Environ Sci Pollut Res DOI 10.1007/s11356-014-3678-x

RESEARCH ARTICLE

Influence of relative humidity on VOC concentrations in indoor air Pawel Markowicz & Lennart Larsson

Received: 26 June 2014 / Accepted: 30 September 2014 # Springer-Verlag Berlin Heidelberg 2014

Abstract Volatile organic compounds (VOCs) may be emitted from surfaces indoors leading to compromised air quality. This study scrutinized the influence of relative humidity (RH) on VOC concentrations in a building that had been subjected to water damage. While air samplings in a damp room at low RH (21–22 %) only revealed minor amounts of 2ethylhexanol (3 μg/m3) and 2,2,4-trimethyl-1,3-pentanediol diisobutyrate (TXIB, 8 μg/m3), measurements performed after a rapid increase of RH (to 58–75 %) revealed an increase in VOC concentrations which was 3-fold for 2-ethylhexanol and 2-fold for TXIB. Similar VOC emission patterns were found in laboratory analyses of moisture-affected and laboratorycontaminated building materials. This study demonstrates the importance of monitoring RH when sampling indoor air for VOCs in order to avoid misleading conclusions from the analytical results. Keywords Indoor air . Volatile organic compounds . Relative humidity . Building environment . Sink effect

Introduction Volatile organic compounds (VOCs) in indoor environments may, to a large extent, stem from the building itself such as from the floor, ceiling, and walls. Some of these VOCs, typically of low molecular weight, may result from fastdecaying emissions that quickly—within a few weeks or months—disappear from the new building materials (Shin and Jo 2013). Such VOCs are unbound to the materials and Responsible editor: Constantini Samara P. Markowicz : L. Larsson (*) Division of Medical Microbiology, Department of Laboratory Medicine, Lund University, Sölvegatan 23, 223 62 Lund, Sweden e-mail: [email protected]

may include various solvents and impurities. Other VOCs result from slow-decaying emissions formed by different processes such as aging and degradation, e.g., due to hydrolysis reactions, sorption processes etc., and are chemically (for example between an adsorbent and adsorbate) or physically (van der Waals or electrostatic forces) bound to the materials (Wolkoff 1999). It is also known (Claeson et al. 2009; Mendell 2007; Wieslander et al. 2007) that slow-decaying emissions may last in indoor environments for a long period of time and create health problems. One example is emissions from linoleum/PVC flooring (Sjöberg and Ramnäs 2007) which may persist during the entire lifetime of the building. Meininghaus et al. (2000) showed that certain materials used indoors can act like buffers for VOCs, thus reducing air concentrations by sorption and diffusion mechanisms; the effect is particularly obvious at low ventilation rates. The mechanisms may be complex and depend upon environmental factors as well as the chemical-physical properties of the emission molecules and the building materials (Wolkoff 1999; Zhang et al. 2002). Analysis of VOCs can be useful in guiding remediation measures of unsatisfactory indoor environments. For example, elevated air concentrations of formaldehyde representing the so-called very volatile organic compounds (VVOCs) may lead to replacement of building materials (Bruinen de Bruin et al. 2008), elevated concentrations of 2-ethylhexanol may lead to floor remediation (Tuomainen et al. 2004), and findings of 1-octen-3-ol may lead to measures taken against molds (Schleibinger et al. 2008). Obviously, it is of utmost importance that the methods used for sampling and analyses of the VOCs are reliable. It is well known that the relative humidity (RH) affects the rate of emissions of VOCs from surfaces. Andersen et al. (1975) already showed, in chamber measurements, that fast-decaying emission rates of formaldehyde at 22 °C doubled when increasing the RH from 30 to 70 %. In addition, Wolkoff (1998) suggested that at high RH, polar

Environ Sci Pollut Res

substances will be desorbed from a surface by the water vapor. Also in field studies, it has been shown that formaldehyde concentration indoors may increase with increased RH (Van Netten et al. 1989). Interestingly, studies on emissions from new materials are recommended to be performed in a climate chamber at defined RH. According to some standards (e.g., by the Swedish Standard Institute in SS-EN ISO 16000–10:2006, 2012), such studies should be performed typically at 50±5 % RH. At the same time, however, monitoring of RH is usually not considered when sampling VOCs in indoor air. In the present study, we analyzed VOCs, both at relatively low and relatively high RH, in one room in a moistureaffected building and in another room in a building not apparently affected by moisture. We also studied, at different RHs, emissions of VOCs from a moisture-affected impregnated sill. In addition, we studied, in a climate chamber, the impact of RH on the emissions of VOCs from three different types of building materials that in laboratory experiments had been exposed to a selected group of VOCs.

Materials and methods Room air samples Two storage rooms were studied. In one room (25 m2, with a wooden floor), there were no complaints on the perceived air quality, whereas in the other room (5.7 m2, with a PVC flooring), there was a disturbing odor. Air samples taken from the latter room had previously showed distinctive amounts of n-butanol, 2-ethylhexanol, and 2,2,4-trimethyl-1,3pentanediol diisobutyrate (TXIB), strongly indicating emissions from the floor (data not shown). To improve the distribution of the water vapor in the room, a small fan was set to operate approximately 1 m from the humidifier (Air-O-Swiss U7145) placed in the middle of the room. In a pre-study, the device in itself had been found not to change air concentrations of the measured VOCs. The temperature and RH were measured by using a Testo® 615 m (accuracy ±0.5 °C and ±3 % RH, respectively). Four samples were taken by pumping air through Tenax® TA tubes (30 min, 100 mL/min) that were then sent to IVL (Stockholm, Sweden) for gas chromatography–mass spectrometry (GC-MS) scan analysis after thermal desorption. The larger room had the mechanical ventilation closed during the experiment. The first sample was collected before the humidifier was switched on (34 % RH). The second sampling started after the RH had reached 58 %, and the third sampling started immediately after the second sampling had been completed. Immediately after the third sampling, at 68 % RH, the humidifier was switched off (5 h after, it had been

switched on). The fourth (final) sampling was performed 16 h later (see Table 1). The smaller room had a ventilation rate (as measured by using a TSI VelociCalc® 9565 air velocity meter) of 0.38 air change/h. Following the same schedule as the previous, four air samples were collected (see Table 1). The first sample was collected before the humidifier was switched on (22 % RH), and two additional samples were taken consecutively when the RH in the room exceeded 58 %. During the third sampling, the humidity peaked at 75 % RH; the humidifier was then shut down automatically to reach 48 % at the end of the sampling (2.5 h after the humidifier had been switched on). As previously mentioned, the last sampling was performed 16 h later. Moisture-affected wood material Five pieces of an impregnated floor wooden sill (approximately 180 g) emitting an unpleasant odor were studied. The material had been collected from a building with severe complaints as regards the perceived indoor air quality. Samples were stored in a ventilated hood at room temperature and a constant air flow rate of 500 l/s before analysis and then transferred to a stainless steel climate chamber (560× 480×400 mm inside, 108 l) with an efficient system for atmosphere and heat distribution (Memmert HCP 108; VWR, Sweden; accuracy of measurements ±0.1 °C and ±1 % RH), adapted for air sampling by having two sampling ports. For purification of the incoming air, a tube containing Anasorb® 747 (SKC Inc., Eighty Four, PA, USA) was used in one of the ports; chamber air samplings were conducted using a tubing passing through the other port. The ventilation rate was 0.06 air change/h during the experiment. The sill samples were kept in the chamber for 2 h (30 °C, 40 % RH) following air sampling and thereafter taken back into the hood. After 24 h, the same samples were again placed in the climate chamber and kept there for 2 h (30 °C, 85 % RH) following air sampling. VOCs were collected by pumping air through Tenax® TA tubes for 30 min at a flow rate of 100 mL/min; Table 1 Timeline of the indoor experiments including sampling events and RH changes in the rooms. RH values and time notations refer to start and finish of each sampling event Sampling

Larger room RH (%)

1st 2nd 3rd 4th

Small room Time [h]

34–35 0.00–0.30 Humidifier switched on 35–58 2.00–2.30 58–68 2.30–3.00 Humidifier switched off 34–35 16.30–17.00

RH (%)

Time [h]

21–22 0.00–0.30 Humidifier switched on 58–73 4.30–5.00 75–48 5.00–5.30 Humidifier switched off 19–21 16.30–17.00

Environ Sci Pollut Res

thereafter, the tubes were sent to IVL (Stockholm, Sweden) for GC-MS analysis in scan mode after thermal desorption. A matched-pairs t test (GraphPad Prism 6 software) was performed to determine if the difference between the air concentrations of the VOCs from the materials when collected at low (40 %) and high (85 %) RH was significant. Laboratory-contaminated building materials A series of experiments were conducted in order to study, at different RHs and time intervals, emissions from building materials intentionally contaminated (in the laboratory) with different VOCs. An aqueous solution containing 1-butanol, 3-methyl-2-butanol, 3-methylbutanol, dimethyl disulfide, toluene, hexanal, 2-heptanone, styrene, anisole, α-pinene, 1-octen-3-ol, benzaldehyde, 2-ethylhexanol, and limonene (purchased from Sigma Aldrich, Schnelldorf, Germany) was prepared by dissolving 5 μL of each of the 14 VOCs in 925 μL of distilled water with a few microliters of detergent (Tween 20, purchased from MP Biomedicals, LLC Solon, USA). These 14 VOCs represent different classes of chemicals with a wide range in polarity and water solubility, molecular weight (74–136), boiling point (109–186 °C), and vapor pressure (0.2–22 mmHg at 20 °C). The studied VOCs, among which toluene, styrene, αpinene, and limonene are the least polar, are commonly reported in damp indoor environments. Ten samples of each of three different types of building materials were subjected to study. Pieces of gypsum drywall (81 cm2) and wood (120 cm2) purchased at a local building store in Lund, Sweden, were used. Disks of 26-cm2 concrete were a kind gift from LTH, Lund, Sweden. One set of samples (A), comprising three pieces of each of the different building materials, was prepared by applying a 1-mL aliquot of the VOC mixture on the material surface using a plastic z-shaped 50×145 mm rod. Each piece was transferred into a 400-mL glass Petri dish that was then covered by Parafilm® and a sheet of an aluminum foil and kept for 24 h at room temperature. Thereafter, the sample was removed from the Petri dish and placed in a ventilating hood for 30 days prior to analysis. A second set of samples (B) belonging to the same batch of the building materials (n=3) was prepared where a piece of the material to be studied was placed in a 1-L glass beaker containing a 50-mL glass beaker with a 1-mL aliquot of the VOC mixture. The larger beaker was then covered with Parafilm® and an aluminum foil and stored as described above. These experiments were designed to illustrate exposure to VOCs in aqueous solution (A) and VOCs in the gaseous (B) phase, respectively. After 30 days of storage, one building material sample at the time was placed in the climate chamber for analysis. The ventilation rate was 0.14 air change/h during the experiments. The first air sampling was performed 2 h after the study

conditions (30 °C and 40 % RH) had been attained; thereafter, the piece was transferred back into the hood. After 24 h, the same piece was again placed in the climate chamber and subjected to a second measurement (30 °C, 85 % RH). The same procedure was repeated for each of the studied building material pieces. All samplings were performed by pumping air through an Anasorb® 747 tube (30 min, 250 mL/min). The same procedure was also applied to each of three building materials (n=4) that had not been exposed to the VOCs (negative controls). Two weeks after the first measurements, during which time the building material pieces had been stored in the hood, they were again analyzed using the procedure described. All analyses were completed within a time frame of 4 weeks. The contents of the Anasorb® 747 tubes were extracted with 1 mL of dichloromethane containing 240 ng of N-octanol (D17) (purchased from Cambridge Isotopes Laboratories, Inc. Andover, USA) as an internal standard. Samples were analyzed by GC-MS in selected ion monitoring (SIM) mode and presented as toluene equivalents. A toluene standard curve was constructed by injecting 2.4–160 pg of toluene and 240 pg of the internal standard. The detection limit was calculated as 0.1 μg/m3. Statistical analyses to determine if the difference between the air concentrations of the VOCs from the materials when collected at low (40 %) and high (85 %) RH was significant were performed as described above. A Varian model 3800 gas chromatograph equipped with a combiPAL autosampler (CTC Analytics AG, Zwingen, Switzerland) and a fused silica capillary column (VF5ms, 60 m× 0.25 mm ID, 1 μm film thickness, Agilent Technologies) coupled to a 1200-L triple quadrupole MS detector (Varian INC. Walnut Creek, CA, USA) was used. Helium was used as a carrier gas at a column flow rate of 1.0 mL/min. The column temperature was programmed to rise from 50 to 200 °C at 7 °C/min where it was held for 4 min. The injector temperature was 200 °C; the transfer line temperature, 250 °C; the ion source temperature, 200 °C; the electron energy, 70 eV; and the filament current, 50 μA. One-microliter injections in the splitless mode were used.

Results Room air samples Results are shown in Tables 2 and 3. Air concentrations of the VOCs collected in the larger room at different RHs were comparable. However, in the smaller room, a 3-fold increase of the air concentrations of 2-ethylhexanol and a 2-fold increase of TXIB were found shortly (approximately 1 h) after increasing the RH. Interestingly, the concentrations of these compounds decreased again according to the measurements at lower RH 16 h later. The room temperatures were 22–24 °C.

Environ Sci Pollut Res Table 2 Air concentrations of VOCs (μg/m3) at different RHs in the larger room Compound

Range of RH 34–35 %

35–58 %

58–68 %

34–35 %

n-Decane

1

1

1

0

α-Pinene n-Hexanal n-Butanol TVOCs

1 5 3 61

1 5 4 67

1 6 4 64

0 7 2 50

Table 4 Air concentrations of VOCs (μg/m3, toluene equivalents) in the climate chamber containing impregnated wood sills exposed to different RHs and p value

Moisture-affected wood material Results are shown in Table 4. Increasing the RH in the climate chamber caused a significant 4-fold increase of the total VOCs (TVOCs) emitted by the wood samples. For example, the concentration of n-butanol rose from 2 μg/m3 (40 % RH) up to 43 μg/m3 (85 % RH); the corresponding increase for trichloroanisole was from 1 to 10 μg/m3. In addition, four compounds, which were not identified, were found only at the higher RH. However, the air concentrations of some other VOCs such as toluene, limonene, or 2-ethylhexanol were unchanged with respect to RH. Laboratory-contaminated building materials The results show that the building material samples 30 days after being exposed to VOCs in an aqueous solution (A) or in the gaseous phase (B) still contained some or all of the VOCs. There was a significant increase in the air concentrations of these VOCs when measuring the material samples at relatively low (40 %) as compared with high (85 %) RH (Tables 5, 6, and 7). None of the unexposed building materials (negative controls) contained any detectable amounts of the studied VOCs. Traces of limonene, benzaldehyde, hexanal, and αpinene were found in air samples of wood exposed at RH 85 %. These trace amounts were considered as background. Air concentrations of TVOCs emitted by the gypsum drywall samples increased 13-fold (A) and 8-fold (B), Table 3 Air concentrations of VOCs (μg/m3) at different RHs in the smaller (damp) room Compound

n-Hexanal n-Butanol 2-Ethylhexanol TXIB TVOCs

Range of RH 21–22 %

58–73 %

75–48 %

19–21 %

5 1 3 8 99

7 1 7 11 108

4 2 9 12 80

3 1 3 7 63

n.d. not detected (

Influence of relative humidity on VOC concentrations in indoor air.

Volatile organic compounds (VOCs) may be emitted from surfaces indoors leading to compromised air quality. This study scrutinized the influence of rel...
184KB Sizes 0 Downloads 7 Views