Non-enantioselective and enantioselective determination of microbial volatile organic compounds as tracer for human exposure to mould growth in buildings

In the industrialized countries people spend 80 to 90 % of their time in indoor areas. About
5 % of the population are known to be sensitive or allergic to the more than 200 mould
species that are found in the indoor environment today.
Identification of mould attacks in the indoor environment, however, is difficult and generally
performed by sampling and counting spores and conidia. These procedures are highly
dependent on seasonal and environmental parameters.
The chemical analysis of indoor mould mainly focuses on microbial volatile organic
compounds (MVOC) in the ambient air. These compounds are present during all stages of the
fungal life cycle, are able to penetrate weak barriers (e.g. wallpaper), and can distribute into
all regions of the indoor environment. The most commonly reported MVOCs are
hydrocarbons (e.g. octane), alcohols (e.g. 2-methyl-1-butanol), aldehydes and ketones (e.g.
octan-3-one), esters (e.g. ethyl acetate), ethers and furans (e.g. 2-methylfuran), terpenes and
terpene derivatives (e.g. geosmin), nitrogen and sulphur compounds (e.g. pyridine and
dimethyl disulfide). A few MVOCs can act as universal fungal signature compounds and sum
up to a characteristic marker pattern.
In this work a method was developed to detect mould within a building by tracing and
quantifying selected MVOCs in the indoor air. The method was supposed to be applicable at
any time and independent of the fungal life cycle. 22 characteristic compounds were chosen
as reference compounds to indicate mould contamination even when other signs of microbial
growth could not be detected.
Sampling was performed by passive sampling to profit from the advantages above active
sampling as simplicity (of field operation), low cost, no need for expensive or complicated
equipment, no power requirement, unattended operation, and time-weighted averaged (TWA)
concentration of the analyte to gain a representative overview of the sampling site. The
passive sampler used throughout this work was the 3M organic vapour monitor (OVM) 3500.
This badge-type sampler is a combination of a diffusion and a permeation sampler with an
activated charcoal adsorbent and an average sampling rate of 30 mL min-1. A sampling period
of 28 days was chosen to ensure sufficient detection limits for the trace components. Solvent
desorption was performed with diethyl ether.
Separation of the selected 22 MVOCs was carried out by conventional non-enantioselective
and enantioselective high resolution gas chromatography (HRGC). The detection was
performed by mass spectrometry (MS), but for method development also a flame ionization
detector (FID) was applied. The detection of the trace concentration required the selected ion
monitoring (SIM) mode of the MS.
Cold on-column injection showed to be most suitable for analyzing the MVOC ether solution.
The non-vaporizing technique reduced the thermal stress of the analytes and ensured a
quantitative transfer of the solutes into the column. A routinely installed retention gap was
used to achieve optimal focussing of the compounds within the capillary column.
Best non-enantioselective chromatographic separation was obtained by a 30 m long DB-Wax
capillary (0.25 mm inner diameter and 0.25 μm film thickness) combined with a 30 min
temperature program.
The enantioselective analysis of the MVOCs should allow the conclusive differentiation
between naturally occurring MVOCs and synthetic/anthropogenic racemates.
11 enantioselective columns from 4 manufacturers were tested with different stationary
phases and/or different combinations of chiral selector and polysiloxane solvent. The best
enantioselective separation was achieved with the BGB-174, a heptakis-(2,3-di-O-acetyl-6-Otert.-
butyldimethylsilyl)-β-cyclodextrin dissolved in BGB-1701 (14 % cyanopropylphenyl
86 % dimethyl polysiloxane). It was able to resolve 13 of the 14 monitored chiral MVOCs.
Quantification of the target compounds accumulated on the passive sampler was performed
using the internal standard method. Of the evaluated standards 1-chlorohexane was best
suitable as internal standard (ISTD) and chlorocyclohexane as recovery standard (RSTD). The
utilized non-enantioselective GC-MS method was validated as a reliable semiquantitative
method for trace analysis of MVOCs. It was linear over a concentration range of 0.01 to
5 ng μL-1 with a coefficient of determination (r²) between 0.96 and 0.99. The recovery rate
ranged from 40 to 127 % for the majority of compounds. The between-runs precision was 2 to
7 %. The limit of detection (LOD) was 1 to 86 pg μL-1 (S/N-ratio 3:1) and the limit of
quantification (LOQ) was 2 to 286 pg μL-1 (S/N-ratio 10:1).
The developed method as described above was successfully able to detect possible fungal
contamination on a real case site. 16 of the 22 MVOC compounds contributed to the
analytical fingerprint pattern and indicated possible fungal contamination. However, some
problems remained. The VOC burden of the indoor air was tremendously high and showed a
great variety of compounds that interfered with the MVOC detection. Furthermore, the
enantioselective GC-MS analysis of the MVOCs in indoor air was unusable for the given
concentration range. The sensitivity of the enantioselective method was too low to
unequivocal differentiate the source of the detected compounds. These problems should be
investigated further before enantioselective GC-MS analysis in combination with passive
sampling of indoor air can become a reliable, easy, and cheap way of detecting indoor mould.