Environ. Sci. Technol. 2009, 43, 1329–1335
Chlorine Dioxide Reactions with Indoor Materials during Building Disinfection: Surface Uptake
TABLE 1. Properties of Chlorine Dioxide property
†
HEIDI HUBBARD, DUSTIN POPPENDIECK,‡ AND R I C H A R D L . C O R S I * ,§ Research Triangle Park, United States Environmental Protection Agency, North Carolina 27711; Department of Environmental Resources Engineering, Humboldt State University, 1 Harpst Street, Arcata, California 95521; and Department of Civil, Architectural, and Environmental Engineering, The University of Texas at Austin, 10100 Burnet Road (R7100, Bldg 133), Austin, Texas 78758
Received July 12, 2008. Revised manuscript received November 27, 2008. Accepted December 15, 2008.
Chlorine dioxide received attention as a building disinfectant in the wake of Bacillus anthracis contamination of several large buildings in the fall of 2001. It is increasingly used for the disinfection of homes and other indoor environments afflicted by mold. However, little is known regarding the interaction of chlorine dioxide and indoor materials, particularly as related to the removal of chlorine dioxide from air. Such removal may be undesirable with respect to the subsequent formation of localized zones of depleted disinfectant concentrations and potential reductions in disinfection effectiveness in a building. The focus of this paper is on chlorine dioxide removal from air to each of 24 different indoor materials. Experiments were completed with materials housed in flow-through 48-L stainless steel chambers under standard conditions of 700 ppm chlorine dioxide inlet concentration, 75% relative humidity, 24 °C, and 0.5 h-1 air changes. Chlorine dioxide concentration profiles, deposition velocities, and reaction probabilities are described in this paper. Deposition velocities and reaction probabilities varied over approximately 2 orders of magnitude across all materials. For most materials, deposition velocity decreased significantly over a 16-h disinfection period; that is, materials became smaller sinks for chlorine dioxide with time. Four materials (office partition, ceiling tile, medium density fiberboard, and gypsum wallboard) accounted for the most short- and long-term consumption of chlorine dioxide. Deposition velocity was observed to be a strong function of chlorine dioxide inlet concentration, suggesting the potential importance of chemical reactions on or within test materials.
Introduction Chlorine dioxide (ClO2) in solution has been extensively used for disinfection of drinking water, biocide treatment of cooling water, sterilization of medical devices, and bleaching of pulp (1). The detailed chemistry of ClO2 in aqueous solutions has been well understood for many decades (2). Chlorine dioxide * Corresponding author phone: (512) 475-8617;
[email protected]. † United States Environmental Protection Agency. ‡ Humboldt State University. § The University of Texas at Austin. 10.1021/es801930c CCC: $40.75
Published on Web 01/29/2009
2009 American Chemical Society
e-mail:
CAS molecular weight boiling point melting point specific gravity (0 °C) solubility in water (20 °C) vapor pressure (20 °C) Henry’s law constant (20 °C)
value and unit
reference/ comment
10049-04-4 67.45 11 °C -59 °C
4 4 4
1.642
4
8 g/L
5
101 kPa 4.01 × 10-2 atm · m3/mol 1.67 m3liq/m3gas
5 6 based on ref 6 and ideal gas law
is a selective oxidant, involving a single electron transfer from an electron donor. In solution, chlorine dioxide is known to react with natural organic matter and metals such as iron and manganese and is rapidly converted to the chlorite ion (ClO2-) (3). In the disinfection of drinking water, approximately 70% of the chlorine dioxide dose is observed as ClO2-, with the remaining 30% typically observed as Cl-, OCl-, and ClO3- ions (3). Acidic solutions (pH < 7) favor the reformation of aqueous ClO2, which exists as a dissolved gas in water. Selective properties of chlorine dioxide are presented in Table 1. The use of gaseous chlorine dioxide as a disinfection/ decontamination agent has been increasing in recent years. In the food/agricultural industry, chlorine dioxide disinfection effectiveness has been shown for the decontamination of apples inoculated with Escherichia coli (E. coli) (7), strawberries containing E. coli and Listeria monocytogenes (8), and carrots containing mesophilic aerobic bacteria, psychotrophs, lactic acid bacteria, and yeasts (9). Chlorine dioxide gas has recently gained attention as a means for treating mold afflicted buildings in the wake of hurricanes in the Gulf Coast region of the United States. However, its use for treatment of mold is not new. Gaseous chlorine dioxide has been used to eradicate mold in libraries for nearly 20 years (10). Wilson et al. (11) inocculated paper filters with four different types of fungi and exposed them to 500-1000 ppm of chlorine dioxide gas for 24 h. The process was effective to a degree at inactivating the fungi, but S. Chartarum remained toxic even after these relatively high gas-phase exposures. Chlorine dioxide gas has also been used to effectively deactivate Bacillus anthracis spores in actual buildings following the contamination of several government and commercial buildings in the fall of 2001. Its effectiveness for deactivating other bacillus bacteria in more controlled settings has also been studied. Buttner et al. (12) tested chlorine dioxide gas to decontaminate an experimental room challenged with endospores of Bacillus atrophaeus. Chlorine dioxide was applied at 1400 ppm in room air for four hours at a relative humidity (RH) of 85% and a temperature of 24 °C. Twenty-four of 27 samples were observed to contain no culturable spores of the bacteria. Han et al. (13) studied the effectiveness of chlorine dioxide gas for the deactivation of Bacillus thuringiensis spores on four materials. The gas was applied at approximately 1800-11 000 ppm for 12 h. The effectiveness of deactivation was observed to be material VOL. 43, NO. 5, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1329
TABLE 2. Test Materials graph label flooring material carpet, LL carpet, NLL aged carpet, NLL carpet, PVC linoleum linoleum w/ polish VCT VCT w/ polish concrete concrete w/ sealer
level loop carpet nonlevel loop carpet nonlevel loop carpet (aged) PVC-backed carpet linoleum linoleum with polish vinyl composition tile vinyl composition tile with polish concrete concrete with sealer
1
1785 1773 1769 1760 1880 2330 2303
other office material medium density fiberboard1 MDF 1 medium density fiberboard with laminate MDF w/ laminate particle board1 particle board office partition2 office partition painted metal (file cabinet)2 filing cabinet paper (50 stacked sheets, copier paper) paper paper (50 stacked sheets, copier paper with toner) paper w/ toner
1875 1793 1833 1923 1365 667 667
Edge and back coated with sodium silicate.
2
Edge coated with sodium silicate.
Experimental Procedures Experiments were completed using a flow-through benchscale experimental chamber system. Standard experiments were conducted at 700 ppm chlorine dioxide inlet concentration, 75% relative humidity, and 24 °C with an air exchange rate of 0.5 h-1. Conditions were chosen to approximate previous full-scale disinfection scenarios. Twenty-four building materials were analyzed using these conditions. Nine additional experiments were completed to explore the effects of environmental parameter variations on select materials. Control (empty) chamber experiments were always completed in parallel to experiments involving material test specimens. Experimental Sequence. Each experiment involved three stages. The first stage (predisinfection) was completed prior 9
1753 1660 1742 1724 1894 1465 967 967 1154 1158
wall and ceiling material gypsum wallboard backing1 GWB backing gypsum wallboard with flat acrylic paint1 GWB w/ flat paint 1 gypsum wallboard with satin acrylic paint GWB w/ satin paint gypsum wallboard with wallpaper (PVC coated)1 GWB w/ wallpaper acoustic ceiling tile1 ceiling tile HVAC duct and liner HVAC duct HVAC duct and liner (aged) aged HVAC duct
dependent, with spores deposited on paper having a much higher survival rate than those deposited on three smoother and less porous materials. The removal of chlorine dioxide gas to indoor materials is important as it may adversely impact the effectiveness of decontamination, cause damage to materials, and possibly produce unwanted and persistent byproducts. However, other than anecdotal observations, little is known about the interactions that occur between chlorine dioxide and interior building materials. Han et al. (13) and references cited therein reported that chlorine dioxide gas injected into the Hart Senate building to deactivate Bacillus anthracis spores was rapidly consumed, presumably through removal to indoor surfaces. A 6-fold increase in chlorine dioxide injection time (i.e., relative to design injection time) to treat mold in a library was attributed to chlorine dioxide reactions with the surfaces of books (10). In this paper, we include the results of a series of laboratory chamber experiments intended to improve the existing knowledge base related to gas-phase interactions between chlorine dioxide and a wide range of indoor materials.
1330
horizontally projected surface area (cm2)
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 5, 2009
to injection of chlorine dioxide into the experimental system. For each experiment, the predisinfection stage lasted nine hours. The predisinfection stage was followed by a 16-h disinfection stage (i.e., time during which chlorine dioxide was injected into the experimental system). The disinfection stage was followed by a 20-h postdisinfection stage, during which time chlorine dioxide concentrations were allowed to decay by chamber ventilation. Test Materials. Twenty-four materials were exposed to chlorine dioxide. Materials and their shortened graph labels are listed in Table 2. All materials were new unless otherwise noted (i.e., marked as “aged”). The aged HVAC duct and liner were removed after use in a building on the University of Texas (UT) campus. The aged carpet was purchased new and placed in a high-traffic entryway of a building at UT for approximately six months prior to testing. New materials were allowed to off-gas for at least three months prior to use. Experimental System. The core of the experimental system was four identical 48-L electropolished stainless steel chambers that were positioned in parallel. This allowed for the analysis of three materials at any given time, with the fourth chamber acting as an empty control chamber. A zero air generator (Perma Pure Zero-Air generator) was used to supply clean air to the system. The relative humidity of the supply air was adjusted through the use of a split stream; one stream was bubbled through a humidification column before being reconnected to the dry stream. A relative humidity probe (TSI, Inc., Q-TRAK model 8551) was used in-line following humidification to measure influent air relative humidity and temperature. Chlorine dioxide was generated using a gas-solid system (CDG gas-solid bench-scale system) that employed chlorine gas at 4% by volume in a nitrogen makeup gas flowing through a bed of solid sodium chlorite pellets. The corresponding reaction of 2 moles of NaClO2 (s) and 1 mole of Cl2 (g) yielded 2 moles of ClO2 (g) (i.e., 8% by volume). The chamber inlet
FIGURE 1. Concentration profiles for several test materials. chlorine dioxide concentration was diluted using zero air to a standard value of 700 ppm. Mass flow controllers (Aalborg GCF171S) were used to regulate air flows to each chamber and to disinfectant monitoring systems. An electronically actuated sampling valve (Vici Instruments, CondyneTM Model 100) was programmed to sequentially sample from each chamber. Chlorine dioxide was analyzed using an Interscan continuous monitoring system (model LD33-2) based on an electrochemical cell. The chlorine dioxide monitor was calibrated by the manufacturer immediately prior to experiments, with confirmation at the approximate midpoint of the experimental program. Data Analysis. Chlorine dioxide consumption was analyzed as a heterogeneous reaction at material surfaces. Corresponding deposition velocities (vd) for control chamber walls and each building material were determined based on a mass balance on each chamber. Previous tracer gas studies and the control chamber concentrations observed in this study indicated that well-mixed conditions existed inside the chambers. Equation 1 was used as the governing equation for chlorine dioxide concentration inside each well-mixed chamber: Ass dC A ) λCo - λC - vdC - vd,ssC dt V V
(1)
where C is the chlorine dioxide concentration inside the chamber (and in chamber exhaust) (mg/m3), Co is the chlorine dioxide concentration entering the chamber (mg/m3), λ is the air exchange rate for the chamber (h-1), vd is the chlorine dioxide deposition velocity for a material (m/h), vd,ss is the chlorine dioxide deposition velocity for the walls of the chamber (m/h) as determined from control experiments, A is the projected exposed area of the material sample (m2), Ass is the exposed area of the stainless-steel walls of the chamber (m2), and V is the volume of the chamber minus the volume occupied by the material sample (m3). For each experiment, the air exchange rate was measured by dividing the average air flow rate through the chamber by the chamber volume. Equation 1 was solved in discrete form for deposition velocity:
2 n [C - Cn+1] + λ[Con+1 + Con - Cn+1 - Cn]∆t Ass vd,ss(t ) tave) [Cn+1 + Cn] V ave vd(t ) t ) ) A n+1 [C + Cn] V (2) where n and n + 1 indicate consecutive data points and tave is the time midway between the times corresponding to data points at n and n + 1. The time varying deposition velocity for each material was obtained by solving eq 2 for each consecutive pair of concentration data points. The deposition velocity associated with chamber walls (vd,ss) was determined by a mass balance on a control chamber in which the walls were considered to be the material of interest; that is, the last term in the numerator of eq 2 was omitted, A in the denominator corresponds to the collective area of chamber walls, and the left-hand-side of eq 2 corresponds to vd,ss. The deposition velocity to chamber walls was observed to be negligible for all experiments.
Results and Discussion Chlorine Dioxide Concentrations. Figure 1 shows the concentration profile for a selected group of materials. The vertical line at hour 9 indicates when the chlorine dioxide was switched on. The vertical line at hour 25 indicates when the chlorine dioxide was switched off. The top set of data points in Figure 1 is the concentration profile for the control chamber. The control chambers for all experiments had similar chlorine dioxide concentration profiles. There was negligible removal of chlorine dioxide to chamber walls, and each control chamber approximated the curve for an ideal continuous-flow stirred-tank reactor. The relative reactivity of each material is seen as the distance of the corresponding chlorine dioxide concentration curve from the chlorine dioxide concentration curve for the control chamber. The least reactive materials have higher concentrations of chlorine dioxide in the chambers than the more reactive materials. For this example, concrete is the least reactive material per unit of horizontally projected surface area. Surface reactions with gypsum wallboard backing and medium density fiberboard consumed the most chlorine dioxide. Deposition Velocity. Deposition velocities for a representative group of materials using a log scale on the vertical axis are shown in Figure 2. Values of deposition velocity VOL. 43, NO. 5, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1331
FIGURE 2. Deposition velocities for several test materials.
FIGURE 3. Geometric mean deposition velocities for the first hour of disinfection (experimental hours 9-10). spanned approximately 2 orders of magnitude, with ceiling tile the most reactive of all materials tested. The scatter in Figure 2 reflects the sensitivity of the deposition velocity calculation to slight changes in chlorine dioxide concentration. Deposition velocities varied with time. For most materials, there was a rapid decay in deposition velocity during the first several hours of the disinfection stage. Following this, the deposition velocity decayed more slowly for the remainder of the experiment. The geometric mean chlorine dioxide deposition velocities during the first hour and last five hours of the disinfection stage for each test material are presented in Figures 3 and 4, respectively. The mean deposition velocity calculated during the first hour of the disinfection stage is used to represent the initial deposition velocity. The mean deposition velocity calculated over the final five hours of the disinfection stage is used to represent the final, relatively constant, 1332
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 5, 2009
deposition velocity. The final four materials listed in Figure 4, office partition, MDF, gypsum wallboard backing, and ceiling tile, appear to account for the majority of the longterm consumption of chlorine dioxide. This is important from a practical standpoint because these materials may comprise a large surface area in buildings thus consuming relatively large amounts of chlorine dioxide and potentially reducing the effectiveness of disinfection. In addition, the persistent high reactivity between chlorine dioxide and both ceiling tile and gypsum wallboard backing has potential implications with respect to difficulties in deactivating biological agents in wall cavities and open plenums. Paper, paper with toner, HVAC duct, and aged HVAC duct had significant first-hour chlorine dioxide consumption (Figure 3), but all were quickly limited in their ability to consume chlorine dioxide. High initial chlorine dioxide reactivity in HVAC ducts may result
FIGURE 4. Geometric mean deposition velocities for the last five hours of the disinfection stage for chlorine dioxide (experimental hours 20-25). For comparison, data from a previous study (see ref 14) for the geometric mean deposition velocities for the last five hours of the disinfection stage for ozone are also shown.
FIGURE 5. Reaction probabilities for several test materials. in longer disinfection times if the HVAC system is used to deliver chlorine dioxide during disinfection. For comparison, ozone deposition velocities from a previous study by the authors are also shown in Figure 4 for similar materials and experimental conditions (14). For the less reactive materials, the ozone deposition velocity was generally much greater toward the end of the disinfection event in comparison with those for chlorine dioxide. However, for the more reactive materials, deposition velocities for chlorine dioxide were comparable to those for ozone and even greater for ceiling tile, particle board, and HVAC duct. Reaction Probability. Deposition velocities depend on both the rate of transport of molecules to the surface and the rate of reaction on the surface (15). The surface reaction rate is governed by the rate of collision of chlorine dioxide molecules with the surface as well as by the reaction probability, defined as the number of molecules that react
with the surface divided by the number of molecules that collide with the surface. The overall resistance to deposition is the inverse of the deposition velocity. This resistance can be further parsed into transport-related resistance (i.e., the resistance associated with transport to the material surface) and surface resistance (i.e., the resistance associated with reactions on the surface). Equation 3 shows the relationship between overall, transport, and surface resistances: RD )
1 1 1 1 4 ) Rt + Rs ) + ) + vd vt vs vt vbγ
(3)
where, RD is the overall resistance to deposition (h/m), Rt is the transport resistance (h/m), Rs is the surface resistance (h/m), vt is the transport-limited deposition velocity (m/h), vs is the surface-limited deposition velocity (m/h), vb is the VOL. 43, NO. 5, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1333
FIGURE 6. Deposition velocities for carpet with PVC backing under varied concentrations.
FIGURE 7. Deposition velocities for ceiling tile, paper with toner, and nonlevel loop carpet under varying relative humidity. mean of the Maxwell-Boltzmann velocity distribution (m/ h), and γ is the reaction probability (unitless). Transport-limited deposition velocities were determined for ozone in a parallel study using the same experimental system, chamber air flow conditions, and nearly identical materials as those discussed here (14). Equation 3 was employed to back-calculate the reaction probability from a measured vd for chlorine dioxide, an inferred vt for chlorine dioxide from vt for ozone, and a calculated (from the molecular theory of gases) vb for chlorine dioxide of 306 m/s at 25 °C. The transport-limited deposition velocity for chlorine dioxide, vt(ClO2), was estimated from vt(O3) as vt(ClO2) ) 0.87vt(O3), where the value 0.87 is the ratio of the molecular diffusion coefficient for chlorine dioxide to the molecular diffusion coefficient for ozone. The transport-limited deposition velocities for chlorine dioxide for this study were on the order of 0.15-0.53 m/h, values at the lower end of typical for actual buildings but reasonably consistent with previous small chamber experiments (15). The reaction probabilities for a set of illustrative materials are shown in Figure 5. Like the deposition velocity it is calculated from, reaction probability decayed with time. In general, when γ is greater than 10-4, deposition can be 1334
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 5, 2009
considered to be mass transport limited (15). When γ is less than 10-6, the rate of deposition is limited by surface reaction kinetics. Between these values, mass transport and surface reaction kinetics are both important contributors to the rate of deposition. For this study, reaction probabilities were less than 10-5. Ceiling tile, GWB backing, MDF, and office partition are the only materials that had deposition velocities greater than 10-6 four hours after chlorine dioxide had been introduced to the chamber. Parameter Variations. Multiple experiments were completed to examine the effects of varied disinfectant concentration and relative humidity. Decreasing the chlorine dioxide concentration by an order of magnitude for level-loop carpet lead to an increase in the final hour deposition velocity by roughly a factor of 2 (Figure 6). Increasing the chlorine dioxide concentration by 40% resulted in a deposition velocity decrease by a factor of 3. This trend was also observed for ceiling tile and to a slightly lesser extent for paper with toner. Building disinfection is typically completed at elevated relative humidity (RH), a condition which enhances the effectiveness of biological deactivation. As such, RH was limited to a lower-bound of 50% in this study. Results involving chlorine dioxide deposition velocity at RH ) 50%
and 75% for three different materials are presented in Figure 7. Lowering the RH from 75% to 50% led to a lower chlorine dioxide deposition velocity for ceiling tile. This result suggests that the build-up of water molecules at the surface of ceiling tile enhances chlorine dioxide uptake, possibly by absorption into surface films or droplets. Because of the high variability at lower deposition velocities, it is difficult to ascertain an RH trend for paper w/ print or carpet NLL. Temporal Variations in Material Reactivity. Deposition velocity and reaction probability for chlorine dioxide generally decreased rapidly for most test materials. For some materials, the uptake of chlorine dioxide appeared to reach a near equilibrium condition after which only small changes in deposition velocity occurred. These results are similar to those observed for ozone for similar materials and experimental conditions (14). The observation that chlorine dioxide removal decreases with time suggests that one or more of several different processes may occur at the external or internal (pore) surfaces of indoor materials, including (A) reactions between chlorine dioxide and molecules at unwetted surfaces of a material, (B) adsorption between chlorine dioxide and sorption sites on/in a material, (C) absorption into water films or droplets at material surfaces, and (D) absorption followed by aqueousphase chemistry. The reaction of chlorine dioxide with molecules at unwetted surfaces could account for reductions in surface reactive sites and reduced deposition velocities. However, as described earlier in this paper, chlorine dioxide is a highly selective oxidant and is unlikely to react appreciably outside of the aqueous phase. Chlorine dioxide might also be adsorbed to unwetted surfaces. Sorptive interactions might explain a rapid decrease in chlorine dioxide removal rates as sorptive equilibrium is approached. Chlorine dioxide will also be absorbed to some extent into surface films or droplets, where it can engage in aqueousphase chemistry and conversion to chlorite ions. If aqueousphase chemistry dominates surface removal processes, then rapid and prolonged conversion to ClO2- would lead to a sustained deposition velocity for chlorine dioxide. Slow conversion to chlorite or other ions would lead to accumulation of aqueous chlorine dioxide and an apparent decrease in deposition velocity with time. Since the extent of conversion of aqueous chlorine dioxide is pH dependent, the acidity of indoor surfaces may play an important role with respect to the uptake of chlorine dioxide. We did not explore this effect by, for example, purposely acidifying the surfaces of test materials. It is recommended that follow-up studies be completed to ascertain whether variation in surface pH affects the temporal variation in ClO2 deposition velocities. It is possible that more than one of the processes described above affects chlorine dioxide removal to indoor surfaces. The relative importance of each process would depend on the nature of a given material and environmental conditions, particularly relative humidity. In the case of buildings afflicted by mold, large material areas may be covered by fungal contamination that interacts with ClO2 with reaction probabilities different than those for the underlying material. Microbial growth may also affect surface properties (e.g., moisture content and pH), which further influence ClO2 chemistry at the surface. An improved understanding of these
processes through additional research is warranted, particularly given the increasing use of chlorine dioxide for residential and commercial mold remediation.
Acknowledgments The authors wish to thank the Technical Support Working Group (TSWG) and the United States Environmental Protection Agency (USEPA) for their funding, review, and guidance. Dr. Hubbard was a doctoral candidate at the University of Texas at Austin (UT) and Dr. Poppendieck was a postdoctoral fellow at UT when the work described herein was completed.
Supporting Information Available Schematic of experimental system, figure showing deposition velocities for duplicate experiments, and table listing transport-limited deposition velocities. This material is available free of charge via the Internet at http://pubs.acs.org.
Literature Cited (1) Gordon, G.; Rosenblatt, A. Chlorine dioxide: the current state of the art. Ozone: Sci. Eng. 2005, 27, 203–207. (2) Gordon, G.; Kiefer, R. G.; Rosenblatt, D. H. The chemistry of chlorine dioxide. In Progress in Inorganic Chemistry, Vol. 15; Lippard, S. J., Ed.; Wiley & Sons: New York, 1972. (3) Carr, M.; Carlson, K.; Gregory, D. Formation of chlorine dioxide under laundry bleaching conditions. J. Environ. Eng. Sci. 2003, 2, 293–297. (4) The Merck IndexsAn Encyclopedia of Chemicals, Drugs, and Biologicals, 13th ed.; O’Neil, M. J., Ed.; Merck and Co. Inc.: Whitehouse Station, NJ, 2001. (5) National Institute for Occupational Safety and Health, International chemical safety cardsschlorine dioxide. ICSC: 1027. http://www.cdc.gov/niosh/ipcsneng/neng0127.html (accessed September 16, 2007), validated October 20, 1999. (6) Staudinger, J.; Roberts, P. V. A critical review of Henry’s law constants for environmental applications. Crit. Rev. Environ. Sci. Technol. 1996, 26, 205–297. (7) Sapers, G. M.; Walker, P. N.; Sites, J. E.; Annous, B. A.; Eblen, D. R. Vapor-phase decontamination of apples inoculated with Escherichia coli. J. Food Sci. 2003, 68, 1003–1007. (8) Han, Y.; Selby, T. L.; Schultze, K. K.; Nelson, P. E.; Linton, R. H. Decontamination of strawberries using batch and continuous chlorine dioxide gas treatments. J. Food Prot. 2004, 67, 2450– 2455. (9) Gomez-Lopez, V. M.; Devlieghere, F.; Ragaert, P.; Debevere, J. Shelf-life extension of minimally processed carrots by gaseous chlorine dioxide. Int. J. Food Microbiol. 2007, 116, 221–227. (10) Southwell, K. L. Chlorine dioxide dry fumigation in special collection librariessa case study. Lib. Arch. Secur. 2003, 18, 39–49. (11) Wilson, S. C.; Wu, C.; Andriychuk, L. A.; Martin, J. M.; Brasel, T. L.; Jumper, C. A.; Straus, D. C. Effect of chlorine dioxide gas on fungi and mycotoxins associated with sick building syndrome. Appl. Environ. Microbiol. 2005, 71, 5399–5403. (12) Buttner, M. P.; Cruz, P.; Stetzenbach, L. D.; Klima-Comba, A. K.; Stevens, V. L.; Cronin, T. D. Determination of the efficacy of two building decontamination strategies by surface sampling with culture and quantitative PCR analysis. Appl. Environ. Microbiol. 2004, 70, 4740–4747. (13) Han, Y.; Applegate, B.; Linton, R. H.; Nelson, P. E. Decontamination of Bacillus thuringiensis spores on selected surfaces by chlorine dioxide gas. J. Environ. Health 2005, 66, 16–20. (14) Poppendieck, D.; Hubbard, H.; Ward, M.; Weschler, C. J.; Corsi, R. L. Ozone reactions with indoor materials during building disinfection. Atmos. Environ. 2007, 41, 3166–3176. (15) Cano-Ruiz, J. A.; Nazaroff, W. Removal of reactive gases at indoor surfaces: combining mass transport and surface kinetics. Indoor Air 1993, 2, 555–560.
ES801930C
VOL. 43, NO. 5, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1335
