Chemosphere 122 (2015) 32–37
Contents lists available at ScienceDirect
Chemosphere journal homepage: www.elsevier.com/locate/chemosphere
Plant leaves as indoor air passive samplers for volatile organic compounds (VOCs) Todd A. Wetzel a,1, William J. Doucette b,⇑ a b
Utah Department of Air Quality, Salt Lake City, UT 84116, USA Utah Water Research Laboratory, Utah State University, 8200 Old Main Hill, Logan, UT 84322, USA
h i g h l i g h t s VOC leaf-air concentrations factors measured using static headspace method. Leaf-air concentration factors related to leaf lipid content and chemical Koa. Leaf concentrations paralleled measured air concentrations during residential study. Plant leaves show potential as cost effective, real-time indoor air VOC samplers.
a r t i c l e
i n f o
Article history: Received 20 August 2014 Received in revised form 20 October 2014 Accepted 21 October 2014 Available online 27 November 2014 Handling Editor: I. Cousins Keywords: Leaf-air concentration factors Passive sampler Indoor air contamination Octanol-air partition coefficients Chlorinated solvents Indoor air quality
a b s t r a c t Volatile organic compounds (VOCs) enter indoor environments through internal and external sources. Indoor air concentrations of VOCs vary greatly but are generally higher than outdoors. Plants have been promoted as indoor air purifiers for decades, but reports of their effectiveness differ. However, while airpurifying applications may be questionable, the waxy cuticle coating on leaves may provide a simple, cost-effective approach to sampling indoor air for VOCs. To investigate the potential use of plants as indoor air VOC samplers, a static headspace approach was used to examine the relationship between leaf and air concentrations, leaf lipid contents and octanol-air partition coefficients (Koa) for six VOCs and four plant species. The relationship between leaf and air concentrations was further examined in an actual residence after the introduction of several chlorinated VOC emission sources. Leaf-air concentration factors (LACFs), calculated from linear regressions of the laboratory headspace data, were found to increase as the solvent extractable leaf lipid content and Koa value of the VOC increased. In the studies conducted in the residence, leaf concentrations paralleled the changing air concentrations, indicating a relatively rapid air to leaf VOC exchange. Overall, the data from the laboratory and residential studies illustrate the potential for plant leaves to be used as cost effective, real-time indoor air VOC samplers. Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction Adults living in North America spend an estimated 80–90% of their time indoors (Dales et al., 2008; Orwell et al., 2004), and concerns about the potential exposure to volatile organic compounds (VOCs) in indoor air have increased as new home construction techniques and improvements in heating, ventilation, and air conditioning (HVAC) efficiency have significantly reduced the introduction of outdoor air (Cohen, 1996).
⇑ Corresponding author. Tel.: +1 435 797 3178; fax: +1 435 797 3663. E-mail addresses: [email protected] (T.A. Wetzel), [email protected] (W.J. Doucette). 1 Tel.: +1 (801) 536 4429. http://dx.doi.org/10.1016/j.chemosphere.2014.10.065 0045-6535/Ó 2014 Elsevier Ltd. All rights reserved.
Concentrations of VOCs in indoor air are generally 5–10 times higher than outdoors, with even higher indoor air concentrations where extreme cold weather conditions exist (Dales et al., 2008). Some of the VOCs identified in indoor air are considered suspected or confirmed carcinogens by the World Health Organization (WHO) and International Agency for Research on Cancer (IARC) and enter indoor environments through internal (e.g. paints, paint strippers, fuels, cleaning supplies, pesticides, building materials, adhesives) and external sources (e.g. vapor intrusion (VI)) from contaminated soil and/or groundwater and ambient air from automobiles and industrial facilities). The use of ornamental plants to reduce indoor air concentrations of VOCs has been studied for decades (NASA, 1989; Cornejo et al., 1999; Liu et al., 2007; Yang et al., 2009). However, stated removals differ widely and the variety of experimental approaches
T.A. Wetzel, W.J. Doucette / Chemosphere 122 (2015) 32–37
used to determine removals complicate comparisons among studies. Depending on the plant and chemical of interest, reported VOC removal mechanisms include stomatal uptake and metabolism (Baur et al., 1997), microbial transformation within plant growth media (Orwell et al., 2004), and sorption to leaves and plant growth media (Orwell et al., 2004; Keymeulen et al., 1997; Bacci et al., 1990). Modeling results suggest that the high plant biomass to air ratio necessary to make meaningful reductions in indoor air VOC concentrations make the use of houseplants as air cleaners impractical in most cases (Girman et al., 2009). However, even if plants do not significantly impact indoor air concentrations, the waxy cuticle of leaves may allow common houseplants to be used as simple, inexpensive indoor air samplers for VOCs. Successful implementation of this approach would reduce costs associated with conventional passive sorbent samplers or active canisters, minimize sampler intrusions into the home, and potentially allow residents to directly participate in the sample collection. Plants have been widely used as passive samplers for semi-volatile organic compounds (SVOCs) in outdoor environments for compounds like PAHs (Lin et al., 2006; Li and Chen, 2009), PCBs (Luca et al., 2006), dioxins (Luca et al., 2006), herbicides, and pesticides (Bacci et al., 1990), and predictive models relating leaf-air concentration ratios (bioconcentration factors) to octanol-air partition coefficients (Koa) have been developed (Bacci et al., 1990; Cornejo et al., 1999). Far fewer studies have examined the effectiveness of plants as samplers for more volatile compounds, especially in indoor environments. Hiatt (1998) investigated the use of plant leaves to sample outdoor air for VOCs such as benzene, toluene, TCE and PCE. It was found that leaf-air concentration ratios were generally well predicted using existing Koa based models except for plant species containing higher amounts of monoterpenes, These species were greatly under predicted, suggesting the importance of lipid character (e.g. Li and Chen, 2014; Chen et al., 2008). It was also reported that VOC uptake by leaves was rapid, and higher vapor pressure compounds reached equilibrium concentrations faster. In a subsequent publication, Hiatt (1999) reported that leaves provided a good indication of early morning exposures when air concentrations were relatively constant, but in the afternoon, pulses of clean air during windy conditions quickly desorbed the VOCs from the leaves. The relatively fast uptake and release of VOCs from plant leaves observed by Hiatt (1999) may limit outdoor sampling applications where air concentrations can change rapidly due to changes in wind speed and direction, but this is likely less important in indoor environments. To investigate the potential of using plant leaves as samplers for indoor air VOC concentrations, a static headspace approach was used to determine leaf-air concentration factors (LACFs) for six VOCs as a function of VOC concentration and plant type. The relationship between leaf and air concentrations was further examined in an actual residence after the introduction of several chlorinated VOC emission sources. Finally, the measured indoor air concentrations were compared to concentrations predicted from laboratory derived LACFs and measured leaf concentrations.
2. Materials and methods 2.1. Plants Four common houseplants were used in both the laboratory and residential experiments: ficus (Ficus benjamina); golden pothos (Epipremnum aureum); spider fern (Chlorophytus comsosum ‘vittatum’); and Christmas cactus (Schlumbergera truncate ‘harmony’). The plants were selected because they are commonly available,
33
are hardy, have relatively low light requirements and vary in leaf morphology. Plants were purchased locally from several different vendors and re-potted in stainless steel planters from Stainless Lux (Livermore, CA) using a 50% peat moss and 50% vermiculate mix. Prior to their use, plants were kept under a 400-W metal halide grow lamp with a 10-h photoperiod and were watered as needed to maintain health. Generally, eight plants of each species were used: four to collect leaves for characterization and for the headspace study, and four were used in the residential sampling study. 2.2. Chemicals Several VOCs commonly identified in indoor air monitoring surveys were selected for the laboratory study: trichloroethylene (TCE), tetrachloroethylene (PCE), 1,2-dichloroethane (1,2-DCA), benzene, toluene and m,p-xylene (e.g. Dawson and McAlary, 2009). Typical indoor air sources of these VOCs include vapor intrusion and consumer product use (e.g. Dawson and McAlary, 2009; Doucette et al., 2010). Known concentrations of the chemicals, dissolved in methanol, were purchased from Ultra Scientific (N. Kingstown, RI, USA). 2.3. Solvent extractable leaf lipid content (waxes), surface area, and water content Leaves were collected from several plants of the same species using gloved hands and ethanol-cleaned scissors. The leaves were weighed using an analytical balance, air-dried for 7–21 d, and then placed in a desiccator until a constant mass was obtained (usually 24–36 h). The water content of the leaves was determined from the difference between the fresh and dry weights. Oven drying was not used in order to minimize organic matter losses through degradation and volatilization. After drying, the tissue was finely shredded using a coffee grinder, and subsamples of known weight (typically one gram) were placed in 25/80 mm cellulose thimbles and Soxhlet extracted for 24 h with ethyl ether. This extraction procedure removes mainly the waxy components of the cuticle (e.g. Chefetz, 2003; Li and Chen, 2014). After extraction, the remaining ethyl ether was evaporated and the flasks containing the extracted lipids were placed in desiccators until a constant mass was obtained. Differences between pre- and post-flask weights were used to determine the mass of lipid extracted. Extraction efficiencies were evaluated by adding known quantities of olive and vegetable oil to selected thimbles. Lipid content was calculated by dividing the extracted lipid weight by the dry tissue weight added to the thimble. Leaf surface areas also were measured using a leaf area meter (LICOR Instruments, Model 6000). 2.4. Static headspace determination of leaf air concentration factors (LACFs) Leaves were collected from the same plants used to characterize the leaf properties. Leaf samples were placed in pre-weighed 20 mL headspace vials and then re-weighed to determine sample mass (typically 0.3–1 g fresh weight depending on plant species). Mixtures of the six VOCs were then added into the vials (in 2 lL methanol) containing the leaf samples and blank vials containing glass slides (approximately same area and volume as leaf samples) to yield six headspace concentrations ranging from 0.25 to 5 lg L 1. The concentration range was selected for analytical convenience and environmental relevance (i.e. screening risk levels, USEPA, 2013). The vials were equilibrated for approximately 24 h at room temperature after adding the VOCs. Triplicate samples
34
T.A. Wetzel, W.J. Doucette / Chemosphere 122 (2015) 32–37
were analyzed at each concentration for each of the four plant species. A Hewlett-PackardÒ 7890A gas chromatograph (GC)/5973C mass spectrometer (MS) operated in Selected Ion Monitoring (SIM) mode and equipped with a CTC PAL auto-sampler configured for headspace sampling was used to measure VOC concentrations. After a 30-min equilibration at 30 °C (the lowest constant temperature consistently attainable in the autosampler), headspace samples were introduced into the GC/MS equipped with a J&W Scientific (Folsom, CA) 122-1334 capillary column (30 m 0.25 mm ID 1.4 lm film thickness), with an automated oven temperature program as follows: 35 °C initial temperature to 105 °C at 10 °C min 1, followed by 50 °C min 1 to a final temperature of 180 °C. The concentrations of VOCs in the leaves (lg VOC kg 1 fresh leaf weight) were determined by subtracting the headspace concentrations measured in the leaf-spiked vials from those in the blankspiked vials and then dividing by leaf weight. Indirectly determining the leaf concentrations was found to be more analytically convenient and reproducible than directly determining the leaf concentrations using a conventional solvent extraction approach, but potentially less accurate. Leaf-air concentration factors (LACFs) in units of L kg 1, were generated from the slope of the regression line obtained by plotting the leaf concentrations by the vial headspace VOC concentration. 2.5. Leaf and air sampling within residence Several consumer products emitting chlorinated VOCs were introduced into a 20-year-old, two-story residence with a basement (sub-ground floor) containing the same four species of plants used in the headspace study. The consumer products included a gun cleaner containing TCE, a hobby glue containing PCE and an injection molded plastic lamp base emitting 1,2-DCA. These consumer products had been previously identified as VOC sources during field vapor intrusion investigations (Gorder and Dettenmaier, 2011). The concentrations of benzene, toluene and m,p-xylene, the aromatic hydrocarbons having the highest indoor air concentrations, originating from gasoline stored in the attached garage were also measured in leaf and air samples. Ethylbenzene, oxylene and other related aromatic hydrocarbons typically found in gasoline were not quantified. The residence was equipped with a forced air heating, ventilation and cooling (HVAC) system and had a total air volume, including the basement, of 600 m3 (estimated from measured room dimensions). Recording thermocouples were placed in HVAC vents to determine the frequency and duration of the HAVC operation with the house thermostat set at 78 °F (25.6 C). The basement had a separate forced air heating system. Three sets of four potted plants, one of each species, were introduced into the residence (one set on each floor) 24–48 h prior to the collection of the first background samples. A fourth set of plants was placed in an outdoor location at the rear of the residence. Background air and leaf (typically 0.5–1 g fresh weight) samples were collected just before the emitting objects were added to the residence. Emitting objects were quickly transported to a 2nd story room, subsequently referred to as the source room, inside an airtight container. After the objects were placed in the source room the door was closed and the HAVC system turned off to minimize mixing between rooms and allow the concentrations in the source room to increase. This mimicked the approach used in the roomto-room sample scheme that was originally used to identify the objects during VI investigations (Doucette et al., 2010). Twenty-four hours after the objects were placed in the source room the door to the room was opened and the HVAC system was turned on to mix the house air. Twenty-four to 48 h after the objects were introduced into the home, the objects were removed. Eight
sampling events took place, beginning with the background samples and continuing for 24 h after the objects were removed. During each air sampling, leaf samples from each of the four plant species were collected at the same time and placed in headspace vials for analysis as described previously, except that the vials were equilibrated at 100 °C to increase the desorption (i.e. extraction efficiencies) of the VOCs from the leaves. At this temperature, spike recoveries were greater than 90% for all six compounds. During each sampling event, sample collectors did not exit or reenter the house until all sampling was completed. This was done to avoid any unusual introduction of outdoor air into the residence. Air samples were collected using two TenaxÓ sorbent tubes (Alltech Associates, Deerfield, Illinois) attached in series to an Air Chek Sampler, SKC Model 224-44XR (Eighty Four, Pennsylvania) following the general approach outlined in EPA Method TO-17. Samples were collected for approximately 10–30 min at known flow rates from 100 to 150 mL min 1 depending on the expected concentrations. An Alltech Digital Flow Check™ model DFC™ Flowmeter (Nicholasville, Kentucky) was used to confirm flow rates at the beginning and end of each sampling event. Only a small number of replicate air samples were collected over the course of the residential study because of the limited number of sampling pumps available. Replicate air samples varied from 1% to 10% depending on the compound and concentration. Sorbent tube samples were introduced into an AgilentÒ 6890/ 5793GC/MS equipped with a J&W Scientific (Folsom, California) DB-624 capillary column (30 m 0.25 mm ID 1.4 lm film thickness) using a Perkin Elmer TurboMatrix ATD Automated Thermal Desorber operating under the following conditions: 5 min trap purge; cryo-trap temperature = 30 °C; TenaxÓ tube desorb = 300 °C for 10 min; cryo-trap temperature program 30 °C initial temperature to 320 °C at 40 °C s 1, transfer line to GC/MS at 225 °C. The moisture control system, traps, and tubes were thermally cleaned between each sample. Chromatographic conditions were as follows: oven temperature program start at 35 °C, hold for 2 min then ramp at 30 °C min 1 to 170 °C, followed by 70 °C min 1 from 170 to 230 °C with a 1 min hold at the final temperature. The MS was operated in selected ion monitoring (SIM) mode. An external standard approach was used to quantify the mass of compounds collected in each trap. Standards were prepared by injecting 1 lL of standards dissolved in methanol onto clean TenaxÓ traps with a micro-syringe. Masses of compound injected on the tubes ranged from 100 to 100 000 pg. New standard curves were prepared if the concentrations of the continuous calibration verification (CCV) standards deviated more than 10% from the expected value.
3. Results and discussion 3.1. Static headspace studies: leaf air concentrations factors (LACFs) Leaf air concentration factors (LACF), defined as the VOC concentration in the leaves (lg VOC per kilogram of fresh leaf, CL) divided by the concentration in the air (lg VOC per liter of air, CA), were generated from the slopes of the linear regression lines obtained by plotting leaf to air concentrations (Fig. 1) and are summarized in Table 1. As shown in Table 1, the LACFs generally increased as the extractable lipid content of the leaf increased (ficus to cactus) and as the Koa of the chemical (benzene to xylene) increased. A trend of increasing LACF for increasing chemical Koa value has been previously observed for semi-volatile compounds (Tolls and McLachlan, 1994; Thomas et al., 1998; Platts and Abraham, 2000; Weiss, 2000). However, since the range of log Koa values exhibited by the six VOCs was relatively narrow, the only significant (p < 0.05) linear relationship between log LACF
35
T.A. Wetzel, W.J. Doucette / Chemosphere 122 (2015) 32–37
Fig. 1. Example plot of leaf versus air concentrations generated using static headspace technique for Ficus and TCE (regression line: y = 41.2x 0.77. R2 = 0.98).
and log Koa was observed for ficus, the plant species with the highest extractable lipid content. Lipid normalization reduced the differences in the LACF values between species. However, since the cuticle consists of several lipid or lipid like components including cutin, cutan, and extractable waxes that exhibit varying affinities for organic contaminants, incorporation of these cuticular fractional amounts is likely necessary to better understand species differences (Barber et al., 2004; Chen et al., 2008). Future work should specifically examine the role of the polymeric lipid fraction (cutin) that was shown by Chen et al. (2008) to be the dominant sorptive fraction for naphthalene and 1-naphthol on tomato and apple cuticles. It is also possible, but less likely to be significant, that VOCs can sorb to the non-lipid fractions of the leaves. Normalizing the LACFs to the surface area of the leaves was also evaluated (data not shown), but it did not reduce the species LACF variation as much as lipid normalization. The observed relationships between the leaf and air concentrations for the six VOCs in the static headspace study showed that plants have the potential to be used as indoor air samplers. This potential was evaluated during the subsequent study conducted within an actual residence. 3.2. Relationship between leaf and indoor air VOC concentrations in an actual residence Measured leaf concentrations paralleled the indoor air concentrations for the chlorinated VOCs as illustrated in Fig. 2 for TCE and ficus in the source room. The corresponding changes in leaf and air
Fig. 2. Ficus leaf (average of triplicate leaf samples with error bars showing standard deviation) and air (single measurement) TCE concentrations in source room over time.
concentrations indicate a relatively rapid exchange of VOCs between leaves and air. For the non-chlorinated hydrocarbons, the leaf concentrations also paralleled the air concentrations. However, because of the nature of the source (fuel stored in an attached garage), the indoor air hydrocarbon VOC concentrations (benzene, toluene and xylene) varied only by factors of 2–3 as compared with factors of 100 for the chlorinated VOCs as shown in Fig. 3 for xylene. 3.3. Estimating indoor air concentrations for laboratory derived LACFs and measured leaf concentrations The parallel relationship between leaf and air VOC concentrations within the residence illustrates the potential for using leaves as real-time air samplers. In practice, it would be ideal if the measured leaf concentrations could be used along with the laboratory derived LACFs to predict indoor air concentrations. To evaluate this possibility, predicted indoor air concentrations, calculated by dividing the measured leaf concentrations by the appropriate laboratory derived LACFs, were compared to the measured indoor air concentrations. As illustrated in Fig. 4 for TCE and ficus, significant relationships (p < 0.05) between estimated and measured air concentrations were found for the three chlorinated VOCs for all plant species except for PCE with cactus and ficus. The indoor air concentrations of PCE, ranging from 0.0002 to 0.006 lg L 1, were the lowest of the three chlorinated VOCs but were still considerably below EPAs screening level carcinogenic risk based concentration of 9.4 lg PCE m 3 or 9400 lg L 1 (a). For the non-chlorinated hydrocarbons (benzene, toluene and m,p-xylene), the relationships between measured air concentrations and those predicted from laboratory derived LCAFs were not significant. The reason for the difference in predictability between the chlorinated and non-chlorinated VOC in not known but may be related to the nature of the source materials. The sources of chlorinated VOCs were introduced into the upper floor of the residence and removed 24–48 h after introduction. This resulted in indoor air concentrations of the chlorinated VOC that varied up to two orders of magnitude. In contrast, the source of the non-chlorinated VOCs (gasoline stored in attached garage) was constant and the indoor air concentrations varied only by factors of 3–5.
Bacci, E., Calamari, D., Gaggi, C., Vighi, M., 1990. Bioconcentration of organicchemical vapors in plant-leaves – experimental measurements and correlation. Environ. Sci. Technol. 24 (6), 885–889. Barber, J.L., Thomas, G.O., Kerstiens, G., Jones, K.C., 2004. Current issues and uncertainties in the measurement and modelling of air–vegetation exchange and within-plant processing of POPs. Environ. Pollut. 128 (1–2), 99–138. Baur, P., Marzouk, H., Schonherr, J., Grayson, B.T., 1997. Partition coefficients of active ingredients between plant cuticle and adjuvants as related to rates of foliar uptake. Agric. Food Chem. 45 (9), 3659–3665. Chefetz, B., 2003. Sorption of phenanthrene and atrazine by plant cuticular fractions. Environ. Toxicol. Chem. 22, 2492–2498. Chen, B., Li, Y., Guo, Y., Zhu, L., Schnoor, J.L., 2008. Role of the extractable lipids and polymeric lipids in sorption of organic contaminants onto plant cuticles. Environ. Sci. Technol. 42, 1517–1523. Cohen, Y., 1996. Volatile organic compounds in the environment: a multimedia perspective. In: Wang, W., Schnoor, J., 1996, Doi, J. (Eds.), Volatile organic compounds in the environment. ASTM STP 1261. American Society for Testing and Materials, West Conshohocken, PA. pp. 7–32. Cornejo, J.J., Munoz, F.G., Ma, C.Y., Stewart, A.J., 1999. Studies on the decontamination of air by plants. Ecotoxicology 8, 311–320. Dales, R., Liu, L., Wheeler, A.J., Gilbert, N.L., 2008. Quality of indoor residential air and health. Can. Med. Assoc. 179, 147–152. Dawson, H.E., McAlary, T., 2009. A Compilation of statistics for VOCs from post-1990 indoor air concentration studies in North American residences unaffected by subsurface vapor intrusion. Ground Water Monit. Remed. 29 (1), 60–69. Doucette, W.J., Hall, A.J., Gorder, K.A., 2010. Emissions of 1,2-dichloroethane from holiday decorations as a source of indoor air contamination. Ground Water Monit. Remed. 30 (1), 67–73. Girman, J.R., Phillips, T.R., Levin, H., 2009. Critical review: how well do house plants perform as indoor air cleaners? In: Proceedings, 9th International Conference and Exhibition on Healthy Buildings 2009, vol. 1 of 3. 13–17 September, Syracuse, NY, USA, pp. 823–826. Gorder, K.A., Dettenmaier, E.M., 2011. Portable GC/MS methods to evaluate sources of cVOC contamination in indoor air. Ground Water Monit. Remed. 31 (4), 113– 119.
T.A. Wetzel, W.J. Doucette / Chemosphere 122 (2015) 32–37 Hiatt, M., 1998. Bioconcentration factors for volatile organic compounds in vegetation. Anal. Chem. 70 (5), 851–856. Hiatt, M., 1999. Leaves as an indicator of exposure to airborne volatile organic compounds. Environ. Sci. Technol. 33 (22), 4126–4133. Keymeulen, R., Bruyn, G.D., Langenhove, H.V., 1997. Headspace gas chromatographic determination of the plant cuticle-air partition coefficients for monocyclic aromatic hydrocarbons as environmental compartment. Chromatography A 774, 213–221. Li, Y., Chen, B., 2009. Phenanthrene sorption by fruit cuticles and potato periderm with different compositional characteristics. Agric. Food Chem. 57, 637–644. Li, Q.Q., Chen, B.L., 2014. Organic pollutant clustered in the plant cuticular membranes: visualizing the distribution of phenanthrene in leaf cuticle using two-photon confocal scanning laser microscopy. Environ. Sci. Technol. 48 (9), 4774–4781. Lin, D., Zhu, L., He, W., Tu, Y., 2006. Tea plant uptake and translocation of polycyclic aromatic hydrocarbons from water and around air. Agric. Food Chem. 54, 3658– 3662. Liu, Y., Mu, Y., Zhu, Y., Ding, H., Arens, N.C., 2007. Which ornamental plant species effectively remove benzene from indoor air? Atmos. Environ. 41, 650–654. Luca, N., Jones, K., Gramatica, P., Papa, E., Gerabolini, B., Guardo, A., 2006. Accumulation of persistent organic pollutants in canopies of different forest types: role of species composition and altitudinal-temperature gradient. Environ. Sci. Technol. 40, 6580–6586.
37
National Aeronautics and Space Administration, 1989. Interior Landscape Plants for Indoor Air Pollution Abatement. Orwell, R.L., Wood, R.L., Tarran, J., Torpy, F., Burchett, M.D., 2004. Removal of benzene by the indoor plant/substrate microcosm and implications for air quality. Water Air Soil Pollut. 157, 193–207. Platts, J.A., Abraham, M.H., 2000. Partition of volatile organic compounds from air and from water into plant cuticular matrix: an LFER analysis. Environ. Sci. Technol. 34 (2), 318–323. Thomas, G., Sweetman, A.J., Ockenden, W.A., Mackay, D., Jones, K.C., 1998. Airpasture transfer of PCBs. Environ. Sci. Technol. 32 (7), 936–942. Tolls, J., McLachlan, M.S., 1994. Partitioning of semivolatile organic-compounds between air and lolium-multiflorum (welsh ray grass). Environ. Sci. Technol. 28 (1), 159–166. USEPA, 2013. U.S. Environmental Protection Agency. Residential Air Risk Screening Table [Internet]. Available from: (Cited 26.03.14). Weiss, P., 2000. Vegetation/soil distribution of semivolatile organic compounds in relation to their physicochemical properties. Environ. Sci. Technol. 34 (9), 1707– 1714. Yang, D.S., Pennisi, S.V., Son, K., Kays, S.J., 2009. Screening indoor plants or volatile organic pollutant removal efficiency. HortScience 44 (5), 1377–1381.
Contents lists available at ScienceDirect
Chemosphere journal homepage: www.elsevier.com/locate/chemosphere
Plant leaves as indoor air passive samplers for volatile organic compounds (VOCs) Todd A. Wetzel a,1, William J. Doucette b,⇑ a b
Utah Department of Air Quality, Salt Lake City, UT 84116, USA Utah Water Research Laboratory, Utah State University, 8200 Old Main Hill, Logan, UT 84322, USA
h i g h l i g h t s VOC leaf-air concentrations factors measured using static headspace method. Leaf-air concentration factors related to leaf lipid content and chemical Koa. Leaf concentrations paralleled measured air concentrations during residential study. Plant leaves show potential as cost effective, real-time indoor air VOC samplers.
a r t i c l e
i n f o
Article history: Received 20 August 2014 Received in revised form 20 October 2014 Accepted 21 October 2014 Available online 27 November 2014 Handling Editor: I. Cousins Keywords: Leaf-air concentration factors Passive sampler Indoor air contamination Octanol-air partition coefficients Chlorinated solvents Indoor air quality
a b s t r a c t Volatile organic compounds (VOCs) enter indoor environments through internal and external sources. Indoor air concentrations of VOCs vary greatly but are generally higher than outdoors. Plants have been promoted as indoor air purifiers for decades, but reports of their effectiveness differ. However, while airpurifying applications may be questionable, the waxy cuticle coating on leaves may provide a simple, cost-effective approach to sampling indoor air for VOCs. To investigate the potential use of plants as indoor air VOC samplers, a static headspace approach was used to examine the relationship between leaf and air concentrations, leaf lipid contents and octanol-air partition coefficients (Koa) for six VOCs and four plant species. The relationship between leaf and air concentrations was further examined in an actual residence after the introduction of several chlorinated VOC emission sources. Leaf-air concentration factors (LACFs), calculated from linear regressions of the laboratory headspace data, were found to increase as the solvent extractable leaf lipid content and Koa value of the VOC increased. In the studies conducted in the residence, leaf concentrations paralleled the changing air concentrations, indicating a relatively rapid air to leaf VOC exchange. Overall, the data from the laboratory and residential studies illustrate the potential for plant leaves to be used as cost effective, real-time indoor air VOC samplers. Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction Adults living in North America spend an estimated 80–90% of their time indoors (Dales et al., 2008; Orwell et al., 2004), and concerns about the potential exposure to volatile organic compounds (VOCs) in indoor air have increased as new home construction techniques and improvements in heating, ventilation, and air conditioning (HVAC) efficiency have significantly reduced the introduction of outdoor air (Cohen, 1996).
⇑ Corresponding author. Tel.: +1 435 797 3178; fax: +1 435 797 3663. E-mail addresses: [email protected] (T.A. Wetzel), [email protected] (W.J. Doucette). 1 Tel.: +1 (801) 536 4429. http://dx.doi.org/10.1016/j.chemosphere.2014.10.065 0045-6535/Ó 2014 Elsevier Ltd. All rights reserved.
Concentrations of VOCs in indoor air are generally 5–10 times higher than outdoors, with even higher indoor air concentrations where extreme cold weather conditions exist (Dales et al., 2008). Some of the VOCs identified in indoor air are considered suspected or confirmed carcinogens by the World Health Organization (WHO) and International Agency for Research on Cancer (IARC) and enter indoor environments through internal (e.g. paints, paint strippers, fuels, cleaning supplies, pesticides, building materials, adhesives) and external sources (e.g. vapor intrusion (VI)) from contaminated soil and/or groundwater and ambient air from automobiles and industrial facilities). The use of ornamental plants to reduce indoor air concentrations of VOCs has been studied for decades (NASA, 1989; Cornejo et al., 1999; Liu et al., 2007; Yang et al., 2009). However, stated removals differ widely and the variety of experimental approaches
T.A. Wetzel, W.J. Doucette / Chemosphere 122 (2015) 32–37
used to determine removals complicate comparisons among studies. Depending on the plant and chemical of interest, reported VOC removal mechanisms include stomatal uptake and metabolism (Baur et al., 1997), microbial transformation within plant growth media (Orwell et al., 2004), and sorption to leaves and plant growth media (Orwell et al., 2004; Keymeulen et al., 1997; Bacci et al., 1990). Modeling results suggest that the high plant biomass to air ratio necessary to make meaningful reductions in indoor air VOC concentrations make the use of houseplants as air cleaners impractical in most cases (Girman et al., 2009). However, even if plants do not significantly impact indoor air concentrations, the waxy cuticle of leaves may allow common houseplants to be used as simple, inexpensive indoor air samplers for VOCs. Successful implementation of this approach would reduce costs associated with conventional passive sorbent samplers or active canisters, minimize sampler intrusions into the home, and potentially allow residents to directly participate in the sample collection. Plants have been widely used as passive samplers for semi-volatile organic compounds (SVOCs) in outdoor environments for compounds like PAHs (Lin et al., 2006; Li and Chen, 2009), PCBs (Luca et al., 2006), dioxins (Luca et al., 2006), herbicides, and pesticides (Bacci et al., 1990), and predictive models relating leaf-air concentration ratios (bioconcentration factors) to octanol-air partition coefficients (Koa) have been developed (Bacci et al., 1990; Cornejo et al., 1999). Far fewer studies have examined the effectiveness of plants as samplers for more volatile compounds, especially in indoor environments. Hiatt (1998) investigated the use of plant leaves to sample outdoor air for VOCs such as benzene, toluene, TCE and PCE. It was found that leaf-air concentration ratios were generally well predicted using existing Koa based models except for plant species containing higher amounts of monoterpenes, These species were greatly under predicted, suggesting the importance of lipid character (e.g. Li and Chen, 2014; Chen et al., 2008). It was also reported that VOC uptake by leaves was rapid, and higher vapor pressure compounds reached equilibrium concentrations faster. In a subsequent publication, Hiatt (1999) reported that leaves provided a good indication of early morning exposures when air concentrations were relatively constant, but in the afternoon, pulses of clean air during windy conditions quickly desorbed the VOCs from the leaves. The relatively fast uptake and release of VOCs from plant leaves observed by Hiatt (1999) may limit outdoor sampling applications where air concentrations can change rapidly due to changes in wind speed and direction, but this is likely less important in indoor environments. To investigate the potential of using plant leaves as samplers for indoor air VOC concentrations, a static headspace approach was used to determine leaf-air concentration factors (LACFs) for six VOCs as a function of VOC concentration and plant type. The relationship between leaf and air concentrations was further examined in an actual residence after the introduction of several chlorinated VOC emission sources. Finally, the measured indoor air concentrations were compared to concentrations predicted from laboratory derived LACFs and measured leaf concentrations.
2. Materials and methods 2.1. Plants Four common houseplants were used in both the laboratory and residential experiments: ficus (Ficus benjamina); golden pothos (Epipremnum aureum); spider fern (Chlorophytus comsosum ‘vittatum’); and Christmas cactus (Schlumbergera truncate ‘harmony’). The plants were selected because they are commonly available,
33
are hardy, have relatively low light requirements and vary in leaf morphology. Plants were purchased locally from several different vendors and re-potted in stainless steel planters from Stainless Lux (Livermore, CA) using a 50% peat moss and 50% vermiculate mix. Prior to their use, plants were kept under a 400-W metal halide grow lamp with a 10-h photoperiod and were watered as needed to maintain health. Generally, eight plants of each species were used: four to collect leaves for characterization and for the headspace study, and four were used in the residential sampling study. 2.2. Chemicals Several VOCs commonly identified in indoor air monitoring surveys were selected for the laboratory study: trichloroethylene (TCE), tetrachloroethylene (PCE), 1,2-dichloroethane (1,2-DCA), benzene, toluene and m,p-xylene (e.g. Dawson and McAlary, 2009). Typical indoor air sources of these VOCs include vapor intrusion and consumer product use (e.g. Dawson and McAlary, 2009; Doucette et al., 2010). Known concentrations of the chemicals, dissolved in methanol, were purchased from Ultra Scientific (N. Kingstown, RI, USA). 2.3. Solvent extractable leaf lipid content (waxes), surface area, and water content Leaves were collected from several plants of the same species using gloved hands and ethanol-cleaned scissors. The leaves were weighed using an analytical balance, air-dried for 7–21 d, and then placed in a desiccator until a constant mass was obtained (usually 24–36 h). The water content of the leaves was determined from the difference between the fresh and dry weights. Oven drying was not used in order to minimize organic matter losses through degradation and volatilization. After drying, the tissue was finely shredded using a coffee grinder, and subsamples of known weight (typically one gram) were placed in 25/80 mm cellulose thimbles and Soxhlet extracted for 24 h with ethyl ether. This extraction procedure removes mainly the waxy components of the cuticle (e.g. Chefetz, 2003; Li and Chen, 2014). After extraction, the remaining ethyl ether was evaporated and the flasks containing the extracted lipids were placed in desiccators until a constant mass was obtained. Differences between pre- and post-flask weights were used to determine the mass of lipid extracted. Extraction efficiencies were evaluated by adding known quantities of olive and vegetable oil to selected thimbles. Lipid content was calculated by dividing the extracted lipid weight by the dry tissue weight added to the thimble. Leaf surface areas also were measured using a leaf area meter (LICOR Instruments, Model 6000). 2.4. Static headspace determination of leaf air concentration factors (LACFs) Leaves were collected from the same plants used to characterize the leaf properties. Leaf samples were placed in pre-weighed 20 mL headspace vials and then re-weighed to determine sample mass (typically 0.3–1 g fresh weight depending on plant species). Mixtures of the six VOCs were then added into the vials (in 2 lL methanol) containing the leaf samples and blank vials containing glass slides (approximately same area and volume as leaf samples) to yield six headspace concentrations ranging from 0.25 to 5 lg L 1. The concentration range was selected for analytical convenience and environmental relevance (i.e. screening risk levels, USEPA, 2013). The vials were equilibrated for approximately 24 h at room temperature after adding the VOCs. Triplicate samples
34
T.A. Wetzel, W.J. Doucette / Chemosphere 122 (2015) 32–37
were analyzed at each concentration for each of the four plant species. A Hewlett-PackardÒ 7890A gas chromatograph (GC)/5973C mass spectrometer (MS) operated in Selected Ion Monitoring (SIM) mode and equipped with a CTC PAL auto-sampler configured for headspace sampling was used to measure VOC concentrations. After a 30-min equilibration at 30 °C (the lowest constant temperature consistently attainable in the autosampler), headspace samples were introduced into the GC/MS equipped with a J&W Scientific (Folsom, CA) 122-1334 capillary column (30 m 0.25 mm ID 1.4 lm film thickness), with an automated oven temperature program as follows: 35 °C initial temperature to 105 °C at 10 °C min 1, followed by 50 °C min 1 to a final temperature of 180 °C. The concentrations of VOCs in the leaves (lg VOC kg 1 fresh leaf weight) were determined by subtracting the headspace concentrations measured in the leaf-spiked vials from those in the blankspiked vials and then dividing by leaf weight. Indirectly determining the leaf concentrations was found to be more analytically convenient and reproducible than directly determining the leaf concentrations using a conventional solvent extraction approach, but potentially less accurate. Leaf-air concentration factors (LACFs) in units of L kg 1, were generated from the slope of the regression line obtained by plotting the leaf concentrations by the vial headspace VOC concentration. 2.5. Leaf and air sampling within residence Several consumer products emitting chlorinated VOCs were introduced into a 20-year-old, two-story residence with a basement (sub-ground floor) containing the same four species of plants used in the headspace study. The consumer products included a gun cleaner containing TCE, a hobby glue containing PCE and an injection molded plastic lamp base emitting 1,2-DCA. These consumer products had been previously identified as VOC sources during field vapor intrusion investigations (Gorder and Dettenmaier, 2011). The concentrations of benzene, toluene and m,p-xylene, the aromatic hydrocarbons having the highest indoor air concentrations, originating from gasoline stored in the attached garage were also measured in leaf and air samples. Ethylbenzene, oxylene and other related aromatic hydrocarbons typically found in gasoline were not quantified. The residence was equipped with a forced air heating, ventilation and cooling (HVAC) system and had a total air volume, including the basement, of 600 m3 (estimated from measured room dimensions). Recording thermocouples were placed in HVAC vents to determine the frequency and duration of the HAVC operation with the house thermostat set at 78 °F (25.6 C). The basement had a separate forced air heating system. Three sets of four potted plants, one of each species, were introduced into the residence (one set on each floor) 24–48 h prior to the collection of the first background samples. A fourth set of plants was placed in an outdoor location at the rear of the residence. Background air and leaf (typically 0.5–1 g fresh weight) samples were collected just before the emitting objects were added to the residence. Emitting objects were quickly transported to a 2nd story room, subsequently referred to as the source room, inside an airtight container. After the objects were placed in the source room the door was closed and the HAVC system turned off to minimize mixing between rooms and allow the concentrations in the source room to increase. This mimicked the approach used in the roomto-room sample scheme that was originally used to identify the objects during VI investigations (Doucette et al., 2010). Twenty-four hours after the objects were placed in the source room the door to the room was opened and the HVAC system was turned on to mix the house air. Twenty-four to 48 h after the objects were introduced into the home, the objects were removed. Eight
sampling events took place, beginning with the background samples and continuing for 24 h after the objects were removed. During each air sampling, leaf samples from each of the four plant species were collected at the same time and placed in headspace vials for analysis as described previously, except that the vials were equilibrated at 100 °C to increase the desorption (i.e. extraction efficiencies) of the VOCs from the leaves. At this temperature, spike recoveries were greater than 90% for all six compounds. During each sampling event, sample collectors did not exit or reenter the house until all sampling was completed. This was done to avoid any unusual introduction of outdoor air into the residence. Air samples were collected using two TenaxÓ sorbent tubes (Alltech Associates, Deerfield, Illinois) attached in series to an Air Chek Sampler, SKC Model 224-44XR (Eighty Four, Pennsylvania) following the general approach outlined in EPA Method TO-17. Samples were collected for approximately 10–30 min at known flow rates from 100 to 150 mL min 1 depending on the expected concentrations. An Alltech Digital Flow Check™ model DFC™ Flowmeter (Nicholasville, Kentucky) was used to confirm flow rates at the beginning and end of each sampling event. Only a small number of replicate air samples were collected over the course of the residential study because of the limited number of sampling pumps available. Replicate air samples varied from 1% to 10% depending on the compound and concentration. Sorbent tube samples were introduced into an AgilentÒ 6890/ 5793GC/MS equipped with a J&W Scientific (Folsom, California) DB-624 capillary column (30 m 0.25 mm ID 1.4 lm film thickness) using a Perkin Elmer TurboMatrix ATD Automated Thermal Desorber operating under the following conditions: 5 min trap purge; cryo-trap temperature = 30 °C; TenaxÓ tube desorb = 300 °C for 10 min; cryo-trap temperature program 30 °C initial temperature to 320 °C at 40 °C s 1, transfer line to GC/MS at 225 °C. The moisture control system, traps, and tubes were thermally cleaned between each sample. Chromatographic conditions were as follows: oven temperature program start at 35 °C, hold for 2 min then ramp at 30 °C min 1 to 170 °C, followed by 70 °C min 1 from 170 to 230 °C with a 1 min hold at the final temperature. The MS was operated in selected ion monitoring (SIM) mode. An external standard approach was used to quantify the mass of compounds collected in each trap. Standards were prepared by injecting 1 lL of standards dissolved in methanol onto clean TenaxÓ traps with a micro-syringe. Masses of compound injected on the tubes ranged from 100 to 100 000 pg. New standard curves were prepared if the concentrations of the continuous calibration verification (CCV) standards deviated more than 10% from the expected value.
3. Results and discussion 3.1. Static headspace studies: leaf air concentrations factors (LACFs) Leaf air concentration factors (LACF), defined as the VOC concentration in the leaves (lg VOC per kilogram of fresh leaf, CL) divided by the concentration in the air (lg VOC per liter of air, CA), were generated from the slopes of the linear regression lines obtained by plotting leaf to air concentrations (Fig. 1) and are summarized in Table 1. As shown in Table 1, the LACFs generally increased as the extractable lipid content of the leaf increased (ficus to cactus) and as the Koa of the chemical (benzene to xylene) increased. A trend of increasing LACF for increasing chemical Koa value has been previously observed for semi-volatile compounds (Tolls and McLachlan, 1994; Thomas et al., 1998; Platts and Abraham, 2000; Weiss, 2000). However, since the range of log Koa values exhibited by the six VOCs was relatively narrow, the only significant (p < 0.05) linear relationship between log LACF
35
T.A. Wetzel, W.J. Doucette / Chemosphere 122 (2015) 32–37
Fig. 1. Example plot of leaf versus air concentrations generated using static headspace technique for Ficus and TCE (regression line: y = 41.2x 0.77. R2 = 0.98).
and log Koa was observed for ficus, the plant species with the highest extractable lipid content. Lipid normalization reduced the differences in the LACF values between species. However, since the cuticle consists of several lipid or lipid like components including cutin, cutan, and extractable waxes that exhibit varying affinities for organic contaminants, incorporation of these cuticular fractional amounts is likely necessary to better understand species differences (Barber et al., 2004; Chen et al., 2008). Future work should specifically examine the role of the polymeric lipid fraction (cutin) that was shown by Chen et al. (2008) to be the dominant sorptive fraction for naphthalene and 1-naphthol on tomato and apple cuticles. It is also possible, but less likely to be significant, that VOCs can sorb to the non-lipid fractions of the leaves. Normalizing the LACFs to the surface area of the leaves was also evaluated (data not shown), but it did not reduce the species LACF variation as much as lipid normalization. The observed relationships between the leaf and air concentrations for the six VOCs in the static headspace study showed that plants have the potential to be used as indoor air samplers. This potential was evaluated during the subsequent study conducted within an actual residence. 3.2. Relationship between leaf and indoor air VOC concentrations in an actual residence Measured leaf concentrations paralleled the indoor air concentrations for the chlorinated VOCs as illustrated in Fig. 2 for TCE and ficus in the source room. The corresponding changes in leaf and air
Fig. 2. Ficus leaf (average of triplicate leaf samples with error bars showing standard deviation) and air (single measurement) TCE concentrations in source room over time.
concentrations indicate a relatively rapid exchange of VOCs between leaves and air. For the non-chlorinated hydrocarbons, the leaf concentrations also paralleled the air concentrations. However, because of the nature of the source (fuel stored in an attached garage), the indoor air hydrocarbon VOC concentrations (benzene, toluene and xylene) varied only by factors of 2–3 as compared with factors of 100 for the chlorinated VOCs as shown in Fig. 3 for xylene. 3.3. Estimating indoor air concentrations for laboratory derived LACFs and measured leaf concentrations The parallel relationship between leaf and air VOC concentrations within the residence illustrates the potential for using leaves as real-time air samplers. In practice, it would be ideal if the measured leaf concentrations could be used along with the laboratory derived LACFs to predict indoor air concentrations. To evaluate this possibility, predicted indoor air concentrations, calculated by dividing the measured leaf concentrations by the appropriate laboratory derived LACFs, were compared to the measured indoor air concentrations. As illustrated in Fig. 4 for TCE and ficus, significant relationships (p < 0.05) between estimated and measured air concentrations were found for the three chlorinated VOCs for all plant species except for PCE with cactus and ficus. The indoor air concentrations of PCE, ranging from 0.0002 to 0.006 lg L 1, were the lowest of the three chlorinated VOCs but were still considerably below EPAs screening level carcinogenic risk based concentration of 9.4 lg PCE m 3 or 9400 lg L 1 (a). For the non-chlorinated hydrocarbons (benzene, toluene and m,p-xylene), the relationships between measured air concentrations and those predicted from laboratory derived LCAFs were not significant. The reason for the difference in predictability between the chlorinated and non-chlorinated VOC in not known but may be related to the nature of the source materials. The sources of chlorinated VOCs were introduced into the upper floor of the residence and removed 24–48 h after introduction. This resulted in indoor air concentrations of the chlorinated VOC that varied up to two orders of magnitude. In contrast, the source of the non-chlorinated VOCs (gasoline stored in attached garage) was constant and the indoor air concentrations varied only by factors of 3–5.
Bacci, E., Calamari, D., Gaggi, C., Vighi, M., 1990. Bioconcentration of organicchemical vapors in plant-leaves – experimental measurements and correlation. Environ. Sci. Technol. 24 (6), 885–889. Barber, J.L., Thomas, G.O., Kerstiens, G., Jones, K.C., 2004. Current issues and uncertainties in the measurement and modelling of air–vegetation exchange and within-plant processing of POPs. Environ. Pollut. 128 (1–2), 99–138. Baur, P., Marzouk, H., Schonherr, J., Grayson, B.T., 1997. Partition coefficients of active ingredients between plant cuticle and adjuvants as related to rates of foliar uptake. Agric. Food Chem. 45 (9), 3659–3665. Chefetz, B., 2003. Sorption of phenanthrene and atrazine by plant cuticular fractions. Environ. Toxicol. Chem. 22, 2492–2498. Chen, B., Li, Y., Guo, Y., Zhu, L., Schnoor, J.L., 2008. Role of the extractable lipids and polymeric lipids in sorption of organic contaminants onto plant cuticles. Environ. Sci. Technol. 42, 1517–1523. Cohen, Y., 1996. Volatile organic compounds in the environment: a multimedia perspective. In: Wang, W., Schnoor, J., 1996, Doi, J. (Eds.), Volatile organic compounds in the environment. ASTM STP 1261. American Society for Testing and Materials, West Conshohocken, PA. pp. 7–32. Cornejo, J.J., Munoz, F.G., Ma, C.Y., Stewart, A.J., 1999. Studies on the decontamination of air by plants. Ecotoxicology 8, 311–320. Dales, R., Liu, L., Wheeler, A.J., Gilbert, N.L., 2008. Quality of indoor residential air and health. Can. Med. Assoc. 179, 147–152. Dawson, H.E., McAlary, T., 2009. A Compilation of statistics for VOCs from post-1990 indoor air concentration studies in North American residences unaffected by subsurface vapor intrusion. Ground Water Monit. Remed. 29 (1), 60–69. Doucette, W.J., Hall, A.J., Gorder, K.A., 2010. Emissions of 1,2-dichloroethane from holiday decorations as a source of indoor air contamination. Ground Water Monit. Remed. 30 (1), 67–73. Girman, J.R., Phillips, T.R., Levin, H., 2009. Critical review: how well do house plants perform as indoor air cleaners? In: Proceedings, 9th International Conference and Exhibition on Healthy Buildings 2009, vol. 1 of 3. 13–17 September, Syracuse, NY, USA, pp. 823–826. Gorder, K.A., Dettenmaier, E.M., 2011. Portable GC/MS methods to evaluate sources of cVOC contamination in indoor air. Ground Water Monit. Remed. 31 (4), 113– 119.
T.A. Wetzel, W.J. Doucette / Chemosphere 122 (2015) 32–37 Hiatt, M., 1998. Bioconcentration factors for volatile organic compounds in vegetation. Anal. Chem. 70 (5), 851–856. Hiatt, M., 1999. Leaves as an indicator of exposure to airborne volatile organic compounds. Environ. Sci. Technol. 33 (22), 4126–4133. Keymeulen, R., Bruyn, G.D., Langenhove, H.V., 1997. Headspace gas chromatographic determination of the plant cuticle-air partition coefficients for monocyclic aromatic hydrocarbons as environmental compartment. Chromatography A 774, 213–221. Li, Y., Chen, B., 2009. Phenanthrene sorption by fruit cuticles and potato periderm with different compositional characteristics. Agric. Food Chem. 57, 637–644. Li, Q.Q., Chen, B.L., 2014. Organic pollutant clustered in the plant cuticular membranes: visualizing the distribution of phenanthrene in leaf cuticle using two-photon confocal scanning laser microscopy. Environ. Sci. Technol. 48 (9), 4774–4781. Lin, D., Zhu, L., He, W., Tu, Y., 2006. Tea plant uptake and translocation of polycyclic aromatic hydrocarbons from water and around air. Agric. Food Chem. 54, 3658– 3662. Liu, Y., Mu, Y., Zhu, Y., Ding, H., Arens, N.C., 2007. Which ornamental plant species effectively remove benzene from indoor air? Atmos. Environ. 41, 650–654. Luca, N., Jones, K., Gramatica, P., Papa, E., Gerabolini, B., Guardo, A., 2006. Accumulation of persistent organic pollutants in canopies of different forest types: role of species composition and altitudinal-temperature gradient. Environ. Sci. Technol. 40, 6580–6586.
37
National Aeronautics and Space Administration, 1989. Interior Landscape Plants for Indoor Air Pollution Abatement. Orwell, R.L., Wood, R.L., Tarran, J., Torpy, F., Burchett, M.D., 2004. Removal of benzene by the indoor plant/substrate microcosm and implications for air quality. Water Air Soil Pollut. 157, 193–207. Platts, J.A., Abraham, M.H., 2000. Partition of volatile organic compounds from air and from water into plant cuticular matrix: an LFER analysis. Environ. Sci. Technol. 34 (2), 318–323. Thomas, G., Sweetman, A.J., Ockenden, W.A., Mackay, D., Jones, K.C., 1998. Airpasture transfer of PCBs. Environ. Sci. Technol. 32 (7), 936–942. Tolls, J., McLachlan, M.S., 1994. Partitioning of semivolatile organic-compounds between air and lolium-multiflorum (welsh ray grass). Environ. Sci. Technol. 28 (1), 159–166. USEPA, 2013. U.S. Environmental Protection Agency. Residential Air Risk Screening Table [Internet]. Available from: (Cited 26.03.14). Weiss, P., 2000. Vegetation/soil distribution of semivolatile organic compounds in relation to their physicochemical properties. Environ. Sci. Technol. 34 (9), 1707– 1714. Yang, D.S., Pennisi, S.V., Son, K., Kays, S.J., 2009. Screening indoor plants or volatile organic pollutant removal efficiency. HortScience 44 (5), 1377–1381.
