Breakthrough Times for Vapors of Organic Solvents with Low Boiling Points in Steady-state and Pulsating Flows on Respirator Cartridges

Transcription

1 Industrial Health, 1996, 34, Breakthrough Times for Vapors of Organic Solvents with Low Boiling Points in Steady-state and Pulsating Flows on Respirator Cartridges Shigeru TANAKA1, Miki HANEDA2~, Masami TANAKA2~, Kazushi KIMURA2~ and Yukio SEKI1~ 1) School of Allied Health Sciences, Kitasato University, , Kitasato, Sagamihara, Kanagawa 228, Japan 2) Koken LTD., 7, Yonban-cho, Chiyoda-ku, Tokyo 102, Japan (Received October 4, 1995 and in revised form January 24, 1996) Abstract: The breakthrough times of five organic solvents at various vapor concentrations were measured in steady-state and pulsating flows on commercially available respirator cartridges. The relationship between the logarithmic vapor concentration and the logarithmic breakthrough time of each organic solvent showed an inverse correlation in both of the flow patterns. The organic solvents with lower boiling points exhibited the shorter breakthrough times in both of the flow patterns. The ratios of the breakthrough times in the pulsating flow to those in the steady-state flow were lower than 0.9 when the vapor concentrations were higher than 600 ppm for ethyl acetate, methyl acetate, acetone and dichloromethane. From the present study, the breakthrough in the pulsating flow tends to occur earlier than in the steady-state flow when using a highly concentrated vapor of an organic solvent with a low boiling point. Key words: Respirator cartridge - Breakthrough time - Pulsating flow - Breathing machine - Organic solvent - Steady-state flow INTRODUCTION The best method suggested to date of evaluating the service lives of organic vapor cartridges has been to test them with a pulsating breathing simulator. Many researchers have investigated the service lives of respirator cartridges using various types of organic solvent vapors '6. Most of their papers presented experimental results measured in a steady-state flow of organic solvent vapors, only a few included results obtained in a simulated pulsating flow'' g). The pattern of a pulsating flow is generally represented by a sine curve in which the instantaneous flow rate at the peaks of the sine curve is about three times the average flow rate9, 1O) There is, however, some speculation that the equilibrium between the organic solvent in the vapor phase and the adsorbed phase on activated carbon is so rapid

2 126 S. TANAKA et al. that little difference would be observed in the breakthrough times between the steady-state flow and the pulsating flow. In our previous study", a breakthrough test of a respirator cartridge was performed using steady-state flows of 46 kinds of organic solvents. The test results indicated that 10 of these organic solvents among them exhibited breakthrough times shorter than that of carbon tetrachloride, which is designated as the standard test vapor for the Governmental Approval Test. Furthermore, the boiling points of these 10 organic solvents were relatively low. This paper presents the comparative breakthrough test results of steady-state and pulsating flows of organic solvents with low boiling points at four levels of vapor concentrations passing through a commercially available respirator cartridge. EXPERIMENTAL Materials A commercially available respirator cartridge filled with 22.5 g (56 ml) of activated carbon, KGC-8 (Koken Co. Ltd., Tokyo), was used for the experiment. The activated carbon layer was 13 mm in height, with a 78 mm inner diameter. This cartridge contained relative little activated carbon compared to other commercially available cartridges. The organic solvents used for the experiment were dichloromethane, acetone, methyl acetate, ethyl acetate and carbon tetrachloride. In a previous test, these four organic solvents were shown to have shorter breakthrough times than that of carbon tetrachloride. Methods The schematic diagram of the test apparatus is illustrated in Figure 1. Two air flows from a compressor were saturated with water vapor and an organic solvent vapor, respectively, by passing them through impingers containing distillated water and an organic solvent. The two air flows were mixed in a mixing chamber with another air flow sent through a flowmeter to adjust the humidity and the concentration of the organic solvent vapor. Vapors of each organic solvent at 100, 300, 600 and 900 ppm, at 20 C and relative humidity of 50%, were generated. The steady-state flow tests were made by setting the respirator cartridge in its holder and passing each test vapor through the cartridge at the flow rate of 30 L/min. The pulsating flow tests were made by placing the respirator cartridge in a Tedler bag, expanding the bag by introducing each test vapor flow and inhaling the test vapor through the respirator cartridge using the breathing machine, SN-480 (Shinano Works, Tokyo). The exhaled gas of this machine was directly released out of the bag. The breathing machine drew 1.5 liters/cycle at 20 cycles/minute and the total flow rate was adjusted to be similar to the 30 L/min rate used in the steady-state flow test. The vapor concentration was measured at the upstream and downstream of the respirator cartridge with a gas chromatograph equipped with an auto-gas sampler and a hydrogen flame ionization detector (GC-8A Shimazu, Kyoto) at five-

3 BREAKTHROUGH OF SOLVENTS IN PULSATING FLOW ON CARTRIDGE 127 Fig. 1. A schematic diagram of the apparatus times of respirator cartridges. for measuring the breakthrough minute intervals. The breakthrough time was defined as the time required until the vapor concentration at the downstream of the respirator cartridge reached 5 ppm. RESULTS AND DISCUSSION The relationships between the logarithmic concentrations of organic solvent vapors in the test vapor flow and the logarithmic breakthrough times in the steadystate flow are illustrated in Figure 2. All organic solvent showed high inverse correlation. Table 1 shows the constants of the regression equations for logarithmic vapor concentration vs. logarithmic breakthrough time. The slopes of the lines for carbon tetrachloride and ethyl acetate were nearly equal to -1.0, while those of the lines for methyl acetate, acetone and dichloromethane were smaller. Nelson et al.3~ reported similar relationships for 9 kinds of organic solvent vapors, revealing that the slopes of the lines were smaller for the solvents with lower boiling points. This can be attributed to the fact that the vapor of a solvent with a low boiling point has such a low adsorption affinity to activated carbon that, at equilibrium, it is adsorbed at smaller than is the vapor of a solvent with a higher boiling point. The relationships between the logarithmic vapor concentrations and the logarithmic breakthrough times as determined by the pulsating flow tests are illustrated in Figure 3. The lines indicated inverse correlations similar to those found in the steady-state flow tests and the constants of these regression equations are shown in Table 2. The slopes of the lines were smaller for the solvents with lower boiling points in the same manner as in the steady-state flow. Table 3 shows the ratios of the breakthrough times in the pulsating flow to those in the steadystate flow. At 100 ppm, the ratio of breakthrough times of dichloromethane was as low as 0.77, while those of the other solvents were higher than 0.9, meaning that there were not large differences in the breakthrough times between the steady-

4 128 S. TANAKA et al. Fig. 2. The breakthrough time as a function of the organic solvent vapor concentration in steady-state flow. Each plot represents a mean of triplicate measurements. Air flow rate, temperature and relative humidity were fixed at 30 L/min, 20 C and 50% relative humidity, respectively. Table 1. The relationships between the concentration of organic solvent vapor and the breakthrough time in the steady-state flow. Regression Line: log (breakthrough time: min) = A log (vapor concentartion: ppm) + B

5 BREAKTHROUGH OF SOLVENTS IN PULSATING FLOW ON CARTRIDGE 129 Fig. 3. Breakthrough time as a function of the organic solvent concentration in pulsating flow. Each plot represents a mean of triplicate measurements. Tests were made at 1.5 liters/cycle, 20 cycles/min (total flow rate, 30 L/min), 20 C and 50% relative humidity. Table 2. The relationships between the concentration of organic solvent vapor and the breakthrough time in the poulsating flow.

6 130 S. TANAKA et al. Table 3. pulsating The flows ratio of the breakthrough times for five organic solvent vapors. between the steayy-state and Fig. 4. The flow patterns of the steady-state and pulsating flows.

7 BREAKTHROUGH OF SOLVENTS IN PULSATING FLOW ON CARTRIDGE 131 state and pulsating flows at 100 ppm. As the concentrations of the test vapor increased, the breakthrough times in the pulsating flow became shorter in comparison to that in the steady-state flow. The flow patterns of the steady state and pulsating flows are illustrated in Figure 4. The pulsating flow of a sine curve shows a maximum peak flow rate of 94.2 L/min, approximately three times that of the steady-state flow10~. The shorter breakthrough time of an organic solvent in the pulsating flow than in the steady-state flow is due to this large peak flow rate. It can be concluded from the present study that the breakthrough in the pulsating flow occurs earlier than in the steady-state flow at higher concentrations of these four organic solvents with low boiling points. REFERENCES 1) Nelson G0, Hodgkins DJ. Respirator cartridge efficiency studies II. Preparation of test atomospheres. Am Ind Hyg Assoc J 1972; 33: ) Nelson G0, Harder CA. Respirator cartridge efficiency studies V Effect of solvent vapor. Am Ind Hyg Assoc J 1974; 35: ) Nelson G0, Harder CA, Bigler BE. Respirator cartridge efficiency studies VI. Effects of concentration. Am Ind Hyg Assoc J 1976; 37: ) Freedman RW, Ferber BI, Harstein AM. Service lives of respirator cartridges versus several classes of organic vapors. Am Ind Hyg Assoc J 1973; 34: ) Wood G0, Moyer ES. A review and comparison of adsorption isotherm equations used to correlate and predict organic vapor cartridge capacities. Am Ind Hyg Assoc J 1991; 52: ) Yoon YH, Nelson JH, Lara J, Kamel C, Fregeau D. A theoretical interpretation of the service life of respirator cartridges for the binary acetone/m-xylene system. Am Ind Hyg Assoc J 1991; 52: ) Nelson G0, Harder CA. Respirator cartridge efficiency studies IV. Effects of steady-state and pulsating flow. Am Ind Hyg Assoc J 1972; 33: ) Emura H, Konishi A. Experimental study on service life of chemical cartridges under conditions of constant and sinusoidal flows. Jpn J Ind Health 1989; 31: 749 (in Japanese). 9) Murata M, Ikezaki Y, Yoshida Y. A mechanical breathing simulator for respirator testing. Health Physics 1976; 11: 51-6 (in Japanese). 10) Nozaki K. Method for studies on inhaled particles in human respiratory system and retention of lead fumes. Ind Health 1966; 4: ) Tanaka S, Kido S, Seki Y, Imamiya S. Service lives of respirator cartridges for 46 organic solvent vapors. Jpn J Ind Health 1993; 35: (in Japanese).