Determination of Ozone Removal Rates by Selected Building Products

May 18, 2001 - National Institute of Occupational Health, Lersø Parkallé 105,. DK-2100 Copenhagen Ø, Denmark. Ozone removal by 16 aged (older than ...
0 downloads 0 Views 84KB Size
Download PDF
Environ. Sci. Technol. 2001, 35, 2548-2553

Determination of Ozone Removal Rates by Selected Building Products Using the FLEC Emission Cell JACOB G. KLENØ, PER A. CLAUSEN, CHARLES J. WESCHLER,† AND PEDER WOLKOFF* National Institute of Occupational Health, Lersø Parkalle´ 105, DK-2100 Copenhagen Ø, Denmark

while fleecy or liquid products are inserted into specially designed test piece holders to ensure a proper seal (13). The airflow is distributed from a circular slit over the material surface into the center, where it exits the FLEC outlet. Under normal test conditions, the air velocity and the area specific ventilation rate over the surface of the building material are comparable to typical indoor settings and the airflow is supposedly laminar (13, 14). The purpose of this study was to develop an operational and reliable method using the FLEC to examine ozone removal by building products, in particular, by older building products with limited primary emissions.

Materials and Methods Ozone removal by 16 aged (older than 1-120 months) but unused building products or materials was studied in a test system that included the field and laboratory emission cell (FLEC). The ozone removal was studied at 50 ( 1 ppb ozone, a relative humidity of 50 ( 5%, a temperature of 21 ( 2 °C, and an air flow rate of 900 ( 10 mL min-1 through the FLEC (air velocity ca. 3 cm s-1). The ozone removal increased rapidly during the first 1-2 min and either remained at a constant level or decreased asymptotically to reach a steady state-like value. The ozone removal profiles for a given material showed good repeatability during replicate experiments. Ozone deposition velocities for the building products were calculated to be between 0.0007 cm s-1 (lacquered ash) and 0.8 cm s-1 (unpainted gypsum board).

Introduction Indoor ozone has received attention recently because of its possible adverse health effects (e.g., refs 1-3). The potential degradation of building products and cultural artifacts by ozone is also of concern (4, 5). In addition, emissions from building products in the presence of ozone can differ from those measured in the absence of ozone; oxidation processes that may not otherwise occur produce volatile organic compounds (6-9) of which some may be odorous compounds that influence the indoor air quality (10) for potentially long periods (11). Understanding of the chemical reactions of ozone within, at the surface of a building product, or in the gas phase with emitted compounds is limited but is important to assess human exposures to ozone, in addition to ozone reaction products. So far, labeling schemes for source control have been concerned with the emission of volatile organic compounds (i.e., primary emissions such as solvents), while secondary influences such as ozone deposition only recently have been an issue of concern. Ozone uptake by building materials has traditionally been studied in decay experiments using medium-sized climate chambers, usually made of stainless steel (see refs in 12). This study uses the field and laboratory emission cell (FLEC), which is a microemission cell made of stainless steel that is positioned upon the building material. The material and the inner surface of the FLEC form a cone-shaped cavity (13). Planar materials are sealed to the FLEC by means of an O-ring, † Department of Environmental and Community Medicine, UMDNJ/Robert Wood Johnson Medical School and Rutgers, 675 Hoes Lane, Piscataway, NJ 08854, USA. * Corresponding author phone: (+45) 39 16 52 72; fax: (+45) 39 16 52 01; e-mail: [email protected].

2548

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 12, 2001

Chemicals. Oxygen (99.999%, N2 99.5%), methanol (>99.8%), and dichloromethane (>99.5%) were from Merck. Building Products and Materials. Sixteen building products or materials were selected for testing. The building products A-O were older than 1 month, see Table 1 and prior to use were stored nonpackaged at 21 ( 2 °C and RH 30 ( 5%. The ozone level in the storage room was close to zero, and the air exchange rate was low. Stainless steel specimens N and O were additionally cleaned as described for the FLEC parts. Office paper P was taken directly from the packing and preconditioned as described below. Preconditioning of the Test Piece. Prior to the measurement, the test piece was preconditioned in pure nitrogen in a FLEC (CHEMATEC, Denmark) for 44 h. This is subsequently referred to as the conditioning FLEC. The flow through the FLEC was 900 ( 10 mL min-1 and the relative humidity (RH) was 50 ( 5%. RH was adjusted by an FLEC air supply (CHEMATEC, DK), in which the flow was split into a dry and a wet channel, respectively (16). RH was measured with a Testo 601 hygrometer (Testoterm GmbH & Co, Germany). The temperature was 21 ( 2 °C. Test Procedure. The test system in Figure 1 consisted of an ozone generator, an FLEC air supply, a test FLEC, and an ozone monitor (API model 400, San Diego, California) interfaced to a computer. The ozone concentration was measured at the outlet of the FLEC with a sample flow of 800 mL min-1 and a measurement cycle time of 8 s. Ozone was generated photochemically in pure oxygen (17) by a mercury lamp in a thermostated lamp housing controlled by a high performance variable power supply (for details, see ref 18). The synthetic air was supplied by mixing humidified nitrogen from the FLEC air supply with dry oxygen passing through the ozone generator to give XO2 ) 0.21 and XN2 ) 0.79. The ozone generator was adjusted manually to supply 50 ( 1 ppb ozone inside the FLEC with a glass plate present. RH was 50 ( 5% and the air flow was 900 ( 10 mL min-1. The temperature at the FLEC was 21 ( 2 °C for all experiments. The test FLEC was slightly modified from the conditioning FLEC to minimize ozone removal by the test system itself. The air distribution channels were hand-polished stainless steel, and the inlet tubes and fittings were replaced by Teflon (PFA) parts. The FLEC including its O-ring and the glass plate was used for each experiment and was cleaned with methanol. Ferules and Swagelok parts exposed to emissions from the test piece were cleaned in CH2Cl2. All parts were vacuum oven dried at 70 °C for 1 h. Prior to each test, the 10.1021/es000284n CCC: $20.00

 2001 American Chemical Society Published on Web 05/18/2001

TABLE 1. Ozone Deposition Velocities of 16 Selected Building Products or Materials (A-P) A B C D E F G H I J K L M N O P

building product or material

age (months)

n

kdbp (cm s-1)a

carpet (nylon fibers and latex backing) carpet (nylon fibers and latex backing) linoleum linoleum oiled beech parquet oiled ash untreated ash lacquered ash painted gypsum board (urethane modified alkyd binding agent) painted gypsum board (vinyl acetate binding agent) painted gypsum board (acrylic binding agent) unpainted gypsum board melamine-coated particle board stainless steel stainless steel (hand polished) office paper, white (80 g/m2)

>12 >12 >12 >120 >12 >12 >12 >12 >3 >3 >1 >14 >1 >1 >1 new

2 2 2 2 2 3 4 2 2 2 3 3 3 1 3 2

0.7 ( 0.4 0.032 ( 0.0043 0.007 ( 0.004 0.004 ( 0.005c 0.0078 ( 0.0027 0.003 ( 0.0007 0.007 ( 0.0015 0.0007 ( 0.0008c 0.030 ( 0.0052 0.67 ( 0.11 0.042 ( 0.0023 0.8 ( 0.4 nmd 0.007 0.010 ( 0.0049 >1.49

slope (ppb min-1)b 0.01 ( 0.23 -0.01 ( 0.17 0.01 ( 0.23 0.01 ( 0.35 0.02 ( 0.23 0.01 ( 0.20 0.03 ( 0.27 0.02 ( 0.34 0.01 ( 0.24 -0.002 ( 0.23 0.03 ( 0.26 0.01 ( 0.27 -0.01 ( 0.22 0.05 0.01 ( 0.17 nae

a Mean of k b The average slope of ozone concentration versus dbp calculated from eq 5, based on n experiments ( the 95% confidence interval. time in the time interval [100;120] minutes ( the 95% confidence interval. c The confidence interval of kdbp includes the value of zero. d Not measurable with this method. e Not available.

FIGURE 1. Experimental set-up for the study of ozone removal by building products. FLEC was placed on a clean glass plate, and the ozone concentration was adjusted to 700 ppb until a steady-state condition was reached with respect to ozone removal. Then, the ozone concentration was adjusted manually to 50 ( 0.5 ppb, still with the glass plate present. The incoming ozone concentration was somewhat higher because the interior of the FLEC, which is stainless steel, and the glass plate both remove ozone. These contributions are compensated for in the calculation of the ozone deposition velocities, see eqs 1-5. To measure the total ozone removal by the FLEC and

the glass plate together, the bypass circuit was opened, and the bypass ozone concentration was monitored until it had stabilized (after approximately 10 min). Finally, the bypass was closed and the ozone concentration in the test FLEC was monitored until it had stabilized at 50 ( 0.5 ppb. Following the period of preconditioning, the test piece was carefully transferred from the conditioning FLEC and placed in the test FLEC. The transfer was carried out in less than 2 s. The ozone concentration was then monitored for a period of 120 min. In all but one case, this procedure was repeated VOL. 35, NO. 12, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2549

two to four times for each test piece, see Table 1. The slope of the ozone concentration versus time was calculated for each experiment in the interval between 100 and 120 min to verify that the ozone removal had reached steady state. Confidence intervals are shown in Table 1. A value of zero is assumed to indicate a quasi-steady-state situation. The ozone generation rate, the air exchange rate in the FLEC, and the surface removal rates (the product of the surface-to-volume ratio and the deposition velocity) determine the ozone concentration at steady state. Gas-phase reactions are suppressed due to a short residence time of 5 s in the FLEC and the tubing. Surface loss on the tubing is negligible relative to the other loss terms. Steady-state ozone concentrations for the glass plate and the building product in question are

Cg ) Cbp )

Ri

(1)

Ex + kdg(A/V) + kdFLEC(A′/V) Ri

(2)

Ex + kdbp(A/V) + kdFLEC(A′/V)

Where Cg and Cbp are steady-state ozone concentrations with the FLEC placed on the glass plate and the building product (or the material), respectively. The values are calculated as mean concentrations in the time interval between 100 and 120 min. Ri is the ozone generation rate (ppb s-1), and Ex is the air exchange rate which equals 0.429 s-1 at an air flow rate of 15 mL s-1 through the FLEC. kdg, kdFLEC, and kdbp are ozone deposition velocities (cm s-1) of the glass plate, the inner surface of the FLEC, and the exposed surface of the test piece. A and A′ are the exposed areas of the test piece and the inner surface of the FLEC, which equal 177 and 180 cm2, respectively. V is the volume enclosed by the FLEC and the top surface of the test piece, which equals 35 mL (13). The microstructure of fleecy or textured materials is not included in the A/V ratio (i.e., the value used for A is the nominal surface area). The deposition velocity of ozone for a particular building product or material can be found by combining eqs 1 and 2.

Cbp Ex + kdg(A/V) + kdFLEC(A′/V) ) Cg Ex + kdbp(A/V) + kdFLEC(A′/V)

(3)

The deposition velocity for glass, kdg, has previously been reported to be 0.001-0.0005 cm s-1 (18). This means that kdg(A/V) is 1% of Ex and justifies the approximation that leads to eq 4.

Discussion

Cbp Ex + kdFLEC(A′/V) ≈ Cg Ex + kdbp(A/V) + kdFLEC(A′/V) Equation 4 can be rearranged to give eq 5.

kdbp )

(

)(

) (

(4)

)

Cg Ex Cg Ex -1 -1 + kdFLEC(A′/A) ≈ Cbp A/V Cbp A/V

(5)

It is not possible to measure kdFLEC with the current method. However, the deposition velocity of stainless steel (kdSS) is reported to be 0.008-0.015 cm s-1 at RH 40-50% (12, 19). The term kdFLECA′/A in eq 5 can be evaluated using kdFLEC ≈ kdSS. With this approximation, we calculate that the omission of the term underestimates the calculated deposition velocities by 9-15% (see Discussion for a further evaluation of this term based on the experimental measured removal of ozone by the FLEC and glass plate).

Results The stability of the ozone concentration during the 120 min test period is illustrated in Figure 2 with the FLEC placed on 2550

9

a glass plate. The ozone concentration in the outlet of the FLEC was 93% of that entering the FLEC. Hence the FLEC, including the glass plate, removes 7% of the ozone. Ozone removal curves for selected building products and materials are shown in Figure 2. During the initial 3 min, the glass plate was in the FLEC. The system was in steady state with respect to ozone removal. The glass plate was then replaced by the test piece, and the ozone measurements were continued for another 120 min. The ozone removal was initially fast and either remained so throughout the experiment or declined to a more moderate rate. The ozone removal of nylon carpet A, painted gypsum board J, unpainted gypsum board L, and office paper P was substantial and quickly reached a constant value. For oiled beech parquet E and the melamine-coated particleboard M, ozone removal peaked briefly before declining. Hand polished stainless steel O showed a gradual shift from rapid to more moderate ozone removal rates. In all experiments, for a given test piece, the ozone removal process showed good repeatability as illustrated by the nearly coincident curves, see Table 1 and Figure 2. The ozone deposition velocities, kdbp, of the 16 slightly aged building products and materials have been calculated using eq 5. All calculations have been based on the average ozone removal in the range between 100 and 120 min of exposure. During this period, the slopes of the ozone concentration versus time were close to zero (see last column of Table 1). The ozone deposition velocities are listed in Table 1; values for different materials differ by 3 orders of magnitude. The lower the final ozone concentration in Figure 2, the larger the deposition velocity. In the case of the building products D and H, the 95% confidence intervals, include the value of zero. The ozone deposition velocity of the melaminecoated particleboard was smaller than that of the glass plate in two out of three experiments. Since the deposition velocity is calculated from the ozone removal of a building product relative to that of the glass plate, the calculated deposition velocity was negative (i.e., not measurable). The analysis of office paper P included two additional analyses. First, VOCs were sampled on Tenax TA at the outlet of the FLEC. Compounds present included alcohols, ketones, and aldehydes with abundances and rate constants that are too small to affect the ozone concentration during its short residence time. Second, the ozone generator was connected downstream of the FLEC to measure any ozone removal due to gas-phase reactions. The gas-phase reactions caused a 1% decline in the ozone concentration.

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 12, 2001

Repeatability and Limitations of the Method. The method is limited to test pieces that remove less than the incoming 50 ppb ozone. This concentration was chosen as a compromise to match a realistic indoor ozone level (3) and to avoid exceeding the mass-transfer limited value of the system. The limit of detection can be calculated from an experiment where nearly all the ozone has been removed. In this case, it is assumed that the standard deviation is equal to a situation where the ozone removal is 100%. The unpainted gypsum board in experiment L removes 45.6 ppb ozone. The standard deviation of the repeated measurements (n ) 3) is 0.90 ppb. If the limit of detection is calculated as 3 times the standard deviation, the maximum measurable ozone deposition velocity is 1.49 cm s-1 according to eq 5. However, this limitation can theoretically be overcome by increasing the incoming ozone concentration. In the case of material M, the ozone removal is smaller than that of the glass plate and hence the calculated deposition velocity becomes negative. A reference material with a smaller ozone removal than melamine must be applied to measure such low deposition velocities. The method also shows its limitation if the test

FIGURE 2. Ozone removal curves for selected building products and materials. The stability measurement on a glass plate illustrates the minimal fluctuations of the system. piece exhibits very little ozone removal and the standard deviation is comparably large. This is the case for materials D and H, since the confidence intervals of their deposition velocities include the value of zero. The deposition of ozone on the FLEC surface including the glass plate was measured to be 7% (of the incoming ozone) in all experiments under conditions similar to the measurements of ozone deposition velocities. The loss is attributed to the FLEC surface since kdSS . kdg. By inserting the 7% loss

in eq 1, kdFLEC can be estimated to be 0.004 cm s-1. Hence, the omission of the term kdFLECA′/A in eq 5 underestimates the calculated deposition velocities by 5%. This, further justifies the omission of kdFLECA′/A in eq 5. Mechanism/Description. The experiments presented here show that, initially, ozone removal is high and either remains high throughout the experiment or declines to a more moderate rate. However, as is apparent from Figure 2, the removal profiles for different materials differ markedly. VOL. 35, NO. 12, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2551

For carpet A and painted gypsum board J, ozone removal quickly reaches a large and constant value, whereas for oiled beech parquet E and the melamine coated particle board M ozone removal quickly levels off to a small rate. Stainless steel O shows a slow shift from rapid to more moderate ozone removal rates. The large initial ozone removals are presumably a consequence of the preconditioning of the building products (i.e., in a nitrogen atmosphere absent ozone). The ozone removal rates measured at times greater than 100 min are likely to be representative of indoor surfaces in real world settings. In the region from 100 to 120 min, the slopes of ozone concentration versus time are close to zero, which implies a steady-state ozone removal, see Table 1. The fact that ozone removal is replicable by the same test piece indicates a reverse aging of the building product during the preconditioning period of 44 h in a pure nitrogen atmosphere at RH 50% (cf. 19-20). The relatively good repeatability allows comparison of ozone removal by different building products. In actual indoor settings, ozone exposure occurs day after day; surfaces quickly become conditioned, and ozone removal is expected to occur in a predictable fashion. Lower I/O ratios for ozone in the winter versus other seasons have been observed (21). The authors speculated that when the ozone levels are low for an extended period the surfaces regain their scavenging efficiency for ozone. This observation is consistent with the ozone removal curves in Figure 2. Comparisons. The deposition velocities of the two carpets with similar fibers and backing (A and B) are significantly different. This is in contrast to the linoleum samples C and D and the stainless steel plates N and O, but it indicates that large differences can be expected within similar building products. The ozone deposition velocity of stainless steel has previously been reported to be 0.008-0.015 cm s-1 (12); the values for specimens N and O in Table 1 are consistent with this earlier study. Lee et al. (1) have measured a mean ozone deposition velocity of 0.049 cm s-1 in living rooms in California. This is 14 to 16 times lower than the deposition velocities for carpet A, painted gypsum board J and unpainted gypsum board L, but 5 to 70 times higher than those for linoleum C and D, wood E, F, G, and H, and stainless steel N and O. Deposition velocities for carpet B and painted gypsum boards I and K are comparable to those measured by Lee et al. (1) in residences and also those measured earlier by Shair (22) in offices. Other examples of indoor ozone deposition velocities (bedrooms, offices, labs, etc.) are reported by Cano-Ruiz et al. (12) and Nazaroff et al. (23); these range from 0.025 to 0.075 cm s-1. Implications. The fact that deposition velocities for the selected building products differ by more than 3 orders of magnitude may imply that surfaces such as linoleum C and D or lacquered ash H may be of minor importance as ozone consumers. Conversely, building products such as carpet A, painted gypsum board J, and unpainted gypsum board L may be responsible for the majority of the decay in the indoor environments. The surface areas must also be taken into consideration. Carpets and walls account for a large fraction of the overall indoor surface, indicating that their contribution to the total removal is expected to be high. In another recent study, it has been shown that the same above-mentioned carpet and painted gypsum board, when exposed to ozone, resulted in a significant change in air quality as perceived by a panel of judges, while linoleum C and melamine coated particle board M did not (10). Of the materials examined in this study, office paper P exhibits the greatest ozone deposition velocity, but it typically accounts for only a minor fraction of the overall indoor surface area. The office paper P was taken directly from the package and preconditioned as described without previous storage. In both experiments, it removed all ozone, i.e., it exceeded the mass-transfer limited value for the system. Gas-phase 2552

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 12, 2001

reactions accounted for 1% of the ozone removal, and hence the large removal is due to surface reactions. The large deposition velocity may be important to ozone scavenging within photocopying machines. Knowledge of the relative importance of the different ozone removal processes is integral to deriving the implications of ozone removal by indoor surfaces. Ozone removal by physical processes has different consequences than ozone removal by chemical processes. The latter is a potential source of odorous aldehydes (11). Therefore, it may be relevant to study the chemical reactions of certain building products with ozone. Ozone deposition may also influence the aging of building products and possibly impact the perception of indoor air quality. At this point, it is unclear to what extent the deposition velocities reported in Table 1 represent the catalytic destruction of ozone at active sites on the different surfaces versus the oxidation of chemicals or dangling carbon bonds associated with such surfaces. This should be a subject for future study, as should be the effect of relative humidity on such processes. The building products examined in this study were more than one month old but had not been exposed to actual indoor environments (i.e., they were unused). Their age implies that their primary emissions were quite low. The ozone removal properties of building products will probably change as the products age, become soiled, or are modified by cleaning processes. For this reason, the measured deposition velocities probably differ somewhat from those obtained by similar product surfaces in actual indoor settings. The method described in this paper can be applied in the field to determine ozone removal rates in actual indoor settings. This allows nondestructive measurements of deposition velocities by authentic building products. Such field studies could provide valuable insights regarding the influence of soiling, prior ozone exposure, humidity, and temperature on ozone deposition velocities. They would also provide knowledge regarding the range of deposition velocities displayed by surfaces that had experienced real world exposures.

Acknowledgments This work was supported by the Danish Research Agency under the program “Center for Development of Indoor Friendly Building Products 1998-2001”. The authors are grateful for the excellent work by Ms. A.B. Olsen. We thank the referees for constructive suggestions.

Literature Cited (1) Lee, K.; Vallarino, J.; Dumyahn, T.; O ¨ zkaynak, H.; Spengler, J. D. J. Air Waste Manage. Assoc. 1999, 49, 1238. (2) Morrison, G. C.; Nazaroff, W. W. Environ. Sci. Technol. 2000, 34, 4963. (3) Weschler, C. J. Indoor Air 2000, 10, 269. (4) Lee, D. S.; Holland, M. R.; Falla, N. Atmos. Environ. 1996, 30, 1053. (5) Cass, G. R.; Druzik, J. R.; Grosjean, D.; Nazaroff, W. W.; Whitmore, P. M. W. C. L. “Protection of Works of Art from Photochemical Smog.”; Final report submitted to the Getty Conservation Institute; Environmental Quality Laboratory, California Institute of Technology, Pasadena, 1988. (6) Moriske, H.-J.; Ebert, G. K.; Konieczny, L.; Menk, G.; Scho ¨ ndube, M. Toxicol. Lett. 1998, 96-97, 319. (7) Morrison, G. C.; Nazaroff, W. W.; Cano-Ruiz, J. A.; Hodgson, A. T.; Modera, M. P. J. Air Waste Manage. Assoc. 1998, 48, 941. (8) Reiss, R.; Ryan, P. B.; Koutrakis, P.; Tibbetts, S. J. Environ. Sci. Technol. 1995, 29, 1906. (9) Weschler, C. J.; Hodgson, A. T.; Wooley, J. D. Environ. Sci. Technol. 1992, 26, 2371. (10) Knudsen, H. N.; Nielsen, P. A.; Clausen, P. A.; Wilkins, C. K.; Wolkoff, P. In Proceedings of Healthy Building; 2000, Espoo; Seppa¨nen, O.; Sa¨teri, J., Eds., SIY Indoor Air Information Oy, Helsinki, August 2000, Vol. 4, p 217.

(11) Wolkoff, P. Sci. Tot. Environ. 1999, 227, 197. (12) Cano-Ruiz, J. A.; Kong, D.; Balas, R. B.; Nazaroff, W. W. Atmos. Environ. 1993, 27A, 2039. (13) Wolkoff, P. Gefahrstoffe - Reinhalt. Luft 1996, 56, 151. (14) Uhde, E.; Borgschulte, A.; Salthammer, T. Atmos. Environ. 1998, 32, 773. (15) Wolkoff, P. Atmos. Environ. 1998, 32, 2659. (16) Wolkoff, P.; Clausen, P. A.; Nielsen, P. A. Indoor Air 1995, 5, 196. (17) Dohan, J. M.; Masschelein, W. J. Ozone: Sci. Eng. 1987, 9, 315. (18) Wolkoff, P.; Clausen, P. A.; Wilkins, C. K.; Hougaard, K. S.; Nielsen, G. D. Atmos. Environ. 1999, 33, 693. (19) Mueller, F. X.; Loeb, L.; Mapes, W. H. Environ. Sci. Technol. 1973, 7, 342. (20) Sabersky, R. H.; Sinema, D. A.; Shair, F. H. Environ. Sci. Technol. 1973, 7, 347.

(21) Weschler, C. J.; Schields, H. C.; Naik, D. V. In Tropospheric Ozone and the Environment II: Effects, Modeling and Control, Air and Waste Management Association, Pittsburgh, PA November 1992, Berglund, R., Ed.; 1992; pp 681-700. (22) Shair, H. ASHRAE Trans. 1981, 87 (part 1), 116-139. (23) Nazaroff, W. W.; Gadgil, A. J.; Weschler, C. J. Critique of the use of deposition velocity in modelling indoor air quality, In Modeling of Indoor Air Quality and Exposure, ASTM STP 1205; Nagda, N. L., Ed.; American Society for Testing and Materials: Philadelphia, 1993; pp 81-104.

Received for review November 28, 2000. Revised manuscript received March 26, 2001. Accepted March 27, 2001. ES000284N

VOL. 35, NO. 12, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2553