3.1. Side by side variation - precision
Side-by-side sampling is a way to evaluate the precision of the entire method, including sampling and laboratory analysis. To do so, five replicas (n = 5) were analyzed at each workstation on the second sampling day at UC and WSP. Using the GLM-ANOVA analysis, no significant difference could be demonstrated between the concentrations of any side-by-side group of samples. This observation is true for the 21 mVOCs tested here. The low variability between the samples taken simultaneously demonstrates the good precision of our ambient air mVOCs method.
3.2. Sampling duration
During the first day of sampling at UC and WSP, no breakthroughs exceeding 5% could be observed in the corresponding backup tube and this is effective for all sampling duration tested (15 min, 30 min, 1 h, and 2 h) and all mVOCs. The maximum breakthrough values obtained were 4.7% at the UC and 2.9% at the WSP, both for cyclohexanone. In order to optimize the sampling protocol, the impact of sampling duration on the measured concentrations were determined. The GLM-ANOVA analysis showed that the sampling duration had no influence on the concentrations of most of the tested mVOCs. A significant impact was only found for pent-2-en-1-ol and octan-3-one concentrations (Table 2). For pent-2-en-1-ol, the difference can be explained by the fact that at low sampling duration, concentrations were below LOQ, but no difference could be observed between the samples that were above LOQ (Fig. 3, panel A). For octan-3-one, a significant decrease of concentration was observed when the sampling duration was increased. The 2-hour sampling duration produced significantly lower concentrations than the 15- and 30- minutes’ durations (Fig. 3, panel B). It is worth mentioning that a 2-hour sampling duration will result in concentrations representing an average of the concentrations over that time and may include high and low levels. This could result in lower reported concentrations if the low concentrations period represents most of the sampling duration. Other explanations could be (i) the presence of other mVOCs interfering with the octan-3-one binding, (ii) releases, (iii) its stability, or (iv) other chemical reactions in the sorbent tubes. However, due to the lack of data in the literature and since the developed analytical method is specific to the 21 selected mVOCs, other mVOCs and degradation products cannot be detected.
Thus, regarding the sampling duration, since no breakthrough nor any major impact of the duration were observed and with the objective of improving the sensitivity of the method, the maximum sampling duration of 2 hours was chosen for the second days of sampling.
Table 2
Statistical results of the sampling duration’s impact on the mVOC concentration (statistically significant) at the WSP during the first sampling day.
mVOC | Duration (min) | N | Mean | SD |
Octan-3-one | 15 | 12 | 1.99 | 0.11 |
| 30 | 12 | 1.79 | 0.18 |
| 60 | 3 | 1.65 | 0.04 |
| 120 | 3 | 1.49 | 0.02 |
Pent-2-en-1-ol | 15 | 16 | 1.70 | 3.77E-08 |
| 30 | 16 | 2.64 | 0.04 |
| 60 | 4 | 2.62 | 0.02 |
| 120 | 4 | 2.65 | 0.02 |
3.3. Temporal variations
Daily variations of contaminants concentrations in a workplace can be attributed to many factors. For example, an increase in contaminants’ concentrations may be due to deficient ventilation, modification in the workload, modifications of treated materials, higher number of sources. The temporal variation was studied by comparing the concentrations measured during four consecutive 2-hours periods at both sites (UC and WSP). The only significant variation observed was for the ethyl acetate at the UC (Table 3). The difference was demonstrated between the two AM periods and the last period of the day which took place from 3 p.m. to 5 p.m (Fig. 4). Surprisingly, ethyl acetate is the only compound in this study with an increase, making it challenging to explain. The emission of VOCs in indoor environments is not well understood, since some compounds have a combination of sources (Rösch et al., 2014). The ethyl acetate increase could be attributed to many factors, one of which could be insufficient ventilation during the daily activities. This is because during the day, regular emissions from building materials are added to the daily anthropogenic emissions from the use of cleaning products, perfumes, or cosmetics. A weakness of this explanation is that building materials and daily anthropogenic activities are sources for other tested compounds such as butan-2-one. Ethyl acetate is used as a solvent for varnishes and in base coats and paints as material sources (National Pollutant Inventory Substance Profile, 2006) and in perfumes and cosmetics as anthropogenic source. Butan-2-one is also present in new building materials as it is used as a solvent, surface coating and in synthetic resins (ATSDR, 1992), but it also has a potential source from cleaning agents which are more related to human activities.
At the WSP, no hourly significant difference between the concentrations was demonstrated for any of the mVOCs measured (Table 3 and Fig. 4, example: Nonan-2-one). This could be explained by a steady average hourly emission of mVOCs, or by efficient ventilation that limits the accumulation and permits proficient dilution of the contaminants maintaining the concentrations at a steady state.
Table 3
Statistical results of the sampling period’s impact on the mVOC concentration (statistically significant) at the WSP during the first sampling day.
mVOC | Period | N | Mean | SD |
Ethyl acetate (University/spring) | AM1 | 10 | 2.67 | 0.10 |
| AM2 | 9 | 2.90 | 0.05 |
| PM1 | 10 | 2.89 | 0.05 |
| PM2 | 10 | 3.05 | 0.06 |
Nonan-2-one (WSP/summer) | AM1 | 10 | 1.93 | 0.21 |
| AM2 | 10 | 2.71 | 0.02 |
| PM1 | 10 | 2.67 | 0.05 |
| PM2 | 10 | 2.57 | 0.07 |
3.4. Variations between sites (UC vs WSP)
Among the 21 selected mVOCs, 8 compounds (2-ethyl-1,6-dioxaspiro[4.4]nonane, 3-methylbutan-1-ol, 5-ethyloxolan-2-one, octan-2-ol, pentan-2-ol, pentan-3-ol, pentyl hexanoate and undecan-6-one) did not produce any result above the limit of quantification (LOQ) nor the limit of detection (LOD) in the studied sites (UC and WSP). The LOQ is the lowest concentration of the compound, capable of producing a quantifiable signal in the real matrix with good reliability. Since the purpose of this study is to quantify mVOCs present in the air of two workplaces and compare their concentrations, LOQ were used.
At the UC, 17 mVOCs were below the LOQ. Pent-1-en-3-ol was detected only on the first sampling day and remained below the LOQ on the second sampling day, leaving only three mVOCs (ethyl acetate, butan-2-one and cyclohexanone) detected and quantified on all days. For these three compounds, the concentrations were significantly higher at the UC than at the WSP (Fig. 5). All mVOCs detected at the UC except for the pent-1-en-3-ol, detected only once, have known non-microbial sources (El Aroussi et al., 2018) and therefore could come from sources other than molds. Given that the campus was newly built, multiple other non-microbial VOCs sources can be responsible for their presence. These could be wall painting, new furniture, building materials, office equipment (printers, copiers), cleansers and disinfectants.
At the WSP, 13 mVOCs were detected and quantified and 10 of them were not detected at the UC (Fig. 5). Four of the detected mVOCs (pent-1-en-3-ol, pent-2-en-1-ol, nonan-2-one and non-3-en-1-ol) have no known non-microbial sources and can possibly be considered specific to the presence of microorganisms and their emissions (El Aroussi et al., 2018).
In addition, different studies have reported concentrations of various mVOCs in composting facilities (Fischer et al., 1998, 2000; Müller et al., 2004; Tolvanen et al., 2005). Tolvanen et al., (2005) measured 18 mVOCs in a composting plant with two of them (ethyl acetate and butan-2-one) in common with our mVOCs. The concentrations measured in the composting hall varied between 20 to 490 µg/m3 for the ethyl acetate and 2 to 560 µg/m3 for the butan-2-one (Tolvanen et al., 2005). The concentrations measured in the present study at the WSP were approximately 12 to 115 times lower (1.64 to 4.27 µg/m3) for ethyl acetate and more than 60 times lower for butan-2-one (8.6 µg/m3). In another study by Müller et al., (2004) in compost plants, the concentrations of 1-octen-3-ol measured were between 0.42 to 0.94 µg/m3 and the ones for 3-octanone were between 0.99 and 2.04 µg/m3 of air. The concentrations found by Müller et al. were approximately in the same order of magnitude as the ones measured at the WSP for 1-octen-3-ol (0.27 to 0.62 µg/m3). However, the one measured for 3-octanone were 30 to 60 times higher in their compost plant than the WSP. A study by Rodriguez-Navas et al., (2012) in different solid waste treatment plants two mVOCs measured were in common with the one measured at the WSP (ethyl acetate and cyclohexanone). Their concentrations of ethyl acetate (1 and 796 µg/m3) were higher than the concentrations measured by Tolvanen et al. and the concentrations of cyclohexanone (0.5 and 7.5 µg/m3) (Rodríguez-Navas et al., 2012) were 2 to 30 times higher than those measured in the present study (0.25 and 0.28 µg/m3). Finally, a study by Gallego et al., (2012) measured the VOCs in indoor air of a waste treatment facility. The measured concentrations of ethyl acetate, the only compound in common, were much higher (varying from 691 ± 199 µg/m3 to 7866 ± 5297 µg/m3 depending on the sampling location) than those at the WSP (1.64 to 4.27 µg/m3). These differences can be attributed to several factors, including different methodology (sampling and analysis techniques such as sampling duration and volume, type of adsorbent tubes, desorption conditions, etc.), good ventilation during sampling days at the WSP (credit to the large open garage doors contributing to indoor air dilution), the absence of organic waste at the WSP, variations in the tonnage of material processed between centres or different statistical analysis used (Mbareche et al., 2021).
3.5. Spatial variation (within site)
At the UC, for the three mVOCs having concentrations above the detection limits (butan-2-one, cyclopentanone, ethyl acetate) (Table 4), the levels found in the two laboratories (geography and information technology IT) were statistically significantly lower than the ones in the classrooms and the cafeteria. For cyclopentanone, the lowest concentrations were measured in the geo-lab, for butan-2-one, the classroom 2061 had significantly the highest concentrations and for ethyl acetate, the concentrations measured in the two laboratories were significantly lower than the ones measured in the two classrooms. One possible explanation for the lower concentrations in both laboratories could be attributed to the better performance of the ventilation systems in the laboratories compared to the basic ventilation in the classrooms. In fact, ventilation systems in laboratories always have 100% fresh air with an air change per hour adjustable according to occupancy; it can reach 10 ACH during peak occupancy periods. A good ventilation system in indoor environment dilutes the air contaminants by introducing and removing the ambient air and can control the risks related to the accumulation of these contaminants (Turcotte and Marceau, 2021).
For the 13 mVOCs measured at the WSP (Table 4) significant differences of ambient air concentrations could be demonstrated for 11 of them, no difference was observed for oct-1-en-3-ol and pent-2-en-1-ol between the five workstations. The concentrations obtained at the reception of the recyclable waste were lower for 12/13 of the mVOCs measured. Butan-2-one is the only one not having the lowest concentrations at this workstation. As indicated previously, butan-2-one is not a specific mVOC and is found in various products (ATSDR, 1992). The lower concentrations at the reception could be due to a very good aeration provided by the open garage doors. On the opposite side, the workers at the three sorting stations are working with the highest mVOCs concentrations in the ambient air. For six mVOCs (butan-2-one, cyclohexanone, cyclopentanone, ethyl acetate, octan-3-one, pentan-2-one) the highest concentrations were measured in the pre-sorting station. For five of them, (decanal, non-3-en-1-ol, oct-1-en-3-ol, octan-1-ol, pent-1-en-3-ol) the highest concentrations were observed in the manual sorting station. And for three mVOCs (non-3-en-1-ol, nonan-2-one, pent-2-en-1-ol), it was in the optic sorting that the highest concentrations could be measured, with the same concentrations measured for nonen-3-ol at the manual and the optic sorting stations (Table 4). Given that the pre-sorting room is the first station of the recyclable treatment process, it is expected to have high levels of dust and therefore high generation of contaminated particles and mVOCs into the air. It has been observed that the concentrations of contaminants may decrease throughout treatment process (Marchand et al., 2007). By proceeding, the materials will be divided and separated which can hypothetically cause some variation of the contaminants’ concentrations for some type of materials. However, it is not surprising to observe an increase in mVOC in the air of all working stations with sorting activities, due to the significant handling of contaminated materials (Salambanga et al., 2022). Many factors can influence the mVOCs’ levels in the air of the different workstations at the WSP, namely the working behaviors, the treatment stage, the system of ventilation and its position, the recyclable microbial contamination, and the sampling location from the source (Madsen et al., 2021).
Table 4
Mean log-concentrations by workstations measured in the ambient air of the WSP and the UC and the standard deviation (SD). According to the Tukey-Kramer multiple comparison test for each mVOC separately, two workstations sharing the same letter (A, B or C) do not have significantly different concentrations, on the opposite two different letters meaning significant differences were demonstrated. ND: not detected.
| | Log-concentrations (ng/m3) |
| | WSP | UC |
| | Reception | Pre-sorting | Optic sorting | Sorting | Ballot | Cafeteria | Classroom 1201 | Classroom 2061 | Geo lab | IT lab |
| N | 5 | 5 | 5 | 5 | 5 | 5 | 5 | 5 | 5 | 5 |
Butan-2-one | Mean | 2.33AB | 2.50B | 2.05A | 2.05A | 2.05A | 2.80B | 2.79B | 2.90A | 2.75B | 2.74B |
| SD | 0.3 | 0.47 | 0.00 | 0.00 | 0.00 | 0.03 | 0.03 | 0.01 | 0.13 | 0.04 |
Cyclohexanone | Mean | 1.24A | 2.42C | 1.70B | 1.67AB | 1.60AB | 2.71B | 2.81B | 2.85B | 2.22A | 2.70B |
| SD | 0.00 | 0.03 | 0.48 | 0.46 | 0.38 | 0.01 | 0.01 | 0.01 | 0.25 | 0.06 |
Cyclopentanone | Mean | 0.51A | 1.52B | 1.15AB | 1.11AB | 0.51A | ND | ND | ND | ND | ND |
| SD | 0.00 | 0.64 | 0.67 | 0.63 | 0.00 | - | - | - | - | - |
Decanal | Mean | 2.91A | 3.18B | 3.28BC | 3.46C | 3.24B | ND | ND | ND | ND | ND |
| SD | 0.14 | 0.24 | 0.14 | 0.09 | 0.09 | - | - | - | - | - |
Ethyl acetate | Mean | 2.32A | 3.42B | 3.00AB | 2.36A | 2.42A | 3.04C | 3.27B | 3.11BC | 2.76A | 2.90AC |
| SD | 0.09 | 0.22 | 0.31 | 1.13 | 0.36 | 0.01 | 0.32 | 0.01 | 0.09 | 0.11 |
Non-3-en-1-ol | Mean | 2.29C | 2.41C | 2.86B | 2.86B | 2.56A | ND | ND | ND | ND | ND |
| SD | 0.03 | 0.22 | 0.11 | 0.07 | 0.06 | - | - | - | - | - |
Nonan-2-one | Mean | 2.18C | 2.49AC | 2.75AB | 2.70B | 2.35AC | ND | ND | ND | ND | ND |
| SD | 0.38 | 0.46 | 0.23 | 0.25 | 0.35 | - | - | - | - | - |
Oct-1-en-3-ol | Mean | 1.65 | 1.72 | 2.03 | 2.10 | 1.71 | ND | ND | ND | ND | ND |
| SD | 0.82 | 0.89 | 0.80 | 0.67 | 0.88 | - | - | - | - | - |
Table 4
| | Log-concentrations (ng/m3) |
| | WSP | UC |
| | Reception | Pre-sorting | Optic sorting | Sorting | Ballot | Cafeteria | Classroom 1201 | Classroom 2061 | Geo lab | IT lab |
| N | 5 | 5 | 5 | 5 | 5 | 5 | 5 | 5 | 5 | 5 |
Octan-1-ol | Mean | 1.81A | 1.99AB | 2.53AB | 2.70B | 2.31AB | ND | ND | ND | ND | ND |
| SD | 0.65 | 0.84 | 0.46 | 0.35 | 0.46 | - | - | - | - | - |
Octan-3-one | Mean | 0.70A | 1.45B | 1.31B | 1.03C | 0.70A | ND | ND | ND | ND | ND |
| SD | 0.00 | 0.07 | 0.14 | 0.34 | 0.00 | - | - | - | - | - |
Pent-1-en-3-ol | Mean | 2.53A | 2.97B | 3.28C | 3.30C | 3.07BC | ND | ND | ND | ND | ND |
| SD | 0.23 | 0.33 | 0.22 | 0.09 | 0.09 | - | - | - | - | - |
Pent-2-en-1-ol | Mean | 1.96 | 2.11 | 2.16 | 2.15 | 2.01 | ND | ND | ND | ND | ND |
| SD | 0.32 | 0.33 | 0.22 | 0.09 | 0.09 | - | - | - | - | - |
Pentan-2-one | Mean | 2.48A | 4.04B | 3.67BC | 3.12AC | 3.07AC | ND | ND | ND | ND | ND |
| SD | 0.59 | 0.11 | 0.18 | 0.79 | 0.53 | - | - | - | - | - |
3.6. Seasonal variations
At the WSP, the GLM-ANOVA demonstrated a significant impact of the seasons on the average concentrations for all but two of the mVOCs (cyclohexanone and non-3-en-1-ol). Of the 11 remaining mVOCs, a significant augmentation of the concentrations could be demonstrated except for the butan-2-one for which a reduction of 50% was observed between spring and summer. Since butan-2-one may have other non-microbial sources and can’t be considered as a good indicator of mold presence, this decrease in concentration between spring and summer is not necessarily contradictory. During the summer, the largest increase observed was for oct-1-en-3-ol with concentrations 30 times higher, followed by cyclopentanone and pent-2-en-1-ol with an increase of about 10 times and octan-1-ol with a ratio summer/spring of 7.76. The rest of the detected mVOCs showed less marked differences although significantly higher concentrations were observed in the summer sampling days with ratios varying from 1.7 to 3.7. Figure 6 shows the concentration ratios between April and June for the quantified mVOCs at the WSP. The increase of the average concentrations during the summer may be due to higher microbial growth related to higher temperature and humidity during this season. Several studies have tested the impact of the seasons on the concentrations of bioaerosols and have shown that higher temperatures were associated with higher concentrations and therefore higher level of exposure to bioaerosols (Mbareche et al., 2022; Madsen et al., 2021, 2020, 2019; Park et al., 2011; Neumann et al., 2002). This could also be due to some changes in the load of the recyclable waste received as well as in the microbial burden of the received waste (Madsen 2021). However, due to the small number of samples, it is not possible to draw conclusions about temporal, spatial and seasonal impacts. To do so, further measurements in different locations will be needed.
3.7. Correlation analysis
The Pearson correlation was performed on the log-transformed concentrations measured on the second sampling days of the campaigns at both UC and WSP. This demonstrates statistically significant relationships between many of the mVOC quantified. Several direct proportional correlations (Table 5) with coefficients above 0.85 showing strong relationship were observed, meaning that these compounds could be produced by different molds that are present at the same location during the sampling. Those correlations are mainly observed between the two mVOCs, believed to be specific to fungi, included in the present study (nonan-2-one and non-3-en-1-ol) and two mVOC indicators of fungal contaminations (oct-1-en-3-ol and octan-3-one) (Ryan & Beaucham, 2013). Decanal and octan-1-ol, although not previously identified as fungal indicators and frequently used in solvents, could be informative mVOCs about the presence of fungi since they show good relationships with specific and indicator mVOCs. These results indicate that many mVOC are concomitantly present in the air and that grouping them to create profiles indicative of fungal incidence might be a promising approach. This suggests that a profile constructed from some of the indicator mVOCs quantified in the air would be necessary to suspect judiciously fungal contamination. However, due to the limited number of sampled locations, the present study can’t define the final profile of mVOCs to be analyzed to identify a contaminated area but can recommend some of them. Hence, multiple, supposedly contaminated, workplaces should be studied in order to be able to draw conclusions.
Table 5
Pearson correlation (R) for mVOCs with concentrations above the LOQ.
| Decanal | Non-3-en-1-ol | Nonan-2-one | Oct-1-en-3-ol | Octan-1-ol | Octan-3-one | Pent-1-en-3-ol | Pent-2-en-1-ol | Pentan-2-one | Cyclo-pentanone | Cyclo-hexanone | Butan-2-one | Ethyl-acetate |
Decanal | 1.000 | 0.915 | 0.957 | 0.790 | 0.902 | 0.641 | 0.995 | 0.710 | 0.656 | 0.525 | -0.665 | -0.889 | -0.210 |
Non-3-en-1-ol | | 1.000 | 0.881 | 0.696 | 0.859 | 0.581 | 0.907 | 0.600 | 0.598 | 0.497 | -0.582 | -0.881 | -0.231 |
Nonan-2-one | | | 1.000 | 0.907 | 0.964 | 0.705 | 0.943 | 0.839 | 0.617 | 0.624 | -0.590 | -0.887 | -0.088 |
Oct-1-en-3-ol | | | | 1.000 | 0.951 | 0.580 | 0.755 | 0.974 | 0.416 | 0.663 | -0.420 | -0.782 | 0.022 |
Octan-1-ol | | | | | 1.000 | 0.584 | 0.874 | 0.890 | 0.502 | 0.608 | -0.531 | -0.917 | -0.089 |
Octan-3-one | | | | | | 1.000 | 0.637 | 0.624 | 0.879 | 0.834 | -0.050 | -0.404 | 0.471 |
Pent-1-en-3-ol | | | | | | | 1.000 | 0.671 | 0.663 | 0.487 | -0.693 | -0.880 | -0.225 |
Pent-2-en-1-ol | | | | | | | | 1.000 | 0.474 | 0.752 | -0.262 | -0.680 | 0.204 |
Pentan-2-one | | | | | | | | | 1.000 | 0.733 | -0.069 | -0.414 | 0.435 |
Cyclopentanone | | | | | | | | | | 1.000 | 0.150 | -0.365 | 0.534 |
Cyclohexanone | | | | | | | | | | | 1.000 | 0.710 | 0.724 |
Butan-2-one | | | | | | | | | | | | 1.000 | 0.297 |
Ethyl-acetate | | | | | | | | | | | | | 1.000 |
3.8. Total mVOCs
Some researchers have used calculated total mVOCs levels, as indicative of mold growth or contamination in indoor air (Araki et al., 2010; Kim et al., 2007; Persoons et al., 2010; Rodríguez-Navas et al., 2012; Sahlberg et al., 2013; Wieslander et al., 2007). Kim et al. have calculated the total concentration of 15 mVOCs in air of schools and outdoors. They found total mVOCs of 0.423 µg/m3 in the air of schools and 0.123 µg/m3 outdoor and the difference was significant (Kim et al., 2007). Wieslander team evaluated mVOCs in control and damp buildings and reported average total concentration for 11 mVOCs, excluding butanol, varying from 0.097 µg/m3 in their control to 0.152 µg/m3 in a damp building (Wieslander et al., 2007). In five solid waste treatment plants, working environments related to the WSP, Rodriguez-Navas et al. reported average total mVOCs concentrations of 93 compounds varying between 77 and 1477 µg/m3 of air depending on the type of waste treated (sludge, compost, plants) (Rodríguez-Navas et al., 2012). Persoons et al. measured the total mVOCs of 27 compounds in two composting plants. The average total mVOCs concentrations varied from 100 to 27 000 µg/m3 in the green waste and from 10 to 40 000 µg/m3 in the bio-waste (Persoons et al., 2010). Müller et al. reported the total average concentration of 27 mVOCs in three different composting facilities which varied between 5.42 (facility A) and 34.89 (facility B) µg/m3 at the compost piles sites, between 0.96 (facility C) and 47.61 (facility B) µg/m3 at the sieving sites and between 6.94 (facility C) and 14.33 (facility B) µg/m3 on the top of biofilters (Müller et al., 2004). The calculated total mVOCs for the 13 compounds measured in the present study are presented in Table 6. It can be difficult to compare total mVOCs between studies, however, the evaluation of total mVOCs in one study can aid identify the environments with the greatest concentrations and perhaps recognize where microbial presence could be a problem. The research team wanted to highlight the potential of total mVOCs as an evaluation parameter when comparing different environments. Comparisons are possible if the protocols are identical, however, they will present serious limitations when the number and type of mVOCs diverge and in totally different settings such as these ones.
In the present study, the pre-sorting station is, definitely, the workplace with the highest mVOCs concentrations, both during spring and summer. As for the two other sorting stations, high concentrations are calculated mostly during the summer campaigns.
This study allows to restrict the choice of the mVOCs that can be considered as indicators of mold exposure in indoor environment. In fact, ethyl acetate, butan-2-one and cyclohexanone should be, if not removed from the list, considered with great care given that they showed higher concentrations at the UC, which is a new building, an environment presumed to be low in mold contamination, making these chemicals likely more associated with VOC emissions from non-microbial sources rather than indicators of mold exposures. A list of 9 remaining mVOCs (octan-3-one, pent-1-en-3-ol, decanal, nonan-2-one, oct-1-en-3-ol, octan-1-ol, cyclopentanone, non-3-en-1-ol, pent-2-en-1-ol) could therefore be considered as best candidates for markers of molds exposure.
Table 6
Total mVOCs calculated for the spring (n = 50, both sites) and summer (n = 50, both sites) campaigns for all the workstations at the UC and the WSP with sampling duration of 2 hours.
| Total mVOCs |
| Spring | Summer |
| µg/m3 of air | µg/m3 of air |
Geo-Lab | 1.96 | 1 |
IT-lab | 2.62 | 1.56 |
Cafeteria | 2.66 | 2.18 |
Classroom-1201 | 2.36 | 4.95 |
Classroom-2061 | 3.35 | 2.74 |
Reception | 3.04 | 3.25 |
Ballot | 7.39 | 6 |
Optic sorting | 7.04 | 18.83 |
Sorting | 5.56 | 19.74 |
Pre-sorting | 12.89 | 25.74 |
This study allows to restrict the choice of the mVOCs that can be considered as indicators of mold exposure in indoor environment. In fact, ethyl acetate, butan-2-one and cyclohexanone should be, if not removed from the list, considered with great care given that they showed higher concentrations at the UC, which is a new building, an environment presumed to be low in mold contamination, making these chemicals likely more associated with VOC emissions in new building rather than indicators of mold exposures. A list of 9 remaining mVOCs (octan-3-one, pent-1-en-3-ol, decanal, nonan-2-one, oct-1-en-3-ol, octan-1-ol, cyclopentanone, non-3-en-1-ol, pent-2-en-1-ol) could therefore be considered as best candidates for markers of molds exposure.
Additional studies will continue in different workplaces (composting centers, agricultural sectors, etc.) to study and measure molds and mVOCs in the air and mVOCs in biological matrices (urine, blood, exhaled air) of exposed workers to further verify the link of specificity between these mVOCs and molds and to confirm their potential use as biomarkers of mold exposure. Moreover, more mVOCs may be added or removed from this list in the future.
3.9. Limitations
This study has certain limitations. Although there are multiple standardized methods for the analysis of VOCs, to date there is none specific for mVOCs in air. Many of the mVOCs selected for this study do not have air concentrations reported in the literature for comparison purposes. It is also complex to select mVOCs that are entirely specific to microbial emissions and the results obtained at UC are proof of this. The distinction between mVOCs and VOCs can therefore be difficult because the sources may be overlapping in the environment. Since it is still not known whether it is the average exposure or the maximum exposure (peak) that produce the health effects, it may be difficult to select the best sampling duration. In order to achieve sufficient quantities of compounds for the laboratory analysis, two-hour sampling duration was selected. However, despite this choice, several compounds remained below the limits of quantification for all samples. This choice also precludes the peak exposure evaluation. Consequently, our analysis is qualitative (marker of exposure) rather than quantitative (measure of individual exposure).