Environ. Sci. Technol. 2002, 36, 846-853
Gas-Phase Organics in Environmental Tobacco Smoke. 1. Effects of Smoking Rate, Ventilation, and Furnishing Level on Emission Factors BRETT C. SINGER* AND ALFRED T. HODGSON E. O. Lawrence Berkeley National Laboratory, Berkeley, California 94720 KARLA S. GUEVARRA, ELISABETH L. HAWLEY, AND WILLIAM W. NAZAROFF University of California, Berkeley, California 94720
We measured the emissions of 26 gas-phase organic compounds in environmental tobacco smoke (ETS) using a model room that simulates realistic conditions in residences and offices. Exposure-relevant emission factors (EREFs), which include the effects of sorption and re-emission over a 24-h period, were calculated by mass balance from measured compound concentrations and chamber ventilation rates in a 50-m3 room constructed and furnished with typical materials. Experiments were conducted at three smoking rates (5, 10, and 20 cigarettes day-1), three ventilation rates (0.3, 0.6, and 2 h-1), and three furnishing levels (wallboard with aluminum flooring, wallboard with carpet, and full furnishings). Smoking rate did not affect EREFs, suggesting that sorption was linearly related to gas-phase concentration. Furnishing level and ventilation rate in the model room had little effect on EREFs of several ETS compounds including 1,3-butadiene, acrolein, acrylonitrile, benzene, toluene, and styrene. However, sorptive losses at low ventilation with full furnishings reduced EREFs for the ETS tracers nicotine and 3-ethenylpyridine by as much as 90 and 65% as compared to high ventilation, wallboard/ aluminum experiments. Likewise, sorptive losses were 4070% for phenol, cresols, naphthalene, and methylnaphthalenes. Sorption persisted for many compounds; for example, almost all of the sorbed nicotine and most of the sorbed cresol remained sorbed 3 days after smoking. EREFs can be used in models and with ETS tracer-based methods to refine and improve estimates of exposures to ETS constituents.
Introduction Environmental tobacco smoke (ETS) is a mixture of gases and particles whose chemical composition varies with environmental conditions because of transformation processes including the selective sorption of specific components to indoor surfaces. Indoor concentrations of all ETS components depend on smoking frequency, dilution volume, and ventilation rate. * Corresponding author phone: (510)486-4779; fax: (510)486-7303; e-mail:
[email protected]. 846
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 5, 2002
Exposure to ETS has been linked to an array of adverse health effects including heart disease, lung cancer, asthma, and impaired respiratory function (1-6). It remains uncertain which ETS components are responsible for each outcome, but recent studies indicate that exposure to filtered (i.e., gasphase) ETS increases tumor rates in A/J mice at the same rate as exposure to whole ETS (7). The potential importance of the gas phase is further suggested by the established toxicity of many volatile and semivolatile organic compounds in tobacco smoke (1); ETS contains many U.S. federally regulated hazardous air pollutants (HAPs) and California state-regulated toxic air contaminants (TACs) (8, 9). Quantifying exposures to gas-phase organic compounds from ETS presents significant challenges. Gas-phase organic concentrations may be measured directly in indoor environments where smoking occurs (10), but ETS contributions must be distinguished from those of other pollutant sources, e.g., formaldehyde from wood products and benzene from gasoline. Measurements of ETS-specific tracer compounds (tracers) can be combined with emission factor ratios to predict exposures to ETS constituents or to attribute measured contaminant concentrations to ETS versus other sources (11). Such tracer-based approaches require target constituents and tracers to have similar dynamic behavior in the relevant environment (12). Tracers of gas-phase ETS constituents include pyridine, 3-ethenylpyridine (3-EP), pyrrole, nicotine, and myosmine (11-17). ETS concentrations also may be estimated using mathematical models of the indoor environment. If smoking frequency, dilution volume, and ventilation rate are known with sufficient accuracy, dynamic concentrations of nonsorbing compounds can be modeled using appropriate ETS emission factors. Sorption can be explicitly included in these models, but the required kinetic rate constants are available for only a small fraction of the compound-surface combinations for which sorption may be significant. Prior measurements of ETS organic compound emission factors have been made in specialized test chambers that do not resemble typical indoor settings. Daisey et al. (18) measured emission factors for 21 gas-phase ETS components from six leading commercial brands and a reference cigarette in an unventilated stainless steel (SS) chamber. Martin et al. (19) also employed an unventilated SS chamber in their study of emission factors from 50 commercial cigarette products. The results from these two studies were consistent with each other and with previous studies to within a factor of 2-3 for most individual compounds (13, 20-22). In contrast, nicotine emissions appeared to be 5 times lower in SS chambers than in a Teflon chamber where little sorption was likely (13). This finding illustrates that, even in the absence of common indoor materials, sorption may strongly alter concentrations of some ETS constituents. Our study provides the experimental basis for an improved approach to estimating exposures to gas-phase ETS constituents. Exposure-relevant emission factors (EREFs), which include the effects of sorption and re-emission over a 24-h period, were calculated by mass balance from measured concentrations of ETS gas-phase organic compounds in a controlled room environment under simulated but realistic indoor conditions. These EREFs can be employed in indoor air quality models to provide more accurate estimates of exposure to ETS constituents. In addition, they can be used to refine ETS tracer-based methods of estimating exposure. Additional experiments were conducted in a SS chamber, under static and ventilated conditions, to provide a link between our results and previous studies. 10.1021/es011058w CCC: $22.00
2002 American Chemical Society Published on Web 01/26/2002
TABLE 1. Matrix of 24 Experiments Conducted in Model Room at Three Furnishing Levels, Three Smoking Rates, and Three Ventilation Levelsa no. of experiments wallboard only
wallboard/carpet
fully furnished
measd vent. rate (h-1)
20 cig
10 cig
5 cig
20 cig
10 cig
5 cig
20 cig
10 cig
5 cig
1.85-2.0 0.56-0.64 0.31-0.37
1
2 2 1
2
1
1 1 2
1 1b
1 1
2 2 2
1
b
a Cigarettes (cig) were smoked in the specified number over 3 h. Six cigarettes were smoked in 1-h period.
TABLE 2. Material Surface Areas for Stainless Steel Chamber and Model Room chamber/surface material
projected area (m2)
Stainless Steel Chamber (20 m3) 304 stainless steel 45.0
surface/vol ratio (m-1) 2.3
Model Room (49.5 m3) painted wallboard 64.2 aluminum or carpet floor 20.4 subtotal 84.6 furnishings wood & veneer furniture 18.7 upholstered chairs 13.9 drapery 10.5 subtotal 43.1
0.4 0.3 0.2 0.9
totalsfully furnished condition
2.6
128
1.3 0.4 1.7
Experimental Methods Twenty-four experiments were conducted in a model room containing typical residential materials. The experimental matrix, shown in Table 1, was chosen to provide independent axes of varying smoking and ventilation rate at each furnishing level. At least one replicate experiment was performed for each furnishing level. The number of combinations was limited by the labor and time required for each experiment. Five additional experiments were conducted in a test chamber with stainless steel surfaces. Model Room. The room had a volume of 49.5 m3 (4.61 m × 4.42 m × 2.43 m height). Table 2 lists surface areas of chamber materials and furnishings with resulting surface to volume ratios. The ceiling and walls were gypsum wallboard finished with low-VOC latex, flat wall paint. The plywood floor was covered with aluminum (Al) sheeting. There was a single door. The chamber was housed within a small building. Chamber air temperature was controlled at 21-23 °C during all experiments by regulating the building temperature. Relative humidity was in the range of 45-55% over the first 24 h for 16 of 24 experiments and was 35-45% in six others; humidity data were unavailable for two experiments. Clean air was supplied to the chamber by drawing outside air though activated carbon to remove VOCs. Ventilation was preset at nominal levels of 0.3, 0.6, or 2 h-1 by controlling the flow of supply air. These air exchange rates represent low, moderate, and high values for typical residences. The actual ventilation rate during each experiment was measured by first-order decay of injected SF6 (Table 1). There was no recirculation of air. Air exited the chamber through an exhaust port with a valve adjusted to maintain slight positive pressure in the chamber. Four small axial fans were placed ∼1 m from the corners, alternately at one-third and two-thirds of the room height, to provide air circulation.
During experiments labeled WB, the floor was covered with Al sheeting only. A second series of experiments labeled WB/CP was conducted with 20.4 m2 of unused but aged residential nylon carpet installed over the Al. A third series of experiments, labeled FF, was conducted after used furnishings were added to the carpeted chamber. These furnishings included cotton draperies, chairs upholstered with polyester fabric, and several wood and wood-laminate pieces (Table 2). Chamber background measurements confirmed that these had negligible emissions of the measured ETS constituents. Stainless Steel Chamber. The 20-m3 SS chamber was previously used and described by Daisey et al. (18). It was operated under both static (unventilated) and dynamic (ventilated) conditions. In static mode, air exchange was 1 m from the wall. The smoking cycle consisted of one 35-cm3, 2-s puff every minute with mainstream smoke exhausted from the chamber. Cigarettes were drawn from a pack and loaded into the carousel just before each experiment. Starting at time t ) 0, cigarettes were smoked by the machine in equal intervals over a 3-h period. Puffing was halted when the heat from the burning end triggered an infrared sensor below the cigarette. Puff count varied from 5 to 8 per cigarette, with most cigarettes receiving 6 or 7 puffs. Total smoking time from ignition to burnout was 7-9 min per cigarette. Weighing of new cigarettes and the butts from several experiments indicated that average tobacco consumption was 0.60 g/cigarette, consistent to within (5% among experiments. We tested the effect of smoking variations on emissions by conducting one of the three unventilated SS chamber experiments with smoldering cigarettes (i.e., no puffing). Target Compounds. Target compounds include regulated toxics (HAPs/TACs), ETS tracers, and other major components of ETS. They encompass very volatile organic compounds (VVOCs) such as 1,3-butadiene, isoprene, and acrolein with vapor pressures of 0.4-3 atm through semivolatile organic compounds (SVOCs) such as methylnaphthalenes, nicotine, and myosmine with vapor pressures of 10-5-10-6 atm. Gas-Phase Organic Compound Sampling and Analysis. Air samples were collected on Tenax-TA sorbent tubes (P/N CP-16251; Varian, Inc.) modified by substituting a 15-mm section of Carbosieve S-III 60/80 mesh (P/N 10184, Supelco Inc., Bellefonte, PA) for Tenax-TA at the outlet end. Sorbent tubes were attached to stainless steel tubing and inserted through ports in the chamber wall so that air was drawn directly into the tubes from the room, 0.6 m away from the wall. Air was sampled at rates of 1.6-7 mL min-1, using peristaltic pumps located outside the chamber. Pump flow rates were measured during each experiment and tracked over time. Prior to use, sorbent tubes were cleaned by helium purge at 275 °C for 30 min. After being sampled, tubes were capped and either analyzed the same day or stored in a freezer until analysis. Organic compounds collected on sorbent samplers were quantitatively analyzed by thermal desorption gas chromaVOL. 36, NO. 5, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
847
TABLE 3. Exposure-Relevant Emission Factors (µg/cigarette) for ETS Compounds Measured at Three Furnishing Levels and Three Ventilation Rates in a 50-m3 Model Room wallboard only compda
CVb
2 h-1
0.6 h-1
wallboard/carpet 0.3 h-1
2 h-1
0.6 h-1
fully furnished 0.3 h-1
2 h-1
0.6 h-1
0.3 h-1
Gas-Phase Organics
1,3-butadiene isoprene acrolein acrylonitrile 2-butanone acetonitrile benzene toluene ethylbenzene m,p-xylene o-xylene styrene 1,2,4-TMB naphthalene 2-Me-naphth 1-Me-naphth phenol o-cresol m,p-cresol
0.20 394 ( 53 314 ( 27 350 ( 67 409 ( 78 288 ( 32 261 ( 35 281 ( 38 394 ( 43 313 ( 35 0.19 2750 ( 370 2480 ( 210 2750 ( 520 2470 ( 470 2540 ( 280 2270 ( 300 1990 ( 270 2420 ( 330c 2830 ( 310 0.30 590 ( 123 453 ( 60 550 ( 163 577 ( 171 478 ( 82 373 ( 78 404 ( 84 557 ( 95 417 ( 71 0.11 215 ( 16 196 ( 9 191 ( 20 190 ( 20 219 ( 13 190 ( 14 161 ( 12 207 ( 13 187 ( 11 0.14 389 ( 38 334 ( 21 277 ( 39 269 ( 37 319 ( 35c 289 ( 28 327 ( 32 315 ( 25 275 ( 22 0.13 1110 ( 100 999 ( 57 942 ( 119 951 ( 121 1140 ( 90c 940 ( 84 858 ( 77 850 ( 140c 812 ( 59 0.12 517 ( 43 438 ( 23 387 ( 45 457 ( 54 432 ( 29 400 ( 33 429 ( 36 435 ( 29 429 ( 29 0.09 997 ( 61 895 ( 34 815 ( 70 899 ( 77 877 ( 44 826 ( 50 821 ( 50 863 ( 43 850 ( 42 0.09 158 ( 10 142 ( 6 125 ( 11 141 ( 12 139 ( 7 126 ( 8 129 ( 8 132 ( 7 128 ( 6 0.09 457 ( 30 404 ( 17 360 ( 33 405 ( 37 392 ( 21 365 ( 24 376 ( 25 383 ( 20 371 ( 20 0.10 87 ( 6 77 ( 4 71 ( 7 78 ( 8 75 ( 5 68 ( 5 74 ( 6 75 ( 5 68 ( 4 0.18 203 ( 25 189 ( 15 172 ( 30 173 ( 31 185 ( 19 132 ( 16 166 ( 21 178 ( 18 144 ( 15 0.16 72 ( 8 72 ( 5 55 ( 9 63 ( 10 67 ( 6 63 ( 7 74 ( 8 64 ( 6 57 ( 5 0.12 54 ( 5 40 ( 2 26 ( 3 42 ( 5 41 ( 3 28 ( 2 34 ( 3 25 ( 2 16.6 ( 1.2 0.14 35 ( 4 24 ( 2 16.3 ( 2.4 32 ( 5 25 ( 2 15.0 ( 1.5 21 ( 2 14.0 ( 1.2 7.8 ( 0.7 0.14 34 ( 3 23 ( 1 15.7 ( 2.3 31 ( 4 24 ( 2 14.5 ( 1.5 20 ( 2 13.0 ( 1.1 7.5 ( 0.6 d 0.21 289 ( 42 185 ( 17 83 ( 17 NA 163 ( 20 91 ( 13 175 ( 26 98 ( 12 49 ( 6 0.13 37 ( 3 23 ( 1 14.1 ( 1.8 32 ( 4 23 ( 2 12.9 ( 1.2 21 ( 2 12.3 ( 0.9 7.1 ( 0.5 0.12 68 ( 6 40 ( 2 23 ( 3 62 ( 7 41 ( 3 19.2 ( 1.6 41 ( 3 22 ( 1 12.3 ( 0.8
pyridine 2-picoline 3-picoline pyrrole 3-EP nicotine myosmine
0.11 489 ( 37 421 ( 20 361 ( 39 450 ( 49 385 ( 24 0.12 154 ( 13 146 ( 8 132 ( 16 156 ( 19 138 ( 10 0.09 368 ( 24 332 ( 14 288 ( 27 354 ( 33 297 ( 16 0.18 362 ( 45 255 ( 20 277 ( 48 392 ( 68 301 ( 30 0.14 744 ( 71 575 ( 35 479 ( 65 634 ( 86 507 ( 40 0.12 3750 ( 330 2040 ( 110 1600 ( 200 3020 ( 380 1540 ( 110 0.18 240 ( 31 106 ( 9 67 ( 12 206 ( 37 92 ( 10
ETS Tracers
351 ( 27 130 ( 11 273 ( 18 197 ( 24 412 ( 39 873 ( 77 50 ( 6
364 ( 28 114 ( 10 264 ( 17 230 ( 28 428 ( 41 1270 ( 110 83 ( 11
348 ( 22 107 ( 7 231 ( 12 231 ( 23 380 ( 30 689 ( 50 46 ( 5
292 ( 18 81 ( 6 179 ( 10 157 ( 16 272 ( 21 396 ( 29 25 ( 3
a 1,2,4-TMB, 1,2,4-trimethylbenzene; 2-Me-naphth, 2-methylnaphthalene; 1-Me-naphth, 1-methylnaphthalene; 3-EP, 3-ethenylpyridine. b Coefficient of variation (CV) was calculated from the ratios of eight pairs of replicate experiments. Uncertainty was calculated as the mean values multiplied by the CV and divided by the square root of n experiments. c Uncertainty was calculated directly from variability of experiments at these conditions, not using global CV. d Not available; value uncertain due to high background concentration.
tography/mass spectrometry (TD-GC/MS) generally following U.S. Environmental Protection Agency (EPA) Method TO-1 (23). Chemical analysis methods were similar to those reported by Daisey et al. (18) except that samples were thermally desorbed and concentrated on a cryogenic inletting system (model CP-4020 TCT; Varian, Inc.). The system was fitted with a Tenax-packed trap (P/N CP-16425; Varian, Inc.) to avoid loss of the most volatile compounds. Desorption temperature was set to 235 °C for 6.5 min. The cryogenic trap was held at -100 °C and then heated to 235 °C for injection. Multipoint calibrations were referenced to an internal standard of 1-bromo-4-fluorobenzene. Most experiments started with the measurement of background organic compound concentrations in the chamber. On occasions when experiments occurred without a break, concentrations from the last day of the earlier experiment were used as the background concentrations for the next experiment. Time-averaged organic compound concentrations were measured in discrete increments, for up to 96 h from the start of smoking (t ) 0). Air sampling typically occurred over the following intervals: 0-4 h, 4-12 h, 12-24 h, 24-48 h, and 48-72 h. Shorter intervals were used to collect lower volume samples for conditions resulting in higher compound concentrations. Duplicate samples were collected during many periods. Concentrations calculated from these collocated samples agreed within (10% in most cases. A possible sampling problem was observed for 1,3butadiene, isoprene, acrolein, and acrylonitrile. In experiments with the lowest overall ETS concentrations, higher VVOC concentrations were sometimes observed in lower volume samples as compared with collocated higher volume samples, possibly indicating sample breakthrough or incomplete recovery. The difference was e50% for isoprene, 848
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 5, 2002
e20% for the other compounds, and most prominent in smoking period samples. When available, we used data only from the lower volume sample to calculate emission factors. Since lower volume samples were not always collected, isoprene and to a lesser extent 1,3-butadiene, acrylonitrile, and acrolein emission factors may be understated for some high-ventilation (2 h-1), low-smoking rate (5 cigarettes day-1) experiments. Additional uncertainties resulted because acrolein, acrylonitrile, and to a lesser extent isoprene, acetonitrile, and 2-butanone masses were below the mass limits of quantitation for many background and post-smoking samples. When this occurred, we calculated emission factors based on the periods for which quantifiable responses were available, using one-half the mass LOQ to estimate concentrations as necessary for background samples. Thus, emission factors for acrolein and acrylonitrile are based on the 4-h smoking period concentrations alone for most of the 2 h-1 experiments; on 8-12 h of sampling data for the 0.6 h-1 experiments; and on 12-24 h for the 0.3 h-1 experiments. Calculation of Emission Factors. Background-corrected organic compound concentrations (µg m-3) were multiplied by the measured ventilation flow rate (m3 h-1) to calculate the mass of each constituent removed from the chamber during each sample period. EREFs were calculated by dividing the total mass removed over 24 h by the number of cigarettes smoked during the initial 3 h. For the static SS chamber experiments, emission factors for compounds that did not sorb appreciably were calculated by direct mass balance (i.e., mass emitted equals mass added to chamber air) using concentrations measured over the first hour after smoking was completed. Samples collected during subsequent hours were analyzed to identify compounds whose concentrations decreased over time, presumably from sorption. Emission
FIGURE 1. Exposure-relevant emission factors (EREFs) of selected compounds in ventilated stainless steel (SS) chamber and in model room at nine conditions. WB, wallboard with aluminum sheeting on floor; WB/CP, wallboard with carpet; FF, fully furnished (see text and Table 1 for details). Ventilation rates are expressed as air changes per hour (ach). Uncertainty is shown as ( one standard error for replicate experiments. factors for these compoundssnaphthalene, methylnaphthalenes, phenol, cresols, 3-EP, nicotine, and myosmines were estimated by regressing the time-dependent results to estimate concentrations and thus emissions at time t ) 0, as described by Daisey et al. (18). All three static SS experiments were used for nonsorbing compounds, whereas only the two experiments with the standard smoking cycle were used for the sorbing compounds. Uncertainty. Reported uncertainty bounds analogous to one standard error were calculated as follows. We first calculated the ratio of emission factor results for each of eight pairs of replicate experiments. A global coefficient of variation (CV) was then calculated for each compound based on these eight ratios. This CV describes the overall experimental variability of our approach and includes uncertainties in sample volumes, variations in the application of the internal standard and TD-GC/MS performance, inadvertent variations in cigarette handling and smoking-machine operation, and other factors. The product of the compoundspecific CV and the mean estimate by condition were divided by the square root of n experiments at the same conditions.
Results and Discussion Exposure-Relevant Emission Factors. EREFs for 26 gasphase ETS organic compounds measured under various conditions in the model room are presented in Table 3. The good precision of EREF measurements is indicated by low
percentage uncertainties, ranging from 10% or less for some compounds (toluene, ethylbenzene, m,p-xylene, 3-picoline) to a maximum of 30% (acrolein for WB/CP at 2 h-1). EREFs as a function of room conditions are illustrated for six representative compounds in Figure 1. At a given furnishing level and ventilation rate, variations in daily smoking rate did not affect EREFs, even for the lowest volatility compounds. As an example, Figure 2 shows nicotine EREFs calculated for three smoking rates at each furnishing level, all at 0.6 h-1. The consistency of EREFs across varied smoking rates suggests that sorptive losses were linear with room air concentrations. Consequently, different smoking rate experiments at each unique combination of furnishing level and ventilation rate were averaged together. The stability or variability of EREFs across experimental conditions provides evidence about the consistency of emissions and the significance of sorption and other interactions with indoor surfaces. Specifically, a compound that is emitted at consistent rates across experiments, that is measured accurately, and that does not interact significantly with surfaces will exhibit consistent EREFs across all measurement conditions. For compounds that sorb strongly (and do not desorb rapidly) or are otherwise removed via surface interactions, emission factors will tend to be lowest for the fully furnished room (more surfaces) and for the lowest ventilation rate (more time for interaction). Fluctuations in EREFs that do not exhibit these features probably reflect VOL. 36, NO. 5, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
849
FIGURE 2. Variation in nicotine exposure-relevant emission factor as function of number of cigarettes smoked in model room ventilated at 0.6 h-1.
FIGURE 4. Effect of elapsed time interval from initiation of smoking on exposure-relevant emission factors (EREFs) of selected compounds in fully furnished model roon at nominal ventilation rates of 0.3, 0.6, and 2 ach.
FIGURE 3. Effect of furnishing level on exposure-relevant emission factors (EREFs) of target compunds. Values are emission factor ratios of fully furnished to wallboard only experiments in model room averaged for three ventilation conditions. either sampling limitations or inconsistent emissions from cigarettes. Many of the compounds exhibited little effect of surface interaction across experimental conditions. Figure 3 illustrates this by presenting the ratio of average EREFs for fully furnished experiments to the corresponding average for wallboard-only experiments. For half of the compounds (1,3-butadiene to 1,2,4-trimethylbenzene in Figure 3), ratios vary from 0.83 to 0.98, indicating that interactions with the furnishings reduced EREFs for these compounds by less than 20% as compared to a bare room. The effects of sorption in the model room were greatest for nicotine (see also Figure 1) and myosmine. Both 850
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 5, 2002
compounds exhibited clear trends, such that for a given furnishing level the EREF monotonically decreased with decreasing ventilation rate (see Table 3). Also, for a given ventilation rate, the EREF monotonically decreased with increasing furnishing level. The overall effect of sorption for these compounds over the range of experimental conditions is striking. Fully furnished, low ventilation conditions yielded EREFs for nicotine and myosmine that were only about 10% of corresponding values under wallboard-only, high ventilation conditions (Table 3). The other 11 compounds (naphthalene through 3-EP in Figure 3) exhibited intermediate behavior. By their construction, the EREFs in Table 3 incorporate sorption losses during active smoking and subsequent reemission over 24 h. Chamber air concentrations of moderately sorbing compounds such as 3-EP returned to background levels within 24-48 h, but concentrations of naphthalene, methylnaphthalenes, phenol, cresols, nicotine, and myosmine often remained higher than the previously measured background levels for several days following a smoking period. Calculated EREFs would thus increase with an increase in the post-smoking sampling interval for these compounds. Figure 4 shows the effects of post-smoking sampling duration on the calculated EREFs for five compounds in FF experiments at the three ventilation rate conditions.
TABLE 4. Emission Factors (µg/cigarette) for ETS Target Compounds in Stainless Steel Chamber Experiments and Comparisons with Published Studies compda
ventilatedb 2 h-1
unventilatedb
