1、Volume 16, Number 2, March 2010An International Journal of Heating, Ventilating,Air-Conditioning and Refrigerating ResearchAmerican Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc.Volume 16, Number 2, March 2010HVAC accepted December 4, 2009This paper is based on findings resul
2、ting from ASHRAE Research Project RP-1243.Musty odors, often associated with damp or water-damaged buildings, originate from therelease of microbial volatile organic compounds (MVOCs) from mold growing on buildingmaterials and construction substrates. Chemical analysis of air samples is a feasible w
3、ay to sup-plement conventional bioaerosol techniques during building investigations. Analytical method-ologies for MVOCs are straightforward; however, development of a scientifically validatedmethod to measure unique MVOCs that indicate with high confidence the presence of hiddenmold regardless of t
4、he amount of mold present remains a challenge.Laboratory studies identified and quantified specific MVOCs associated with various moldspecies and MVOCs generated by specific molds growing on selected building materials in sim-ulated, realistic conditions.This research determined that numerous MVOCs
5、are released from active mold growth andare dependent on both the type of mold and the host substrate. MVOC profiles generated by 32combinations of various molds and materials were determined, but only a few of these com-pounds demonstrated effectiveness in field/building studies. Certain mold-selec
6、tive MVOCswere identified as potential indicators for specific mold, including methoxybenzene for Stachy-botrys chartarum and benzothiazole and menthol for Chaetomium globosum. These studies pro-vided a firm foundation for continued research of mold-specific MVOC markers as indicators ofhidden mold
7、and as predictors of potential mold sources in problem buildings.INTRODUCTIONIt is well known that fungal growth produces emissions as a result of secondary metabolicprocesses (Horner and Miller 2003). These microbial volatile organic compounds (MVOCs)represent a variety of chemical classes includin
8、g alcohols, amines, aldehydes, ketones, sulfides,and many other hydrocarbons (Claeson et al. 2006; Lancker et al. 2008). Therefore, in order toeffectively apply MVOC analysis to building investigations, MVOCs that are identified asmold-indicators must be unique. In other words, the MVOCs should not
9、be among the hundredsof common chemicals that emit from building materials and consumer products but should bespecific indicators of mold growth. While the literature contains studies that have been per-formed to identify MVOC emissions on building materials contaminated with mold, many ofthem do no
10、t reference the use of un-inoculated materials (negative controls); thus, it is not clearStephany Mason is technical director and W. Elliot Horner is principal consultant at Air Quality Sciences, Inc. in Mar-ietta, GA. Don Cortes is laboratory director for STAT Analysis in Chicago, IL.01_Mason.fm Pa
11、ge 109 Monday, March 1, 2010 8:49 AM 2010, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in HVAC previous studies suggest that indoor emissions may be diluted tooquickly to reliably detected (Schleibinger et al. 2005; Schleibinger et al.
12、2008). Identification ofspecific target compounds may provide an opportunity to increase the sensitivity of the analyti-cal method and detect MVOCs when even a minimal amount of mold is present. Hence,extremely important and novel components of this research are the determination and validationof an
13、 MVOC sampling and analysis method.The objectives of this research were (1) to develop a database of MVOCs that are associatedwith types of mold growth found in problem building environments and that would be useful indetermining the presence of hidden mold growing in indoor environments and (2) to
14、accuratelydetermine MVOC emissions from building materials inoculated with mold and exposed undersimulated realistic environmental conditions (temperature, relative humidity, and ventilation airchange rate).METHODSFirst, to identify and quantify specific MVOCs associated with certain organisms grown
15、 ondifferent materials, selected species were grown and isolated in laboratory glass vessels forstatic studies. Following growth and incubation of the molds, air samples were obtained fromthe glass vessels using a passive volatile-organic-compound (VOC) collection technique andanalyzed using thermal
16、 desorption-gas chromatography/mass spectrometry (GC/MS). Specificemissions were identified using a mass spectrometric database of common indoor contaminantsand MVOCs. Potential marker compounds were chosen for specific mold types based on theuniqueness and levels of the identified MVOCs.Second, dyn
17、amic chamber studies were used to accurately determine MVOC emissions frombuilding materials inoculated with mold and exposed under realistic environmental conditions(temperature, relative humidity, and ventilation air change rate). Inoculated materials wereplaced into environmentally controlled cha
18、mbers operating under dynamic conditions. Chamberair was sampled using an active VOC collection technique and was analyzed using the VOCanalysis method previously described.Procurement of Mold-Contaminated Building MaterialsStatic Studies and Dynamic Chamber StudiesSix mold species, typical to water
19、-damaged buildings, were selected for use in static studies:Stachybotrys chartarum, Cladosporium sphaerospermum, Chaetomium globosum, Eurotiumamstelodami, Aspergillus versicolor (tested in duplicate), and Aspergillus sydowii. S. chartarumis widely publicized as a mold of great concern from a health
20、standpoint, and C. globosum iscommonly found in water-damaged environments. Furthermore, the two molds are often foundtogether in the environment. Based on these factors and on results from the static studies, only S.chartarum and C. globosum were used in subsequent dynamic chamber studies.Cultures
21、were freshly obtained from a commercial laboratory conducting analysis of air sam-ples on mold-colonized building samples. Isolated cultures of each organism were plated ontoappropriate growth media. S. chartarum, C. globosum, and C. sphaerospermum were cultivatedon PDA. CY20S was used for A. sydowi
22、i and E. amstelodami. A. versicolor was grown onB-malt. The inoculated plates were incubated at 25C for 5 to 7 days or until after sufficient col-onization and sporulation. Each organism was harvested and suspended in 0.01% Tween 80 with0.1% peptone (PepTween). Suspensions were vortexed for 2 min an
23、d then centrifuged for 20min at a minimum 3500 rpm. The pellet in each tube was washed twice and was suspended insterile 0.9% NaCl. The number of spores per milliliter of stock suspension was counted using a01_Mason.fm Page 110 Monday, March 1, 2010 8:49 AM 2010, American Society of Heating, Refrige
24、rating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in HVAC thismaterial was not chosen for subsequent dynamic chamber studies. While ceiling tile is often asource of mold growth in buildings, its installation and use often is easily accessible for visualinspection of mold on eit
25、her side of this material. Since one purpose of MVOC sampling istargeted at identifying hidden mold growth, then the use of ceiling tiles as substrates for fur-ther study did not seem reasonable. Kraft paper and gypsum wallboard met two key criteriafor MVOC study: 1) they readily support mold growth
26、 in problem indoor environments, and2) they are installed and used in buildings in a manner that results in the subsequent moldgrowth being concealed. Based on the prevalent use of gypsum wallboard in buildings, itspotential to account for mold coverage (in square footage) in problem environments is
27、 signif-icant. Since it has the potential to be a significant host for hidden mold growth, gypsum wall-board was chosen as the subject substrate for the dynamic chamber studies.For the static studies, all test materials were cut to fit into each 3 in. by 5 in. test container, aglass jar with a Teflo
28、n-lined lid. The sample coupons ranged in size from 1 in. by 1 in. to 2 in. by 3 in. Materials were sterilized in dry heat for 2 hr at 121C and inoculated separately with 2 mL ofa 7.0105to 2.0106spores/mL inoculum of each separate fungal culture. Thirty-two test envi-ronments were created; sample se
29、ts of each of four types of building materials were moistenedand inoculated separately with each of six test organisms (one in duplicate) plus one un-inocu-lated control and were then incubated for three weeks at 25C. Methods from the static studies were optimized for application to dynamic chamber
30、studies.In preparation for these dynamic studies, test materials were evaluated weekly to determine thegeneral rate of fungal colonization per test organism and material type. Materials were visuallyassessed and growth was rated according to the following scale: “0” = no visual growth observed,“1” =
31、 trace amount of growth observed (less than 10% coverage of material surface), “2” = lightgrowth observed (1130% coverage of material surface), “3” = moderate growth observed(3160% coverage of material surface) and “4” = heavy growth observed (61100% coverage ofmaterial surface). Periodic visual ins
32、pection of the test containers ensured that they containedsufficient moisture to sustain growth of the fungal species; 0.9% NaCl was added to containersas needed.For the dynamic chamber studies, 1.2 cm thick (nominal -inch) commercial gypsum wall-board panels were cut into coupons approximately 30 c
33、m 30 cm. Five sets of wallboard cou-pons were prepared with each set consisting of five replicates (a total of 25 pieces). All panelswere wetted by submersion in 18 megohm purified water for 30 min and were subsequentlyair-dried for 30 min prior to use (to minimize drainage of inoculum from the surf
34、ace). Foursets were then inoculated with aqueous suspensions of spores (1-2105CFU/mL) of either C.globosum, S. chartarum, or a mixture of the two molds. Panels were sprayed with a manualatomizer. Inoculated panels were incubated at room temperature in a humid chamber (approx-imately 100% RH) for eit
35、her nine days (light growth) or 16 days (heavy growth) before load-ing into the dynamic environmental chambers for further evaluation. The control(uninoculated) panels were wetted and dried in an identical manner to the inoculated panels,immediately prior to loading into the chambers. The five treat
36、ment sets (5 coupons each) were01_Mason.fm Page 111 Monday, March 1, 2010 8:49 AM 2010, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in HVAC the other end of the tube was sealed with a cap.For population verification, a sample from each
37、test container was removed at the appropriatetime period and used to determine the fungal population per weight of material. Organisms fromeach material were harvested in PepTween, serial-diluted, plated onto appropriate media, andincubated at 25C for four to seven days.Exposure controls were establ
38、ished for the static studies. In order to better distinguish build-ing material emissions from true MVOCs, the six species (plus one uninoculated control) werealso cultivated on appropriate growth media in specially prepared “French Squares,” wherebymedia was applied to four sides of this square ves
39、sel. Samples acquired from these vessels wereconsidered positive controls. A sample of the atmosphere in these vessels was measured aftereight days of exposure.Sampling for VOCs and Target MVOCs fromDynamic Environmental ChambersThe exposed area of inoculated wallboard was approximately 1 m2resultin
40、g in a mate-rial-to-air-volume-loading ratio of approximately 10 m2/m3. Environmental chamber operationand control measures complied with ASTM Standards D 5116-97 (ASTM 1997). Using a com-bination of vapor phase and solid media filtration, dehumidified supply air to the chamber wasstripped of formal
41、dehyde, VOCs, particles, and other contaminants so that any contaminantbackgrounds in the empty chamber were below strict levels of 10 g/m3TVOC, 10 g/m3total particles, 2 g/m3formaldehyde, and 2g/m3 for any individual VOC. Air supply to thechambers was maintained at 23C 2C and 50% RH 5% RH. The back
42、ground from the empty chamber was measured prior to loading panels into thechamber. After an overnight equilibration period set at a minimum air change rate of 0.4 ACH toconcentrate the sample for analysis of the headspace, duplicate air samples were collected ontoTenax sorbent tubes. Then the air c
43、hange rate was adjusted to 0.8 ACH followed by a 4-hrre-equilibration period in the chamber. The 0.8 ACH rate was chosen to represent a typical com-mercial building ventilation rate. Duplicate air samples were collected at 0.8 ACH. Repeat sam-ples were collected at both 0.4 ACH and 0.8 ACH rates.Cha
44、mber air was sampled over an approximately 90-min period for a total volume of approx-imately 18 L or 24 L. Both low-volume (18 L) and high-volume (24 L) air samples were col-lected on two consecutive days of growth. Different volumes were used to allow collection ofVOCs across a broad range of vapo
45、r pressure.Analytical Methodology for VOCsCollected samples were thermally desorbed into the GC/MS. Instrumentation included a Per-kin-Elmer Turbo Matrix ATD or ATD 400 Thermal Desorption System, a Hewlett-Packard 5890Series II or 6890 Series Gas Chromatograph, and a Hewlett-Packard 5971 or 5973 Mas
46、s SelectiveDetector GC/MS.The sorbent collection technique, separation and detection analysis methodology was adaptedfrom techniques presented by the USEPA and other researchers. The technique followed EPAMethod IP-1B, generally applicable to C5-C16 organic chemicals with boiling points ranging01_Ma
47、son.fm Page 112 Monday, March 1, 2010 8:49 AM 2010, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in HVAC Bruner et al. 1978; Mangani et al. 1982; Winberry et al.1990). The detection limit was 0.5 g/m3for most individual VOCs and total VO
48、Cs (TVOCs).The detection limit for MVOCs was 0.2 g/m3. The relative standard deviation of duplicate sam-ples was acceptable if it was less than 20% and typically it fell between 10 and 15%. Analyticalerror was estimated at less than 20% based on historical accuracy and precision data. Chamber andlab
49、oratory blanks are routinely analyzed as part of the overall laboratory quality assurance pro-gram. Spiked sorbent tubes with a second sorbent tube connected in series were evaluated toensure that all MVOCs were retained on the collection media at the flow rates used and at the con-centration levels expected.Individual VOCs (IVOC) were separated and detected by GC/MS. VOCs were calibrated toauthentic standards for the MVOC quantitation procedure and to toluene for the IVOC quantita-tion procedure.The data was analyzed first using a target list approach for 23 kno
