.461 0.537 0.495 0.498 3255 3255 3255 0.155 0.149 0.144 0.003 0.004 0.004 0.162 0.157 0.153 0.611 0.593 0.577 0.594 4715 4715 4715 0.216 0.214 0.214 0.007 0.007 0.007 0.232 0.230 0.229 0.604 0.597 0.596 0.599 We compared the monitor results with charcoal tubes taken on Day 1 and Day 3 and can see excellent agreement (table 8). TABLE 8 Diffuison Monitor vs Charcoal Tube ComDound Isopropanol n-butanol Monitor ('own') 0.55 +/- 0.07 0.18+/-0.008 Charcoal Tube Tppm'l 0.48+/-0.11 0.17+/-0.03 3M 110735 CONCLUSIONS In summary, we can see that the accuracy's ranged from 10% to 17% which are well within the recommended 25%. These experiments show that diffusion monitors can be used for extended sampling periods of one week to one month and still provide acceptable accuracy. When volatile compounds like acetone and methylene chloride are anticipated, it is essential to use the 3M 3520 organic vapor monitor with a backup layer of sorbent to achieve accuracy of 25%. SUMMARY OF LONG TERM LOW CONCENTRATION EXPERIMENTS COMPOUND EXPOSURE TIME (DAYS) CONC (PPM) AMOUNT FOUND AMOUNT ACCURAC EXPECTED Y% Toluene 1,1,1-TCE Acetone Methylene Chloride Toluene/MEK Toluene/MEK Toluene/MEK 5 3/4 8 5 5 15 21 28 2.3 0.45 2.86 1.9 1.091 mg 0.86 mg 1.767 mg 1.088 mg 0.83 mg 1.97 mg 0.173/0.028 0.174/0.029 0.173/0.029 438.2/68.1 ug 614.9/99.8 ug 850.5/135.7 ug 438/64 ug 618/93 ug 827/125 ug 12.1 14.1 17.4 13.8 9.8/16.8 Comparison of boiling points vs. observations on reverse diffusion leads to the following general recommendation (table 9): the use of a monitor with backup section for those compounds with boiling points below 60 C and recommending the use of a monitor without a backup for those with boiling points above 90 C. For those compounds with boiling points between 60 and 90 C, a monitor with backup would provide higher accuracy with less risk of loss. 3M 110736 TABLE 9 BOILING POINT C N-BUTANOL TOLUENE ISOPROPANOL MEK 1,1,1 -TRICHLOROETHANE ACETONE METHYLENE CHLORIDE 118 111 82 80 75 56 40 REFERENCES 3M Organic Vapor Monitor Sampling Rate Validation Protocol Anders, L. W., andH. E. Mullins: Comparison ofDiffusional Organic Vapor Monitors with Charcoal Tubesfor Sampling Laboratory Challenges to Contaminant Mixtures Anders, L. W, H. E. Mullins and P. L. Sullivan: Organic Vapor Monitor with Backup Section Pieper, RichardM., Donald J. Larsen, and Patricia A. Ishaug: STEL Sampling Using Diffusional Monitors, American Chemical Society Meeting, Boston, MA, April, 1990. Epstein, Paul S., etal: Experiences Using Passive Monitors to Measure Volatile Organic Compounds During Indoor and Ambient Air Quality Surveys, American Industrial Hygiene Conference, 1990. Cohen, Martin A., et al: The Validation ofa Passive Samplerfor Indoor and Outdoor Concentrations of Volatile Organic Compounds, J. Air Waste Manage. 40:993(1990). 3M 110737