In this study we assessed quantitatively airborne exposures generated from cleaning tasks performed under controlled work environment conditions. Several exposure measures such as TVOC, 2-butoxyethanol (2-BE) and ammonia were assessed with selected measurement methods. Our results show that VOC exposures remain airborne even after the cessation of cleaning tasks, suggesting potential exposure to anyone entering the room shortly after cleaning. Additionally, the results indicated that 2-BE peak concentrations from cleaning can approach occupational exposure limits, warranting further workplace investigations. The quantitative exposure measurements reported here contribute to the limited workplace exposure data in the literature related to cleaning. The main conclusions are discussed below:
1) The measurement methods we applied for assessing exposures from cleaning tasks are useful for future studies, with limitations.
We utilized several measurement methods for quantifying different exposure metrics including integrated sampling and analytical method of EPA TO-17. This method was selected because it provides detection of VOC at low concentrations and grants collection of compounds with a wide range of volatilities (e.g. ethanol BP = 78°C; 2-BE BP = 218 °C) by utilizing multi-media sorbent sampling tubes. However, for the 10 minute sampling only 2-BE was detectable with the method. The initial list of target volatile ingredients from the MSDS data included ethanol, 1-methoxy 2- propanol, and ethanolamine. Ethanolamine was removed from the target list because it was not amenable to the method. Ethanol &1-methoxy 2- propanol concentrations were lower than the LOD, for the 10 minute sampling period. For longer sampling periods (such as 20 minutes, data not shown) ethanol, 1-methoxy 2- propane were detectable with the EPA method. Given that our goal of capturing a range of VOC with this method was not achieved (either because of short term sampling or low product concentrations of target compounds), specific measurement of individual compounds may be more feasible to apply for workplace exposure assessment. For example, 2- BE can be measured using the NIOSH 1430 method.
TVOC measured with the DRI- PID underestimated exposures from cleaning activities. Given that the TVOC metric represents the sum of the volatile compounds of the mixture, including 2-BE, one would expect that the value of TVOC would be higher than the single ingredient. Because the 2-BE ionization potential (IP) is lower than the IP of the PID lamp (IP for 2-BE is 8.6 eV vs. 10.6 eV), we expected that 2-BE would be measured by the PID. However, our data indicate a clear 2-BE exposure underestimation by the PID. A possible explanation may be related to the differences in sampling methods between integrated sampling, which is based on active sampling; and the real time sampling, which is based on diffusion. Aerosol particles generated during product spraying may be captured by active sampling and not by the PID, therefore producing higher 2-BE levels by integrated sampling.
The same underestimation of VOC from the PID was observed by Coy et al. [25
]. This study compared PID results with integrated sampling during simultaneous measurements from the same solvent mixtures. The authors suggest that PID response underestimation is related to: a) different ionization potential of the individual compounds of the mixture; b) non-linearity of the PID response for high concentrations (2000 ppm); c) the size of the ionization chamber. Consistent with our findings, this study showed high correlation of the PID response with integrated sampling measures.
Due to the observed underestimation, we recommend that TVOC -DRI measurement for cleaning mixtures be conducted only when the limitations are taken into account. DRI can be used for: a) initial screening of TVOC concentrations; b) for evaluating exposure control strategies; and c) for identifying exposure peaks and exposure dynamics, which can be useful for prioritizing activities for further and more precise quantitative measures.
2) Quantification of airborne exposures from cleaning requires investigation of other exposure metrics and a variety of sampling and analytical techniques.
In addition to the exposure metrics considered here (TVOC, 2-BE, ammonia) other metrics can be considered for a comprehensive quantitative exposure assessment strategy. These include assessments of additional chemical agents with important health relevance for respiratory irritation and sensitization, such as quaternary ammonium compounds (quats), ethanolamines, and phenols. Quantitative assessment of these ingredients cannot be achieved with one single method, given that these chemicals have different chemical and physical properties. For example quats can be measured with Ion Chromatography [26
] and ethanolamines can be measured with GC/FID NIOSH 2007 method.
Further considerations for quantitative assessment can include aerosol exposure characterizations. Product spraying, a common activity during cleaning, generates liquid aerosols of variable chemical composition, including non-volatile compounds such as quats that have been associated with asthma symptoms in several case reports [27
]. However, current literature lacks the evidence on size distributions of particles generated from spraying during cleaning. Determination of respirable or ultrafine particle concentrations from spraying may provide a better understanding of cleaning related health effects. Additionally, there are no studies to date that focus on assessment of aerosol dust particles present in indoor environments as potential carriers of volatile and nonvolatile ingredients from cleaning. Secondary emissions generated from cleaning chemicals reaction with ozone, which have been investigated by experimental studies [17
], may also be important to consider when developing quantitative workplace exposure assessment strategies.
3) The quantitative findings for airborne TVOC and 2-BE suggest that common cleaning tasks contribute to poor indoor quality and may present a risk of adverse health effects.
The highest TVOC peak concentrations (approximately 11 ppm) and 10 minute average concentrations (approximately 6 ppm) were measured when the general purpose cleaner was used in the small unventilated bathroom. Although occupational and environmental standards for indoor air TVOC have not been established, Molhave et al. [28
] proposed indoor TVOC concentrations of increasing concern for health effects as follows: a comfort range (< 0.2 mg/m3
), a multi factorial exposure range (0.2-3 mg/m3
); a discomfort range (3- 25 mg/m3
); and a toxic range (> 25 mg/m3
). Our peak TVOC concentration data converted to mg/m3
(isobutylene equivalent) ranged from 0.66-26 mg/m3
. Concentrations we recorded for most of the tasks fall into the discomfort range. These results suggest that cleaning can make a significant contribution to the poor indoor air quality. Additionally, Molhave et al. [29
] recommended that if a direct reading detector indicates concentrations above 0.3 mg/m3
, further detailed exposure assessments for health effects evaluations are essential.
TVOC concentrations have been measured in several indoor environments including offices, schools, homes, and hospitals [29
]. In these settings, the TVOC ranged from 1-25 mg/m3
and were expressed as the average values for different time durations, from hours to days of air sampling. Even though these studies recognize the possibility of higher short-term TVOC exposures, peak TVOC- activity specific data which are important for asthma assessment, have not been evaluated [23
]. Because the degree of cleaning contribution to the short term peak exposures is unknown, further assessments in the workplace are needed.
Airborne concentrations from cleaning 2-BE may be a concern in the workplace. Concentrations of 2-BE measured here ranged widely among the tasks, with the highest values obtained when the general purpose cleaner with 5-7% 2-BE by weight was used in the small bathroom, approximately 21 ppm. California Proposition 65 has set the Reference Exposure Limit (REL) for 2-BE at 2.9 ppm for one hour of exposure. Our 2-BE results suggest that application of a general purpose cleaner continuously for several consecutive tasks in the workplace can easily result in worker's exposure higher than the California REL limit.
Several laboratory emissions studies have measured 2-BE concentrations from cleaning products. Slightly lower air concentrations than ours were reported by Zhu et al. [35
] in an experimental study that determined 2-BE emission factors using a field and laboratory emission cell (FLEC). One hour concentrations of 2-BE ranged from 2.8-62 mg/m3
(0.57- 12.6 ppm). Singer et al. [18
] investigated emission profiles of 2-BE from several cleaning products and reported concentrations of 0. 33-2.3 mg/m3
over one hour of exposure.
There are very limited workplace exposure data of 2-BE from cleaning. Occupational standards for 2-BE such as OSHA Permissible Exposure Limit (PEL) of 8 hr TWA is 50 ppm and NIOSH Recommended Exposure Limit (REL) for 10 hr TWA is 5 ppm (24 mg/m3
). Vincent and coworkers in 1993 [16
] assessed 2-BE workplace exposures for 29 cleaning workers, which ranged from 0.1-7.33 ppm for 8 hour TWA. 2-BE exposures from cleaning may meet OSHA regulations, however, compliance does not always imply that workers are protected from respiratory irritation symptoms from short term peak exposures [23
]. Our findings of concentrations as high as 21 ppm, although not directly comparable with the occupational standards, warrant further assessment of 2-BE from cleaning in the workplace.
Ammonia concentrations from the tasks performed (0.01-2.8 ppm) were low compared to OSHA -PEL 8 hour TWA of 50 ppm and NIOSH short term exposure limit (STEL) 15- min TWA of 35 ppm. Several studies have associated inorganic gases such as ammonia and chlorine with irritation symptoms reported among cleaning workers [17
]. Concentrations of ammonia reported by Ramon and coworkers range from 0.6-6.4 ppm with peaks over 50 ppm during domestic cleaning tasks [15
]. Lower concentrations were reported by Fedoruk et al. [37
] when assessing airborne ammonia from a window and a bathroom tile cleaner. This study concluded that standard cleaning solutions are unlikely to produce significant ammonia exposures, but the authors advise that application of more concentrated products (e.g. > 3%) in poorly ventilated areas may be of concern.
4) Concentrations of TVOC measured after cleaning suggests that exposures may affect not only workers involved in cleaning but also other building occupants.
Real time TVOC concentration profiles after the cessation of cleaning tasks indicated that it takes more than 20 minutes after cleaning for exposures to decline to background levels. This finding relates to a single application of one product used during one task, especially in the small unventilated room. It would be expected that multiple tasks performed consecutively would generate higher exposure concentrations requiring longer decay times. These results suggest that not only workers involved with cleaning, but others, who are present in the room after cleaning, are potentially exposed. Several emissions studies conducted in laboratory chambers have suggested that ingredients in cleaning products such as glycol ethers are slowly released in the air even hours after product applications [17
]. These experimental results indicate that there is a potential risk for exposure to other building occupants not involved with cleaning. Further quantitative investigation in real world scenarios is critical to evaluate airborne exposures after cleaning.